Tests - bowling August 10, 2010

The fearsome fifteen: a look at the great bowlers

A detailed statistical look at the greatest Test bowlers
168

Muralitharan: statistically the best bowler © AFP

As I had mentioned in my recent article on Muralitharan, this is the right time to have an in-depth look at the great bowlers of our times and before. We have just seen the retirement of one of the all-time great bowlers and the next active bowler is, in Formula-1 parlance, three laps behind.

I have selected the following fifteen bowlers for analysis. Readers might have their own favourites. However I believe that this collection contains the best ever bowlers, barring personal preferences. I have also tried hard to have as wide a selection as possible, covering all the major test playing countries. I did a Test bowlers analysis last year. However this one has removed all the warts from that and is also at the career-end stage for all the fifteen bowlers. I have a major advantage in doing this analysis. Since the analysis is limited to 15 bowlers, I do not need to worry about the wide disparity which existed between around 60 bowlers who were part of my previous analysis.

Australia:      Shane Warne
Glenn McGrath
Dennis Lillee
West Indies:    Malcolm Marshall
Curtley Ambrose
Joel Garner
Pakistan:       Imran Khan
Wasim Akram
Waqar Younis
India           Anil Kumble
Kapil Dev
Sri Lanka:      Muralitharan
New Zealand:    Richard Hadlee
South Africa:   Allan Donald
England:        Fred Trueman.

A few notes on the selections. I did not select any pre-WW2 bowlers since their figures would have distorted the numbers considerably, especially Sydney Barnes. The three Australian bowlers select themselves. For West Indies I selected their three best bowlers over the past 30 years. Holding for Garner could be an option. However Garner has a bowling average of 20.98 against Holding's 23.69. First I had selected only Imran Khan and Wasim Akram for Pakistan. Then I realized that Waqar Younis led in one key measure (strike rate) and there was no way could miss him. I did not want to drop Wasim Akram since he is the only left-hander in this group.

Kumble selects himself while Kapil Dev, although he has average overall figures deserves a place since he was the leading wicket-taker for quite some time and changed the face of Indian fast bowling. Shaun Pollock is the only one who could challenge Donald's place. I selected Fred Trueman to represent the period between 1947 and 1969. He is also the best modern English bowler. I also followed the basic principle that any bowler who was the best in a key category (wkts/strike rate/rpo) cannot be left out. One could find justifications for including Walsh, Holding, Shaun Pollock, Willis et al. However this is my selection and not a bad one at that. Let me also add that this is not necessarily the top bowlers list. I am sure Holding, Roberts, Pollock et al would be placed higher than a few in this list. However I wanted to have as wide a representation as possible and restricted one country to 3 bowlers.

Now for the measures on which the rating work is done. First let me clarify two fundamental differences to the way I have done such exercises before.

The first is that I have selected ten measures and given equal weight to all. That way I reduce the chances of subjective valuations.

The next is that for each measure, the best gets the maximum points, viz., 10.0 and the others get proportionate points. This reduces the possibility of differential weights and ensures a fair allocation of points.

The following measures are used.

1. Wickets captured.
2. Bowling strike rate (Bpw).
3. Bowling accuracy (Rpo).
(These two are components of the Bowling average, but have been considered
independently).
4. Quality of wickets captured (Average of dismissed batsmen's batting averages)
5. Away bowling average.
6. % of top order wickets captured.
7. Team load borne by bowler (balls bowled and wickets captured).
8. Ratio of bowling average to peer bowling average - all teams.
9. Ratio of bowling average to peer bowling average - other bowlers of own team.
10.Win index (Combination of two ratios)- (% of win wickets to career wickets
and % of win wickets to team win wickets).

Just to summarize, Muralitharan leads in 2 measures (Wickets and Team load factor). The other 8 measures are led by Ambrose (RpO), Garner (Away bowling average), Imran Khan (Dismissed batsman quality), Waqar Younis (Bowling strike rate), Marshall (Peer comparison to all bowlers), Hadlee (Peer comparison to own team bowlers), McGrath (Top order wickets ratio) and Warne (Win index). A very fair distribution of the top positions with nine bowlers leading in one or more measures. There is no domination by one bowler.

There is one point worth mentioning here. There have been a number of comments about the wickets captured by Muralitharan against Bangladesh and Zimbabwe. I have thought long and hard and decided not to do anything about it. What is the definition of a weak team. India against Trueman in 1952 were much worse than Bangladesh against Muralitharan. England in 1984 were the rabbits against Marshall. How do we value the English wickets against Australia in 2006. It pains me to say this, what about Pakistan during the past few months. And finally where do we place Sri Lanka itself during its first 10 years. Let us not forget that Warne did not bowl against the strong Australian batsmen, nor Kumble against the great Indian line-up and so on. However the fact that the bowlers in the strong batting lineups did not bowl against their own lineups does not make them any less bowlers. Let Muralitharan not get penalized for playing in a weaker team.

If Muralitharan captured wickets against a team including the Flower brothers, Campbell and Goodwin, how can anyone downgrade these wickets. Where do we draw the line. A test wicket is a test wicket. There are enough measures built in to take care of wickets of batsmen of lower quality. If we start down-valuing performances against Bangladesh, what about Tendulkar's recent 105 out of 243, one of his best ever. For that matter, Gilchrist's 144 or Inzamam's 138 were all truly great match-winning innings. So let us put to rest this red herring, once and for all.

A final point to ponder for those doubting Thomases. Zimbabwe have won 8 of their 83 tests played. India won 8 out of their first 83 tests and New Zealand won 4 of their first 83 tests. Nice to remember that the first away test India won was past the 100th test they played (during 1968). Both Zimbabwe and Bangladesh have won away tests well before this number.

I understand that this analysis favours the fast bowlers slightly. This is primarily because fast bowlers' strike rates are lower and they have a better chance of capturing top order wickets. However we have to consider the contribution to team causes and taking top order wickets and having lower strike rates are very essential to the team cause. As far as Strike rates and Rpo are concerned there is no need to do any adjustment since these are all great bowlers. Only two of these bowlers, Kumble and Kapil Dev have strike rates exceeding 60. Muralitharan and Warne have strike rates comparable to the pace bowlers. Surprisingly the bowler with the best Rpo figure is Ambrose. This clearly shows that there is no need to do any special adjustment.

Now for the tables.

1. Career wickets captured

Bowler            Value   Points

Muralitharan M 800 10.00 Warne S.K 708 8.85 Kumble A 619 7.74 McGrath G.D 563 7.04 Kapil Dev N 434 5.43 Hadlee R.J 431 5.39 Wasim Akram 414 5.18 Ambrose C.E.L 405 5.06 Marshall M.D 376 4.70 Waqar Younis 373 4.66 Imran Khan 362 4.53 Lillee D.K 355 4.44 Donald A.A 330 4.12 Trueman F.S 307 3.84 Garner J 259 3.24

This table is self-explanatory.

2. Bowling strike rate (Bpw)

Bowler            Value   Points

Waqar Younis 43.49 10.00 Marshall M.D 46.77 9.30 Donald A.A 47.03 9.24 Trueman F.S 49.44 8.72 Hadlee R.J 50.85 8.42 Garner J 50.87 8.41 McGrath G.D 51.95 8.18 Lillee D.K 52.02 8.17 Imran Khan 53.75 7.79 Ambrose C.E.L 54.58 7.62 Wasim Akram 54.66 7.60 Muralitharan M 55.05 7.52 Warne S.K 57.49 6.99 Kapil Dev N 63.92 5.61 Kumble A 66.00 5.16

Waqar Younis is the bowler with the lowest strike rate. It is not surprising to see that the three spinners and Kapil Dev prop up the table, although Muralitharan and Warne are not too far from Wasim Akram.

3. Bowling accuracy (Rpo)

Bowler            Value   Points

Ambrose C.E.L 2.31 10.00 Garner J 2.47 9.12 Muralitharan M 2.48 9.10 McGrath G.D 2.50 8.98 Imran Khan 2.55 8.74 Wasim Akram 2.59 8.49 Trueman F.S 2.62 8.35 Hadlee R.J 2.63 8.29 Warne S.K 2.65 8.18 Marshall M.D 2.69 7.99 Kumble A 2.70 7.95 Lillee D.K 2.76 7.61 Kapil Dev N 2.78 7.49 Donald A.A 2.84 7.19 Waqar Younis 3.25 5.02

One would have expected a spinner like Murali to head this table. It is a surprise that the often-underrated Ambrose is on top with a Rpo value of 2.31. Waqar Younis has topped the previous table and is propping up this table.

4. Average quality of wickets captured

Bowler            Value   Points

Imran Khan 25.76 10.00 Lillee D.K 24.58 9.54 Kapil Dev N 23.83 9.25 Hadlee R.J 23.64 9.18 Marshall M.D 22.97 8.92 Garner J 22.18 8.61 Kumble A 20.51 7.96 Waqar Younis 20.34 7.89 Muralitharan M 20.27 7.87 Donald A.A 20.06 7.79 Ambrose C.E.L 20.01 7.77 Wasim Akram 19.56 7.59 McGrath G.D 19.22 7.46 Warne S.K 18.48 7.17 Trueman F.S 17.80 6.91

Imran Khan's wickets were of high quality as proved by his high placing here. Warne's and McGrath's low placing is the result of many late order wickets and the fact that they played for a very strong batting lineup. Muralitharan is also pulled down by the number of wickets against the weaker teams. However his figures are still higher than those of Warne.

5. Away wicket bowling average

Bowler            Value   Points

Garner J 20.33 10.00 Ambrose C.E.L 20.39 9.98 McGrath G.D 21.23 9.70 Marshall M.D 22.25 9.35 Hadlee R.J 22.37 9.31 Donald A.A 23.51 8.93 Wasim Akram 23.97 8.77 Lillee D.K 24.18 8.70 Warne S.K 25.27 8.33 Trueman F.S 26.16 8.04 Imran Khan 26.29 7.99 Waqar Younis 26.29 7.99 Muralitharan M 27.02 7.74 Kapil Dev N 32.69 5.83 Kumble A 37.36 4.26

Garner has captured his away wickets at an incredible average of 20.33, closely followed by Ambrose with an away average of 20.39. In fact Garner, Ambrose, McGrath and Warne have an away bowling average which is better than their home bowling average. Kumble's away bowling average was abysmal. Look at the huge daylight which exists between Murali and Kapil/Kumble.

6. % of top order wickets captured

Bowler            Value   Points

McGrath G.D 0.501 10.00 Donald A.A 0.497 9.92 Kapil Dev N 0.494 9.84 Ambrose C.E.L 0.475 9.46 Lillee D.K 0.468 9.33 Imran Khan 0.462 9.21 Trueman F.S 0.453 9.04 Marshall M.D 0.445 8.87 Hadlee R.J 0.439 8.75 Waqar Younis 0.437 8.72 Garner J 0.406 8.09 Wasim Akram 0.394 7.86 Kumble A 0.383 7.64 Muralitharan M 0.350 6.99 Warne S.K 0.318 6.35

Just over half of the wickets captured by McGrath are top order wickets. The three spinners prop up the table, with figures around a third of the total wickets. This is to be expected.

7. % of team load borne

Bowler            Value   Points

Muralitharan M 0.361 10.00 Hadlee R.J 0.298 8.26 Kumble A 0.296 8.21 Warne S.K 0.281 7.77 Lillee D.K 0.275 7.62 Imran Khan 0.267 7.40 Donald A.A 0.253 7.01 Wasim Akram 0.250 6.92 Marshall M.D 0.249 6.89 McGrath G.D 0.247 6.85 Ambrose C.E.L 0.244 6.76 Garner J 0.243 6.74 Trueman F.S 0.242 6.71 Waqar Younis 0.239 6.63 Kapil Dev N 0.222 6.14

This is a composite value taking into account the work load in terms of balls bowled and the resultant wicket captures. Muralitharan is way out on top having shared captured 39% of his team wickets and bowling 33% of his team balls. 6 of the pace bowlers fall below 25%.

8. Peer comparison of average to all peer bowlers

Bowler            Value   Points

Marshall M.D 1.550 10.00 McGrath G.D 1.527 9.84 Ambrose C.E.L 1.514 9.76 Garner J 1.509 9.73 Muralitharan M 1.478 9.53 Hadlee R.J 1.447 9.33 Trueman F.S 1.435 9.25 Donald A.A 1.433 9.24 Imran Khan 1.410 9.09 Waqar Younis 1.363 8.79 Wasim Akram 1.361 8.77 Lillee D.K 1.332 8.59 Warne S.K 1.295 8.35 Kumble A 1.107 7.14 Kapil Dev N 1.085 7.00

This is a comparison of the bowler's bowling average with that of his peers, exactly measured from his first test to last test. Marshall, McGrath and Ambrose stand at over 150%. Kumble and Kapil Dev just about stay either side of 110%.

9. Peer comparison of average to own team peer bowlers

Bowler            Value   Points

Hadlee R.J 1.534 10.00 Muralitharan M 1.418 9.25 Imran Khan 1.377 8.98 Ambrose C.E.L 1.233 8.04 Donald A.A 1.197 7.80 Lillee D.K 1.188 7.74 Kapil Dev N 1.181 7.70 Marshall M.D 1.170 7.63 Trueman F.S 1.169 7.62 Garner J 1.156 7.54 Wasim Akram 1.151 7.51 McGrath G.D 1.149 7.49 Kumble A 1.125 7.33 Waqar Younis 1.123 7.32 Warne S.K 1.017 6.63

This is a comparison of the bowler's bowling average with that of his peers, from his own team, exactly measured from his first test to last test. As expected, the two so called weak-team bowlers, Hadlee and Murali are on top, Hadlee exceeding 150%. Warne just about goes past 100%, which is understandable. A surprise is Kumble's somewhat low figure. This is a reflection of Kumble's somewhat high bowling average nearing 30. This is not too different from his contemporary bowlers.

10. Performance in won matches

Bowler            Value   Points

Warne S.K 0.500 10.00 McGrath G.D 0.492 9.84 Marshall M.D 0.486 9.73 Muralitharan M 0.477 9.54 Lillee D.K 0.451 9.01 Waqar Younis 0.440 8.81 Donald A.A 0.426 8.51 Trueman F.S 0.419 8.38 Ambrose C.E.L 0.414 8.27 Kumble A 0.401 8.01 Hadlee R.J 0.398 7.96 Garner J 0.393 7.86 Wasim Akram 0.384 7.68 Imran Khan 0.364 7.27 Kapil Dev N 0.198 3.96

This is done by doing a composite calculation involving two ratios. The first is the ratio of the bowler wickets in won matches with his own career wickets. The other is the ratio of the bowler wickets in won matches against the team wickets in won matches. This rewards success and as expected, Warne and McGrath lead the pack. Not so surprisingly Kapil Dev is last, indicating that the Indian wins were probably crafted by the spinners during Kapil's era. The presence of Hadlee, Akram, Imran and Kapil in the last five also indicates that their contributions to their team wins was also with the bat. These four are the best batsmen in this elite group of bowlers.

Now for the summary table.

Total  Wkts BowSR  RpO  WkQty AwAvg TORto  TmLd Peer1 Peer2 WinId

Muralitharan M 87.54 10.00 7.52 9.10 7.87 7.74 6.99 10.00 9.53 9.25 9.54 McGrath G.D 85.38 7.04 8.18 8.98 7.46 9.70 10.00 6.85 9.84 7.49 9.84 Hadlee R.J 84.89 5.39 8.42 8.29 9.18 9.31 8.75 8.26 9.33 10.00 7.96 Marshall M.D 83.36 4.70 9.30 7.99 8.92 9.35 8.87 6.89 10.00 7.63 9.73 Ambrose C.E.L 82.72 5.06 7.62 10.00 7.77 9.98 9.46 6.76 9.76 8.04 8.27

Imran Khan 81.00 4.53 7.79 8.74 10.00 7.99 9.21 7.40 9.09 8.98 7.27 Lillee D.K 80.76 4.44 8.17 7.61 9.54 8.70 9.33 7.62 8.59 7.74 9.01 Donald A.A 79.75 4.12 9.24 7.19 7.79 8.93 9.92 7.01 9.24 7.80 8.51 Garner J 79.34 3.24 8.41 9.12 8.61 10.00 8.09 6.74 9.73 7.54 7.86 Warne S.K 78.63 8.85 6.99 8.18 7.17 8.33 6.35 7.77 8.35 6.63 10.00

Trueman F.S 76.86 3.84 8.72 8.35 6.91 8.04 9.04 6.71 9.25 7.62 8.38 Wasim Akram 76.37 5.18 7.60 8.49 7.59 8.77 7.86 6.92 8.77 7.51 7.68 Waqar Younis 75.84 4.66 10.00 5.02 7.89 7.99 8.72 6.63 8.79 7.32 8.81 Kumble A 71.40 7.74 5.16 7.95 7.96 4.26 7.64 8.21 7.14 7.33 8.01 Kapil Dev N 68.24 5.43 5.61 7.49 9.25 5.83 9.84 6.14 7.00 7.70 3.96

It does not matter which route is taken, be it the high road or low road, be it the country road or the highway, be it the scenic route or the road through the concrete jungle, there is only one destination. The one which proclaims Muralitharan as the best amongst the equals. Let no one forget the X-factor here, the impact Muralitharan has had on Sri Lankan cricket. Glenn McGrath comes in next, a well-deserved place for this outstanding metronome. Richard Hadlee, the legendary fast bowler from down under clocks in the third position. Malcolm Marshall, that fearsome Bajan fast bowling "giant" is in fourth position, closely followed by the other Caribbean giant, this time physically also, Curtley Ambrose.

Imran Khan follows next, a well-deserved position for this wonderful bowler, batsman and born leader of men, despite playing in quite a few tests as a batsman. Dennis Lillee, one of the all-time greats, follows next. He is the highest placed bowler who does not lead in any category. Alan Donald is in seventh position, closely followed by the other West Indian giant, Joel Garner. Now comes Shane Warne, a well-deserved top-10 position for this magician.

The third section of the table is led by the fiery Englishman, Trueman. Now we have arguably the best fast bowling pair ever, Wasim Akram and Waqar Younis, paired together here also. The two Indian bowlers come in next. First, Anil Kumble, whose contributions to Indian cricket can be matched by only one player, the little master. One could say the same of Kapil Dev. The fact that these two Indian stalwarts come in at the end indicates that India has traditionally relied on its batting. However these two are definitely India's best bowlers ever.

Postscript:

In response to the readers' requests I have added Holding, Botham, Underwood and Shaun Pollock (my choice) and presented the revised table below. Surprisingly, it is Shaun pollock who comes into the table in the 11th position, followed by Holding in 12th position. Underwood comes in the 16th position and is followed by Botham.

I have since added Gibbs and Walsh in response to a few readers' requests. They have not exactly moved the world. Walsh splits the Pakistani pair while Gibbs has taken over the last place from Kapil dev. I have steadfastly not included SF Barnes since he changes the whole dynamics since he leads in couple of meaasures. This rounds off such requests.

Total  Wkts BowSR  RpO  WkQty AwAvg TORto  TmLd Peer1 Peer2 WinId

Muralitharan M 87.54 10.00 7.52 9.10 7.87 7.74 6.99 10.00 9.53 9.25 9.54 McGrath G.D 85.38 7.04 8.18 8.98 7.46 9.70 10.00 6.85 9.84 7.49 9.84 Hadlee R.J 84.89 5.39 8.42 8.29 9.18 9.31 8.75 8.26 9.33 10.00 7.96 Marshall M.D 83.36 4.70 9.30 7.99 8.92 9.35 8.87 6.89 10.00 7.63 9.73 Ambrose C.E.L 82.72 5.06 7.62 10.00 7.77 9.98 9.46 6.76 9.76 8.04 8.27 Imran Khan 81.00 4.53 7.79 8.74 10.00 7.99 9.21 7.40 9.09 8.98 7.27 Lillee D.K 80.76 4.44 8.17 7.61 9.54 8.70 9.33 7.62 8.59 7.74 9.01 Donald A.A 79.75 4.12 9.24 7.19 7.79 8.93 9.92 7.01 9.24 7.80 8.51 Garner J 79.34 3.24 8.41 9.12 8.61 10.00 8.09 6.74 9.73 7.54 7.86 Warne S.K 78.63 8.85 6.99 8.18 7.17 8.33 6.35 7.77 8.35 6.63 10.00 Pollock S.M 78.05 5.26 6.91 9.52 7.86 8.21 8.82 6.42 9.28 8.18 7.58 Holding M.A 77.60 3.11 8.40 7.45 9.81 9.19 9.14 6.37 8.59 6.95 8.58 Trueman F.S 76.86 3.84 8.72 8.35 6.91 8.04 9.04 6.71 9.25 7.62 8.38 Wasim Akram 76.37 5.18 7.60 8.49 7.59 8.77 7.86 6.92 8.77 7.51 7.68 Walsh C.A 76.07 6.49 6.91 8.80 8.02 8.20 8.77 6.59 8.42 6.94 6.92 Waqar Younis 75.84 4.66 10.00 5.02 7.89 7.99 8.72 6.63 8.79 7.32 8.81 Underwood D.L 71.74 3.71 3.52 11.06 8.97 7.62 8.34 6.59 7.78 7.71 6.43 Kumble A 71.40 7.74 5.16 7.95 7.96 4.26 7.64 8.21 7.14 7.33 8.01 Botham I.T 70.06 4.79 7.10 6.38 8.76 7.08 8.08 6.67 7.24 6.84 7.11 Kapil Dev N 68.24 5.43 5.61 7.49 9.25 5.83 9.84 6.14 7.00 7.70 3.96 Gibbs L.R 65.08 3.86 0.48 11.68 7.79 6.30 6.33 7.00 7.12 6.94 7.57

One final request. Please restrict yourself to comments relevant to the article.

Anantha Narayanan has written for ESPNcricinfo and CastrolCricket and worked with a number of companies on their cricket performance ratings-related systems

Comments have now been closed for this article

  • Harsh Thakor on August 31, 2010, 18:09 GMT

    It is a great effort ,Anantha but I am still convinced that mere statistics does not tell the true story.In the fast bolwing section I would still rate Marshall and Lillee the greatest.They were more agressive than any of the great fast bowlers,with great versatality,considerable speed and ability to intimidate the opposition in alltypes of conditions.Marshall's incredible strike rate at such a low cost,while Lillee's fantastic rate of 5 wickets per test are statistical evidence that suppport it.Adding the Packer and Rest of the World Series would have reflected his greatness more accurately.Mcgrath and Hadlee had greater control,but were not as versatile or menacing in all conditions than Lillee or Marshall.Imran,a great match-winner did not posess Lillee's repertoire or Marshall's unpredicatability.Statistics does injustice to Andy Roberts,morally the greatest rival to Marshall and Lillee,who was rated by opponents to have been the best paceman of his day.

  • Meety on August 31, 2010, 5:36 GMT

    Interesting read. I haven't read all the comments but I was wondering what you think about the sugestion that the the best and worse score gets dropped? (Like in Olympic judging) - just interested to see what it throws up - I think it would make Murali's gap larger & bring Lillee up the list? [[ Andrew I will have to buy additional insurance if I drop the best performance. In Olympics this is done to cater to biased judging. Here great performances have to be recognized. Ananth: ]]

  • raja on August 28, 2010, 15:30 GMT

    gr8 work anantha.u said that u didnt included sf barnes knowingly.do u think that he was the greatest bowler ever, even better than murali?in my opinion he was the don bradman of bowlers.plz reply. [[ Raja Probably yes although you must consider the fact that the bowling conditions were totally different pre-1914. Ananth: ]]

  • wb on August 27, 2010, 10:01 GMT

    At one time, Indian cricket team had a spin quartet just like the West Indies had a pace quartet. Chandrashekhar, Venkatraghavan, Bishen Bedi and Prasanna managed the bowling for India even on non-Indian pitches and gave test wins for india in india and outside India. The spin quartet also had to bat against the fast bowlers of those times with little protective equipment compared to today's and no restrictions on bouncers to be bowled in an over. There was always a risk that any of the 4 spinners would get retired hurt by a nasty bouncer from Holding or Roberts-remember that 1974 test at the Chidambaram Stadium where Andy Roberts was on fire and Gundappa vishwanath managed to get 97 unbeaten using support from Chandrashekhar and tail enders. And chandrashekhar bowled legspin and even batted with a polio affected right hand-U seem to have forgotten this spin quartet?

  • RogerC on August 24, 2010, 0:28 GMT

    Ananth- kudos on an exemplary compilation. The inquiring mind does wonder about two incremental scenarios. One is the performance of these bowlers in matches where the stakes are raised (win required to square or win the series, win in first test of a series, etc). Another is the performance measure after a team loss. Would either of these alter your standings. Thanks.

  • Sumit Sanghai on August 21, 2010, 9:03 GMT

    Hello Ananth,

    I am not sure if the numbers that you have used are entirely accurate. a) I cant believe that the average of all batsmen dismissed can be in the 20s. Most teams average close to 300.

    b) Can you verify that you have used correct away averages? I checked for Marshall's and that didnt look correct. [[ Sumit Pl send me your Marshall compilation. Pl remember taht this is the average of ALL wickets. The late order wickets, wickets against Zim/Bang all lower the averages considerably. See the late order wickets of Warne and Murali and the Zim/Bang wickets of Murali. Also remember that raina does not give the bowler 91 value. Only a fraction of that. Just an example. Ananth: ]]

  • Ravindra Marathe on August 20, 2010, 15:32 GMT

    Abhi, I quite agree that Laker's 19 for 90 are a 1-off and so its obviously unfair to compare him with Bradman (Ananth, thanks for the correction on Laker's 10 wickets, not 20 in the tour game) That Laker is less celebrated comes from a more general reading of cricket articles. What I meant by the comment on Ananth's article was that the pre-WW1 bowlers, Barnes in particular, would mount that Best Bowler pedestal on several measures, if included in such analyses. Again nothing against Ananth's or anybody's lists, just Barnes' misfortune that he played in conditions prevalent then which aren't directly comparable today. This 'what-if' has a parallel in physical chemistry- J Willard Gibbs work on energetics could have won more than 1 Nobel Prizes but to his misfortune his work preceded Alfred Novel's birth! Sorry for this digression- I am a cricket-loving chemical engineer.

  • Abhi on August 20, 2010, 3:52 GMT

    Ananth, Though Ravindra Marathe has raised some good points there is one huge difference between what Bradman has done and what Laker etc achieved.

    This cardinal difference is that Bradman's 99.94 was a "career" achievement. Reflecting an entire career of unmatched brilliance. The 19 wickets,400,200 etc are by contrast mere "one-offs"...and especially the 400,200 will definitely be passed because all it takes is essentially a good batsman having a "single" good outing and everything "clicking" into place.

    This cardinal difference must always be kept in mind when comparing Bradmans freakish average over an entire career over "one off" incidents.

  • Ravindra Marathe on August 19, 2010, 13:44 GMT

    Ananth, On the topic of leaving out outstanding performances, we've so far kept Bradman's average as the yardstick to measure everybody against. However this article being on bowling, do you think it's fair to exclude Grimmett, Barnes Lohmann? Is is because they were truly exceptional or because we (more recent followers) cannot associate with having somebody from a bygone era dominate the bowling charts- not the number of wickets, but the economy, strike rate et al. I do understand that the wickets and also the rules of the game (8-ball overs, 4 for a 0-bounce boundary hit etc) were different then. I believe credit must be given where and when due. Also why is Laker's achievement considered a freakish one- he had captured all 20 wickets in a tour game that same series. I always feel Laker is under-recognised. Speaking of fearsome bowlers, was Nissar+Amarsingh India's first fast pair? Sutcliffe or Hobbs (not sure who) once said he could not hit the bal coz he could not see it! [[ Ravi You have raised quite a few points. 1. I said Laker's 19 wkts was freaky only in a very positive sense. Let me say this differently. I think the 400 would be beaten, Ithe 200* would be overtaken, however I do not think anyone would ever take 20 wickets in a match. In both the innings of the same match, the bowlers, umpire and even the batsman should co-operate. i don't think that would ever happen. 2. Laker did not take 20 wickets in a match. No one has ever done that. What Laker did was to capture the 10 wickets in an innings a second time against Australians during 1956. 3. The only reason why I excluded Barnes was because I personally feel that the bowling conditions during pre-ww1 period were totally different. Hence the figures are not comparable. This is proved in my period analysis article. The pre-ww1 figures are around 30% better than the overall average. Let me say that Barnes would probably be at the top. Lohmann has taken only 103 wickets. Grimmett was as good as Warne as far as figures are concerned but had taken a third of the wickets. 4. Please contrast this with the Bradman era. He was that much better than the others. the period was no different to any other periods. The bowlers were good. Anyhow Bradman scored most of the runs against England which had Larwood, Voce, Bowles, Verity et al. Ananth: ]]

  • Mark on August 19, 2010, 0:08 GMT

    By the way - I'm a bit disappointed this article got hijacked by people commenting on batsmen and all-time XI's.

    It's about the bowlers which makes for a nice change. [[ Mark Very true. Despite my requests. I will try and see that this soes not happen in future. Ananth: ]]

  • Harsh Thakor on August 31, 2010, 18:09 GMT

    It is a great effort ,Anantha but I am still convinced that mere statistics does not tell the true story.In the fast bolwing section I would still rate Marshall and Lillee the greatest.They were more agressive than any of the great fast bowlers,with great versatality,considerable speed and ability to intimidate the opposition in alltypes of conditions.Marshall's incredible strike rate at such a low cost,while Lillee's fantastic rate of 5 wickets per test are statistical evidence that suppport it.Adding the Packer and Rest of the World Series would have reflected his greatness more accurately.Mcgrath and Hadlee had greater control,but were not as versatile or menacing in all conditions than Lillee or Marshall.Imran,a great match-winner did not posess Lillee's repertoire or Marshall's unpredicatability.Statistics does injustice to Andy Roberts,morally the greatest rival to Marshall and Lillee,who was rated by opponents to have been the best paceman of his day.

  • Meety on August 31, 2010, 5:36 GMT

    Interesting read. I haven't read all the comments but I was wondering what you think about the sugestion that the the best and worse score gets dropped? (Like in Olympic judging) - just interested to see what it throws up - I think it would make Murali's gap larger & bring Lillee up the list? [[ Andrew I will have to buy additional insurance if I drop the best performance. In Olympics this is done to cater to biased judging. Here great performances have to be recognized. Ananth: ]]

  • raja on August 28, 2010, 15:30 GMT

    gr8 work anantha.u said that u didnt included sf barnes knowingly.do u think that he was the greatest bowler ever, even better than murali?in my opinion he was the don bradman of bowlers.plz reply. [[ Raja Probably yes although you must consider the fact that the bowling conditions were totally different pre-1914. Ananth: ]]

  • wb on August 27, 2010, 10:01 GMT

    At one time, Indian cricket team had a spin quartet just like the West Indies had a pace quartet. Chandrashekhar, Venkatraghavan, Bishen Bedi and Prasanna managed the bowling for India even on non-Indian pitches and gave test wins for india in india and outside India. The spin quartet also had to bat against the fast bowlers of those times with little protective equipment compared to today's and no restrictions on bouncers to be bowled in an over. There was always a risk that any of the 4 spinners would get retired hurt by a nasty bouncer from Holding or Roberts-remember that 1974 test at the Chidambaram Stadium where Andy Roberts was on fire and Gundappa vishwanath managed to get 97 unbeaten using support from Chandrashekhar and tail enders. And chandrashekhar bowled legspin and even batted with a polio affected right hand-U seem to have forgotten this spin quartet?

  • RogerC on August 24, 2010, 0:28 GMT

    Ananth- kudos on an exemplary compilation. The inquiring mind does wonder about two incremental scenarios. One is the performance of these bowlers in matches where the stakes are raised (win required to square or win the series, win in first test of a series, etc). Another is the performance measure after a team loss. Would either of these alter your standings. Thanks.

  • Sumit Sanghai on August 21, 2010, 9:03 GMT

    Hello Ananth,

    I am not sure if the numbers that you have used are entirely accurate. a) I cant believe that the average of all batsmen dismissed can be in the 20s. Most teams average close to 300.

    b) Can you verify that you have used correct away averages? I checked for Marshall's and that didnt look correct. [[ Sumit Pl send me your Marshall compilation. Pl remember taht this is the average of ALL wickets. The late order wickets, wickets against Zim/Bang all lower the averages considerably. See the late order wickets of Warne and Murali and the Zim/Bang wickets of Murali. Also remember that raina does not give the bowler 91 value. Only a fraction of that. Just an example. Ananth: ]]

  • Ravindra Marathe on August 20, 2010, 15:32 GMT

    Abhi, I quite agree that Laker's 19 for 90 are a 1-off and so its obviously unfair to compare him with Bradman (Ananth, thanks for the correction on Laker's 10 wickets, not 20 in the tour game) That Laker is less celebrated comes from a more general reading of cricket articles. What I meant by the comment on Ananth's article was that the pre-WW1 bowlers, Barnes in particular, would mount that Best Bowler pedestal on several measures, if included in such analyses. Again nothing against Ananth's or anybody's lists, just Barnes' misfortune that he played in conditions prevalent then which aren't directly comparable today. This 'what-if' has a parallel in physical chemistry- J Willard Gibbs work on energetics could have won more than 1 Nobel Prizes but to his misfortune his work preceded Alfred Novel's birth! Sorry for this digression- I am a cricket-loving chemical engineer.

  • Abhi on August 20, 2010, 3:52 GMT

    Ananth, Though Ravindra Marathe has raised some good points there is one huge difference between what Bradman has done and what Laker etc achieved.

    This cardinal difference is that Bradman's 99.94 was a "career" achievement. Reflecting an entire career of unmatched brilliance. The 19 wickets,400,200 etc are by contrast mere "one-offs"...and especially the 400,200 will definitely be passed because all it takes is essentially a good batsman having a "single" good outing and everything "clicking" into place.

    This cardinal difference must always be kept in mind when comparing Bradmans freakish average over an entire career over "one off" incidents.

  • Ravindra Marathe on August 19, 2010, 13:44 GMT

    Ananth, On the topic of leaving out outstanding performances, we've so far kept Bradman's average as the yardstick to measure everybody against. However this article being on bowling, do you think it's fair to exclude Grimmett, Barnes Lohmann? Is is because they were truly exceptional or because we (more recent followers) cannot associate with having somebody from a bygone era dominate the bowling charts- not the number of wickets, but the economy, strike rate et al. I do understand that the wickets and also the rules of the game (8-ball overs, 4 for a 0-bounce boundary hit etc) were different then. I believe credit must be given where and when due. Also why is Laker's achievement considered a freakish one- he had captured all 20 wickets in a tour game that same series. I always feel Laker is under-recognised. Speaking of fearsome bowlers, was Nissar+Amarsingh India's first fast pair? Sutcliffe or Hobbs (not sure who) once said he could not hit the bal coz he could not see it! [[ Ravi You have raised quite a few points. 1. I said Laker's 19 wkts was freaky only in a very positive sense. Let me say this differently. I think the 400 would be beaten, Ithe 200* would be overtaken, however I do not think anyone would ever take 20 wickets in a match. In both the innings of the same match, the bowlers, umpire and even the batsman should co-operate. i don't think that would ever happen. 2. Laker did not take 20 wickets in a match. No one has ever done that. What Laker did was to capture the 10 wickets in an innings a second time against Australians during 1956. 3. The only reason why I excluded Barnes was because I personally feel that the bowling conditions during pre-ww1 period were totally different. Hence the figures are not comparable. This is proved in my period analysis article. The pre-ww1 figures are around 30% better than the overall average. Let me say that Barnes would probably be at the top. Lohmann has taken only 103 wickets. Grimmett was as good as Warne as far as figures are concerned but had taken a third of the wickets. 4. Please contrast this with the Bradman era. He was that much better than the others. the period was no different to any other periods. The bowlers were good. Anyhow Bradman scored most of the runs against England which had Larwood, Voce, Bowles, Verity et al. Ananth: ]]

  • Mark on August 19, 2010, 0:08 GMT

    By the way - I'm a bit disappointed this article got hijacked by people commenting on batsmen and all-time XI's.

    It's about the bowlers which makes for a nice change. [[ Mark Very true. Despite my requests. I will try and see that this soes not happen in future. Ananth: ]]

  • Mark on August 19, 2010, 0:02 GMT

    Anath: Yes I did see your table and I'm certainly in no way criticising your analysis as I'm about as much a statistician as I am a cat. My point was just that you (or anyone for that matter) can use whatever criteria you like and it probably won't severely impact on who is in the Top 5, it will probably just shuffle them around a bit.

    In an attempt to not appear to be a too biased or patriotic kiwi I'll stick with your analysis as it is and leave Hadlee at #3 :)

    Again - a interesting piece of work to think about.

  • Abhi on August 18, 2010, 13:37 GMT

    And to continue a bit! Though i get the jist of your argument , I also feel that the "top 20%" or so of the greats (with as usual the exception of bradman) will actually be more or less equal. At least the modern guys- Tendulkar,Lara,Ponting etc. [[ Abhi/Alex Somewhere there can we give this a miss although I agree that it was Alex who introduced a top-20 idea into a bowler article. Ananth: ]]

  • Abhi on August 18, 2010, 13:34 GMT

    Alex. Also, your anu malik/rehman comparison – though funny, is off the mark is several aspects. One is that those musicians are “standalone” artists. Batsmen are finally just members of a “Team”. A more apt comparison would be a Cello player who hits every note just absolutely perfectly in one concert and then muffs up the next 4 concerts. Another player also hit the perfect notes but with perhaps a slightly different style- but this guy will do it practically everytime/every concert. Would you give “extra marks” for that one better performance? Who’s the better player?

  • Abhi on August 18, 2010, 13:14 GMT

    Also, the guys who score big innings tend ,in general, to do so fairly briskly. Odd though that may sound.The thing is that the mistake/great ball will inevitably come , so score as many runs in the interim as possible.I remember Sehwag saying something to that effect. So, when the attacking players are in the zone- not only are they going to score fast;but the fast scoring will result in the bigger scores over a longish career.Again, this is assuming a certain “base” quality. The more "defensive" batsman will either get bored out or make that mistake with a fewer number of runs on the board. In any case, the "peaks" will depend very much on the "type" of batsman. Given 2 batsmen of equal quality- the more aggresive type of batsmen seem to have greater peaks over a long career. So, if you accord 20% to both "peaks" and "troughs" it may be ok.

    But then again- what's the point?! That is perhaps the whole reason that the "batting average" though an incomplete and vagueish indicator of batting quality still finds prominence since it is still among the best parameters available to indicate both the “peaks” and the “troughs”.

  • Abhi on August 18, 2010, 13:04 GMT

    Alex, You are essentially,in a roundabout way,arguing that 20% of a batsman's career is more important than the remaining 80%.

    This may be so depending on the "type" of batsman. But does the argument in general hold water and make any sense?

  • Alex on August 18, 2010, 12:47 GMT

    Ananth - this is a philosophical take on criterion to decide best batsmen or bowler, and hope you permit me to state it. Longevity, # tests, #runs, #wkts is all fine but these are measurements. Just because we have not measured something, that does not mean it does not exist. Talent, mental & physical preparation are legitimate aspects that we cannot measure using scoreboards but can estimate by simply observing the process of bowling and batting ... it is not a player's fault that he was ready with these but we did not observe very many times. - Despite posting better numbers and clutch performances, it is said that Sutcliffe "knew" that his ability was weaker than that of Hobbs & Hammond and hence thought them better than him (as did most of his contemporaries). - So, it is understandable if someone rates Pollock much higher than SMG on these & hence declares him to be a greater batsman overall (not that I necessarily do).

  • Alex on August 18, 2010, 10:18 GMT

    Ananth - peak performances merit a place. Else Annu Malik can rate higher than AR Rehman (or an obscure publication obsessed mathematician higher than Grigory Perelman) due to the volume of mediocre stuff. Even the best will fail 40% of the time. The real qns are when did they fail (quality of failures) and how well did they perform in the remaining 60% (quality of success)?

    Troughs should also be counted but already other metrics do so: Consistency 1 & Consistency 2 (and average).

    Longevity is a useful indicator and has obvious financial merits for the cricket boards etc. but does one really need to play (or observe someone play) 200+ tests for X to be better than Y?

  • Ramesh Kumar on August 18, 2010, 8:47 GMT

    Alex,

    Peak performance is a bit unfair. Hussey will have high peak performance and can offset his dip in performance. In bowlers, Mendis had excellent figures in his 1st year. If we use peak performance, we need to offset it for long career players by removing 15% worse performance for both bowlers & batsmen. I feel a study has to be taken on career peaks & troughs(next project for Ananth?) across eras across formats for bowlers & batsmen with the end objective of coming out with an extrapolating factor for players who played less no. of matches,in less no of venues(eg pre ww2) and in less no of farmats(pre ODI). [[ Ramesh Last year I did an extensive analysis on peaks and troughs. Pl have a look at that. Ananth: ]]

  • Boll on August 18, 2010, 5:17 GMT

    ...sorry, must stay on topic! Anyway, back to Paddles. He certainly doesn`t suffer from Kallis syndrome - did it in Aus, did it pretty much anywhere he played, and I`d agree that it`s very hard to argue with his place somewhere in the Top 5.

  • Boll on August 18, 2010, 5:14 GMT

    @Mark. G`day mate, and good to see a Kiwi on here. Just re.your comment about Hadlee being under-rated outside NZ, not sure about elsewhere, but it`s certainly not the case in Aus. He was never a favourite of Aussie fans, but no-one who saw him bowl over there ever doubted his absolute quality.

    I`ve mentioned in comments re.other articles of Ananth`s that it`s only natural for people`s perceptions of players to be clouded by their performances in one`s own country-either matches we watched at the ground or live on TV. (And it`s easy to forget how difficult it was to get live coverage or highlights even a decade ago.) I remember making a comparison between VVS and Kallis a while back, one of whom has played some of the most memorable innings ever seen in Aus (and vs Aus in India for that matter!) and averages 55 against them, the other averages 40. Probably unfairly, these sort of figures have somewhat reduced Kallis` reputation in Aus, and for many of us put VVS close to the top.

  • Abhi on August 18, 2010, 3:53 GMT

    Alex the "Problem" with "peaks" is that the "peaks" are simply one aspect of performance. They are very often negated or even outnumbered by the "troughs". So, if 20% is accorded to "peaks" how much should you give for "troughs"? For the premier batsman in a side to score a "matchwinning" double hundred and then follow it up (or have it preceded)with a string of low scores in the next 4 Tests may actually be doing a disservice to the team.

    The "peaks" are a good indicator of how dangerous a batsman can get when totally on song and when the going is good for him.But the true test of an elite sportsman is how he performs over a period of time, through thick and thin, good and bad...not just how good he is when things "click".

    As Ananth says batsmen like Sehwag ( and to a lesser extent Lara,Laxman,Richards etc) will see huge jumps in points if 20% is awarded to them.Whereas batsmen like Gavaskar,Dravid,Kallis,Boycott will see declines . Guys like Tendulkar and Ponting will probably be unchanged. But what does it really tell us? Just that when Sehwag,Lara ,Richards and Lax are on- they are really deadly.

    But this is almost wholly a function of “Batting style’…i.e when things “click” , an intrinsically aggressive batsman is bound to produce some stunners given a reasonable number of matches. Pure probability.

  • Alex on August 18, 2010, 3:48 GMT

    Ananth - I meant top 20% in terms of most runs scored. However, you could choose a more reasonable metric of your choice (significant innings or Wisden calculations). Using most runs scored as the basis gives us the following: 1. Bradman becomes a greater outlier (but his record is that good). 2. SRT has to score almost 14,700 @ave=55 to be rated greater than Lara in tests - a reasonable albeit debatable conclusion. 3. Players like Weekes are viewed in flattering light --- Weekes's otherwise average record is inflated by massive scores against India's weak bowling attacks.

  • Mark on August 18, 2010, 2:58 GMT

    I've had a bit of a look at a range of ways to rank the stats rather than a straight "adding them up". It seems that the Top 5 (as I call them) are the same no matter what with just a bit of variation as to the order within the top 5.

    Interestingly, the Top 5 are in the top 5 for each category 5 or 6 times each which I presume indicates consistant performance. Even more interesting is that Murali ranks in the bottom 5 three times which is more than the other Top 5 bowlers.

    While I understand the analysis did not rate the categories in terms of importance its also interesting to note that Murali's SR, AwAvg and TOR are where he performs worst (i.e. bottom 5). Does that mean that what Murali did well he did very well and what he did badly he did very badly??? I don't know, but I guess that's what many critics alude to when arguing he's not the greatest bowler ever.

    It doesn't change the fact that, statistically he can be considered 1 of the Top 5 bowlers ever!!

  • Mark on August 18, 2010, 2:07 GMT

    First, I'm from NZ so tend to have a bias towards all things kiwi. That aside, I think this is a brilliant piece of work.

    The only thing I'd question is the "No. of Wickets" as opposed to "Wickets per Test". However, there are so many variables that impact on individual bowler's performance such as playing for a weaker team, having 2 quality bowlers in the same team, bowling against quality vs rubbish batsmen etc that there will always be arguments and counter arguments about how you've gone about it.

    At the end of it all, if you look at the players that statistically are in the top 5, they are all players who you watch waiting for something to happen, waiting for the next wicket. If I were to pick 4 strike bowlers from your list they would probably be Ambrose (absoutely fearsome!), McGrath (same place every time), Hadlee (possibly a bit underrated by those outside Kiwi Land) and either Murali or Warne (I prefer Murali as the "nicer" of the 2).

    Overall - nice piece of work!! [[ Mark You probably missed the table I put in as a response to comments by Chris/Brent. I replaced the Wickets parameter with Wkts/Test parameter and Hadlee moved to the top. Muralitharan moved to second place. If you take the trouble of locating that comment you will see that i have mentioned that I have no problems at all with this sequence. Ananth: ]]

  • Alex on August 17, 2010, 22:06 GMT

    Ramesh - Currently, Ananth has columns on #runs and average. I suggest adding a column on peak performances, e.g., ave in top 20% innings, and split the net wt given to the original two columns equally (0.33) into these three. The other columns are to be left untouched although I feel a column of SR would be nice. [[ Alex By top-20% do you mean the top one-fifth in terms of scores. If so, Bradman will move up still further and only Sehwag, and to a lesser extent, Lara would be up there. Unfortunately I don't have the world rankings. Otherwise this can be replaced by the scores against the top teams at the time of playing the innings. Ananth: ]]

    Roughly, that means SRT can catch the Don (score of approx. 70) only after scoring 24K runs at ave=56 (since he doesn't score very big 100's). Poetically, the Don scored almost equal # runs as his rival (Hammond) at almost twice the ave while SRT has to do the other way with Lara.

    Hope Ananth will soon release the revised list of great batsmen.

  • Alex on August 17, 2010, 21:36 GMT

    Ramesh and Abhi - SMG's oft-repeated lament as a player was "I cannot score off good balls" and it was valid! Even during the CNN-IBN joint interview that celebrated SRT overtaking Lara's run tally, he said "I wanted to attack, play like Sehwag but I could not ... maybe there was a mental block." That was the truth and not modesty. To be fair, he adapted very well to ODI's after leading India to the '85 championship trophy in Aus.

    As a batsman, Dravid is SMG version 1.1, sans the media hype. He can attack well in tests (e.g., 177 vs SL) but did so rarely. So, I use the Ponting over Dravid logic to place Chappell over SMG. Both are great but I prefer a youtube clip of Chappell's batting.

  • Abhi on August 17, 2010, 13:49 GMT

    Ananth, Re. the sehwag century hoopla- I know you never incorporate centuries in your stats (rightly so) but it just goes to show how much a "hundred" means to players- whatever they may say later to the cameras and media. A 'macthwinning or saving Hundred" always has a certain "ring" to it than a corresponding 99 simply does not- Though "logically" we all know that there is no difference. And I don't buy the argument that it happens only in the subcontinent- It is a universal phenomenon. Even recently Watson's 90s blues were quite evident. [[ Abhi Much ado about something minor. I liked Ajay Jadeja's comment "If Sehwag had been the bowler he would have done the same thing". This very effectively proves my point that 100s are over-rated. I have always said that a 99 is as good as a 100 and the 100 is only a single run, the one which is taken when one is on 99. The sheen on Sehwag's innings remains the same whether it was 99 or 105. Ananth: ]]

  • Ramesh Kumar on August 17, 2010, 5:58 GMT

    Hi,

    On Alex's comments..There is a point in Lara/Ponting comparison as both had a long career and many tests. But if we do Ponting/Pollock comparison and try to lower the weightage of no. of runs, it may be unfair to Ponting. Unless we have a mathematical model on career progression backed by cricketing logic is done, we can't solve this problem. May be we should determine models for peaks and troughs and extrapolate for players with lower tests. On SMG vs chappell/Pollock:ability to attack-yes, superior talent:debatable, technique:no, at the best Chappell is equal and not better. SMG is an opening batsmen and he should be evaluated based on his role. Two innings-101 in 1974 old trafford and 96 in Bangalore minefield would give testimony to SMG's technique on moving ball and in spinning track. SMG was a compulsive stroke player in his initial years(difficult to believe). He changed his stance in 1974 due to a finger injury.Ofcourse not to take away the greatness of Chappell/Pollock

  • Abhi on August 17, 2010, 5:02 GMT

    Alex "but Chappell & Pollock did have superior talent, technique, & ability to attack (SR) - even SMG admits."....you kidding? Almost all these parameters are highly subjective: 1)"Talent": This is only a clear differentiator when considering the most outrageously talented sportsmen.Not when they are just about equal. And of course , sport is not just about “physical talent”..”Mental attributes” is the key differentiator between the good and the Great. 2)"Technique" : Not many batsmen had a better technique than Gavaskar. 3)"Ability to attack" : This is completely and utterly different from the "willingness" to attack. Gavaskar's central role was to establish a presence and anchor one end, see off the new ball etc. No to go out there and up the ante regardless of the state of the game. Srikant and co. used to that – in part because they simply did not even know any other way to play the game- Fixed in one gear, regardless of the state of the game. Over a period of time,however, a “slow” approach does seem to become a habit ( eg. the way Gavaskar carried the same approach over to that notoriously slow ODI game) Gavaskar, as opener was fulfilling his particular role in the team. Your statement is almost as ridiculous as saying Sehwag's "ability to attack" is greater than Tendulkar's simply because of the different roles they play in the team- wherein one batsman is licensed to attack , the other is more the fulcrum and requires to consolidate etc. “Ability” and “willingness” or “requirement” etc are different things.

  • Alex on August 16, 2010, 21:33 GMT

    Ananth - your analysis is "return on investment" type, a good approach; i.e. score will likely increase the longer one plays. But I feel too much weight is given to #runs.

    - E.g., Ponting of today is, IMO, just a little bit farther from Lara than Ponting of June '09 but has, in fact, added approx. 0.8 to his June '09 score of 48.2 simply because he scored 800+ runs more despite lower ave & poorer consistency. (I cite Ponting because he is one of my favorites.) - I like the metrics but might assign slightly different wts. E.g., slightly reduce wt given to #runs, add a small wt to peak performances (top 10% or 20% innings), etc. - Pollock was robbed of # tests but Chappell over SMG is not far fetched: he might pip SMG with a rating of 45 if the Packer series is included (& this really validated your analysis for me). Now, this is beyond stats but Chappell & Pollock did have superior talent, technique, & ability to attack (SR) - even SMG admits.

  • Peter M on August 16, 2010, 13:11 GMT

    While I think Viv Richards is the best bat I have seen (I have watched cricket since late 70s)I think the argument for the best after Bradman is as difficult as the one for best bowler. I can't see how you could be confident that Tendulkar or Lara are that much better than the other contenders, Sobers being the first that comes to mind. [[ Peter My comments were based on the extensive analysis which has been done by me and presented in this space during the past two years. Ananth: ]]

  • love goel on August 16, 2010, 12:05 GMT

    I was no way deriding Murali or any other bowler. Who am I do so. Of course murali is a great bowler. Nobody can deny that and neither did I tried to do so. He did got 800 wickets, and that is the final statement.

    Saqlain Mushtaq was one of the finest bowlers in the game till he went away. There is a match of which I remember practically nothing, except the fact that in that match, Saqlain bowled an absolutely peach of delivery to get the batsmen bowled( I think Nathan Astle). Only he could have bowled that one, not Murali or even Warne.

    And yes ,today Graeme Swann can claim to be best spinner in the world. And if he does so, I will be the first person to vote for him.

    There was no thought of deriding Murali or any other. That is all I can say. [[ Goel The statement "Off-spinners are not to get-out to,but to be hit outside the park." was quite clear. Hence my response. Anyhow you are entitled to have your views. Regards Ananth: ]]

  • Vinish on August 16, 2010, 10:00 GMT

    I agree with your XI except for Akram. I guess you pick him for the variety since he is the best left-arm fast bowler to have played test cricket.

    However, if that is not the 'preferred criteria', I would have Hadlee rather than Akram.

    All other 10 players pick themselves automatically. Cheers!

  • Alex on August 16, 2010, 6:25 GMT

    Ananth - In this article, two academia researchers use basic stat analysis using no insight. Their sample set of 50 top batsmen over the last 60+ years includes Kambli!

    I feel your article on the best batsmen is the best available stat-based analysis although I feel it gives too much weight to volume, leading to surprising conclusions such as SMG is better than Greg Chappell & Pollock, for instance. I feel SRT has now surpassed Lara on pure test level also. If he maintains this form for 2+ years, he must be rated higher than anyone bar Bradman. [[ Alex All of us have our own, often quirky, views. Else why should you be so surprised that a batsman, with limited support who scored 10000+ runs at an average of 51+ is rated above one from a stronger team (and better support) who scored 7000+ runs at 53+ average or one who scored 2000+ runs at 60+ average. If I had got, say, Azharuddin, ahead of Chappell, yes, you would have a point. Ananth: ]]

    Sehwag is unique and in the prime of his career: best to evaluate him _after_ his career is over. I don't think he really cares about who is to bat after him. He will play the same way irrespective of whether it is SRT-Dravid-Laxman or a bunch of kids. Hence, I believe Gambhir is a better opener in many ways even though I fully support Sehwag.

  • Abhi on August 16, 2010, 4:13 GMT

    Boll, Yes. Just read it after reading your comment. Again, there is a faint foul odour as rgds.the date used. Why use a date 2006 for a study carried out in August 2010? If you've read any of the previous blogs (and I'm sure you have!) you would realise that 2006 was just about "Tendulkar- termanilis" time. "Endulkar" had just wrapped up his worst 3/4 yrs and Ponting,Dravid,Lara,Kallis and co. had conversely completed their 4 best yrs.

    So, to repeat, why on earth should a study done "Today" use 2006 as the end date? [[ Abhi/Boll I have not read the article yet. However there is a report in Times and frankly does leave me cold. First, if they cannot use current data they should not come out anything. Secondly, the cricketing credentials seem to be lacking in the study. My recent study places Tendulkar as the undisputed top batsman across formats. However in Tests, Bradman is the almost certainly the unassailable king. This is not too difficult a task. To place Tendulkar in a way-down position is the act of crickeing idiocy. They seem to have the study onl\y now as evidenced by the relatively high placing of Sehwag. In that case 2006 figures should not have been used. Quantitatively, qualitatively, subjectively or objectively, the following is what any sane analyst would come out with., Tests: 1. Bradman. 2. Lara/Tendulkar (an year back I placed Lara ahead of Tendulkar. Now, seeing Tendulkar's recent form, 6 quality centuries in 10 matches, I feel Tendulkar would have edged out Lara to the second spot). Then come the rest. ODIs: 1. Tendulkar. 2. Richards. Then come the rest. Anything else smacks of a poor model being used, lack of cricketing sense and hollow research. Ananth: ]]

  • Boll on August 16, 2010, 2:00 GMT

    I`m sure many of you have read reports of a qualitative analysis of the Top 50 batsmen of all time, recently published in the Journal of Quantitative Analysis in Sports. Apologies for adding the link to an examination of the great bowlers, but thought it would make an interesting comparison. cheers

    http://www.bepress.com/cgi/viewcontent.cgi?context=http%3A%2F%2Fwww.bepress.com%2Fjqas&article=1201&date=&mt=MTI4MTkyMjk4MA%3D%3D%0D%0A&access_ok_form=Continue

  • Alex on August 15, 2010, 23:42 GMT

    Ananth - I am not a fan of fantasy cricket but it would be nice if you can do a series of 11 matches across world-wide venues.

    Viv's position is an issue. Opening with him makes way for Weekes/Border/Miandad/Ponting. 1. But Hutton-Sehwag looks good and, given Sehwag, Viv's threat is best at #5. However, both Chappell & Pollock preferred #4 & #5. 2. If Sehwag is cut and Viv opens, the middle order easily reads Ponting, Pollock, Chappell, Kallis.

    Barry Richards is left out for the obvious unfortunate reason while Hutton, who should have made the first team itself, is a must. Now, the batting advantage of Bradman at #3 is negated by that claimed by the Imran-Sanga combo at their positions. Ambrose vs Hadlee is a close call but I felt Ambrose was the better bowler. Also, he brings in a vastly different dimension; Hadlee was quite similar to Lillee.

    ODI's are tougher: you can create 6-7 teams (not just 2) that are almost evenly matched!

  • metman on August 15, 2010, 17:59 GMT

    Very good analyses!I am a West Indian,and used an analysis like yours( not as indepth as yours though)and also came up with Marshall,Garner and Ambrose as my 3 pacemen for the WI all time XI.That jury was particulary bias and only used the statistics when it suited them,and other times likes/dislikes and favourites were heavily used.

  • love goel on August 15, 2010, 16:41 GMT

    For me Warne was always better than Murali. For god's sake, Murali was an off-spinner! Off-spinners are not to get-out to,but to be hit outside the park. How Murali got 800 batsmen out is beyond me. But may be, may be,that's how he got those 800 wickets. But he did got them

    But then to me, Ambrose was better than anybody in his time, including Mcgrath. [[ Goel From any one else I would have trashed this comment, However you are one whose comments have been sound. hence have published the same. I think you have done a great disservice to a whole type of bowlers starting from Jim Laker, onwards to Gibbs, Tayfield, Prasanna, Murali, Saqlain Mushtaq and now Harbhajan by consigning the off-spinners to the dust-heap. These are great bowlers, exponents of a subtle bowling type and stood in equal comparison to the BEST LEG_SPINNERS AND LEFT ARM SPINNERS OF THEIR AND OTHER AGES. Even when Abhi was arguing vehemently for Warne, he never lost sight of Murali's claim to equal greatness. He did not like the daylight which exieted between Murali and Warne, that is all. Pl read his last comment. That clarifies all. Even today Swann can lay claim to be the best spinner in the world. Just ruminate over what you have written. Ananth: ]]

  • Nish on August 15, 2010, 14:48 GMT

    After he got to 300 wickets, Kapil should have quit bowling and focused on his batting. He was talented enough to have easily played 40-50 tests in the middle order (#5 or 6) and maintained a 45+ average (specially on fast wickets abroad).

    What he did with his career in his last few years meant that, his bowling stats went way down, and he lost out on an opportunity to improve his batting states (which he could have had he played purely as a batsman).

    And as usual, this is an excellent article from a stats perspective. Thank you for your excellent efforts, Anantha!

  • Abhi on August 15, 2010, 12:55 GMT

    Alex Passion is the only thing really...The rest follows. As for Chappeli this was the same chap who wanted Tendulkar to retire( Greg's brother you know)...in any case Ian seems to have forgotten that one of the primary reasons for Sehwags carefree attitude and hell for leather approach is that not only has he been given full license to "go for it" but he also knows at the back of his mind that he has the greatest batsman of all time at no.4 as "backup".

    In any case Chappeli has always had a soft spot for the swashbucklers.If you read any of his articles his whole attitude towards the guys who don't go out there swinging is negative.

    So,I suggest dont bother. But as rgds this article -heck both warney and murali were absolute geniuses. How do you separate two geniuses? in any field?

  • Alex on August 15, 2010, 9:33 GMT

    Abhi - I have nothing against anybody. Passion is great but I don't think this article meant to compare, which is bad idea anyways.

    Yet, even the experts indulge in comparisons & in "leveling the field": today's Ian Chappell's article praising Sehwag takes digs at SRT's 100's in BD. Was it necessary and did he witness the 105*? Ian C clearly did this leveling just to belittle SRT. As Ananth said, where do you draw the line between minnows and non-minnows, between what is significant and what is not? Best to avoid that issue!

  • asif on August 15, 2010, 7:03 GMT

    Ananth, I think wpt is not a good measure simply because it doesnot tell us who bowls more innings and more overs. So it gives us a distorted picture. Basically it actually shows the bowler's strike rate which is the most basic measure. I feel that the basic measures should be given more weightage to this kind of analysis instead of the supporting measure. And to my opinion, basic measures should be number of wichekts, strike rate and rpo. This should constitute at least 60% of the weight of which strike rate should get maximum weight. Among other parameters, quality of wickets, team load and win should get more weight. I still dont understand if we count quality of wickets considering batsman's avrg, what is the point is counting top order wicket. Also, in peer 1, already own team bowler's are counted. Hence what is the point to count them again in peer2 specially when we are counting team load. These are personal opinion with no intention to pull someone up or down.

  • Alex on August 14, 2010, 23:52 GMT

    Ananth - you replied to Peter M (August 13, 2010 2:23 PM) that the Packer series are not available in public domain. I checked it and lo and behold: http://en.wikipedia.org/wiki/World_Series_Cricket_Player_Records; even the score-card are now available on Wiki and linked to Cricinfo. (However, the ODI score-boards are sometimes missing on # balls faced.)

    If you have no objection to the "unofficial" Packer series, perhaps you could include those now on, if only as an Appendix. [[ Alex Too much of a work for me. Personally I do not think these should be considered, even though the quality of the 16 matches was far higher than the official tests going on. Unfortunately my entire data base is a completely integrated one and suddenly I do not want Lillee to have 422 wickets. It is like IPL. I do not consider IPL as official T20 matches unless otherwise the client specifically wants this way for their own customized analysis. You will notice that in "It Figures" I do T20-Intls and IPL work separately. Ananth: ]]

  • Alex on August 14, 2010, 22:52 GMT

    Ananth - my rival XI (restricted to post-WW2) reads Hutton, Sehwag, G Chappell, G Pollock, V Richards, Kallis, Sangakkara, Imran (c), Warne (sorry, you can't have Murali and Warne too!), Lillee, Ambrose.

    I think, save Bradman, it is matched well man to man with your team while Imran was, perhaps, a better skipper than Bradman. It can bat & bowl well anywhere. Sehwag and Viv give options in the spin department. I thought of Viv as captain with Davidson replacing Imran to match what Akram can do in your team but think this team will perform better. [[ Alex Aditya's XI bears a striking resemblance to your XI. I will let you have your pick on spinner. I have no problem with either of the two greats. For some other work I recently got my ODI simulation suite of programs perfected. unfortunately my Test simulation programs used for the Times simulation are 2002 vintage and it would take quite some effort for me to get those going. Otherwise I could run a series of simulations on these XIs. Ananth: ]]

  • rahul on August 14, 2010, 17:16 GMT

    sir please make a team of all time 11........it should be the best team at all poaaible 11 positions...... 5 bat 4 bolers 1 all rounder and a wicket keeper...... please repli

  • Aditya Jha on August 14, 2010, 13:55 GMT

    Dear Anantha - here's my team to take yours on: Hutton, Sehwag, Headley, Pollock, Richards, Miller, Walcott, Imran, Hadlee, Holding and Lillee. Test matches at Kensington Barbados, Lords, Wanderers, Eden garden and MCG. [[ Aditya That is an excellent team. No wonder you opted for MCG instead of SCG with the best spinner being Richards. Ananth: ]]

  • Alex on August 14, 2010, 10:44 GMT

    Ananth - I guess the article objective is to present the findings summarized in the 10 columns while the columns were added up simply to enforce an order on the bowlers. If that is mistaken as "this analysis shows X is better than Y", perhaps the following elementary stuff can get the point across: 1. Delete statements such as "Murali is best among equals" (but you are entitled to state anything in your own blog). 2. Replace the photograph of Murali with a montage of all 15/19/21 bowlers. 3. Ask the doubters: If you believe Warne is as good as Murali, do you believe Murali should look like Warne as well? If not, why insist that the metrics on which they differ are not valid? Also, isn't "leveling the field" a subjective process, often used to reach the desired conclusion (and common in financial scams)? [[ Alex There cannot be any phrase better than "Best amongst equals". It is the same as "First amongst equals". Also why is it that in other comparisons the first or best is accepted without arguments, but not here. The photographs are not my suggestions, they are Cricinfo's. After 5 days of running the article nothing is gained by changing these. Ananth: ]]

  • Abhi on August 14, 2010, 10:36 GMT

    Xolile. Thank you sooo very much. I was being made out to be the villain of the piece here- apparently because I simply dared to say warne was as good as murali- and look for holes in the model to find out why such a difference exists.

    Alex. Pardon me. But EVERY single "comparative" analysis we attempt between players from different teams and eras is some version of "pigs flying,wombats etc"....But we still go about it with gusto and a great deal of fun and enthusiasm ,dont we?

  • Xolile on August 14, 2010, 8:13 GMT

    …continued

    Overall, if you exclude the minnows, there is little to choose between the two in the bowling department. Murali has a slightly better average (24.88 v 25.41), but Warne has a better SR (57.67 v 58.65) and has taken more wickets (691 v 624).

    This brings us to the other departments: captaincy, team morale, batting and fielding. It is these aspects of the game where Warne is clearly more capable than Murali. And that is why I would pick Warne ahead of Murali, both in Asia and outside Asia, every time.

  • Xolile on August 14, 2010, 8:13 GMT

    Ananth, I am enjoying the Murali v Warne debate. I suspect it is one that will rage for many more years. The consensus appears to be that Murali is craftier on the flat, dry, slow and low wickets of the subcontinent, and that Warne is more adapt to the bouncier, grassier wickets outside Asia. I took a closer look at the numbers and was surprise to see that there is no apparent statistical basis to support the consensus view. In fact, the numbers suggest Warne is in fact the craftier bowler on the often lifeless pitches of the subcontinent. Here are the averages: in India: Warne 43.11; Murali 45.45 in Sri: Warne 20.45; Murali 22.22 (excluding minnows) 19.56 (including minnows) in Pakistan/UAE: Warne 19.35; Murali 24.96

    continued…

  • Alex on August 14, 2010, 7:48 GMT

    Ananth - Glad that your XI coincides with mine but I daresay a rival XI exists that will give it a hard run for money.

    Murali/Warne arguments on this blog now perfectly illustrate "What if pigs could fly?" syndrome. By the same token, we could say that this analysis (or any) is not good enough because it gives a different result depending on whether, in 1959, Marshall were: 1. born in WI; or 2. born in Sri Lanka; or 3. born Aus as a kangaroo or as a wombat.

  • P.F on August 14, 2010, 5:03 GMT

    Would Murli have churned out the stats he did if he had played for the rampaging Windies teams of yore? Or even the till recently dominant Aussie teams?

    No Chance.

    [[ Probably not. But then if he had been an Austtralian, he and Warne would have competed for place, bringing down both numbers. But he would have won many more matches. And Lloyd might not have selected him at all. For that matter, Lloyd would not have selected Warne if he had been a West Indian. You could continue this fruitless exercise by saying if Marshall was born a Sri Lankan, he would not have been fearsome. And so on. Ananth: ]]

  • Chris Bolton on August 14, 2010, 4:22 GMT

    Thanks for considering Scott & my line of thinking on this Ananth, I understand the hounding you must get from all angles when you produce something like this so the fact that you also went through the process of crunching the numbers again with our preference is much appreciated.

    I like forward to seeing your next piece of work.

    Aside from the numbers, who's your personal favourite out of the "Fearsome Fifteen"? [[ Chris I would not mention a single bowler. However if I have to nominate a 4 bowler attack to support Hobbs, Gavaskar, Bradman, Lara, Tendulkar, Sobers, Gilchrist, that would be Marshall, McGrath, Wasim Akram and Murali in the subcontinent and Warne outside. Ananth: ]]

  • Peter M on August 13, 2010, 14:23 GMT

    I'd be interested to see what would happen if World Series Cricket was included in the analysis. A number of the greats (Lillee, Richards, Imran for instance) involved all state that it was the most competitive cricket they had played. [[ Peter This is wishful thinking since no reliable records are available in the public domain. Ananth: ]]

  • Vasu on August 13, 2010, 13:36 GMT

    I see a big debate mainly because Warne’s placed lower than Murali in this study. The 2 are amongst the finest in the game, so there is no need to slander one to praise the other. The arguments in favor of Warne are: 1. His tally was divided with his colleagues 2. Murali played weak teams more. 1: With McGrath, Lee, Gillespie to support thru Warne’s career, batsmen were choked, attacked the spinner, and created more chances for Warne. So, the fact that he divided his wickets tally is balanced by the high quality support he had at the other end. It is possible for good batsmen to play Murali out without much risks, & score off the other end. No such luxuries against Oz! 2. Eng, WI, & SA up until mid 2000s were clueless against spin, & Warne played them a lot. Bangladesh plays spin lot better than England. So even that factors balances out. If Murali played for Oz, & Warne for SL, Murali would have bowled fewer overs, had lesser wickets, but a better strike rate & average!

  • Aditya Jha on August 13, 2010, 11:14 GMT

    Dear Anantha - most of the main line of arguments seem to have been covered. Here are 2 simple insights from the table - (a) Obviously this is a list of the top bowlers and the rankings can change depending on how we play with your base table. For example, i removed the team load and peer 2 numbers (i do think that they favor bowlers in the weaker team and are not compensated by the win index) - and McGrath moved comfortably to the top, followed by a close cluster of Marshall, Murali and Ambrose. (b) The reason Warne is so low down is driven by 2 things - he had an excellent pace attack ahead of him which means that he suffers in both top order wickets and quality of batsmen (comparatively) and being a spinner meant that he does lose out on strike rate (however fantastic it may be for a spinner. So does Murali, but Warne is losing to the fast bowlers in relative ranking.

  • Abhi on August 13, 2010, 5:54 GMT

    Alex I fully second your comment. It is just that the very idea behind these statistical "Comparisons" should be to effectively "level the playing field". If it turns out that though the pressure on a good player in weak team may be more but other stats are actually enhanced due to the weak team - then there is no level playing field- since stats are what we are using here. [[ Abhi It is human tendency to support the analysis which supports one's view (such as the one which put SRT clearly at the top of the batsmen list across formats) and find holes in the ones which do not necessarily meet one's expectations. Ananth: ]]

  • Alex on August 13, 2010, 5:27 GMT

    Ananth - subjective "who is better than whom" arguments are rarely fully settled by stats. Best to avoid those by saying: these are raw numbers & metrics ... feel free to infer what you please! [[ Alex That is obvious and has to be understood by all. The readers are not the read-in-2-minutes, send-comment-in-1-minute and forget-about-it readers. They have a lot of insight into the game and this blog's methodology. I am also tired of people never letting go. How many explanations do I give. Ananth: ]] This was one your better analysis pieces. By adding the unusual column on minutes/over (and hence minutes/wkt), I think the results reward the spinner a bit better (and stay quite practical).

  • Abhi on August 13, 2010, 4:26 GMT

    Warne vs. SL in matches involving Murali: In Aus : 4 inn. 11 @ 26 ; SR 61 In SL : 12 inn. 34 @ 20 ; SR 42 (Warnes figures in SL are vastly superior to those in Aus vs. SL)

    Murali vs. AUS in matches involving Warne: In Aus : 4 inn. 8 @ 63 ; SR 110 In SL : 12 inn 44@ 25 ; SR 51 I don’t think it can be said that the SL batting was poor vs. spin. But Warne has fantastic figures vs. the SL batsmen in SL. Murali has more wickets in SL , again perhaps because he had a bigger slice of the pie. The Australian batting lineup cannot be said to be better players spin than the Slankan. Jayasuriya, Attapatu, Jayawardene, Ranatunga and co. are as good if not better players of spin than Slater, Langer, Waughs, Ponting…..This is the best “head to head” available since this is one of the few times where both bowlers will be bowling in almost identical pitches/conditions. [[ Abhi I will also stipulate that 45 wickets in 16 innings at 21 is far superior to 52 wickets in 16 innings at 30 if it will help close this. Let us forget that Muralitharan was taunted , teased and virtually hounded out. Frankly he should not have made that second disastrous trip knowling what would happen. Ananth: ]]

  • Abhi on August 13, 2010, 3:30 GMT

    Ananth, Again, to stress forcefully,there is no question of attempting to “bring murali down” or some such. That is the way you have interpreted it. It is more like attempting to “bring warne up”- although even that is not correct. To have a model show up murali and other bowlers as more than 10% ahead of warne simply indicates a flaw in the model- to me at least. I am pointing out what I feel are flaws in the Model- not flaws in any individual bowler. Also pls note we require to compare spinner vs. spinner. Fastbowlers and spinners are like chalk and cheese. A lot of parameters will heavily favour fastbowlers. Murali is the only spinner in the Top 9- and in my opinion the reason for outperformance over warne is that he was the outstanding bowler in a poor bowling attack. Warne was the outstanding “spinner” in a much better bowling attack. Noone is denying that it might not be a “bed or roses being top performer in a poor team” etc etc… The “pressure” may be more- but then a great bowler in a poor team would be statistically compensated by scoring a lot higher on several of these parameters by virtue of the relatively poor team. So, all the other points are essentially being given for the “extra pressure”. As we know “pressure” is a highly subjective term. It can easily be said that other cricketers face more pressure for other reasons such as public expectations---but we are essentially giving all those extra points here for perceived onfield pressure. The question is how much of a leeway or bonus points for “perceived pressure” should be given here which is not already “balanced” out by other stats such as more total wickets, wkt/match etc. [[ Abhi I will stipulate the following. 1. Warne is far ahead of Muralitharan. 2. He had a great problem bowling with equal bowlers in a stronger line-up. Muralitharan could walk in and pick up wickets. 3. The analysis is not correct. Let us close this line of thinking and move on. Ananth: ]]

  • Alex on August 12, 2010, 22:38 GMT

    Ananth - I feel Gibbs got unduly bashed (by myself too) on this blog. Over-rate is a legitimate aspect. Also, while his SR=87 compares poorly with SR<65 of Garner, Roberts, etc., he takes only about 3.5 minutes/over whereas a Garner would take 4.5. So, it terms of minutes/wkt, he is on par with these greats. He is the best on RPO. Also, apart from in NZ where he had 3 poor tests @ave=45, he averages 33 in Aus and <28 in all other countries. So, all things considered, he was unheralded but pretty special.

    The attack of Marshall, Roberts, Holding, & Ambrose certainly has enough variety since they all are different from one another but Gibbs merits a serious consideration, esp. on a spinning wicket. I feel any great team is that much better if it has a great spinner in it.

  • Kartik (the old one) on August 12, 2010, 17:34 GMT

    Alex,

    since the WI greats were effective on turners as well, the only justification is to meet the 90 overs/day requirement.

    Understood, but Sobers can fill in a few overs here and there.

    I still think it is unthinkable to choose an 87 balls/wkt bowler (Gibbs) ahead of a 51 balls/wkt bowler (Garner). The gap is just too large.

    Few teams would last more than 90 overs against Marshall + Holding + Ambrose + Garner (average : 50 balls/wkt).

  • Abhi on August 12, 2010, 16:56 GMT

    Further to Raghav's comment I also feel that point no.s 1,7,9 are inextricably linked/overlapping...leading to a form of double counting. [[ Abhi You have left out the double counting in 10, this time benefiting the stronger team bowlers. Ananth: ]]

  • Raghav on August 12, 2010, 16:34 GMT

    "It is ironic that you are objecting to the Peer2 because it discriminates against spinners while the others, are saying it is unfair to the bowlers in stronger teams."

    The others are correct in objecting to Peer 2 as it is harsh on spectacular bowling units like the WIs of 80s and Aussies during Mcgrath-Warne era. They are right in their views.

    The issue I have is that Warne could end up suffering in Peer 2 ratio even if he played alongside Stuart Clark, Mitchell Johnson and Doug Bollinger. Taking nothing away from the 3 pacemen (they form a good attack) there is no comparison with Warne. A spinner suffers in Peer 1 and Peer 2, but Peer 2 can be worse in good pace attacks. At least Peer 1 lets you make up due to inferior pacemen in other teams being included as well.

    If you are convinced, then remove Peer 2 and break up Away Bowling Average into Away Strike Rate and Away RPO.

    Raghav [[ Raghav I am trying to balance between bowler types, strong/weak team bowlers and periods. It is impossible to do a perfect balancing job pleasing everyone. I have got a 80-90% acceptance here which is almost the highest reached. Peer2 to a certain extent balances WinIdx. I am sick of Murali/Warne comparisons. So let me do a Hadlee/McGrath comparison. What Hadlee loses by being in a weaker team in terms of WinIdx he gains through Peer2 and vice versa. There is no way I am going to touch Peer2. It is as important as Peer1. Do what Alex did, combine Peer2 and WinIdx. The only factor which is blatantly favouring one type is TopOrderWickets. It favours pace bowlers only. When it comes to TeamLoad, only Muralitharan is way ahead. Otherwise all the other bowlers are reasonably similar. And this is a reflection of both overs bowled and wickets captured. Strike rate might favour Pace bowlers but look at Murali/Warne's strike rates. RpO should favour spinners, but look at Ambrose's RpO figure. QualityofWickets is a neutral measue. The spinners could get the higher valued middle order wickets. Please note that the Away average is a clear composite of Away strike rate and Away Rpo. There may be some overlapping between measures. However that is inevitable. Ananth: ]] PS: When copy pasting table into excel everything is pasted into a single column (Raghav, what is the relevance).

  • Raghav on August 12, 2010, 15:02 GMT

    I quite like the adjustment you have made in the wickets parameter. However, Warne out of the top ten does not seem justified whatever the parameters. This I feel is because of a bias against spinner in the parameters numbered 2. 4. & 6. Number 9 is also slightly discriminatory as a spinner is being compared to the other 3 pacemen in the team.

    Also away performances should be given more weight. Some factors chosen overlap. I would keep only one peer factor and break up the away bowling average into away Strike Rate and away RPO (needed to save games). Australia won many games on day 5 because the batsman creamed runs at 4.5 an over. In fact Sehwag has singlehandedly changed draws to wins by virtue of scoring run a ball (England at Chennai, vs Srilanka at Mumbai and vs SAF in Kolkata) Kolkata especially because the extra 30 overs consumed by Amla were saved by Sehwag. If Strike rate of batsman win matches conversely RPO of bowlers can save them as well. [[ Raghav All good points. I myself am not happy with Warne out of the top-10. Both Murali and Warne lost their 700+ wickets impact. Murali's 6+ WpT helped him. Warne's is a good, nut not outstanding, 4.88. Also "Average Batsman average" is not against spinners. It is not often that the openers are the best batsmen. The spinners have as much, or better chance against the stronger middle order batsmen. And more often than not the tail is cleaned up by the fast bowlers. One thing I could is to do the Peer comparisons against like bowlers. It is ironic that you are objecting to the Peer2 because it discriminates against spinners while the others, are saying it is unfair to the bowlers in stronger teams. Let me finally say that Warne should, in any table, be in the top-5. But it does not mean that I am going to fiddle or fudge ???!!! Ananth: ]]

  • Abhi on August 12, 2010, 14:11 GMT

    Ananth another thing! ha! last comment dont worry....dont need to post it? I dont understand why ppl shouldnt state who their favourites are and who they think is the best? if i think warne and murali are as good - why should i put on a cloak and fake veil of "neutrality" and pretend otherwise. why should everyone behave in some monotonic robotic style.

    I have always felt Tendulkar is by far the best modern day batsman...why should i try to disguise this? to appease others?

    why shouldnt you accept and post comments/stats which seem to favour others and negate your particular hypothesis? "friction" is the only way we are ever going to come up with anything "new' Otherwise we are stuck with our "conventional wisdom"...like the seeming fallacy that stats suffer for good player in weak team. [[ Abhi This is the one post which has allowed almost all dissenting posts to be published. I have no problem at all with posting your contra-views. You should be the last person to complain seeing the number of Tendulkar posts from you, in diverse articles, to be published. Kindly see what I did with comments of Chris and Scott. I gave my own explanations. However i felt that the least I had to do was to do an analysis with WpT instead of Wkts and Hadlee moved to the top. You have gone on various themes to pull down Murali including convoluted thinking to say that a bowler in a stronger bowling team has the more difficult task. Yes, in taking wickets, but he would win many more matches. Do you think Warne would exchange 10 of his team wins for 100 wickets. No way. Also I will not publish batsmen-related posts in this article, however valid these are. One last point, Abhi. Don't forget that there are 9 bowlers between Murali and Warne in my 21-player list. Finally I will assure you that as long as the comments are within the article's scope, I will post your comments. Ananth: ]]

  • Vasu on August 12, 2010, 13:36 GMT

    @Kartik,Alex: Nice points! Gibbs was a validly debatable, though not fully wrong pick for WI XI.To me Holding & Marshall are 1st choice bowlers. For 3rd,4th you have to decide if its 2 quicks, or 1 quick + 1 spinner. For a hypothetical match in 2010 conditions:1 bouncer / over, 90 overs rule enforced, flat pitches even in WI; playing Gibbs makes sense. If playing conditions were 1980s, then it should be 4 quicks. Even then, an attack with different kinds of bowlers is always more effective. Garner was outstanding, but Ambrose was a faster & scarier version who took big 5/6 wicket hauls. To play an excellent spinner, or a different kind of bowler (Roberts / Croft / Walsh) would be a better option than playing 2 of the same kind. An ideal attack would need one each of a kind - a choker (Ambrose), an express seam / swinger (Holding), a master of seam & cut (Marshall),& a spinner to rush thru overs. Thanks Ananth for adding Holding. The 3 quicks form the greatest test attack ever!

  • Abhi on August 12, 2010, 13:17 GMT

    Ananth the "biomechanical" stuff is the dynamic between individual batsmen and bowler. This will vary considerably between batsmen vs. bowler. I.e one batsman may feel one particular bowler is the most difficult because of his technique vs. the unique bowler technique. the comment is NOT an insinuation on muralis action- Why does one always have to spell out very clearly every utterance???- i dont know.

    In any case- the stats do show that a player in a weaker team will probably end up with the superior stats...for a batsman- runs,avg, hundreds etc.(not other way round- as per conventional wisdom)

    So, a player is adequately "compensated" for being in a weak team (for reasons unclear).To compound this is an error.

  • Boll on August 12, 2010, 12:51 GMT

    Fantastic stuff as always Ananth, another of the ones I needed to print out, grab a cold one one from the fridge for and spend an hour or two perusing. Lots of memories just thinking about these great bowlers who, perhaps apart from Kapil, I saw at close to their best. No great insights from me today, but if I had to choose one who I who least like to face I would have to say Ambrose (the man I would least like to bowl to would have to be Viv)- a delivery stride to make you want to take the pads off and become an umpire!

  • Ramesh Kumar on August 12, 2010, 11:53 GMT

    Ananth,

    Let me try to rephrase my question on last decade bowling performance.

    1. Do we have some bowlers who made their debut after Pollock/Murali/Mcgrath who have the makings of figuring in this list?(you might exclude the total wickets as they may have some more years to go)-Steyn could be one, I don't see any others. We have clear 15 years to have seen a bunch of bowlers. They don't seem to belong to this list.Your comments pls [[ Ramesh My feeling is that Dale Steyn, if he maintains his current form for another 4 years, has the chance to get into an all-time top-20. I feel Harbhajan's figures are not great and are not likely to imrove. For a fast bowler, Lee's figures are worse than a spinner's. Ananth: ]]

    2. Record wicket takers like Mcgrath,Kumble,Warne & Murali bowled in 90s and in the last decade. Any pattern in their figures across these two decades? Just to see whether 2000-09 decade is truly batsmen dominated one and to see whether chances are good for seeing another set of bowlers who can match the figures in this list. [[ The average batting average for 2000-2009 is only 5% is higher than the overall average. This is not an outlier situation. Also the bowling average is about 10% higher. The bowling strike rate is quite good and is better than the previous decade. However the RpO takes a beating. How do you then account for excellent bowlers like Murali/Warne/Steyn/Polloc et al. My feeling is that at the other end there are many bowlers with far worse bowling averages than in other decades. Ananth: ]]

  • on August 12, 2010, 11:05 GMT

    Sorry to add to Scott's debate but as you've mentioned that wickets/test is biased as it penalizes players in teams with other great bowlers so you've excluded it, then your rationale for including the total number of wickets in a career is what? It just adds bias toward players that played in an era where a lot of games were played and where they were the only go to guy in the team.

    The other stats you've produced are truly quality tested, peer assessment etc., but the total wickets taken is not.

    You've expressed different views and fair enough as this is a great piece of work you've put together, something to be proud of. [[ Chris/Scott WpT is one of my favourite measures. I have to be fair to myself and all you informed readers. So I did a recalculation with only one change. Instead of Wickets I did WpT (with the meaximum of 6.015 for Murali). Lo and behold, there IS a change at the top. Murali still retains his 10.0 points but Hadlee brings down the big gap which existed previously (8.33 instead of 5.39) and moves to the top, as shown below. I myself am quite comfortable with this list since it has a good basis, not the Murali-Warne logic which is floating around.

                    Total   WpT 
    

    Hadlee R.J 87.83 8.33 Muralitharan M 87.54 10.00 Marshall M.D 86.38 7.72 McGrath G.D 85.89 7.55 Lillee D.K 84.75 8.43 Ambrose C.E.L 84.53 6.87 Garner J 83.53 7.42 Imran Khan 83.31 6.84 Donald A.A 83.24 7.62 Holding M.A 81.39 6.90 Trueman F.S 80.64 7.62 Pollock S.M 79.27 6.48 Waqar Younis 78.31 7.13 Warne S.K 77.90 8.12 Wasim Akram 77.82 6.62 Walsh C.A 76.12 6.54 Underwood D.L 73.77 5.74 Botham I.T 71.51 6.24 Kumble A 71.46 7.80 Kapil Dev N 68.33 5.51 Gibbs L.R 67.72 6.50

    Ananth: ]]

  • vishal naik on August 12, 2010, 9:49 GMT

    nice work mr ananta .but i would like to see performances era wise . see if for eg from 1990 till now. or from 1970 till 1989. so that conditions of playing will be same . also we can see how the bowlers in one era performed against the bowlers of other era

  • Abhi on August 12, 2010, 9:45 GMT

    And it all started from my utter conviction that murali and warne are just about as good. Ive (and im sure all of us in this blog have) watched both murali and warne for almost 20 yrs. So ,to me,it absolutely defies all logic to say murali is better than warne- or vice versa. The ripping off breaks, doosra, and variations are easily matched by warneys roaring legbreaks, desire to win, and his variations too. The only difference between them for most batsmen would be due to “biomechanical” reasons and geometry(rightie ,leftie)etc. I find it weird when I see stats like this with murali some 10% ahead of warne. [[ Abhi I thought your complaints were academic. Now I see the only purpose was to somehow put Warne at least equal to Murali. It loses lot of validity with me. To end with the sly digs at Murali is not becoming of you. Ananth: ]]

  • Alex on August 12, 2010, 8:27 GMT

    Kartik (the old one): since the WI greats were effective on turners as well, the only justification is to meet the 90 overs/day requirement. In Lloyd's great days, the WI pace battery regularly bowled less than 80. Lloyd was glad to pay the fine but maybe the Cricinfo committee wasn't! (This wasn't the only blunder: Weekes, Kanhai, and Dujon were left out as well. Asking Viv to open gives a sensible choice: Greenidge, Viv(c), Headley, Lara, Weekes/Kanhai, Sobers, Dujon, Marshall, Holding, Roberts, Ambrose.)

  • Kartik (the old one) on August 12, 2010, 5:07 GMT

    have forgotted who the pacemen were, but whoever was left out would probably have been a much better option to Gibbs

    Marshall, Holding, and Ambrose were (correctly) chosen as the 3 pacemen, but I think it is shocking that Gibbs was chosen ahead of Garner, or otherwise Roberts/Bishop/Croft.

    Now if Gibbs had a strike rate in the Murali/MacGill range, that would be another story. But there is NO justification for an 87 str.rate bowler (Gibbs) to be selected ahead of a 51 str.rate bowler (Garner). The goal of a bowling attack is to take 20 wickets. Period.

    Clive Lloyd knew this. The Cricinfo panel somehow does not.

  • Kartik (the old one) on August 12, 2010, 4:49 GMT

    Gibbs' low ranking proves a point that has been troubling me for weeks.

    In the Cricinfo All-Time XI feature, I was shocked that Gibbs could be in the West Indies All-Time XI, just for being a specialist spinner (apparently Sobers was not sufficient as a spinner).

    Gibbs strike rate was 87 balls/wicket, vs 51 balls/wicket for Garner and comparable strike rates for Roberts, Croft, Bishop, etc. Under NO circumstances should an 87 strike rate bowler be preferred over a 50 strike rate bowler, just due to some imagined need for a spinner (a belief Clive Lloyd famously invalidated).

    Furthermore, Marshall has often done well even on spinner-friendly wickets. I will wager that there is no track, no matter how tailored to Gibbs, where Gibbs would take more wickets than Marshall.

    I am less concerned whether Garner, or Bishop/Roberts/Croft is chosen as the 4th bowler, as long as it is one of these quicks and not Gibbs.

    The Cricinfo WI All-Time XI has made a grave error. [[ I am sure Sobers+Richards would have provided the spoinning options. I have forgotted who the pacemen were, but whoever was left out would probably have been a much better option to Gibbs Ananth: ]]

  • Alex on August 12, 2010, 4:35 GMT

    Ananth - re your August 12, 2010 3:07 AM comment, a few names exist who were arguably better than some of the bowlers in this list. Many old-timers rate O'Reilly, Grimmett, Lindwall, and Lohmann highly. E.g., Bradman rated O'Reilly to be the greatest bowler he ever saw; and Bradman, judging by his appraisal of Hobbs & Trumper, was hardly a romantic! Likewise, Gupte was possibly the greatest spinner since O'Reilly till Warne & Murali arrived. Of course, due low low # wkts, these bowlers will rank lower than even Gibbs in this analysis.

  • Ananth on August 12, 2010, 3:07 GMT

    I have since added Gibbs and Walsh in response to a few readers' requests. They have not exactly moved the world. Walsh splits the Pakistani pair in the third part of the table while Gibbs has taken over the last place from Kapil dev. I have steadfastly not included SF Barnes since he changes the whole dynamics by leading in couple of meaasures. I don't believe there exists one bowler outside this collection of 21, barring Barnes, who could be in anybody's list of top bowlers. This rounds off such requests.

  • Alex on August 11, 2010, 21:53 GMT

    Ananth - I think a few comments view this article as a means to decide who was better than whom whereas, perhaps, it is best used to understand how the game evolved and not to rank McGrath over Lillee, for example. Had Lillee played for Aus in McGrath's time, he possibly could have replicated something near McGrath's numbers.

    If the sample set includes Gibbs as well, you will have all bowlers since 60's who, at some point in their careers, became world record holders in terms of # wkts taken. Just like Walsh, Gibbs was under-rated but highly capable and durable.

    I doubt if anybody ever kept a track of world-record holders in average, SR (someone mentioned Bond), & RPO categories. [[ Alex If you read the title it does not say "best bowlers". It is only the people's own ideas. I have always talked of this "first amongst equals" list. Ananth: ]]

  • kamran khalid on August 11, 2010, 19:45 GMT

    Reading the comments has made me want to just say that, For All Bond, Akhter could not do was keep themselves fit. For whatever reason, If Bond needs to be included then what about Akhter. But then they are not the greatest of bowlers of all time. It takes endurance too. Bowling or batting, it is the hunger that drives a person. Murli may not be most loved or championed bowler still he is the best statistically. I do not like his style of bowling consider him a chucker but thats a personal feeling. Statistics should not be based on personal feelings. If they were Waqar Younis would be the best bowler in the world. As for murli & zim/bd issue Pakistan & England played as badly against Warne & Mcgrath as bd & zim played against Murli.

  • Bala Yugandar on August 11, 2010, 18:34 GMT

    Ananth-Really insightful analysis and stands to scrutiny. For all the people who are picking on Kapil's inclusion I can only say they are either ignorant or prejudiced or both. Remember Kapil was the fastest to 2000 runs and 200 wickets and was a genuine strike bowler until 1983 home series against Windies. He had a knee operation then and was a stock bowler until his retirement which did injustice to his figures. On an unrelated note just on sheer talent alone he's better than other 3 allrounders.

    I thought, like always, my favorite Macko would come on top whatever may be the measure....may be you should have included something like fear factor and Macko would get 10/10. How is quality of the wickets different than top order wickets? I guess the sole reason Imran leads in this measure is 82-83 series against India when he repeatedly got Vishy and pretty much everyone else in Indian batting order. Does Waqar's strike rate represent his whole career or at 200 wks milestone?

  • Asha1 on August 11, 2010, 18:24 GMT

    How great would it be to have a match between two teams with the first team having the top 5 bowlers, and the second team fielding numbers 6 to 10. They will have a great balance, two of the greatest spinners, greatest all rounders, as well as the fiercest fast bowlers to play the game. Also, just to compare the great fast bowling all rounders - can we also analyse the other peer fast bowling all rounder: keith miller? comparing him to imran and hadlee will be great!

  • dr somnath gore on August 11, 2010, 17:22 GMT

    fabulous work anant. no doubt contraversies, opinions will arise regarding the best among them as everybody have their own choices. but dont forget that, it is done considering many parameters. murali will be always best. nice to see the giant ambrose up in the order. simply unplayable according to sachin too. thanks again. pls do the sama for top twenty test and odi batsmen too with min. qualification of 3000 TO 5000 RUNS.

  • Mohammad Asad on August 11, 2010, 17:07 GMT

    Mohammad Asad from USA.........

    Very very good job !!!! Great analysis ...... Thank you Anand !!!! Keep it up ....

  • Glen on August 11, 2010, 13:38 GMT

    Great analysis,

    However,I have a few points esp in case of richard hadlee. Hadlee dint have the benefit of bowling to a bangladesh or zimbabwe during his time.He almost had to always bowl to the best batsmen. I wonder how he woul have fared if he had bowled to his own teammates. Although India and NZ took 83 odd tests to win certain number of tests their loss ratio during the period wasnt as appalling as bangladesh and Zimbabwe.

    Another problem is how does one decide the quality of the opposition batsman. The batting avg of a naseer hussain or atherton is in the mid 30s is partly bcos of McGrath and Donald and not the other way round. Ponting and waugh played against murali,wasim,waqar and lost I believe only 1 test each against them from 1995 to 2009. Does that make them the best batsmen?

    Finally, I am from India and I think Kapil deserves o be in the list. Not sure about Kumble

  • Scott- part 2 on August 11, 2010, 12:39 GMT

    Comment 2- The decision to include total wickets penalises the bowlers from an earlier era who due to both less tests per year, and shorter careers due to the game being only semi-professional lead to shorter careers. Bowlers such as Hadlee/Lillee/Imran/Garner and of course Trueman suffer by comparison. Bearing this out is the simple fact that Lillee, Truman,Hadlee each held at their retirement the WOrld Record for number of test wickets...ie) they eeked out every possible test and wicket available to them in the game as it stood in their day. To even this playing field it is imperative that wickets per test is utilised and NOT total wickets. Overall as always though great and interesting and debate-firing work. Many many thanks as always [[ Scott There is no need to be 100% fair to bowlers across the ages. That way it would be impossible to do any comparisons. After all wickets comprise 10% of the total and probably means 3/4 points which can always be made up through other measures. Incidentally on a WpT basis, Murali is first comfortably. His 6 WpT is exceeded only by Barnes and Lohmann.

    Ananth: ]]

  • Ron on August 11, 2010, 12:37 GMT

    I second the advocates of the team impact problem. The idea should be to eliminate team impact to determine intrinsic bowler worth.Not accentuate team impact. Murali may be said to be More Valuable to Sri Lanka than Warne was to Australia. It simply cannot be said that Murali was the better bowler.

    The interpretations from the tables are wrong.

    Murali will be able to get more top order wickets simply because he did not have the strike bowlers to take them out. This will also increase the quality of wickets taken. This will also lead to a higher % of team wickets taken/load borne- since the other bowlers were less capable.And in a way lead to a higher tally of wickets.Warne will also lose out on the peer comparison to team bowlers. The notion that Warne will improve his stats since he will then bowl to more lower batsmen is negated by Muralis Bangladesh wickets. Only the performance in won matches will be higher for Warne simply because of the larger number of matches won.

    Overall,Murali coming out as the top tweaker in this type of format is a no brainer. [[ Ron These facts are fine. How then do you explain Murali's outstanding performance (nt just wkts but bowling average) against other countries such as SA, England, WI, to a certain extent India. Warne against india was poor. It does not make him a lesser bowler. I think everyone seem to forget that it is not exactly a bed of roses to be a top performer in a weak team. Tendulkar had Dravid/Laxman/Sehwag. Who did Lara have. Similarly Murali had one support bowler throughout. Many of these comments seem to imply that Murali had to go and bowl over after over and the wickets would come. How then do you explain the low average and low strike rate. Ananth: ]]

  • scott on August 11, 2010, 12:34 GMT

    Great analysis and yet another prompt for debate aplenty- lets face it we'd all love any of these bowlers representing our country at any point in time. The only two queries are the QUALITY OF WICKETS & the impact of thee Zimb/Bang' wickets. -Quality of Wickets- this is a self-fulfilling prophechy. The reason that the Windies bowlers of the 80's and the Warnes/McGrath's couldn't further increase their ratings i this area is that the batsman's averages were subject to/a result of the performance of these very bowlers; a viscious and misleading circle. Re Murali's Zimb/Bang' wickets: it would nonetheless be interesting to see these removed. Over the course of his career they were "consistently" poor, and he excelled against them. This consistency of their poorness and the number of tests/wickets he gained take these team's impacts on this analysis to a new level, and thus warrant a second analysis ex-them.No doubt Murali would still be near or at the top, but the gaps may shorten.tbc [[ Scott If I take away Murali's wickets I have to take away many other wicket hauls. All bowlers against Bng/Nzl, Trueman's againsi India/Nz, Other bowlers against Slk during 1992-2001, the 5-0 Ashes drubbing ??? Where do we draw the line. Ananth: ]]

  • Raj on August 11, 2010, 12:28 GMT

    Its an excellent article. For those who are complaining against Murali for taking wickets against weaker team; Australia looked like a dream team for almost 10 years because it had more than 6 great players in the team at the same time. So all of Warne's and McGrath's wickets should be counted as wickets against weaker teams. Also to those who say "it's just numbers", "you will get it if you play long"; its not easy to play long. Look at Bond and look at how Mendis started his career and how he is performing now? These are just few names from present, there are countless examples from the past as well. It takes incredible commitment to play continuously for more than 10-15 years, i think it has to be valued greatly. Its easy to write that Kapil prolonged his career for other reasons, just take the performance of Kapil in his last 15 tests and please give the names of Indian fast blowers who performed better than him in the past 20 years, i bet you wont get more than 2 names. [[ Raj You are making a valid point. There was a time when Warne had to bowl one over of leg breaks and Cullinan would go. When Australia won 16 consecutive tests, what about the opposition. Were all the matches close. Ananth: ]]

  • Taimur Khan on August 11, 2010, 11:53 GMT

    Ananth your response to my query makes so much sense that I would argue for another type of analysis based on it. You said "Give Kumble credit for getting over 600 wickets after being branded after a series as a non-spinning spinner. And credit Kapil, bowling his heart out on the flat Indian pitches."

    Agreed. Therefore, a logical follow up should be to assign factors for pitches. Kapil could have 150 more wickets had he played most of his cricket in England...remember that away swinger! Lillie wanted to be buried in faisalabad! India will produce great bowlers if you give them responsive pitches.

  • Subhranil Deb on August 11, 2010, 11:42 GMT

    The fact that you are an India-baised is evident! How can Kapil Dev be there in this list? You dropped Holding, Croft and Shaun Pollock, Botham. I dont know what was your criteria to select 3 from some countries and 2 from some. I guess there would be many who would rate higer than Kapil. [[ Subhranil Thanks for calling me biased towards India. I normally get called the other way. You are making your comments without reading the article fully. Kindly re-read and let me know if Holding, Pollock and Botham are not there. Ananth: ]]

  • Chris Bolton on August 11, 2010, 11:26 GMT

    Like Brent I have my doubts when using the overall wickets taken, essentially you are punishing players like Trueman for not playing enough games when those games did not exist! Although you don't want to diminish the huge workloads and strains borne by the likes of McGrath and Murali the counter argument is always that they were bowling to batsmen who were going through equally strenuous playing schedules.

    You have selected a sample of players from different eras so it would be odd to use a stat that blatantly favours one era over all others. Some may argue that it was easier to bowl in earlier eras but the existence of only one true 'anomaly' in SF Barnes throughout the history of the game would tend to quell that view.

    Anyway, this is an interesting way of looking at some of the greats of the game, top work Anantha!

  • Anshuman on August 11, 2010, 10:56 GMT

    Ananth..as usual you vl justify your statistics in every possible way..n there will be doubters...

    so..as it should be..my two cents on this issue or the batsmen list you hv done before...

    instead of number of total wickets taken or toatl runs scored which skew it in favor of murali and sachin..it would be really prudent to take wickets per match an druns per innings as one of teh attributes in doing teh final selection..

    anyways its ur take on it.. [[ Anshuman Pl see my reply to Scott. Ananth: ]]

  • Sayan on August 11, 2010, 10:15 GMT

    Excellent article, and I'm quite happy to see how it's turned out. I'm a skeptic in general when it comes to quantitative cricket analysis but this takes the cake, as do most of your articles.

    I agree, however, with the decision to include Kapil maybe not being a great one. As much as he had a huge impact on Indian cricket, who's to say that a Kiwi wouldn't say the same about vettori? Or any other country, really. I think he was out of his depth here, especially as an allrounder.

    I wouldn't be sure about this, but maybe something can be accounted for match situation? for instance, surely the wickets taken on a dustbowl in india aren't worth as much (weren't as toiled for) on average than wickets taken on, I dunno, any one of the recent India/Sri Lanka tests? Maybe something along the lines of percentage of wickets taken out of total wickets in the match?

    All in all, fantastic article, and I hope to reads some sort of sequel to it soon!

  • Max Anandappa on August 11, 2010, 10:11 GMT

    For those who devalue Murali on the basis of wickets against Bangladesh and Zimbabwe here is a pragmatic way to look at. His wickets against Zimbabwe included a stronger team that included the likes of Flower brothers etc. But still let us discount this. In 26 tests against Zim and Bangla Murali had taken 176 wickets. Against England in 16 tests Murali had taken 112 wickets @ 7 wickets per match. Instead of Zim and Bangladesh had Murali played 26 more matches against England he would take 26x7= 182 wickets, theoretically more than the wickets he took against Zim and Bangladesh

  • Priyank on August 11, 2010, 9:52 GMT

    A good analysis but I think you should include one very important factor which is the overall fielding ratings of the team represented. If someone comes up with similar analysis after 10 years, then Mhd Aamer and Mhd Asif would not even make the cut due the the number of catches pakistan team has managed to drop.

  • Waseem Khemji on August 11, 2010, 9:06 GMT

    Excellent work. Very Impressive.

    No Complaints on the quality of the content but just wanted to add that this fact needs to be mentioned and considered that the final standings on the list is not only the bowler itself but the fielding support that it got as all the three Pakistani Bowlers would have had a lot more wickets if their fielders have supported them.

    Anyways great statistics.

  • Vasu on August 11, 2010, 8:54 GMT

    @Sanjeev: Good point. My idea was not to belittle Gavaskar or Ponting. If Gavaskar scores 200 against an attack that has Marshall, Roberts, etc plus Gomes, it is fantastic even if he scores 0 against Marshall. Keeping out the best bowlers is great in itself. I was referring to matches where the attack itself was week. For e.g. in Gavaskars's famous debut series in 1971, WI had a very weak bowling attack –Roberts, Holding, Garner, Croft, & Marshall were not even around. I don't want to deviate from THIS article, but my idea was to square off the best batsmen Vs. best bowlers of his time and see who comes on top. For e.g. I would GUESS Chappell (WSC comes to mind), Sachin, Lara are be top batsmen in last 3 decades purely based on quality of bowlers in the opposition attack.(e.g. Ponting did not face Warne, McGrath), just like Hadlee, Imran etc. figure highly in quality of wickets taken. I think Anantha fully understands and like he said, it’s the mother of head to head!

  • Yogi on August 11, 2010, 8:52 GMT

    Simply marvellous!

  • praveen on August 11, 2010, 8:48 GMT

    great article...

  • RC on August 11, 2010, 8:15 GMT

    First, I am glad that you have not downgraded wickets against weak 'teams'. What Warne supporters who claim that he is better than Murali often forget is that against high quality in-form spinners, England and New Zealand are as 'weak' as Zimbabwe and arguably worse than Bangladesh. Second, I think you should have given points for captaincy( with Imran Khan getting 10). This might be unfair to those who never were captains or were for only a short time, but captaincy can have as detrimental an effect on the performance of bowlers as on that of batsmen. With both batsmen and bowlers, many greats were not able to be effective captains. Third, you should have included Vettori(I have no idea at whose expense). He would provide a superb case for studying a situation where one bowler is very important to his team(like Hadlee or Murali) while facing a modern workload and having captaincy duties.

  • Nataraj on August 11, 2010, 8:07 GMT

    Ananth, Excellent Analysis. Again proves Glenn Mcgrath is always there wit the best. Also admire Imran, though bowling on subconstinent wickets mostly has better figures than Botham and comparable figures to Hadlee(conditions in Eng and Aus where Botham played most were helpful to bowlers).

  • Abhi on August 11, 2010, 8:03 GMT

    Ananth Noooo. I am not implying any starting bias. I am saying that with the methodology used the outcome will always be similar. As rgds. Hadlee the main thing is wickets taken where he lags behind Murali by a full 5 points.(Though of course you have to give credit there) Murali played 230 inn vs Hadlees 150. Fast bowlers generally have to get their wickets in lesser overs and do not have the luxury of bowling 40/50 overs a day. They also tend to get injured more. So, comparing a fast bowler to a spinner is in any case fraught with peril.

    Like i said there is NO way I am implying ANY sort of staring bias. It is just that with the parameters used Murali is almost certain to trump Warne by a whopping margin. Also, I would add that out of the Great triumvirate of Murali/Warne/Kumble…Kumble was a notch below in any case.

  • KiwiRick on August 11, 2010, 7:31 GMT

    The weighting scheme you have chosen has in fact introduced bias between measures, as the range of values in some measures is much lower than the range in others. This gives bias towards two measures - no 1 and no 7 - namely the two Murali is way ahead of the rest in.

    If we weight on the range from best to 15th value, rather than best to zero, we get a more even distribution per measure. However the result is very similar with Murali just edging out Hadlee with McGrath third and Marshall 4th.

    If we assign ranking points per measure (1= best and 15 = worst), we get a very different picture - Hadlee edging McGrath and Marshall, with Murali a distant 5th behind Ambrose.

    Either way, Murali, Hadlee, McGrath and Marshall in any order is a great top 4. The Kiwi in me would choose Hadlee as with him NZ won tests and beat Oz, before and since we struggle.

    The big surprise is Lillee's low position given the reverence he has in Oz --- could this be due to WSC? [[ Rick As you have observed, think it is a good idea to consider these bowlers in groups of 5. Murali, Hadlee, McGrath, Marshall and Ambrose, irrespective of the order represent an outstanding collection any captain would love to have. Ananth: ]]

  • Ali on August 11, 2010, 7:21 GMT

    Great article. I think it puts an end to the Murali/Warne spat once and for all. So clearly Murali is way ahead. Those who still have doubts "I am really sorry for them"

  • brent on August 11, 2010, 6:58 GMT

    perhaps might have been better to use wickets per test rather than overall wickets taken. how many players have played 100 or more tests in the last 25 years as opposed to those before. Trueman or Garner would never had the chance to play 100 tests as Kumble, Warne, Kapil or Murali did. [[ Brent In my earlier analysis I had WpT. It was considered to be unfair to the top bowlers in strong bowling sides like West Indies, Australia and Pakistan. Ananth: ]]

  • Dilin on August 11, 2010, 6:58 GMT

    Anath,

    Excellent article. Its great to see all the figures down in black and white to see. You've also hit the nail on the head when you talk about removing all wickets taken against Sri Lanka from their frist test, along with New Zealand and even England in the 90's.

    This list thoroughly shows the importance of Murali in Sri Lanka's efforts to gain the #1 ranking (even though they weren't able to).

    Maybe another aspect you could add would be how many games were won by each team without each bowler?

  • Xolile on August 11, 2010, 6:42 GMT

    Ananth, Could you please also include Ntini? He performed consistently in an era dominated by the bat. If he played on the bouncier tracks and against the weaker batsmen of previous eras he probably would have averaged low 20s. He also did amazing work in breaking down racial barriers in SA. He is a true ambassador to the game. Courney Walsh would also be a welcome addition to the list. [[ X Ntini's figures are comparable to Kapil Dev's. Hence they lose out when compared to Donald's or Pollock's figures. Walsh is a definite possibility although i have already taken 4 West indian bowlers and Roberts is still there. Ananth: ]]

  • Alex on August 11, 2010, 5:51 GMT

    Ananth - Weighted bowling average, i.e., "bowling average/ WktQty" reveals a lot when all other conditions are the same. I get the following numbers --- Group 1: (Imran: 22.8, Marshall: 23.6, Hadlee: 24.1, Garner: 24.4, Holding: 24.5, Lillee: 25.1); Group 2: (Ambrose: 27.7, McGrath: 28, Donald: 28.5, McGrath: 28, Akram: 29.7, Waqar: 30.2). Murali: 28, Trueman: 30, Warne: 35, Kumble: 38.

    The big difference in Group 1 and Group 2 numbers is probably because the Group 2 bowled a lot in 90's, the decade in which batting averages dropped --- TORto is 9.5 for Ambrose and 9.2 for Imran but WkQty is only 7.7 for Ambrose while it is 10 for Imran.

    Discussions on all-time great fast bowlers often focus on Lillee, Marshall, Hadlee, McGrath, Akram, and Ambrose; somehow Imran, Holding and Roberts get forgotten. Your analysis shows that Imran clearly belongs in those discussions.

  • Ramesh Kumar on August 11, 2010, 5:33 GMT

    Ananth,

    excellent analysis. enjoyed the work which has gone in. Alex's further intrepretation make it very insightful. I have still one unanswered question about these bowlers in batsmen dominated decade. We seem to have seen a lot of runs in the last 10 years and have seen bowlers creating records in wicket hauls. If we do a similar one for 90s decade and the last decade, will we get poorer ratings for bowlers in the last 10 years? Are there atleast 5 bowlers who made a debut after 2000 who will compete for places here(excluding the total wickets, ofcourse)?

  • Anonymous on August 11, 2010, 5:15 GMT

    The article and the stats just blew me away. Although seeing the W's near the end made me a bit :(. But stats are stats. Great piece. [[ Would be a better idea to give a name. Normally I would have deleted it. But I also wanted to send a message for readers to give their names. Ananth: ]]

  • Imran Khan on August 11, 2010, 4:55 GMT

    Nice Comments By all People. But i think Wasim And Waqar were the best bowlers in the world.Especially wasim bowled all types of delieveries at the time of requirement.

    In future AAmir will be the shining star of Pak Team.

  • Abhi on August 11, 2010, 4:50 GMT

    Ananth, Ok, “modified” comment! Once, the "poor team" part is taken care of(ideally with a single parameter)...if we keep on allocating more and more points for parameters which are effectively a "Function" of the same thing...Murali will always come out on top no matter "which route is taken, be it the high road or low road, be it the country road or the highway, be it the scenic route or the road through the concrete jungle, there is only one destination. The one which proclaims Muralitharan as the best amongst the equals."

    To compare bowlers from different teams either we limit the number of “team ratio” parameters of we give more weightage to the “pure bowling” parameters- strike rate, etc. [[ Abhi Two things. One is that this analysis was never done to push one player on top. Pl don't imply that. The second is that there are only two parameters which might be of benefit to a player in a weak team. The Peer2 and the %TeamLoad. The other parematers infact act against such players. If Murali is on top it is not because of his playing for a weak team but because he was able to overcome the shortcomings of playing for a weak team. Pl compare the numbers of two spinners, Murali and Kumble, both playing in sub-continent, both having no great bowling support. They are chalk and cheese. If Murali's average was 27 and strike rate was 62 he would not be first. Let me also throw another point. Hadlee had all what Murali had in terms of playing for a weak team. In addition he had the benefit of a better strike rate, chance to capture top order wickets, much better away average. Why did not Hadlee then come first. Ananth: ]]

  • DI on August 11, 2010, 4:30 GMT

    F A N T A B U L O U S Anathana Narayanalysis.. One of the best i've read in the bowling department. I have always ranked the Pigeon - Glenn Donald Mcgrath higher than Wasim Akram and you've confirmed that.

    Some of the measures that could been considered, - Dubious decisions - Referrals

  • Murali k on August 11, 2010, 4:20 GMT

    Nice approach. Wonder how the results look if the weights are assigned as percentiles based on gaussian normalization instead of simple ratios which can sometimes have inherent bias [[ Murali You have the numbers. Can you do a gaussian normalized table for our benefit. Ananth: ]]

  • Kartik (the old one) on August 11, 2010, 4:09 GMT

    First, Anil Kumble, whose contributions to Indian cricket can be matched by only one player, the little master. One could say the same of Kapil Dev.

    Doesn't the first sentence contradict the second?

    At any rate, my biggest takeaway from the table is the conspicuous lack of bowlers produced by India, relative to other countries, particularly given India's population. [[ Kartik I really mean that the achievements of these two players can only be matched by one other player, the little master. Maybe it did not come out like that. Ananth: ]]

  • AD on August 11, 2010, 2:37 GMT

    There was an article by Rob Steen on Muralitharan a while back- " The Ultimate MVP". That should be the article for this analysis too. Muralitharan is the Most Valuable Player relative to the others in his team.Not relative to other bowlers in other teams. Most of the points considered show how well Muralitharan has done relative to his team mates. It is incorrect to title this "Best Bowler" or the like and compare it directly to other bowlers in other teams with entirely different dynamics. What it does show is that Muralitharan was relatively more important to the fortunes of his team than others.

  • Elton Barkoye on August 11, 2010, 2:18 GMT

    Under the circumstances, I would say this is a fair assessment. Yet, a true analysis of bowlers is very difficult given the evolving state of the game. For instance, the technology available today so that bowlers can study batsmen and how to get them out was not available in early years. Additionally, there were times when teams contained the no. 1 test batsman yet the team failed to win matches - B.C.Lara, A. Flower, etc. Hence, bowlers may have come up against top ranked batsmen but not in top ranked teams.

  • Des on August 11, 2010, 2:08 GMT

    There are 2 problems with this analysis.The omission of Sydney Barnes does distort all of this because his record is incomparable - it is not possible to have the great bowlers when the greatest is omitted. Pre-WW11 also leaves out Grimmett. The second is the exclusion of Shane Bond for NZ with a strike rate unequalled in modern cricket - and Dale Steyn for SA, not far behind Bond. [[ Des 1. All the 15+4 bowlers have completed their careers. It would be unfair to include Steyn who is at the mid-point of his career. 2. If I include Syd Barnes all sorts of adjustments have to be done. How can you otherwise compare a bowler who bowled 100 years back on uncovered pitches with today's bowlers who bowl on protected pitches at well-protected batsmen. 3.Do you seriously want Bond who captured 87 wickets in 18 tests with people who have captured 4-9 times the wickets. I don't think you yourself would like it. Ananth: ]]

  • Sanchez on August 11, 2010, 2:04 GMT

    I think away average counts for alot when it comes to true greatness. I mean, the ability to bowl superbly on all wickets, in all conditions, that is true brilliance. Both Warne (in India especially) and Murali (in Australia especially) struggled a little bit in this regard.

    I would not change the analysis at all, you have done a great job. I am just not convinced, much like Ian Brooke's comment, that Murali was at the top ahead of McGrath, Marshall and Hadlee.

    You mention the impact Murali had on Sri Lankan cricket. You could argue almost all of the bowlers had a seismic impact on their nations. Australia were decent in the early 90s when Warne and McGrath started, but not the world conquerors they turned out to be. Hadlee dragged NZ into the winners circle. Would South Africa be where they are without Donald?

  • Abhi on August 11, 2010, 2:00 GMT

    Ananth My point was that the “good bowler in weak team” has more parameters where he will probably score higher : 1) Career wickets captured (more per match). 2) Average quality of wickets captured. 3) % of top order wickets captured. 4) % of team load borne 5) Peer comparison of average to own team peer bowlers. i.e 5 out of 10 parameters “favour” the good bowler in weak team.

    For the good bowler in good bowling team: It is perhaps just 1) Performance in won matches

    So, with even bowlers who are equally good – the arrangement and choice of parameters will always put a Murali in SL over a Warne in Aus.

    I wonder if there is some way to “equalise” this. i.e though a good bowler in weak bowling team is given the benefit of these extra points these should be balanced out and not overwhelm the good bowler in good team.

  • Amit on August 11, 2010, 1:28 GMT

    Ananth - Fantastic Analysis. And I am thrilled to See McGrath show up at #2. To me he was the most 'effective' bowler in the last 40 years I have been following cricket and made a 'very good' Australian team into one of the three best ever. Now, the part that wasn't convincing to me was the - Quality of wickets captured. it is inherently biased toward bowlers from weaker teams (kapil in the top 3 and McGrath almost at the bottom). Not sure why McGrath should be Penalised, especially since he may be responsible in bringing the average down of the batsman. Also, it a common complaint that batting was easy in 2000s and batting averages are inflated by about 10%. It also means that bowlers avarages should be lowered by about 10% in 2000s. The modern bowlers have had it worse than bowlers in the past and that should be accounted for. I am not sure it would change the ranking, but I am trying to do my bit to push McGrath up to #1! As an Indian fan, McGrath worried me more than anyone else!

  • Lee on August 11, 2010, 1:12 GMT

    Wonderful, Ananth.

    For a country so obsessed with stats it is a wonder that the Yanks don't get cricket!

  • fahad khan on August 11, 2010, 0:05 GMT

    I wonder if Fazal Mahmood would make it. He too like murali and hadlee was a one man bowling attack, leading pakistan to its most famous wins. His stats in wins were awesome. Average and wickes per match also compare favorably with the latter pakistani fast bowlers. I would put him ahead of wasim and waqar for impact. [[ Fahad Fazal was one of the best of his time. However I am very uncomfortable with the types of pitches on which he bowled in india and Pakistan, most of them matting. He was virtually unplayable. Ananth: ]]

  • Yasser on August 11, 2010, 0:05 GMT

    Great Article. In my humble opinion, one way it can be improved is to also give more weightage to bowlers who have toiled relentlessly on dead sub-continent pitches. Imran, Wasim, Waqar and Kapil played most of their cricket there and did great while a great bowler like Lillee suffered badly in Pakistan. I do not think that Trueman played any matches (i could be wrong) in the sub-continent. [[ Yasser To compensate for their toiling away on lifeless pitches they are given higher credit when they go to the livelier wickets in the form of "away average". Ananth: ]]

  • kamran khalid on August 10, 2010, 23:00 GMT

    Well Kapil was to India what Headlee was to Nz. Murli still had a partner in crime in the name of Vaas. Not the greatest but still respectable. Helped him a lot. Wasim had Waqar and vice versa, Ambrose had Walsh and Marshal. Imran had the company of Sarfaraz and then Wasim. So although it makes their figures diminish a little yet it was also a help for them as keeping it tight at both ends.

  • Alex on August 10, 2010, 22:18 GMT

    Ananth - great. The columns reveal a lot more than their sum. (i) I wouldn't add the columns to get the total. E.g., this penalizes Marshall who has a "Peer1" score of 10 but "Peer2" score of only 7.6 just because his team-mates were great (leading to a low "TmLd" score as well). (ii) The product "WkQty*TmLd*Peer2" reveals how irreplaceable the bowler was for his team. Hadlee ranks 1st on it, Murali 2nd, and Imran 3rd. (iii) The product "TORto*TmLd" reveals how potent he was, in the context of his team, to the opponent top order. Hadlee ranks 1st, Murali 2nd, McGrath 3rd, and Imran 4th (W&W, Pollock, and Underwood are the bottom 4 on this metric). (iv) The product "BowSR*WkQty" reveal the true danger to top class batsmen in terms of wicket taking ability. Scoring 83 on this metric, Marshall & Holding are the top 2 and daylight is 3rd, followed by Lillee, Imran, McGrath, & Hadlee bunched together @76. [[ Alex Trust you to extract more from the tables than was intended. Your combinatiuons for creating these sub-tables are excellent. Ananth: ]]

  • Kartik (the old one) on August 10, 2010, 21:51 GMT

    I feel it necessary to point out that a country with more than twice as many people as all other Test nations combined, still has not produced a bowler who can crack the Top-15 of the final table.

    This is despite the incentives derived from the fact that such a bowler would accrue unimaginable fame and wealth.

    If you think Tendulkar has extreme wealth and fame in India, imagine what a cricketer such as McGrath, Akram, or Ambrose would get if he hailed from India.

    Yet, that huge population cannot produce a bowler who can make the Top 15 of the final table of 19.

    Some Indian billionaire cricket fan needs to put up a prize, saying that the first Indian bowler who can take 200 Test wickets at a strike rate better than 55/average better than 23, will receive a prize of $1 Million.

    This would change things...

  • Fahid on August 10, 2010, 21:48 GMT

    Brilliant stats analysis, stats never show full skill of a bowler but uptill some aspects 1-Wasim Akram 2-Imran Khan 3-Lillee 4-Garner 5-Hadlee 6-Holding 7-Pollock 8-Warne 9-Trueman 10-Waqar

  • Jason on August 10, 2010, 21:01 GMT

    Just one last thought Ananth - I wonder whether there could be a measure to see how each bowler did 'against the best', see what their performance against the leading team of their era was or perhaps the most powerful batting lineup of his time. That is a common argument against Murali - his record against and particularly in Australia. Perhaps the measures you have already cover a bit of it, but I just thought it may be worth further investigation.

    One final suggestion: do this study but for the top 20 bowlers of each era. Pick the top 10 wicket takers of the time, plus the guys with the 10 best averages from wicket takers #11-30 to get your 20 bowlers. I can think of 9 eras to do: 1876/77-1896,1896/97-1914,1920-1939,1946-early 60s,early 60s-1979 (end of WSC),1979/80-1995 (end of WI reign) and 1995/96-now. You may need to combine the 2 Pre WW1 eras to beef up the data, and there will be some overlap for some bowlers across eras I guess, but it would be fun to see how it looks. [[ Jason Excellent idea but a lot of work. Will get into it when I can find more time. Ananth: ]]

  • shmulik on August 10, 2010, 20:51 GMT

    why no mention of dale steyn? [[ Mulik All the 15+4 bowlers have completed their careers. It would be unfair to include Steyn who is at the mid-point of his career. Ananth: ]]

  • Arsalan Khan on August 10, 2010, 19:22 GMT

    Thank you for the insight on these great bowlers. However, I was wondering do you ever take the surfaces and conditions for bowlers into account? for example like Murali may have played more in the sub continent - the pitches are more spinner friendly.. likewise even if Kapil and Wasim are lower down the order they have had to bowl on the graveyards of lahore and chennai. Maybe that is a reason why these greats prop up in the lower order.

  • rahul on August 10, 2010, 19:10 GMT

    hi sir.......... great article i m very impressed......

    Sir u please do one thing for all of us....

    with regards to ur great cricket statistics i m giving u the most difficult job u would have ever imagined....

    my question is can u pick an all time world eleven in both ODI and TEST....

    you should keep in mind that the players should be selected only based on their potential... should not consider the total no of runs or wicket to select batsman or bowler....

    it may be possible that one with less runs have more impact on winning match for his country... EXAMPLE: michesl beven(7000 runs ) is better than dravid and many others....

    there should be 5 bats 1 all rounder 1 WK and 4 bowlers....

    and please repli me and print it on cricinfo page and let the whole world know which is the best all time world eleven... thanks...... [[ Rahul Possible but has to be approached with lot of caution. There are too many contrasting opinions floating about. Ananth: ]]

  • JHALMREST on August 10, 2010, 18:57 GMT

    Sometimes a bowling pair is much more effective that just one single bowler. With this in mind, where do Wes Hall and Charlie Griffith from the West Indies stand?

    Considering much more cricket is being played today than previously, (like in the 1960's when they played), the amount of wickets actually taken would give today's bowlers a distinct advantage.

    I am still to be convinced that there was ever a more feared and dectructive pair than Hall and Griffith!

    .......And I haven't even mentioned the beguiling spin of the great Lance Gibbs yet.

  • TheBigFatFlapjack on August 10, 2010, 17:57 GMT

    Hi Anand,

    Awesome work here. This list shows Murali is the best Test bowler in history, statswise at least. I particularly like your inclusion of other facts such as quality of wickets captured etc as they give a more complete assessment of a bowler's ability.

    But i have to mention something which has already been brought up here - in my opinion kumble and kapil dev weren't really as good as the other bowlers on here. i'm not saying this to undermine their bowling - i know kumble has been consistent and kapil dev has also had his fair share of success but to me javagal srinath, chaminda vaas, shoaib akhtar and saqlain mushtaq are far better bowlers from the subcontinent.

    that said, this last bunch (possibly with the exception of mushtaq) are again not in the same class as holding, pollock etc. however, i do understand your point about not wanting to populate this table with w. indian/aus bowlers but it goes to show these countries have produced a disproptionate no. of good bowlers

  • Samir on August 10, 2010, 17:50 GMT

    Hope you're right ananth, my comment was however tinged with sadness, I used to feel that Agarkar could have amounted to somehing more than he ended up as if someone had taken him in hand. He did seem to be a rare natural, rare by India standards, perhaps that's why we don't have good pacers or good finds, is it the coaching? i wonder??

  • Sanjeev Chandran on August 10, 2010, 17:46 GMT

    @Vasu- while home and away stats would be interesting I am not sure what value head -to head would be. Let's take the case of Gavaskar's 236 against WI which had an attack of Marshall, Roberts, Holding, winston Davis and Roger Harper and Gomes. Gavaskar scoerd more than half of india's runs and Harper and Gomes (the unknown spinners) gave more than a 3rd of the runs India scored. So, it's a good chance that a huge proportion of Gavaskar's runs came against the spinners. So what? If a man is able to get a through pace attack like the above and keep scoring against the spinners, that too is great. In fact, to get to the extreme, if a player can completely farm out the best bowlers, not get out and score all his runs against the weakest bowlers- it's brilliant -both strategically and tactically.

  • Samir Yeshwant Hajarnis on August 10, 2010, 16:58 GMT

    Dear Vignesh Devaraj, what have you been drinking, mate?? Even Boycott's mum could have hit agarkar all over the park .. sad, but true. [[ samir i get the feeling that was satirical comment by Vignesh. Ananth: ]]

  • Abhi on August 10, 2010, 16:57 GMT

    To elaborate: If you had a Murali clone. One in the SL team and one in the Aus team. You would get vastly different rankings as per this system. The method would only be accurate if “both” the Muralis are on par… [[ Abhi This is an argument for the sake of argument. If you had a Murali clone in the Australian team, ny no stretch of imagination would their figures be identical. the Murali clone would bowl fewer overs, capture fewer wickets and win more matches. He would not be bowling at the stronger Australian batsmen. Ananth: ]]

  • Abhi on August 10, 2010, 16:52 GMT

    Still not fully convinced. Even with 2 exactly identical bowlers: the bowler in a weaker bowling lineup will almost surely get a bigger crack at “ Average quality of wickets captured” , “% of top order wickets captured” ,”% of team load borne” in any case…"Peer comparison of average to own team peer bowlers"...just keeps adding on to this mix.

    I agree that the “load” on a bowler is perhaps more when bowling in a weaker team……but does this qualify him as a “better” bowler?

  • P.N.Sudarshan on August 10, 2010, 16:29 GMT

    Ananth A good analysis. I was wondering whether there should be another factor in terms of 5 WPI or 10 WPM, which typically translate to match-winning performances from a bolwer perspective [[ Sudarsan Don't you think these innings landmarks are being given too much importance. 4 for 25 is probably any day better than 5 for 100. It is the same way I never take 100s and 50s in any oif my analysis. Ananth: ]]

  • Ananth on August 10, 2010, 16:19 GMT

    In response to some readers' valid requests I have added Holding, Botham, Underwood and Shaun Pollock (my choice) and presented the revised table at the end. Let me also add my thanks to the readers on the mature comments which are being made. The excellent comments of Kamran Khalid sums it all. The way in which the unfortunate low placement of the two great Paistani bowlers is taken in the right spirit indeed wonderful. Let me add my two-bit. If my life depended on a bowling attack dismissing England below 118 at Birmingham, I would unhesitatingly choose Wasim Akram and Waqar Younis and they might very well have saved my life.

  • Abhi on August 10, 2010, 16:13 GMT

    Ananth Im not sure that both the “ Peer comparison of average to own team peer bowlers” and ““% of team load borne” are good choices. They also sort of double count each other. .the better bowler in a weak bowling lineup will have a higher count of both. “Peer comparison of average to all peer bowlers” is more accurate and better than the above 2. Since we are primarily concerned with bowler in team A to bowler in team B. (since even fast bowler vs. spinner in same team will give inacurrate results because of quality of wickets taken) If you deduct both ““ Peer comparison of average to own team peer bowlers” and ““% of team load borne” you get : Murali :68.29 Warne:64.23 The difference in “reality” between Murali and Warne is miniscule( if at all)…the current tables give figures of 87.54 to Murali and 78.63 to Warne. Something not quite right somewhere. The other parameters seem ok. [[ Abhi The Peer comparisons to own team, to a certain extent compensates for playing in a weaker team. Murali might score in this measure (and it so happens in the other also) while Marshall scores in the other one and not in this one. Finally it was my intention to make sure that the most important measures, bowling average and peer comparisons carried addotional weights. Ananth: ]]

  • raja dutta on August 10, 2010, 15:51 GMT

    amazing work can u also do such a comparison for batsmen hope Indians will be a the top there.

  • unni on August 10, 2010, 15:12 GMT

    After you took break, I stopped following this blog for sometime.... forgot that you are back.. need to catch up all the missed ones !!

    As usual I see you put lot of effort for this. Nice read. I agree with Sid and Asif in using just 'quality of wickets' instead of both quality and top order wickets. However, the numbers give a different story above. The order doesn't match for both.

    Anyway, the quality column was bit fascinating. I don't remember if you ever published this in one of your old artiles. Imran, Lillee and Kapil being on top is really a new insight. Eventhough I'm not a fan of Kapil being in this list, this measure would give enough reasons for your decision. I would like to see a list including all major bolwers for this list. [[ Unni I missed your incisive comments. The reason I have left both "similar shades of gray" measures is for the reason that the values and order are quite different indicating that these are genuinely different measures. Ananth: ]]

  • Jim on August 10, 2010, 14:35 GMT

    Undeniably a great list of bowlers. They certainly stand out as the greats according to their career stats. If career longevity wasn't a factor, Fred Spofforth and Jeff Thomson in their respective eras would fit the criteria of the title of your article, "Fearsome". It's reported that Spofforth had pace unequalled at the time, and Thommo, well he had that action and I'm sure he never new what the ball was doing at times. What hope did the batsmen have?

  • Hitman57 on August 10, 2010, 13:51 GMT

    Interesting comparison. Only one thing, maybe one thing that could make Warne better than murali is that warne had to share his wickets with mcgrath, lee etc while there was no other noteworthy bowler in SL with whom murali had to share wickets.

  • Vasu on August 10, 2010, 12:38 GMT

    Hi Anantha, Not sure if this was studied earlier, but your list makes me think about this: Is it possible to study batsmen, purely based on the quality of the opposition bowlers they faced through their careers? For example, Ambrose, Wasim, Waqar, Warne, McGrath, Murali, Donald etc played thru the 90s, and Sachin & Lara apart, not many batsmen shined against them over a period. Ponting has similar stats to Sachin, but faced much weaker bowling. You could take certain periods dominated by strong bowling where every run you scored was worth a 100 today, and study batsmen who did well. Say for example mid 70s thru late 80s dominated by WI quicks, and early 90s thru the end of the decade dominated by the list above. Maybe you could dive deep into the stats – home, away, runs against specific bowlers etc. For e.g. Gavaskar scored a lot of runs against Windies, but how many of them were against Roberts, Holding, and Garner, as opposed to some unknown spinner? Thanks. Vasu [[ Vasu The mother of head-to-head analyses. Let me look at it. Ananth: ]]

  • Ganesh on August 10, 2010, 12:20 GMT

    Ananth As usual a marvellous analysis. There is no doubt that these are the 15 of the all time best bowlers after WWII. Although I am slightly sad that Kapil lags in the table ( I am one of those who saw all of his 29 wickets in the 83 series against a West Indies that was formidable) , his true worth comes in "% of top order wickets captured " and " Average quality of wickets captured " where he is 3rd punching far above his weight even among such an elite company

  • asif on August 10, 2010, 12:14 GMT

    I agree with Sid, if we calculate the quality of the wickets (average of the batsmen dismissed), i see no point in % of top order wickets calculation as the former is the better measure. Also, after having point no 7 (sharing team load) and point 8 (performance against all peer, i think point no 9 (performance against team peer only) becomes redundant. again thanks to ananth for another great analysis. [[ Asif Subtle differences. Team load is Balls bowled (a Work ethic measure) and Wickets captured (a Result achieved measure). Peer comparisons are based on Bowling average (a performance measure) Ananth: ]]

  • Vasu on August 10, 2010, 12:09 GMT

    Hi Anantha, I’m a regular reader of your columns & admire your work. Like you said, whatever route you take & parameter weightages you use, Murali would be on top, and the top 10 would be more or less the same. But I was disappointed with your shortlist. Of course it is your analysis & your rules, but the rule of only 3 per country spoilt the piece for me. Holding is far better than any bowler India has ever produced – Kumble included. Kumble and Kapil are amongst my favorite cricketers of all time, but Kapil does not deserve to be in top 15 bowlers of all time. Top 5 all rounder’ of all time yes, but not a top 15 bowler. So if not for the rule, such worthy bowlers as Steyn, Bedser, Holding, O’Reilly, just to name a few, could have figured But great job anyway. Its a reminder to an India that aspires to maintain its #1 ranking, that without any top bowlers, they wont stay there for long. Similarly, Sri Lanka is going to stay hard to win matches without Murali. Cheers! Vasu [[ Vasu I have given my comments on Kumble and Kapil earlier. While Holding might be a great fast bowler, he had three other greats competing with but also providing great pressure at the other end. Please look at the way Kapil and Kumble toliled year after year. And Kapil's manfully shouldering the pace bowling load on tough Indian pitches. Your last point is very valid. Even now Kumble and Kumble are at the end of the table. WHAT NEXT. What will india do. Or Sri Lanka do. Ananth: ]]

  • Ian Brooke on August 10, 2010, 12:05 GMT

    It shows Muralitharan skills that in the only two measaures where 2nd place is not in the 9s, he is leading. That being said, when evaluating with respect to placement, ranking from 1-15 for each measure he comes 5th because of his bad placing in the Bowling strike rate (Bpw), Away wicket bowling average and % of top order wickets captured, where he places 12th, 13th and 14th respectively. This measure of ranking shows Hadlee(57), McGrath(58) and Marshall(59) the top three showing how consistent they are across all rankings, with Murali 5th (64) This is my interpretation of the results, Murali dominating 2 areas with Hadlee, McGrath and Marshall consistent across all measures of performance [[ Ian I fully agree with your evaluation and it would be your prerogative to re-draw the lists based on the positions held. I initially thought of 1-15 (or at that time 1-10 since I started with 10 bowlers). However I think that is unfair in that the absolute diffrential will not be incorporated. It will be like saying Bradman gets 10, Graeme Pollock gets 9 and Headley 8 and so on. Also don't forget that all players benefit. Look at the cushion Waqar and Marshall get in Strike rate. I again agree that your suggestion is quite a good method. Ananth: ]]

  • sid on August 10, 2010, 11:51 GMT

    Ananth, one point: Quality of wickets and % of top order end up pointing to the same thing as you will have a higher batting average in the top order. This would skew the result in favor of the quicks (murali being the execption). [[ Sid Subtle difference. The Aussie or Indian top order would add to both while the Bangladeshi top order would only add to one. I agree that there is a bit of double counting. This would possibly benefit the spinners as and when they capture the middle order wickets. Ananth: ]]

  • Suhaib Ahmed on August 10, 2010, 11:34 GMT

    I wish there was a measure of the types of pitches the bowlers bowled on... may be average runs per batsman on that ground. The subcontinent bowlers should get some credit for bowling on the batting paradises of the region.

  • Jason on August 10, 2010, 11:29 GMT

    I like this analysis. It supports my standard arguments that McGrath, Hadlee and Imran are underrated, while Wasim Akram in particular is a bit overrated by the cricket public. Talented bowler for sure, but for every magical delivery he bowled there seemed to be some listless and inconsistent bowling and hence stats show him not quite as effective as lesser lauded bowlers.

    And I imagine I won't be the first one to ask, so lets gets the hypothetical question out of the way: How does Murali finish if Bang and Zim are not included? In this case he has benefited a bit from SLs low international standing, meaning that they got to play the weaker sides more often. McGrath and Warne by comparison barely got to face them, just based on scheduling abnormalities between boards. I'm not sure whether this has occurred in the past where the best sides were too 'busy' to play the minnows or new teams.

    Anyway, all that aside - good stuff! [[ Jason As I have mentioned in my article, if I take away Murali's wickets against Bang/Zim away, i have to take out quite a lot of other wickets also. Ananth: ]]

  • Faraz on August 10, 2010, 11:23 GMT

    It is a wonderful , wonderful analysis. Shows the importance of each bowler and his worth not only to his team but to the Cricketing fraternity of his/her own respective generation. A tremendous work I must say and well worth praising.

  • Simon on August 10, 2010, 10:54 GMT

    Good work but one comment. Kapil is a poor pick and is out of his depth in this company. That he changed the face of Indian fast bowling is neither here nor there. A very good bowler, not a great one. This is backed up by your results. Derek Underwood? Jim Laker? [[ Simon Only an Indian would understand the impact of Kapil Dev, poor though he may be in this exalted company. He, like Botham, completely spoiled his career figures by his misguided attempt to overhaul Hadlee's 431 wicket mark. If One takes away the last 15 tests or so his figures are much better. Laker would suffer a lot because of his sub-200 wickets. There is a case for Underwood to be considered. Ananth: ]]

  • Mukarram Bashir on August 10, 2010, 10:54 GMT

    Simply splendid. I was particularly pleased to see McGrath at his well deserved position; often overlooked by the Warne bandwagon. In fact, to me, three automatic bowling choices in a TEAM for all times, are Marshall, McGrath and Wasim, in the same order. Thank you for the wonderful treat this article is.

  • Rafay Iqbal on August 10, 2010, 10:47 GMT

    Interesting analysis. I suppose things like these will always be subjective but the measure seems fair enough. Would love to see an analysis of guys like Barnes, Grimmett and Lohmann who have got amazing stats which are way above their peers... [[ Rafay Would do the special pre-WW1 work and post, in a day or two. Ananth: ]]

  • Shani Lala on August 10, 2010, 10:30 GMT

    Superb Article! It really is a great work and all the mentioned factors are rightly considered. But one thing is ignored as Spin bowlers bowl almost twice or thrice the number of overs than fast bowlers most of the time. This might have the tendency to mislead some of the facts presented in the section "% of team load borne" and "Peer comparison of average to own team peer bowlers". DO you agree??? [[ Shani Let us not forget that when a spinner bowls 30 overs a day, HE HAS REALLY TAKEN THAT LOAD. Not easy and this should partly make up for the higher strike rates. Ananth: ]]

  • Taimur Khan on August 10, 2010, 10:21 GMT

    Good analysis overall. Some issues. First, Wasim & Waqar & also Marshall, Garner & Ambrose were together in the side which also included Walsh, Roberts, Croft & Clarke, so in comparisons with teammates' bowling stats, they suffered. Second,the performance in won matches stat is wonky since it is skewed in favor of winning teams. Comparison against contemporary bowlers is unfair for bowlers in 70s & 80s since we had many great bowlers operating at the time (8 out of your 15). Career wickets captured is not relevant. Wickets per match is more pertinent as a comparison for this group. In that case, Kapil & Kumble would be amongst the last (in the first category). Kapil & Kumble dont belong here. I appreciate the desire to represent all, but how can we leave out Holding or Roberts for Kapil (purely as a bowler) and Kumble just played too many tests. His away stats and strike rate are just not good enough. I wonder if you looked at Underwood & Gibbs against Kumble's stats. [[ Taimur First check out Murali's win index. Quite high for some one not in a winning team. Second give Kumble credit for getting over 600 wickets after being branded after a series as a non-spinning spinner. And credit Kapil, bowling his heart out on the flat Indian pitches. I don't think you really want 6 West Indians and 6 Australians in an analysis of this type. Underwood is one strong possibility. I will give his figures soon. Ananth: ]]

  • vignesh devaraj on August 10, 2010, 10:12 GMT

    I am surprised to see one big name missing “Indian Ambrose” Agarkar.He was also one of the most feared fast bowlers of all time.He along with Debashis mohanty formed a lethal combination.They always reminded me of legendary walsh and ambrose..

  • kamran khalid on August 10, 2010, 9:43 GMT

    Explains a lot. Murli for all his legal issues still was the best bowler by a mile. Which is to say i right. Interesting to see that Pakistani fast bowlers together actually cancel each other out in a lot of ways but as they were playing together hence pakistan was winning more in those days. Also explains the Kapil, Kumble and Headlee's lone ranger thing, wish they had a partner, would have been good for the team but not so good for them. Waqar and Wasim give the best example of hunting in pairs. [[ Kamran Great comment. I really appreciate the mature thinking in evaluating the Indian/Pakistani bowlers. Ananth: ]]

  • jaufer on August 10, 2010, 9:22 GMT

    It is almost a true reflection of the status of the bowler. One more factor could have considered the period these bowlers played. As we know, the rpo scored by these days is much higher than that was in good old days.

  • Nilotpal on August 10, 2010, 9:14 GMT

    This is one of the best stats analysis write-ups on this site ever. Take a bow, Anand. a v accurate and balanced call, in my view. Tough to argue against any of the parameters set/ notions based on.

  • Chinar Samant on August 10, 2010, 9:12 GMT

    Awesome list of bowlers! In the quality of wickets list, I expected the modern guys to come on top given that batting averages are generally higher nowadays. Unless I've not understood the methodology correctly. Also, you mentioned that 'Not so surprisingly Kapil Dev is last, indicating that the Indian wins were probably crafted by the spinners during Kapil's era'. Wouldn't the more pertinent reason for this be that India won very few matches during Kapil's time and thus his number of wickets in matches won is low? [[ Chinar The win index is not just wins but the bowler's share in the winning wickets. See Murali's numbers for this. Ananth: ]]

  • Grant on August 10, 2010, 9:04 GMT

    Interesting analysis. I would observe that giving all your criteria equal weighting might not bt entirely appropriate, but I realise that any attempt at such weighting would introduce a lot of bias and personal preference in an otherwise rigorous analysis. I do wonder if Botham had fallen under a bus in 1982 rather than suffer his long, overweight and injured decline whether he might have made the 15. [[ Grant Both Kapil and Botham prolonged their careers for different, eminently avoidable, reasons. Ananth: ]]

  • No featured comments at the moment.

  • Grant on August 10, 2010, 9:04 GMT

    Interesting analysis. I would observe that giving all your criteria equal weighting might not bt entirely appropriate, but I realise that any attempt at such weighting would introduce a lot of bias and personal preference in an otherwise rigorous analysis. I do wonder if Botham had fallen under a bus in 1982 rather than suffer his long, overweight and injured decline whether he might have made the 15. [[ Grant Both Kapil and Botham prolonged their careers for different, eminently avoidable, reasons. Ananth: ]]

  • Chinar Samant on August 10, 2010, 9:12 GMT

    Awesome list of bowlers! In the quality of wickets list, I expected the modern guys to come on top given that batting averages are generally higher nowadays. Unless I've not understood the methodology correctly. Also, you mentioned that 'Not so surprisingly Kapil Dev is last, indicating that the Indian wins were probably crafted by the spinners during Kapil's era'. Wouldn't the more pertinent reason for this be that India won very few matches during Kapil's time and thus his number of wickets in matches won is low? [[ Chinar The win index is not just wins but the bowler's share in the winning wickets. See Murali's numbers for this. Ananth: ]]

  • Nilotpal on August 10, 2010, 9:14 GMT

    This is one of the best stats analysis write-ups on this site ever. Take a bow, Anand. a v accurate and balanced call, in my view. Tough to argue against any of the parameters set/ notions based on.

  • jaufer on August 10, 2010, 9:22 GMT

    It is almost a true reflection of the status of the bowler. One more factor could have considered the period these bowlers played. As we know, the rpo scored by these days is much higher than that was in good old days.

  • kamran khalid on August 10, 2010, 9:43 GMT

    Explains a lot. Murli for all his legal issues still was the best bowler by a mile. Which is to say i right. Interesting to see that Pakistani fast bowlers together actually cancel each other out in a lot of ways but as they were playing together hence pakistan was winning more in those days. Also explains the Kapil, Kumble and Headlee's lone ranger thing, wish they had a partner, would have been good for the team but not so good for them. Waqar and Wasim give the best example of hunting in pairs. [[ Kamran Great comment. I really appreciate the mature thinking in evaluating the Indian/Pakistani bowlers. Ananth: ]]

  • vignesh devaraj on August 10, 2010, 10:12 GMT

    I am surprised to see one big name missing “Indian Ambrose” Agarkar.He was also one of the most feared fast bowlers of all time.He along with Debashis mohanty formed a lethal combination.They always reminded me of legendary walsh and ambrose..

  • Taimur Khan on August 10, 2010, 10:21 GMT

    Good analysis overall. Some issues. First, Wasim & Waqar & also Marshall, Garner & Ambrose were together in the side which also included Walsh, Roberts, Croft & Clarke, so in comparisons with teammates' bowling stats, they suffered. Second,the performance in won matches stat is wonky since it is skewed in favor of winning teams. Comparison against contemporary bowlers is unfair for bowlers in 70s & 80s since we had many great bowlers operating at the time (8 out of your 15). Career wickets captured is not relevant. Wickets per match is more pertinent as a comparison for this group. In that case, Kapil & Kumble would be amongst the last (in the first category). Kapil & Kumble dont belong here. I appreciate the desire to represent all, but how can we leave out Holding or Roberts for Kapil (purely as a bowler) and Kumble just played too many tests. His away stats and strike rate are just not good enough. I wonder if you looked at Underwood & Gibbs against Kumble's stats. [[ Taimur First check out Murali's win index. Quite high for some one not in a winning team. Second give Kumble credit for getting over 600 wickets after being branded after a series as a non-spinning spinner. And credit Kapil, bowling his heart out on the flat Indian pitches. I don't think you really want 6 West Indians and 6 Australians in an analysis of this type. Underwood is one strong possibility. I will give his figures soon. Ananth: ]]

  • Shani Lala on August 10, 2010, 10:30 GMT

    Superb Article! It really is a great work and all the mentioned factors are rightly considered. But one thing is ignored as Spin bowlers bowl almost twice or thrice the number of overs than fast bowlers most of the time. This might have the tendency to mislead some of the facts presented in the section "% of team load borne" and "Peer comparison of average to own team peer bowlers". DO you agree??? [[ Shani Let us not forget that when a spinner bowls 30 overs a day, HE HAS REALLY TAKEN THAT LOAD. Not easy and this should partly make up for the higher strike rates. Ananth: ]]

  • Rafay Iqbal on August 10, 2010, 10:47 GMT

    Interesting analysis. I suppose things like these will always be subjective but the measure seems fair enough. Would love to see an analysis of guys like Barnes, Grimmett and Lohmann who have got amazing stats which are way above their peers... [[ Rafay Would do the special pre-WW1 work and post, in a day or two. Ananth: ]]

  • Mukarram Bashir on August 10, 2010, 10:54 GMT

    Simply splendid. I was particularly pleased to see McGrath at his well deserved position; often overlooked by the Warne bandwagon. In fact, to me, three automatic bowling choices in a TEAM for all times, are Marshall, McGrath and Wasim, in the same order. Thank you for the wonderful treat this article is.