THE CORDON HOME

BLOGS ARCHIVES
SELECT BLOG
September 27, 2008

Allrounders

The world's best all-rounder

Anantha Narayanan
Garry Sobers hits out 1973, England v West Indies, Lord's
 © Getty Images
Enlarge

Possibly the best responses in the first 12 hours itself.

Based on these responses I have decided that my follow-up post will be to do a far more rigorous and in-depth analysis of only the very best 10 or so all-rounders, excluding the also-rans like Vettori/Vaas/Hooper/Shastri et al. Then I can be very strict and demanding in my parameters since I will be looking at the best. There is no need to worry about very low batting or bowling averages of these pretenders, upsetting the balance of algorithms.
Many thanks.

Who is a Test all-rounder? There prevails a peculiar idea of all-rounders. A bowler who can bat a bit (Abid Ali) or a batsmen who can turn his arm a bit (Sehwag) or a bowler who chances his way to a hundred (Agarkar), at various times have been dubbed as all-rounders.

That is a very low-level expectation of an all-rounder. Let us raise the bar substantially. An all-rounder should be capable of winning matches consistently with his batting or bowling. Since this is a subjective statement, let us lay down some rules to be used as the basis for our analysis.

He should have scored a minimum of 2000 Test runs at an average of 20.00 or above. The limit of 20.00 is necessary to exclude long-career bowlers such as Warne and Kumble getting into the All-rounders list. Much as I admire their batting skills I am not ready to accept them as all-rounders.

He should have taken a minimum of 100 Test wickets. There is no need to have a limit of average since the all-rounder with the worst bowling record among this lot, Carl Hooper, with a bowling average of 49.43 is still considered as a genuine all-rounder. If I incorporate a cut-off limit of 40.00 for bowling average, Ravi Shastri and Hooper go out.

The rationale behind these two cut-off numbers is that, on an average, it takes 25-30 Tests to score 2000 runs and take 100 wickets. So we are looking at players who have played these many Tests at the minimum. 21 players qualify under these criteria. Wally Hammond misses out based on this citeria. Jayasuriya just misses out by two wickets. Steve Waugh also misses out by a few wickets.

There is a piquant situation what with Vettori, Vaas and Akram vaulting over the bar meant for all-rounders. Well, we cannot question the numbers. Vettori has a higher batting average than Craig Spearman while Vaas and Akram have acceptable 23+ and 22+ batting averages.

How do we analyse all-rounder performances? Once we set the minimum criteria and select the players it becomes easy to classify them. This time I have anticipated readers' comments and got the analysis done under the following three classifications. Finally I have a composite Index determination process based on these three classifications.

1. Performance based
2. Longevity based
3. Individual match performances.

1. Career Performance based:

The simplest and a very effective method of evaluating player performances is by measuring their averages. The batting average has to be as high as possible

and the bowling average has to be as low as possible. So we subtract the bowling average from the batting average and arrive at, what we call, an All-rounder

Index 1. The higher this index is, the more effective the all-rounder is. Let us now see the complete table, on this criteria.

No  Player         LBt LBw Ctry BatAvg Bow  BowAvg  ARIdx1

1. Kallis J.H Saf 55.46 RFM 31.23 24.23 2. Sobers G.St.A ~ ~ Win 57.78 LM 34.04 23.75 3. Imran Khan Pak 37.69 RF 22.81 14.88 4. Miller K.R Aus 36.97 RF 22.98 14.00 5. Pollock S.M Saf 32.32 RFM 23.12 9.20 6. Goddard T.L ~ ~ Saf 34.47 LFM 26.23 8.24 7. Greig A.W Eng 40.44 RFM 32.21 8.23 8. Botham I.T Eng 33.55 RFM 28.40 5.15 9. Hadlee R.J ~ Nzl 27.17 RFM 22.30 4.87 10. Cairns C.L Nzl 33.54 RFM 29.40 4.13 11. Rhodes W ~ Eng 30.19 LSP 26.97 3.23 12. Kapil Dev N Ind 31.05 RFM 29.65 1.41 13. Bailey T.E Eng 29.74 RFM 29.21 0.53 14. Flintoff A Eng 32.35 RFM 32.21 0.14 15. Mankad M.H ~ Ind 31.48 LSP 32.32 -0.84 16. Wasim Akram ~ ~ Pak 22.64 LFM 23.62 -0.98 17. Benaud R Aus 24.46 RLB 27.03 -2.58 18. Shastri R.J ~ Ind 35.79 LSP 40.96 -5.17 19. Vaas WPUJC ~ ~ Slk 23.97 LFM 29.31 -5.35 20. Vettori D.L ~ ~ Nzl 26.65 LSP 34.44 -7.79 21. Hooper C.L Win 36.47 ROB 49.43 -12.96 Note: ~ indicates Left handed batsman/bowler.

No real surprises here. Kallis is one of the most under-rated players ever. He comes in, does his job in a quite manner and walks away. However he has an

outstanding batting average of 55+ (dropped recently because of his disastrous series in England) and a very acceptable bowling average of 31+. Thus the

difference is 24+. Gary Sobers is the supreme all-rounder and his index value is around 23. Imran is in third position, by virtue of his Batting Average of

37.69 (8 more than Srikkanth!) and a very low Bowling Average of 22.81 (7 fewer than Brett Lee!). Imran is followed by the mercurial Miller and the

under-rated Shaun Pollock.

At the other end of the table, the spinning all-rounders occupy the low positions, led by Hooper who has a difference of nearly -13.

As an alternative, we could divide the Batting Average by the Batting Average and arrive at, what we call, an All-rounder Index. The higher this index is,

the more effective the all-rounder is. This table is almost similar to the first table and is not shown.

2. Longevity based:

As a second alternative, we normalise all performances to a common base, say, Runs. Using a commonly accepted norm of a wicket as equivalent to 20 runs, we

derive a table of Total Runs scored + Runs derived. Then we rank these players.

No  Player         LBt LBw Ctry   Runs  Bow  Wkts ARIdx2

1. Kallis J.H Saf 9761 RFM 240 14561 2. Kapil Dev N Ind 5248 RFM 434 13928 3. Botham I.T Eng 5200 RFM 383 12860 4. Sobers G.St.A ~ ~ Win 8032 LM 235 12732 5. Pollock S.M Saf 3781 RFM 421 12201 6. Hadlee R.J ~ Nzl 3124 RFM 431 11744 7. Wasim Akram ~ ~ Pak 2898 LFM 414 11178 8. Imran Khan Pak 3807 RF 362 11047 9. Vaas WPUJC ~ ~ Slk 2996 LFM 348 9956 10. Hooper C.L Win 5762 ROB 114 8042 11. Vettori D.L ~ ~ Nzl 2745 LSP 257 7885 12. Cairns C.L Nzl 3320 RFM 218 7680 13. Flintoff A Eng 3494 RFM 206 7614 14. Benaud R Aus 2201 RLB 248 7161 15. Shastri R.J ~ Ind 3830 LSP 151 6850 16. Greig A.W Eng 3599 RFM 141 6419 17. Miller K.R Aus 2958 RF 170 6358 18. Mankad M.H ~ Ind 2109 LSP 162 5349 19. Goddard T.L ~ ~ Saf 2516 LFM 123 4976 20. Bailey T.E Eng 2290 RFM 132 4930 21. Rhodes W ~ Eng 2325 LSP 127 4865 Note: ~ indicates Left handed batsman/bowler.

This is a tribute to the longevity of the all-rounders. Kallis is again in the top position, follwed by Kapil Dev, Botham, Steve Waugh, Sobers and Pollock.

The olden day all-rounders who have not played in too many matches are at the lower end of the table.

3. Individual match performances:

Here we take a simple yardstick. Since we are analysing individual match all-round performances we have to look at a measure which brings out the all-round

ability of the player. We need both runs and wickets. One cannot compensate the other. I have defined an 'A' level all-round performance as one in which a

player scores a minimum of 100 runs and captures 5 wickets and a 'B' level all-round performance as one in which the player scores 75-99 runs and captures 4

wickets. I understand that match conditions, pitch conditions, quality of opposition, match result et al are relevant factors. However that will complicate

the issue and we are only looking at all-round performances here. So I will limit myself to the runs scored and wickets captured.

No  Player        Ctry   A-Perf   B-Perf
Tests    Tests

1. Sobers G.St.A Win 7 8 2. Botham I.T Eng 7 4 3. Greig A.W Eng 4 4 4. Cairns C.L Nzl 2 9 5. Kallis J.H Saf 2 6 6. Miller K.R Aus 2 5 7. Mankad M.H Ind 2 4 8. Imran Khan Pak 2 3 9. Hooper C.L Win 2 2 10. Flintoff A Eng 1 5 11. Pollock S.M Saf 1 4 12. Kapil Dev N Ind 1 3 13. Goddard T.L Saf 1 2 14. Benaud R Aus 1 2 15. Vettori D.L Nzl 1 2 16. Wasim Akram Pak 1 1 17. Vaas WPUJC Slk 1 1 18. Hadlee R.J Nzl 0 7 19. Shastri R.J Ind 0 4 20. Rhodes W Eng 0 1 21. Bailey T.E Eng 0 0 Note: A-Perf: 100 or more runs and 5 or more wkts in same test. B-Perf: 75-99 runs and 4 wkts in same test.

Sobers stands supreme with 15 stand-out performances. Botham is equally good. Greig shows that he is a much under-rated all-rounder as does Chris Cairns.

Kallis seems to have a more even distribution of run scoring and wicket capturing.

4. Final All Rounder Rating Index:

Finally, as we normally do always, a composite calculation involving the five key factors with suitable weightings. The formula is explained below.

				  Weightage    Computation methodology

Batting average 30.00 The maximum for an average of 60.00 and above Runs scored 15.00 The maximum for a Runs scored value of 10000 and above Bowling average 30.00 The maximum for an average of 15.00 and below Wickets taken 15.00 The maximum for a Wickets taken value of 500 and above. Test perfs 10.00 One point per 'A' performance + 0.4 point for 'B' performance.

Total 100.00

It can be seen that the two longevity measures carry only 30% of the total and the other 70% are allotted to performance nased measures. The final table is

given below.

No  Player        LBt LBw Ctry  Runs BatAvg  Wkts BowAvg   ARIdx

1. Sobers G.St.A ~ ~ Win 8032 57.78 235 34.04 69.15 2. Kallis J.H Saf 9761 55.46 240 31.23 67.74 3. Botham I.T Eng 5200 33.55 383 28.40 61.27 4. Imran Khan Pak 3807 37.69 362 22.81 60.81 5. Pollock S.M Saf 3781 32.32 421 23.12 58.95 6. Hadlee R.J ~ Nzl 3124 27.17 431 22.30 56.71 7. Miller K.R Aus 2958 36.97 170 22.98 54.05 8. Kapil Dev N Ind 5248 31.05 434 29.65 53.98 9. Wasim Akram ~ ~ Pak 2898 22.64 414 23.62 50.88 10. Cairns C.L Nzl 3320 33.54 218 29.40 49.49 11. Greig A.W Eng 3599 40.44 141 32.21 48.24 12. Goddard T.L ~ ~ Saf 2516 34.47 123 26.23 45.27 13. Vaas WPUJC ~ ~ Slk 2996 23.97 348 29.31 44.01 14. Flintoff A Eng 3494 32.35 206 32.21 43.39 15. Benaud R Aus 2201 24.46 248 27.03 42.74 16. Rhodes W ~ Eng 2325 30.19 127 26.97 40.83 17. Mankad M.H ~ Ind 2109 31.48 162 32.32 40.04 18. Bailey T.E Eng 2290 29.74 132 29.21 38.06 19. Vettori D.L ~ ~ Nzl 2745 26.65 257 34.44 37.52 20. Shastri R.J ~ Ind 3830 35.79 151 40.96 33.81 21. Hooper C.L Win 5762 36.47 114 49.43 28.67 Note: ~ indicates Left handed batsman and bowler.

Let us come to a conclusion.

The top 5 all-rounders of all time are Sobers, Kallis, Botham, Imran Khan and Pollock. Hadlee and Kapil Dev run these 5 close. No surprises except that

Kallis is so close to Sobers at the top.

Sobers stands supreme at the top, helped by 8032 runs at an average of 57.78, 235 wickets at an average of 34.04 and 15 outstanding performances in

Tests. Kallis would have to perform at this high level for couple of more years to overtake Sobers.

Kallis has come first in two of these measures. He has been a vastly under-rated all-rounder. However one cannot question his credentials - 9761 runs

at an average of 55.46 and 240 wickets at an average of 31.23. Independently these figures would be considered great as a batsman and acceptable as a bowler.

Botham is deservedly in the third position helped by his match-winning performances. Imran Khan's reasonably high batting average and very low

bowling average have propelled him to the fourth position. Pollock is another all-rounder not normally given his due. He has got a batting average

higher than Greame Hick and a bowling average 2.5 below Shoaib Akhtar.

At the other end, the three spinning all-rounders are there. All have barely acceptable batting and bowling averages. Vettori's high bowling average lets him

down.

The Australians coined a new definition of an all-rounder, viz., the wicketkeeper. I wanted to do an analysis of the wicketkeepers. However I have decided

to do a separate piece on that for two reasons. The first is that this article has become quite long, but more importantly, the players who have the toughest

job in cricket deserve their own special article. Hence that will be covered in a later article.

PS: Anticipating readers' requests and in order to have a more complete coverage I have lowered the bar to 1500 runs, 20.00 Batting Avge and 75 wickets and

presented a concluding report. Initially I thought of lowering the bar only for pre-1970 players but decided to enlarge the scope. So we now have Steve

Waugh, Walter Hammond, Mushtaq, Faulkner, Jayasuriya et al in the Top-20. 17 new all-rounders have come in. I re-iterate that my suggested cut-off is still

the earlier one and this table has been presented only for information.

No  Player         LBt LBw Ctry  Runs BatAvg Bow Wkts BowAvg ARIdx

1. Sobers G.St.A ~ ~ Win 8032 57.78 LM 235 34.04 69.15 2. Kallis J.H Saf 9761 55.46 RFM 240 31.23 67.74 3. Botham I.T Eng 5200 33.55 RFM 383 28.40 61.27 4. Imran Khan Pak 3807 37.69 RF 362 22.81 60.81 5. Pollock S.M Saf 3781 32.32 RFM 421 23.12 58.95 6. Hadlee R.J ~ Nzl 3124 27.17 RFM 431 22.30 56.71 7. Miller K.R Aus 2958 36.97 RF 170 22.98 54.05 8. Kapil Dev N Ind 5248 31.05 RFM 434 29.65 53.98 9. Waugh S.R Aus 10927 51.06 RFM 92 37.45 53.83 10. Wasim Akram ~ ~ Pak 2898 22.64 LFM 414 23.62 50.88 11. Hammond W.R Eng 7249 58.46 RFM 83 37.81 50.18 12. Cairns C.L Nzl 3320 33.54 RFM 218 29.40 49.49 13. Greig A.W Eng 3599 40.44 RFM 141 32.21 48.24 14. Mushtaq Mohammad Pak 3643 39.17 RLB 79 29.23 46.59 15. Faulkner G.A Saf 1754 40.79 RLB 82 26.59 46.30 16. Goddard T.L ~ ~ Saf 2516 34.47 LFM 123 26.23 45.27 17. Jayasuriya S.T ~ ~ Slk 6973 40.07 LSP 98 34.35 44.09 18. Vaas WPUJC ~ ~ Slk 2996 23.97 LFM 348 29.31 44.01 19. Flintoff A Eng 3494 32.35 RFM 206 32.21 43.39 20. Benaud R Aus 2201 24.46 RLB 248 27.03 42.74 21. Lindwall R.R Aus 1502 21.15 RF 228 23.03 42.45 22. Noble M.A Aus 1997 30.26 ROB 121 25.00 41.76 23. Rhodes W ~ Eng 2325 30.19 LSP 127 26.97 40.83 24. Armstrong W.W Aus 2863 38.69 RLB 87 33.60 40.25 25. Mankad M.H ~ Ind 2109 31.48 LSP 162 32.32 40.04 26. Reid J.R Nzl 3428 33.28 RFM 85 33.35 38.58 27. Streak H.H Zim 1990 22.36 RFM 216 28.12 38.32 28. Bailey T.E Eng 2290 29.74 RFM 132 29.21 38.06 29. Vettori D.L ~ ~ Nzl 2745 26.65 LSP 257 34.44 37.52 30. McMillan B.M Saf 1968 39.36 RFM 75 33.83 37.26 31. Woolley F.E ~ ~ Eng 3283 36.08 LSP 83 33.92 36.54 32. Shastri R.J ~ Ind 3830 35.79 LSP 151 40.96 33.81 33. Illingworth R Eng 1836 23.24 ROB 122 31.20 31.83 34. Prabhakar M Ind 1600 32.65 RFM 96 37.30 30.51 35. Abdul Razzaq Pak 1946 28.62 RFM 100 36.93 30.50 36. Klusener L ~ Saf 1906 32.86 RFM 80 37.91 28.78 37. Hooper C.L Win 5762 36.47 ROB 114 49.43 28.67 38. Emburey J.E Eng 1713 22.54 ROB 147 38.41 25.24

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

RSS Feeds: Anantha Narayanan

© ESPN Sports Media Ltd.

Posted by Harsh Thakor on (July 1, 2009, 12:27 GMT)

I don't mean any discredit to Kallis,the best allrounder today undoubtedly,perhaps statistically the best allrounder of all and the best allrounder batsman after Sobers.However he has not been as great a match-winner as Sobers,Imran or Botham.Sobers dominated great bolwing to a far greater extent,on more difficult pitches.He was perhaps second only to Bradman as a batsman.It is significant that Ian Botham(Remember the 1980 Jubillee test 13 wickets and a century and the 1981 Ashes) has performed the greatset ever performance in a single test match and single test series.Had Ian Botham mantained his form of his early career he could well have given Sobers a run for his money.Imran Khan, I rate superior to Botham as he performed more consistently and his captaincy took Pakistan to the top .He also outperformed Botham in both the series staged on English soil in 1982 and 1987.He also performed brilliantly in West Indies.His performancse in 1982-83 and 1987-88 are unforgettable.

Posted by Harsh Thakor on (July 1, 2009, 12:08 GMT)

There can be no doubt of one fact that Gary Sobers was the greatest allrounder of all time.True his bowling figures were not great but he bowled with other great bowlers.Noone has changed the complexion of a game to such an extent with bat and ball.(Remember his performance in England in 1966 and at Kingston against England in 1968-69).Even if Kallis matches Gary statistically I don't think morally he ranks in the same class.Imran Khan to me is Sobers's closest challenger followed by Ian Botham,who at his peak was the best match-winner.Kapil Dev had the best natural ability, Richard Hadlee was the best bowler and Imran Khan,was the greatest skipper of the allrounders. Gary Sobers was an immortal to the game-the equivalent of a prophet sent to play the game.It would have been fascinating to have had Mike Proctor playing test Cricket.

Posted by patrick cozier on (January 31, 2009, 16:41 GMT)

when using the averages factoring approach, one needs to consider that there is a natural differential between the averages of fast bowlers and those of slow bowlers, a differentail of about 4-5. This therefore needs to be adjusted for in comparisons. For bowlers like Sobers who did both extensively you could use a 50/50 miix, while for Rhodes you need to make full adjustment

Posted by kasi on (December 26, 2008, 5:00 GMT)

yes fantastic analysis and i want to share my congratulation for the efforts made on this

Posted by Sancheeban on (December 6, 2008, 23:16 GMT)

Nice one! J.Kallis should be rated as No.1 in all-rounder list. He is the one who opened the bowling with S.Pollock in few matches. Not so many No.3 batsmen opened the bowling for their country. But, if u rated all-rounder list for an ODI, Kallis will be in No.1 spot!!!

Posted by Kunal Sen on (November 2, 2008, 17:48 GMT)

@ Ananth

Brilliant, stunning article- I think the factors used by you are very fair- though opinions on weightages could vary

Just wanted to understand your reason for taking 20 runs as a proxy value for runs earned for every wicket taken. Don't know if this is accurate. I feel this is too low and arbitrary and consequently, favors the batting allrounders (ones whose primary skills lay in batting than bowling)- hence Sobers and Kallis at top perhaps What's the empirical Average Runs per Wicket (all time batting average in the game of cricket) throughout history? Is it 20? If not, why should it not be used instead of 20?

My top 3- Imran Khan, Ian Botham, Shaun Pollock, Andrew Flintoff, Kapil Dev

Posted by cozico on (November 2, 2008, 16:51 GMT)

This ia a good piece. I myself divide batting by bowling using the same premise. There are however two refinements that I would respectfully suggest, 1. slow bowlers tend to have an average of about 4.00 -5.00 higher than corresponding fast bowlers, therefore one has to adjust slow bowlers averages to establish an equivalency with their fast counterparts. In the case of Sobers one should assume 1/2 fast and 1/2 slow and adjust bowling average by between 2.00-2.5 points. 2. In measuring longevity we should look more at duration of career rather than amount of matches. Reason being that it is biased in favour of recent players because they play more often. It is important in any such measurement to remove as much subjectivity as possible and try to utilise objective criteria as much as possible. Good work.

Posted by bvtesh on (November 1, 2008, 22:38 GMT)

two comments: All stats need to be broken down 2 ways: 1. pre-helmet vs. current, and 2. New test playing nations in mix. During Sobers' time, there were no new entrants (Pakistan was technically a regular test playing nation split away from India). Helmets have increased batting averages by at least 15 (Mukul Kesavan once an analysis of this) - once factored, Sobers is streets ahead of everyone else. Next about Imran Khan - sheer career averages do not tell the whole story - I would split his career into pre-1981 vs post-1981. In the latter, consisting of over 70 tests, he averaged over 55 with the bat and under 22 with the ball. Sobers and Khan are the greatest and inspirational leaders. Others were either dominant either in bowling or batting but rarely both (Botham pre-1981 was tracking better than all - it then dropped dramatically)

Posted by Christoph on (October 30, 2008, 4:17 GMT)

In cricket the numbers don't lie. Kallis batted at 3 or 4 against statistically some of the best bowlers ever. He then took an old ball after Donald, Pollock and Ntini invariably took 20 overs of shine off, and in many cases did not even bowl due to the strength of the SA bowling attack. An analysis of his wickets will find that they are predominantly top and middle order batsmen. The ball was never tossed to him to finish off the easy tail wickets. There is also a decidedly different psyche involved in carrying the innings from 4, than coming in at 6 or 7 with the best bowlers off and the ball old. Talk is cheap in the world of cricket, but runs and wickets buy the whiskey. If Kallis gets 10,000 and 300 history might judge him as the best cricketer (and not just the best all-rounder) of all time.

Posted by Tayyab on (October 25, 2008, 11:08 GMT)

While commenting about A and B performances, you are more inclined towards a batting performance. I think you should set 50 runs and 6 wickets or 75 runs and 5 wickets for class A and 50 runs and 5 wickets or 75 runs and 4 wickets for class B performance. Scoring 50 or 75 runs does the trick for an all rounder.

Comments have now been closed for this article

ABOUT THE AUTHOR

Anantha Narayanan
Anantha spent the first half of his four-decade working career with corporates like IBM, Shaw Wallace, NCR, Sime Darby and the Spinneys group in IT-related positions. In the second half, he has worked on cricket simulation, ratings, data mining, analysis and writing, amongst other things. He was the creator of the Wisden 100 lists, released in 2001. He has written for ESPNcricinfo and CastrolCricket, and worked extensively with Maruti Motors, Idea Cellular and Castrol on their performance ratings-related systems. He is an armchair connoisseur of most sports. His other passion is tennis, and he thinks Roger Federer is the greatest sportsman to have walked on earth.

All articles by this writer