October 16, 2012

Top bowling pairs at work: a fresh look

A detailed analysis of various performance aspects of the top bowling pairs in Tests
155

Wasim Akram and Waqar Younis captured 15 or more wickets in a match on ten occasions
Wasim Akram and Waqar Younis captured 15 or more wickets in a match on ten occasions © Getty Images

A few years back I had taken an across-the-years look at the Test bowling pairs. I had analysed all the pairs and came out with a series of tables. That article merely set the stage. I have taken the next significant step in this article. Using that type of analysis as the base, I have selected 10 top bowling pairs and analysed these individually and as a pair. I am confident that the selected pairs would include most of the top bowlers though there may be other top bowlers who operated alone or in combination with these pairs. This has been a very difficult article for me to compile and it would be equally difficult for the readers to go through and grasp all details. Hence I request the readers to kindly take time. Do not rush.

How did I select these pairs? Using the re-created tables, barring two exceptions, I have selected the top pair for each of the top 8 countries based on number of wickets captured. In all cases it is required that operating as a pair, each bowler must have taken at least 100 wickets. This leaves us with the following pairs, in order of total wickets captured:

Warne & McGrath            1001 wkts
Muralitharan & Vaas         895
Walsh & Ambrose             762
Wasim Akram & Waqar Younis  559
Willis & Botham             476
Donald & Pollock            397
Chandrasekhar & Bedi        368
Hadlee & B Cairns           337.

I have used my discretion only in two cases: South Africa and India. Pollock and Kallis have captured more wickets than Donald & Pollock. However I felt that I needed a real top quality bowling pair and have opted for the latter combination. Pollock & Ntini was another possibility. But I have gone with my heart in this selection as well as that of the Indian pair.

As far as India is concerned, I have not plumped for Kumble & Harbhajan against Chandrasekhar and Bedi even though the former pair has captured well over 100 wickets more. This is both nostalgia and a recognition, that these bowlers turned India into a force for the first time, albeit only at home. I am sure no reader would have any serious objections to the selection of this wonderful duo. My apologies to Kumble & Harbhajan.

Now for the two remaining pairs.

I decided that I would select an additional pair each from the two most successful bowling countries in history, viz., West Indies and Australia. For West Indies, I wanted a pair from the eighties even though Sobers & Gibbs are second, in terms of wickets. Ignoring repeats, the next best was Holding & Garner. But I bit the bullet and went for Garner & Marshall since I cannot think of any bowler analysis without Marshall. Anyhow Garner & Marshall have captured 322 wickets between them as against Holding & Garner, with 331.

Australia presented a few problems. Combinations involving Warne or McGrath were ruled out automatically. That presented me with two pairs: Benaud & Davidson with 346 and Miller & Lindwall with 345. It was an easy decision to select Miller & Lindwall, the pair from late-40s. That completes the top-10 bowling pairs.

Lillee, Thomson, Grimmett, O'Reilly et al did not meet the 100-wkts criterion. There was a case for Kumble & Harbhajan or Prasanna & Bedi or Pollock & Ntini, but the reasons have already been given. Imran & Qadir had 332 wickets but were way behind the great W-pair. Holding and Davidson just missed out.

I have considered each pair separately and presented a huge collection of data elements. It is going to be impossible to comment on all. I will only make comments on the special characteristics of each pair. The readers should download the huge table and do their own sub-analysis. I am sure there would be a comment or two that all these are available in Cricinfo and there is nothing new. That is fine.

Whoever made the debut earlier is named as the first bowler in each pair. All Tests between that debut and the retirement Test of latter constitute the overall range for various analyses. For the pair level analyses the range is strictly between the first and last Tests the pair operated together and only those in which they played together. Each pair has one graph and an exhaustive table. I have also distributed the comments across the pairs.

I have introduced two new concepts here. It is better that I explain these two now before moving on to the Pair details. Like the Form-Dip in the Lara article, these are the starting points and can be refined as we go on. Comments on these two will be very valuable to will fine-tune the ideas.

The first one is an analysis of the Pair-successes. I have considered only those innings in which the team captured 5 or more wickets. I consider that the pair has succeeded if they capture 50% or more of the team wickets. Examples are 3+ out of 5 & 6, 4+ out of 7 & 8 and 5+ out of 9 & 10. The idea is that since these were the premier pairs in their respective teams, they are expected to contribute at least 50% of the team wickets. Fairly subjective I agree, but there are some interesting results. The bar is set quite high but that is the way it should be.

The second is how the individuals performed within these pairs. For this analysis, I have considered only those innings in which the pair captured 5 or more wickets. The dominant bowler within the pair is identified as the one who leads 4-1 (and obviously 5-0) when 5 wickets are captured by the pair. Similarly, in case of 6 wickets captured by the pair, the dominant bowler is the one with 4-2 split (and the wider gaps).

This can be extended to 5-2 (and wider) for 7, 6-2 (and wider) for 8, 6-3 (and wider) for 9 and 7-3 (and wider) for 10 wickets. Again early days. But the idea is worth exploring.

The % of top-order wickets (1 to 6) is very close for all pairs and oscillates on either side of 60%. It is amazing that the % of top order wickets (1-6) is around 60% which is 6/10. I expected this figure to be around the 65% mark. Let me finish this discussion here itself. Only two pairs have this % figure around 66%: Donald & Pollock and Miller & Lindwall. However the big surprise is that Wasim Akram and Waqar Younis have a figure of 57.2%, the lowest amongst these pairs. Seems quite intriguing.

Similarly I will complete the discussion on the team against which the pairs captured their highest number of wickets at this point. Barring two pairs, the other eight pairs have captured their highest number of Tests against England, indicating the frequency of Tests played against England. Warne & McGrath lead with 267 wickets against England. The two other pairs are Akram and Younis, who reserved their best against West Indies and Willis & Botham, who, for some strange reason I cannot comprehend, failed to capture a single wicket against England, reserving their best for Australia.

The graphs, 10 in all, cannot be too big and show all information. Hence I have concentrated on showing relative average values and wins, especially away from home.

Warne & McGrath

P01. Warne 513 @ 24.88 & McGrath 488 @ 21.38 Total: 1001 @ 23.17
-------------------------------------------------------------------
Warne       : 1181(1992)-1826(2007) 513 @ 24.88 (195 @ 26.83) 92.7%
McGrath     : 1235(1993)-1826(2007) 488 @ 21.38 ( 75 @ 23.35) 91.6%
Pair Career:  1235(1993) to 1826(2007)
All matches:          177      Wins: 111 (62.7 %)
Pair matches:         104      Wins:  71 (68.3 %)
Pair Home matches:     54      Wins:  43 (79.6 %)
Pair Away matches:     50      Wins:  28 (56.0 %)
Non-Pair matches:      73      Wins:  40 (54.8 %)
Non-Pair Home matches: 43      Wins:  25 (58.1 %)
Non-Pair Away matches: 30      Wins:  15 (50.0 %)
Matches on their own  Warne: 41  McGrath: 20
Pair   Home wkts:  504 @ 24.40  Away wkts:  497 @ 21.93
Others Home wkts:  472 @ 27.24  Away wkts:  340 @ 32.29
Team Wkts:        1813 @ 25.94  WpT: 9.62  Pair %: 55.2
TeamPeerWkts:      812 @ 29.35
AllPeerWkts:     14765 @ 31.22
Top Order Wkts:    597 (59.6%)  Low Order Wkts:  221 (22.1%)
Top Order Home Wkts: 308 (61.1%)  Top Order Away Wkts: 289 (58.1%)
Highest vs England:  263 @ 21.97
Team wkts-5+: 183  Pair-Successes (50% of Team or more): 118 (64.5%)
Pair wkts-5+: 118 Shared: 74 Warne-dominant: 23 McGrath-dominant: 21

Most figures are self-explanatory. Let me explain Warne's "513 @ 24.88 (195 @ 26.83) 92.7%" further. Warne captured 513 wickets at 24.88 when playing with McGrath and 195 wkts @ 26.83 without him. 92.7%, i.e. a figure below 100%, indicates that Warne's average improves when he plays with McGrath.

The pair-successes % stands at 64. A perusal of the remaining pairs shows that there are other pairs with better figures. That is understandable since those pairs might have constituted over 75% of the bowling strength of their respective teams, which is not true of Warne & McGrath. They also seem to have been fairly similar in the bowler-dominance numbers.

As expected this pair has highs in the overall win % and wickets per Test. Both of them have performed similarly, at just over 90%, when operating individually, indicating the good support available elsewhere. It is also surprising, but not unexpected, that the pair has performed better away than at home. They have also out-performed their own team-mates by well over 50%, away from Australia.

It is indeed very surprising that in their illustrious career Warne & McGrath have never captured all 10 wickets in an innings. Their best innings performance has been Test # 1590 against South Africa at Wanderers during 2002 when they captured 9 for 65 (4 for 44 & 5 for 21). Their best match performance has been 18 wickets (11 and 7) in Test # 1558 against England at Oval during 2001.

Conundrum of Test # 1313. I know there is a story behind it. There are a few instances of Miller not bowling a single ball in a Test when playing with Lindwall and that is understandable. But how does one explain the Hobart Test against Pakistan during 1995. In a match in which Australia bowled 150 overs, Warne did not bowl a single ball. Mark Waugh bowled 20 overs. Is it possible that Warne was injured since he did not bat in the second innings?

Summary of bowling career of Shane Warne and Glenn McGrath
© Anantha Narayanan

Muralitharan & Vaas

P02. Muralitharan 586 @ 22.05 & Vaas 309 @ 28.01 Total: 895 @ 24.11
-------------------------------------------------------------------
Muralitharan: 1195(1992)-1964(2010) 586 @ 22.05 (209 @ 24.40) 90.4%
Vaas        : 1267(1994)-1927(2009) 309 @ 28.01 ( 46 @ 40.11) 69.8%
Pair Career:  1270(1994) to 1909(2009)
All matches:          155      Wins:  59 (38.1 %)
Pair matches:          94      Wins:  41 (43.6 %)
Pair Home matches:     50      Wins:  27 (54.0 %)
Pair Away matches:     44      Wins:  14 (31.8 %)
Non-Pair matches:      59      Wins:  18 (30.5 %)
Non-Pair Home matches: 25      Wins:   5 (20.0 %)
Non-Pair Away matches: 34      Wins:  13 (38.2 %)
Matches on their own  Muralitharan: 36  Vaas: 16
Pair   Home wkts:  534 @ 20.97  Away wkts:  361 @ 28.76
Others Home wkts:  317 @ 35.86  Away wkts:  226 @ 44.92
Team Wkts:        1438 @ 29.97  WpT: 9.52  Pair %: 62.2
TeamPeerWkts:      543 @ 39.63
AllPeerWkts:     16239 @ 31.40
Top Order Wkts:    544 (60.8%)  Low Order Wkts:  187 (20.9%)
Top Order Home Wkts: 310 (58.1%)  Top Order Away Wkts: 234 (64.8%)
Highest vs England:  139 @ 23.22
Team wkts-5+: 145  Pair-Successes (50% of Team or more): 115 (79.3%)
Pair wkts-5+: 116 Shared: 69 Muralitharan-dominant: 40 Vaas-dominant: 7

As expected, this is a pair with wide differences, especially in bowling averages. Muralitharan has dominated their successful innings in a big manner: 40-7. But as a pair they have succeeded a huge 79% of the times, the highest in this analysis.

Readers can see how much they have out-performed their own team bowlers in almost every measure. The bowling averages show a 60% difference. The win % drops substantially. As a pair, they captured 62% of the team wickets: the highest in this analysis. But here the story also is also about how dependent was Vaas on Muralitharan. With Murali, he averaged 28 and without Murali the average went up to 40. Murali himself seems to have gone his own merry way, with or without Vaas, with averages of 22 and 24.

Murali & Vaas have captured all 10 wickets in the innings no fewer than 4 times. Their best innings performance has been Test # 1755 against West Indies at SSC during 2005 when they captured 10 for 51 (6 for 36 & 4 for 15). Their best match performance has been 17 wickets (11 and 6) in Test # 1567 against West Indies at Gale during 2001. They also captured 17 wickets (3 and 14) in Test # 1572 against West Indies at SSC during 2001. Both Tests were played during the series with the unlikely combination of 3-0 and 688 losing runs. The series is also a special one in a way: almost the only series in which Vaas out-performed Muralitharan.

Summary of bowling career of Muttiah Muralitharan and Chaminda Vaas
© Anantha Narayanan

Walsh & Ambrose

P03. Walsh 373 @ 24.29 & Ambrose 389 @ 21.12 Total: 762 @ 22.67
-------------------------------------------------------------------
Walsh       : 0997(1984)-1544(2001) 373 @ 24.29 (146 @ 24.82) 97.9%
Ambrose     : 1095(1988)-1509(2000) 389 @ 21.12 ( 16 @ 17.88) 118.1%
Pair Career:  1095(1988) to 1509(2000)
All matches:          142      Wins:  59 (41.5 %)
Pair matches:          95      Wins:  42 (44.2 %)
Pair Home matches:     50      Wins:  24 (48.0 %)
Pair Away matches:     45      Wins:  18 (40.0 %)
Non-Pair matches:      47      Wins:  17 (36.2 %)
Non-Pair Home matches:  8      Wins:   3 (37.5 %)
Non-Pair Away matches: 39      Wins:  14 (35.9 %)
Matches on their own  Walsh: 37  Ambrose: 3
Pair   Home wkts:  379 @ 23.06  Away wkts:  383 @ 22.29
Others Home wkts:  400 @ 31.41  Away wkts:  329 @ 32.53
Team Wkts:        1491 @ 27.19  WpT: 8.02  Pair %: 51.1
TeamPeerWkts:      729 @ 31.91
AllPeerWkts:     10257 @ 30.32
Highest vs England:  287 @ 21.92
Top Order Wkts:    474 (62.2%)  Low Order Wkts:  153 (20.1%)
Top Order Home Wkts: 233 (61.5%)  Top Order Away Wkts: 241 (62.9%)
Team wkts-5+: 147  Pair-Successes (50% of Team or more): 87 (59.2%)
Pair wkts-5+: 90 Shared: 56 Walsh-dominant: 14 Ambrose-dominant: 20

This is not necessarily the most successful West Indian pair. However they did capture over 750 wickets.

Ambrose has dominated the pair slightly, as shown by the 20-14 division in the dominance analysis. They have also succeeded nearly 60% of the times, indicating a lack of great bowling support.

That they bowled during the waning period for West Indies is shown by the win % values, which are all in the 40+ region. However as a pair they averaged 22.67 which is the third best in this analysis. Ambrose picked almost all wickets, barring 17, while bowling with Walsh. Their WpT is a not-too-great 8 and they have captured around half the team wickets.

Walsh & Ambrose have captured all 10 wickets in the innings once. This was in Test # 1188 against South Africa at Bridgetown at Wanderer's during 1992 when they captured for 10 for 65 (6 for 31 & 6 for 34). Their best match performance has been 16 wickets (5 & 11) in Test # 1257 against England at POS during 1994.

Summary of bowling career of Courtney Walsh and Curtly Ambrose
© Anantha Narayanan

Akram & Younis

P04. Akram 282 @ 21.33 & Younis 277 @ 22.93 Total: 559 @ 22.12
-------------------------------------------------------------------
Akram       : 1010(1985)-1584(2002) 282 @ 21.33 (132 @ 28.51) 74.8%
Younis      : 1127(1989)-1637(2003) 277 @ 22.93 ( 96 @ 25.39) 90.3%
Pair Career:  1127(1989) to 1584(2002)
All matches:          143      Wins:  54 (37.8 %)
Pair matches:          60      Wins:  28 (46.7 %)
Pair Home matches:     22      Wins:  10 (45.5 %)
Pair Away matches:     38      Wins:  18 (47.4 %)
Non-Pair matches:      82      Wins:  26 (31.7 %)
Non-Pair Home matches: 22      Wins:   7 (31.8 %)
Non-Pair Away matches: 60      Wins:  19 (31.7 %)
Matches on their own  Akram: 43  Younis: 26
Pair   Home wkts:  223 @ 20.27  Away wkts:  336 @ 23.35
Others Home wkts:  142 @ 32.46  Away wkts:  272 @ 32.68
Team Wkts:         973 @ 26.58  WpT: 9.32  Pair %: 57.5
TeamPeerWkts:      414 @ 32.60
AllPeerWkts:     11463 @ 30.37
Top Order Wkts:    320 (57.2%)  Low Order Wkts:  145 (25.9%)
Top Order Home Wkts: 130 (58.3%)  Top Order Away Wkts: 190 (56.5%)
Highest vs West Indies:  104 @ 21.22
Team wkts-5+:  96  Pair-Successes (50% of Team or more): 67 (69.8%)
Pair wkts-5+: 66 Shared: 45 Akram-dominant: 12 Younis-dominant: 9

Ah we now come to probably the most colourful bowling pair which ever took the red cherry. Two totally different individuals and bowlers, yet together they became arguably the most fearsome and feared bowling pair ever.

The first thing that strikes is the difference in the two bowlers when they bowled together and separately. Waqar was quite effective bowling alone, operating at around 90%. However Wasim was lost without Waqar. Most of these matches were during the early years. Wasim was acting at only around 75% of his pair averages when he was bowling alone. In fact it can be seen that this is by far the lowest figure amongst all pairs for one bowler.

The pair captured had a success rate of nearly 70%. Akram was slightly more dominant than Younis by a 12-9 margin. Their win %s were all below 50. They had an excellent WpT figure of well over 9 together.

Akram and Younis have captured all 10 wickets in the innings twice. Their best innings performance has been Test # 1257 against Sri Lanka at Kandy during 1994 when they bowled unchanged on the opening day and captured 10 for 66 (4 for 32 & 6 for 34). Their best match performance has been 17 wickets (8 & 9) in Test # 1207 against New Zealand at Hamilton during 1993. In addition, they have captured 16 wickets thrice and 15 wickets 6 times. Imagine: 15 or more wickets in a match captured an astonishing 10 times!

Summary of bowling career of Wasim Akram and Waqar Younis
© Anantha Narayanan

Willis & Botham

P05. Willis 222 @ 24.63 & Botham 254 @ 25.67 Total: 476 @ 25.18
-------------------------------------------------------------------
Willis      : 0676(1971)-0991(1984) 222 @ 24.63 (103 @ 26.44) 93.2%
Botham      : 0806(1977)-1190(1992) 254 @ 25.67 (129 @ 33.78) 76.0%
Pair Career:  0806(1977) to 0991(1984)
All matches:          223      Wins:  61 (27.4 %)
Pair matches:          60      Wins:  25 (41.7 %)
Pair Home matches:     33      Wins:  17 (51.5 %)
Pair Away matches:     27      Wins:   8 (29.6 %)
Non-Pair matches:     162      Wins:  36 (22.2 %)
Non-Pair Home matches: 91      Wins:  16 (17.6 %)
Non-Pair Away matches: 71      Wins:  20 (28.2 %)
Matches on their own  Willis: 30  Botham: 41
Pair   Home wkts:  288 @ 23.43  Away wkts:  188 @ 27.87
Others Home wkts:  227 @ 29.69  Away wkts:  224 @ 29.03
Team Wkts:         927 @ 27.22  WpT: 7.93  Pair %: 51.3
TeamPeerWkts:      451 @ 29.36
AllPeerWkts:      4559 @ 30.16
Top Order Wkts:    299 (62.8%)  Low Order Wkts:   93 (19.5%)
Top Order Home Wkts: 180 (62.5%)  Top Order Away Wkts: 119 (63.3%)
Highest vs Australia:  183 @ 25.24
Team wkts-5+:  92  Pair-Successes (50% of Team or more): 55 (59.8%)
Pair wkts-5+: 55 Shared: 28 Willis-dominant: 9 Botham-dominant: 18

Unlike Akram, Willis' figures do not drop off at the start of the career possibly because he had a good set of bowlers bowling with him, led by Snow during his early years. However Botham, without Willis at the other end, was as spent a force as Akram. His average increases by over 30%. Their pair-success value is around 60. The RpT value is below 8. However Botham was the more dominant bowler by a big margin of 18-9 in those situations where they succeeded.

Willis & Botham have captured all 10 wickets in the innings no fewer than three times. Their best innings performance has been Test # 826 against Pakistan at Lord's during 1978 when they captured 10 for 60 (2 for 26 & 8 for 34). Their best match performance has been 16 wickets (5 and 11) in Test # 830 against New Zealand at Lord's during 1978.

Summary of bowling career of Bob Willis and Ian Botham
© Anantha Narayanan

Donald & Pollock

P06. Donald 208 @ 22.02 & Pollock 189 @ 21.66 Total: 397 @ 21.85
-------------------------------------------------------------------
Donald      : 1188(1992)-1590(2002) 208 @ 22.02 (122 @ 22.66) 97.2%
Pollock     : 1312(1995)-1860(2008) 189 @ 21.66 (232 @ 24.31) 89.1%
Pair Career:  1312(1995) to 1582(2002)
All matches:          155      Wins:  72 (46.5 %)
Pair matches:          47      Wins:  23 (48.9 %)
Pair Home matches:     28      Wins:  17 (60.7 %)
Pair Away matches:     19      Wins:   6 (31.6 %)
Non-Pair matches:     107      Wins:  49 (45.8 %)
Non-Pair Home matches: 30      Wins:  13 (43.3 %)
Non-Pair Away matches: 77      Wins:  36 (46.8 %)
Matches on their own  Donald: 25  Pollock: 60
Pair   Home wkts:  242 @ 20.79  Away wkts:  155 @ 23.49
Others Home wkts:  217 @ 29.14  Away wkts:  148 @ 34.64
Team Wkts:         762 @ 26.41  WpT: 8.45  Pair %: 52.1
TeamPeerWkts:      365 @ 31.37
AllPeerWkts:      6800 @ 30.22
Top Order Wkts:    265 (66.8%)  Low Order Wkts:   80 (20.2%)
Top Order Home Wkts: 163 (67.4%)  Top Order Away Wkts: 102 (65.8%)
Highest vs England:  116 @ 22.76
Team wkts-5+:  78  Pair-Successes (50% of Team or more): 46 (59.0%)
Pair wkts-5+: 44 Shared: 36 Donald-dominant: 3 Pollock-dominant: 5

Look at how close Donald's and Pollock's figures are. The averages are all around 22, the best in this analysis. Donald bowled as effectively without Pollock while Pollock needed the pace of Donald a bit more. Pollock has played many more matches without Donald than any other bowler in this 20 bowler collection. Look at the dominance analysis. They have the highest shared % of all pairs. Very few innings in which either bowler has dominated: only 3-5. A very high level of top order batsmen scalps. And very impressive away averages for the pair.

Donald & Pollock have captured all 10 wickets in the innings twice. Their best innings performance has been Test # 1471 against England at Wanderer's during 1999 when they captured 10 for 69 (6 for 53 and 4 for 16). Their best match performance has been an unprecedented 19 wickets (10 and 9) in the same Test. This is the only instance of one of these pairs capturing 19 wickets in a match.

Summary of bowling career of Allan Donald and Shaun Pollock
© Anantha Narayanan

Chandrasekhar & Bedi

P07. Chandrasekhar 184 @ 28.71 & Bedi 184 @ 27.22 Total: 368 @ 27.96
-------------------------------------------------------------------
Chandrasekhar: 0552(1964)-0851(1979) 184 @ 28.71 ( 58 @ 33.05) 86.9%
Bedi        :  0612(1967)-0854(1979) 184 @ 27.22 ( 82 @ 32.05) 84.9%
Pair Career:   0612(1967) to 0837(1978)
All matches:           87      Wins:  21 (24.1 %)
Pair matches:          42      Wins:  11 (26.2 %)
Pair Home matches:     20      Wins:   7 (35.0 %)
Pair Away matches:     22      Wins:   4 (18.2 %)
Non-Pair matches:      45      Wins:  10 (22.2 %)
Non-Pair Home matches:  9      Wins:   2 (22.2 %)
Non-Pair Away matches: 36      Wins:   8 (22.2 %)
Matches on their own  Chandrasekhar: 16  Bedi: 25
Pair   Home wkts:  191 @ 25.09  Away wkts:  177 @ 31.07
Others Home wkts:  109 @ 36.02  Away wkts:  123 @ 47.26
Team Wkts:         600 @ 33.38  WpT: 8.76  Pair %: 61.3
TeamPeerWkts:      232 @ 41.98
AllPeerWkts:      5835 @ 30.23
Top Order Wkts:    224 (60.9%)  Low Order Wkts:   77 (20.9%)
Top Order Home Wkts: 111 (58.1%)  Top Order Away Wkts: 113 (63.8%)
Highest vs England:  162 @ 26.40
Team wkts-5+:  60  Pair-Successes (50% of Team or more): 46 (76.7%)
Pair wkts-5+: 45 Shared: 33 Chandrasekhar-dominant: 7 Bedi-dominant: 5

The romantic era of the 60s-70s when spin was king, queen and everything else.

Indian wins were not too many during this era and this pair did well, especially at home. They both bowled very well together and fell off by 15% when bowling separately. Look at their very decent away performance: an average of 31. And look at the awful away figures of the other bowlers even though there was one Prasanna in that group. And when India captured 5 wickets, these two were successful a huge 75%. They also seem to have shared the wickets fairly equitably amongst themselves.

Each was lost without the other. In three matches, Bedi had figures of 1 for 163, 1 for 152 and 1 for 146 without Chandra. He did quite well when Chandra joined him in between.. In between the latter two expensive 1-wicket hauls, this time Chandra on his own, captured 2 for 102.

Chandra & Bedi have captured all 10 wickets in the innings no fewer than three times. Their best innings performance has been Test # 812 against Australia at MCG during 1978 when they captured 10 for 110 (6 for 52 & 4 for 58). In the same match they also captured 18 wickets (12 and 6).

Summary of bowling career of Bishen Singh Bedi and BS Chandrasekhar
© Anantha Narayanan

Hadlee & Cairns

P08. Hadlee 207 @ 22.02 & Cairns 130 @ 32.58 Total: 337 @ 26.10
-------------------------------------------------------------------
Hadlee      : 0710(1973)-1147(1990) 207 @ 22.02 (224 @ 22.55) 97.7%
Cairns      : 0730(1974)-1031(1985) 130 @ 32.58 (  0 @  0.00)  0.0%
Pair Career:  0730(1974) to 1031(1985)
All matches:          100      Wins:  22 (22.0 %)
Pair matches:          42      Wins:  12 (28.6 %)
Pair Home matches:     23      Wins:   9 (39.1 %)
Pair Away matches:     19      Wins:   3 (15.8 %)
Non-Pair matches:      58      Wins:  10 (17.2 %)
Non-Pair Home matches: 11      Wins:   1 ( 9.1 %)
Non-Pair Away matches: 47      Wins:   9 (19.1 %)
Matches on their own  Hadlee: 44  Cairns: 1
Pair   Home wkts:  188 @ 23.48  Away wkts:  149 @ 29.40
Others Home wkts:  137 @ 32.07  Away wkts:  127 @ 39.46
Team Wkts:         601 @ 30.28  WpT: 8.02  Pair %: 56.1
TeamPeerWkts:      264 @ 35.63
AllPeerWkts:      7911 @ 30.35
Top Order Wkts:    204 (60.5%)  Low Order Wkts:   71 (21.1%)
Top Order Home Wkts: 118 (62.8%)  Top Order Away Wkts:  86 (57.7%)
Highest vs England:   72 @ 27.76
Team wkts-5+:  59  Pair-Successes (50% of Team or more): 38 (64.4%)
Pair wkts-5+: 38 Shared: 22 Hadlee-dominant: 14 Cairns-dominant: 2

This is the only pair in which one bowler, Cairns, has got all his wickets, bowling with the other bowler. He bowled, by himself, in one match (#758) and got 0 for 44. As expected Hadlee dominates by a huge factor of 14-2. I myself am interested in knowing about those two occasions. In match # 958, Cairns took 7 for 74 and Hadlee 0 for 44. A few matches later Cairns captured 7 for 143 and Hadlee, 2 for 97. That is all.

They did quite well at home and held their own away from home. There is no doubt Hadlee was the mainstay but Cairns played the supporting role very effectively.

Hadlee & Cairns have never captured all 10 wickets in an innings. Their best innings performance has been Test # 875 against West Indies at Christchurch during 1980 when they captured 9 for 143 (3 for 58 & 6 for 85). Their best match performance has been 14 wickets (11 and 3) in the earlier match of that series, Test # 873 at Dunedin. A rare win for New Zealand over the mighty West Indies.

Summary of bowling career of Richard Hadlee and Lance Cairns
© Anantha Narayanan

Garner & Marshall

P09. Garner 152 @ 21.38 & Marshall 170 @ 22.02 Total: 322 @ 21.72
-------------------------------------------------------------------
Garner      : 0797(1977)-1072(1987) 152 @ 21.38 (107 @ 20.40) 104.8%
Marshall    : 0837(1978)-1175(1991) 170 @ 22.02 (206 @ 20.06) 109.8%
Pair Career:  0880(1980) to 1072(1987)
All matches:          117      Wins:  56 (47.9 %)
Pair matches:          36      Wins:  20 (55.6 %)
Pair Home matches:     18      Wins:  12 (66.7 %)
Pair Away matches:     18      Wins:   8 (44.4 %)
Non-Pair matches:      81      Wins:  36 (44.4 %)
Non-Pair Home matches: 14      Wins:   5 (35.7 %)
Non-Pair Away matches: 67      Wins:  31 (46.3 %)
Matches on their own  Garner: 22  Marshall: 45
Pair   Home wkts:  161 @ 21.09  Away wkts:  161 @ 22.35
Others Home wkts:  149 @ 27.46  Away wkts:  140 @ 27.13
Team Wkts:         611 @ 24.36  WpT: 8.94  Pair %: 52.7
TeamPeerWkts:      289 @ 27.30
AllPeerWkts:      4687 @ 31.27
Top Order Wkts:    192 (59.6%)  Low Order Wkts:   69 (21.4%)
Top Order Home Wkts: 102 (63.4%)  Top Order Away Wkts:  90 (55.9%)
Highest vs England:  141 @ 19.30
Team wkts-5+:  60  Pair-Successes (50% of Team or more): 39 (65.0%)
Pair wkts-5+: 38 Shared: 25 Garner-dominant: 5 Marshall-dominant: 8

The first of the two special selections. These two greats might very well have warranted the top selection on their own. Look at their numbers. All pair values are around the 22 mark, almost the same as Donald & Pollock. The surprise is the overall win % mark. I would have expected a slightly higher figure, somewhere around the 60-65% mark. But does not matter. A wonderful away average and an out-of-the-world performance against England mark this pair's achievements.

As expected, Garner & Marshall have never captured all 10 wickets in an innings. Their best innings performance has been Test # 991 against England at Leeds during 1984 when they captured 9 for 90 (2 for 32 and 7 for 53). Their best match performance has been 13 wickets (5 and 8) in Test # 988 against Australia at Kingston during 1984.

Summary of bowling career of Joel Garner and Malcolm Marshall
© Anantha Narayanan

Miller & Lindwall

P10. Miller 150 @ 23.69 & Lindwall 195 @ 22.48 Total: 345 @ 23.00
-------------------------------------------------------------------
Miller      : 0275(1946)-0430(1956) 150 @ 23.69 ( 20 @ 17.65) 134.2%
Lindwall    : 0275(1946)-0487(1960) 195 @ 22.48 ( 33 @ 26.30) 85.5%
Pair Career:  0275(1946) to 0430(1956)
All matches:           78      Wins:  44 (56.4 %)
Pair matches:          51      Wins:  28 (54.9 %)
Pair Home matches:     26      Wins:  17 (65.4 %)
Pair Away matches:     25      Wins:  11 (44.0 %)
Non-Pair matches:      27      Wins:  16 (59.3 %)
Non-Pair Home matches:  4      Wins:   1 (25.0 %)
Non-Pair Away matches: 23      Wins:  15 (65.2 %)
Matches on their own  Miller: 4  Lindwall: 10
Pair   Home wkts:  180 @ 20.84  Away wkts:  165 @ 25.36
Others Home wkts:  278 @ 26.02  Away wkts:  218 @ 30.72
Team Wkts:         841 @ 26.00  WpT: 6.76  Pair %: 41.0
TeamPeerWkts:      496 @ 28.08
AllPeerWkts:      3406 @ 29.41
Top Order Wkts:    229 (66.4%)  Low Order Wkts:   63 (18.3%)
Top Order Home Wkts: 119 (66.1%)  Top Order Away Wkts: 110 (66.7%)
Highest vs England:  174 @ 22.39
Team wkts-5+:  85  Pair-Successes (50% of Team or more): 33 (38.8%)
Pair wkts-5+: 32 Shared: 17 Miller-dominant: 4 Lindwall-dominant: 11

Lindwall & Miller have very good figures but these were good years for bowling. Good win % values indicate the superiority of Australia during the 1946-55 period. Their Wickets per Test is the lowest amongst these pairs and does not even reach 7. This is also reflected in the low % of pair successes. They have always been quite consistent at home or away.

Since Miller & Lindwall were never the dominating bowling pair in the attack, they have never captured all 10 wickets in an innings. Their best innings performance has been Test # 408 against West Indies at Kingston during 1955 when they captured 8 for 171 (6 for 107 & 2 for 64). Their best match performance has been 14 wickets (7 and 7) in Test # 341 against West Indies at SCG during 1952.

Summary of bowling career of Ray Lindwall and Keith Miller
© Anantha Narayanan

Please peruse the following few lines. This is worthy of a special comment. Two wonderful bowlers, almost working like Siamese twins together. I wish I had got them in. At least I have given them a belated salute.

Statham 141 @ 25.72 & Trueman 143 @ 25.76 Total: 284 @ 25.74
-------------------------------------------------------------------
Statham    : 0332(1951)-0596(1965) 141 @ 25.72 (111 @ 23.74) 108.3%
Trueman    : 0351(1952)-0592(1965) 143 @ 25.76 (164 @ 17.93) 143.7%
Pair Career: 0380(1954) to 0547(1963)
Pair wkts-5+: 34 Shared: 28 Statham-dominant: 3 Trueman-dominant: 3

The summary table with key data elements is displayed here. No special comments are needed.

Bowler 1WktsAvgeBowler 2WktsAvgeP-WktsP-AvgeTestsWins% WinsP>=50%
           Twkts>4
Warne51324.88McGrath48821.38100123.171047168.364.5
Muralitharan58622.05Vaas30928.0189524.11944143.679.3
Walsh37324.29Ambrose38921.1276222.67954244.259.2
Akram28221.33Younis27722.9355922.12602846.769.8
Willis22224.63Botham25425.6747625.18602541.759.8
Donald20822.02Pollock18921.6639721.85472348.959.0
Chandrasekhar18428.71Bedi18427.2236827.96421126.276.7
Hadlee20722.02Cairns13032.5833726.10421228.664.4
Garner15221.38Marshall17022.0232221.72362055.665.0
Miller15023.69Lindwall19522.4834523.00512854.938.8
Statham14125.72Trueman14325.7628425.74351337.163.5

To download the complete table, the widest Excel sheet I have seen for a long time (75 columns!), please CLICK HERE.

Finally, let me confess that I might have missed many more but here is a list of stand outs revealed by this analysis:

1. The absence of a big gap between Warne & McGrath and other peer bowlers indicating the abundance of quality. The near-70% success rate of Warne & McGrath.
2. The impact Murali at the other end had on Vaas' bowling.
3. The astounding level of success Murali & Vaas had when their team captured 5 wickets or more.
4. The extent of dependency Wasim Akram had on the support given by Waqar Younis.
5. The unbelievable 10 occasions when Wasim and Waqar captured 15 or more wickets in a match.
6. The closeness of numbers between Donald & Pollock confirming the extent of under-estimating of Pollock.
7. Their unique 19-wickets capture in this elite group. I know that there have been 20-wickets captured in a match by a pair (Trumble/Noble, Hirst/Blythe, Faulkner/Vogler, Laker/Lock, Massie/Lillee et al), but those bowlers are outside this group.
8. The extraordinary variation in the pair's away performance and other bowlers' away performance for Murali & Vaas and Chandra & Bedi.
9. The closeness of Home/Away performances for the two West Indian pairs. In fact it is unbelievable that Walsh & Ambrose have performed better away than at home.
10.The confirmation that Botham was three-fourth the bowler he was when Willis was absent.
11. The 22 level averages for the two West Indian pairs and Donald & Pollock.

Finally my usual request. Give respect to this wonderful collection of great bowlers. The only way this can be done is to dedicate your comments to the bowlers and let them take centre stage. Do not bring in frivolous comparisons, using vague numbers that batsman A played P&Q better than B or that C was shaking/shaky against X&Y or that S&T were ineffective against batsman D. I will be ruthless and cut off such comments right at the beginning. There is so much data provided here and I have just skimmed the top. Look for hidden gems here instead of indulging in fantasies.

Let me add that I have set up the Bowling pair analysis program such that I have a dummy pair which can be replaced by any duo and this pair can be analysed instantly. So you are welcome to ask summary details for any meaningful pair. Of course, if you ask for Hirwani & Arshad Ayub or Cuffy & Ramnarine or Allott & Pocock, it is unlikely that I would oblige.

Thank you.

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

  • Ariz khan on November 24, 2012, 20:48 GMT

    Hi Anantha, just checking your blog after a long time. Busy time. Glad to see you still kicking.

    "However the big surprise is that Wasim Akram and Waqar Younis have a figure of 57.2%, the lowest among these pairs. Seems quite intriguing." This is not a surprise, a well known fact but yes, intriguing. A plain look at the wickets of Akram, the batters whom he has dismissed most would make it clear, many tailenders and also many lowly placed batsmen. A reason why I was so furious when he was selected ahead of so many better bowlers. Possibly because of him being a left-armer. (I never believe these lists) One of the reasons is their inability to deliver with the new balls (with few exceptions, of course).

  • Dr. talha on October 30, 2012, 4:59 GMT

    Pithches have changed or not can be debatable (apart from Windies,it has surely changed) But the quality of quick bowlers since 2000 has certainly gone down. Atleast for a couple of countries. How can u compare: Ambrose, Walsh, Bishop with Roach, Edwards, Rampaul Wasim & Waqar with Gul & Wahab With the introduction of Bangladesh, and now teams touring places like srilanka more often, the batting averages have definitely improved.

  • shrikanthk on October 30, 2012, 3:00 GMT

    scoring has definitely gone up, I imagine changes in pitch conditions is a factor in that.

    I think the reason is a simpler and more proximate one. We just had better batsmen post 1990 around the world. One doesn't need to bring pitches in.

    Why do I say batting standards improved post 1990 vis-a-vis 1970-1990. In order to neutralize for pitches and bowling quality, just look at the First-class records around the world for both periods. If you look at Aus cricket in the 90s and 00s, you had probably a dozen batsmen who scored very very heavily and consistently in FC cricket. Indicative of their superior batting quality (all other things being equal). The Aus batsmen of the 70s/80s just weren't in the same league. Not just in test cricket but even in FC cricket! The likes of Dyson, Laird, Wood had sub-40 FC averages!

    Same thing holds for India. Our batting crop simply got richer in the 90s/00s as compared to the 80s.

  • Alex on October 29, 2012, 20:10 GMT

    @Wasp:

    1. Incidentally, the # bowlers who took at least 45 wkts _and_ averaged less than 32 is 49 for the 1970-1990 era and 65 for the 1991-2011 era. [Note: The 1991-2011 era had twice as many tests and at least 4 more regularly playing nations (SA/SL/Zim/BD).]

    2. So, as it happened with batsmen averaging 50+ (increase of 53%: from 13 to 20), this number should have increased from 49 to 75 but reached 65 only, i.e., fell short by 14%.

    3. Fielding standard was generally superior in 1991-2011 than in 1970-1990. So, this anomaly of 14% indicates that, compared to 1970-1990, 1991-2011 era had either poorer bowling quality or easier pitches or both.

  • Alex on October 29, 2012, 14:33 GMT

    @Wasp: You should restrict yourself to the period of interest only, i.e., for example, 1970-1990 and 1991-onwards.

    1. The 1970-1990 era had about 460 tests whereas the 1991-2012 era had nearly 930 tests ... the frequency of # tests/year has doubled!!

    2. 2. Triple hundreds are a rarity. Better to set 200's as the standard for the big scores. Then, the '70-'90 era had 63 double hundreds whereas the 1991-2012 era had 158. Since the # tests/year itself has doubled, there are barely 20 extra 200's to account for and that is possibly due to Zim/BD and possible due to some pitches.

    3. Let's round off 49.XX to 50. Then, batsmen avg 50+ over '70-'90 are not 4 but fully 13: Chappell, SMG, Viv, AB, Miandad, Boycott, Amiss, Azhar, Dean Jones, Shoaib Mohammad, Mark Taylor, Sobers. For 1991-2012, this number is 22 ... this might be due to somewhat easier pitches.

  • Waspsting on October 29, 2012, 11:13 GMT

    scoring has definitely gone up, I imagine changes in pitch conditions is a factor in that.

    Gooch's 333 in '90 was the first triple in 17 years, after Lawrence Rowe's.

    In 23 years since, you've had 14 triples - 2 from Lara, Gayle, Sehwag and 1 from Jayasuriya, Jayawardena, Taylor, Hayden, Clark, Inzamam, Younis Khan, Amla - and only 1 of those was against Zimbabwae or Bangladesh.

    Look at the # of players ave. 50 up from mid 90s on. Hayden, Ponting, Waugh, Yousuf, Younis, Tendulkar, Dravid, Sehwag, Jayawardena, Sangakarra, Kallis, Lara - as well as Samaraweera, Gambhir, Inzie, Pietersen, Cook, de Velliers, Chanderpaul.

    From 70-90, there were 5 total - Chappell, Viv, Gavaskar, Miandad, Border.

    Some combo of batting getting better, bowling getting worse and pitches getting better probably the reason.

    I'd guess pitches is probably the biggest chunk of the lot

  • shrikanthk on October 27, 2012, 12:53 GMT

    Get well soon Ananth. Health is paramount. Don't rush into writing again ;)

    Perhaps thats the reason why, apart from Steyn u see nobody who will end his career at an avearge of under 25.

    Well, I object to this notion that pitches of today are much flatter. Steyn bowls on some extremely responsive wickets half the time as anyone who watched the 2-test series against Aus last year would testify.

    Also even Lillee (who is often painted as a poor bowler on flat wikets) had some of his best returns at the MCG which generally had this reputation as a flat wicket during those days.

    I don't think pitches have gotten flat at all. Maybe they turn less than they used to (especially pitches like Sydney). But I doubt if they assist seam bowling any less than they used to, though it's hard to generalize either way.

  • dale on October 25, 2012, 17:32 GMT

    Ananth,I wish you a speedy recovery! [[ A common thanks to all. With one arm in a sling, my comments are going to be minimal. Ananth: ]]

  • Dr. talha on October 25, 2012, 12:21 GMT

    @Alex. "In 20 tests leading to the Pak series, Greenidge had averaged mere 39." How can u say that 39 is "mere"? During that time 39 was a pretty good average. Its just like somebody averages around mid 40's in today's cricket. U have to consider the quality of bowlers(especially new ball bowlers)and,the quality of pitches.

  • Boll on October 25, 2012, 10:46 GMT

    I`ve really enjoyed reading the comments here, but have refrained from commenting too much. While bowler-focused analyses generally seem to attract less comment, I think the comments that are made are often more thoughtfully written and less agenda-driven. Discussion of batsmen tends to bring out our partisan/nationalist tendencies a little more - mine included.

    Great reading (from pillar to post) and best wishes for the surgery Ananth - I hope the nurses are beautiful, the anaesthetic strong, and the recovery swift. Unfortunately, as seems to happen everywhere, I imagine the food will be crap. Get well soon.

  • Ariz khan on November 24, 2012, 20:48 GMT

    Hi Anantha, just checking your blog after a long time. Busy time. Glad to see you still kicking.

    "However the big surprise is that Wasim Akram and Waqar Younis have a figure of 57.2%, the lowest among these pairs. Seems quite intriguing." This is not a surprise, a well known fact but yes, intriguing. A plain look at the wickets of Akram, the batters whom he has dismissed most would make it clear, many tailenders and also many lowly placed batsmen. A reason why I was so furious when he was selected ahead of so many better bowlers. Possibly because of him being a left-armer. (I never believe these lists) One of the reasons is their inability to deliver with the new balls (with few exceptions, of course).

  • Dr. talha on October 30, 2012, 4:59 GMT

    Pithches have changed or not can be debatable (apart from Windies,it has surely changed) But the quality of quick bowlers since 2000 has certainly gone down. Atleast for a couple of countries. How can u compare: Ambrose, Walsh, Bishop with Roach, Edwards, Rampaul Wasim & Waqar with Gul & Wahab With the introduction of Bangladesh, and now teams touring places like srilanka more often, the batting averages have definitely improved.

  • shrikanthk on October 30, 2012, 3:00 GMT

    scoring has definitely gone up, I imagine changes in pitch conditions is a factor in that.

    I think the reason is a simpler and more proximate one. We just had better batsmen post 1990 around the world. One doesn't need to bring pitches in.

    Why do I say batting standards improved post 1990 vis-a-vis 1970-1990. In order to neutralize for pitches and bowling quality, just look at the First-class records around the world for both periods. If you look at Aus cricket in the 90s and 00s, you had probably a dozen batsmen who scored very very heavily and consistently in FC cricket. Indicative of their superior batting quality (all other things being equal). The Aus batsmen of the 70s/80s just weren't in the same league. Not just in test cricket but even in FC cricket! The likes of Dyson, Laird, Wood had sub-40 FC averages!

    Same thing holds for India. Our batting crop simply got richer in the 90s/00s as compared to the 80s.

  • Alex on October 29, 2012, 20:10 GMT

    @Wasp:

    1. Incidentally, the # bowlers who took at least 45 wkts _and_ averaged less than 32 is 49 for the 1970-1990 era and 65 for the 1991-2011 era. [Note: The 1991-2011 era had twice as many tests and at least 4 more regularly playing nations (SA/SL/Zim/BD).]

    2. So, as it happened with batsmen averaging 50+ (increase of 53%: from 13 to 20), this number should have increased from 49 to 75 but reached 65 only, i.e., fell short by 14%.

    3. Fielding standard was generally superior in 1991-2011 than in 1970-1990. So, this anomaly of 14% indicates that, compared to 1970-1990, 1991-2011 era had either poorer bowling quality or easier pitches or both.

  • Alex on October 29, 2012, 14:33 GMT

    @Wasp: You should restrict yourself to the period of interest only, i.e., for example, 1970-1990 and 1991-onwards.

    1. The 1970-1990 era had about 460 tests whereas the 1991-2012 era had nearly 930 tests ... the frequency of # tests/year has doubled!!

    2. 2. Triple hundreds are a rarity. Better to set 200's as the standard for the big scores. Then, the '70-'90 era had 63 double hundreds whereas the 1991-2012 era had 158. Since the # tests/year itself has doubled, there are barely 20 extra 200's to account for and that is possibly due to Zim/BD and possible due to some pitches.

    3. Let's round off 49.XX to 50. Then, batsmen avg 50+ over '70-'90 are not 4 but fully 13: Chappell, SMG, Viv, AB, Miandad, Boycott, Amiss, Azhar, Dean Jones, Shoaib Mohammad, Mark Taylor, Sobers. For 1991-2012, this number is 22 ... this might be due to somewhat easier pitches.

  • Waspsting on October 29, 2012, 11:13 GMT

    scoring has definitely gone up, I imagine changes in pitch conditions is a factor in that.

    Gooch's 333 in '90 was the first triple in 17 years, after Lawrence Rowe's.

    In 23 years since, you've had 14 triples - 2 from Lara, Gayle, Sehwag and 1 from Jayasuriya, Jayawardena, Taylor, Hayden, Clark, Inzamam, Younis Khan, Amla - and only 1 of those was against Zimbabwae or Bangladesh.

    Look at the # of players ave. 50 up from mid 90s on. Hayden, Ponting, Waugh, Yousuf, Younis, Tendulkar, Dravid, Sehwag, Jayawardena, Sangakarra, Kallis, Lara - as well as Samaraweera, Gambhir, Inzie, Pietersen, Cook, de Velliers, Chanderpaul.

    From 70-90, there were 5 total - Chappell, Viv, Gavaskar, Miandad, Border.

    Some combo of batting getting better, bowling getting worse and pitches getting better probably the reason.

    I'd guess pitches is probably the biggest chunk of the lot

  • shrikanthk on October 27, 2012, 12:53 GMT

    Get well soon Ananth. Health is paramount. Don't rush into writing again ;)

    Perhaps thats the reason why, apart from Steyn u see nobody who will end his career at an avearge of under 25.

    Well, I object to this notion that pitches of today are much flatter. Steyn bowls on some extremely responsive wickets half the time as anyone who watched the 2-test series against Aus last year would testify.

    Also even Lillee (who is often painted as a poor bowler on flat wikets) had some of his best returns at the MCG which generally had this reputation as a flat wicket during those days.

    I don't think pitches have gotten flat at all. Maybe they turn less than they used to (especially pitches like Sydney). But I doubt if they assist seam bowling any less than they used to, though it's hard to generalize either way.

  • dale on October 25, 2012, 17:32 GMT

    Ananth,I wish you a speedy recovery! [[ A common thanks to all. With one arm in a sling, my comments are going to be minimal. Ananth: ]]

  • Dr. talha on October 25, 2012, 12:21 GMT

    @Alex. "In 20 tests leading to the Pak series, Greenidge had averaged mere 39." How can u say that 39 is "mere"? During that time 39 was a pretty good average. Its just like somebody averages around mid 40's in today's cricket. U have to consider the quality of bowlers(especially new ball bowlers)and,the quality of pitches.

  • Boll on October 25, 2012, 10:46 GMT

    I`ve really enjoyed reading the comments here, but have refrained from commenting too much. While bowler-focused analyses generally seem to attract less comment, I think the comments that are made are often more thoughtfully written and less agenda-driven. Discussion of batsmen tends to bring out our partisan/nationalist tendencies a little more - mine included.

    Great reading (from pillar to post) and best wishes for the surgery Ananth - I hope the nurses are beautiful, the anaesthetic strong, and the recovery swift. Unfortunately, as seems to happen everywhere, I imagine the food will be crap. Get well soon.

  • Ravi on October 25, 2012, 8:50 GMT

    Ananth, you've shouldered a big responsibility of running a successful and popular blog for many years now. About time the shoulder needs attention, surgery and rest. Hope you get well soon and enlighten us again from the next week!

    Ravi

  • Nitin Gautam on October 25, 2012, 7:39 GMT

    Wish you a speedy recovery. Get well soon

  • Dr. talha on October 25, 2012, 6:48 GMT

    @Ananth. Get well soon brother!!

  • Dr. talha on October 25, 2012, 6:30 GMT

    @Alex. Its not about distorting things in preference to Pak players. When there was an article by Ananth on wicketkeepers, and he rated Kamran Akmal very highly, i was the first one to object on that. Greenidge came into discussion,when u said that,W's played against Zim/Srl as well..so my reply to that was holding & lillie bowled to few Eng batting line ups that was even weaker than that of Zim & Srl in the 90's. I agree with u that every batsman (may be bowler as well) struggles vs one nation or may be two. Greenidge was used as an example because he has been a top player of fast bowling before and after that series. In early 90's, Waqar got 35 wickets in 6 tests against the mighty WI (top team of the era) and many of the windies batsmen sruggled against him. I saw the 1990 Pak/WI series live, believe me it was not massively bowling friendly pitches, but the quality of bowling attacks were surely massively destructive.

  • Som on October 25, 2012, 0:27 GMT

    Ananth - Wish you well and a speedy recovery. [[ Thanks for all wishes. Ananth: ]]

  • Dinesh on October 24, 2012, 19:45 GMT

    Get well soon Ananth.

  • Alex on October 24, 2012, 16:51 GMT

    @Ananth: Best wishes for the surgery.

    @Dr Talla: I think you tend to distort things in preference to Pak players. Note this:

    1. Every batsman struggles vs one nation, and it was Pak for Greenidge: his avg is 31 vs Pak over 14 tests, and 17 in Pak!!

    2. He scored 366 in his final series vs Oz at avg=46 but that had a truly great innings of 226 which is like the brightness of a lamp before its extinction ... otherwise, he scored a pitiful 140 runs in 8 innings in that series.

    2. In 20 tests leading to the Pak series, Greenidge had averaged mere 39. The Pak series was massively bowler friendly featuring fast pitches --- it had these avg: 14 (Akram), 17 (Ambrose), 19 (Waqar), 19 (Bishop), 28 (Marshall & Walsh). Only 4 batsmen scored more than 180 runs in that 3-test series despite playing all 6 innings, and no one managed even 50 runs/innings.

  • Waspsting on October 24, 2012, 11:41 GMT

    Incidentally, Garner has a rep for being very good against the tail (no idea how the stats back this up - i tend to not take "reputations" as proof).

    If that is so, that'd account for his very low overall average (assuming he's about as good against top order as others)

    Marshall stands supreme - with a poor rep against low order, he must've been extra successful against the top to keep his overall average so low. --- i was looking at all pairs of Roberts, Holding, Garnger, Croft, Marshall, and here are the results (average, w/ min. 100 wickets)

    Roberts (w/Holding)25.22 Holding (w/Roberts)26.09 Holding (w/Garner) 23.60 Holding (w/Marshall) 25.27 Garner (w/Roberts) 21.93 Garner (w/Holding)20.42 Garner (w/Croft) 21.24 Garner (w/Marshall) 21.38 Croft (w/Garner) 23.74 Marshall (w/Holding) 20.77 Marshall (w/Garner) 22.02

    Very crude analysis, certainly no conclusions possible.

    Garner has 100 wickets plus with everyone.

    Roberts w/Holding and vice versa two of the worst averages there [[ I am off to the hospital for a minor shoulder surgery tomorrow morning and will be away from base for between 2 and 3 days. While I will have my laptop with me, there is no guarantee that I will either get the network or the shoulders to co-operate. Hence my responses might be delayed. Bear with me. Ananth: ]]

  • Waspsting on October 24, 2012, 11:30 GMT

    @Ananth, Alex, re: stats - I agree with Ananth that bowling average is the neatest of all statistical measurements, and also with Alex that if SR and ER are provided, it adds something on top of that [[ Problem, WS, is that there are major problems in showing three measures for every option: as against one. Think of this article. How could I have shown the S/R and RpO separately for the 50 or so classifications. One reason why I stick to Bowling average. Ananth: ]] re: late order wickets, intuitively, i'd fall back on "what you lose on the roundabouts you gain on the swings" to make sense of it re: the Ws.

    If they're SR is about the same as other great bowlers and they're better at removing the tail, then it follows logically that the others have an edge at removing the top order (compared to Ws).

    In the case of Waqar at any rate, his SR is NOT the same as other greats - its significantly better.

    My guess would be stats on top order vs late order wickets would show that Waqar was at least as good against top order as anyone (maybe even marginally better) and definitely better against late order - accounting for his overall SR superiority

    Wasim - probably slightly less good against top order (relative to say Lillee, Hadlee, Holding etc.)

  • Dr. talha on October 24, 2012, 9:01 GMT

    @Alex."Greenidge's age was 39 when he played Waqar". Greenidge got 366 runs (including a DOUBLE hundred) in a series after the one against Pak that i mentioned.And those runs were scored at an average of 45.75, against a pretty good bowling. Even for three series before the one against Pak he averaged around 45, which is close to his career avg. @Ananth."Peak period is difficult to define". I agree with you. but i guess in waqar's case it was pretty clear.Waqar before 95 (back injury)and after 95. After his first couple of series, he took 180 wickets in 28 consecutive tests(and 156 wickets in 85 ODI's)at an average of UNDER 20 and a SR which was the 2nd best in cricket history. So u need to be lucky & get as many matches as u can when u r at your prime. Barring 1996 Eng tour & 1998 SAF, he was never the same Waqar. Not to forget that he had some serious "issues" with Akram (the skipper)in late 90's, due to which his place in the team, was not certain. Well thats Pak cricket for u!!

  • Alex on October 24, 2012, 3:33 GMT

    @Ananth: I only made an observation on the % of late order wkts for different pairs. Clearly, every wicket matters.

    1. However, it is likely that a good bowler with SR=55 might have SR=66 vs top order batsmen and SR=40 vs lower order. Have you done that exercise? That needs too much data but the other way is relatively easy ... just look at the #deliveries faced per innings by every batsmen. [[ I do not know how many times I am going to tell about the non-availability of ball-data. Henceforth one phrase "NO BALL-DATA" will suffice. The number of deliveries faced by late order batsmen will only say the avge balls faced by late order batsmen per innings. That cannot be linked to the bowlers. One bowler might have bowled 50 balls to a no.9 batsman and gone wicket-less and another bowler might have taken his wicket in his first over. Finally what is the objective??? Ananth: ]] 2. So, if Holding is not getting even 60 balls at the tail on a consistent basis, he loses out on an extra wkt every innings: just 4 more overs, and he would take that wkt. That is a disadvantage of too good an attack ... Murali/Hadlee had it easy there. Even W&W had quality support: at least one of Imran/Qadir/Saqlain/Mushtaq was almost always present with them.

  • dale on October 24, 2012, 1:13 GMT

    @Waspsting - Wasim Akram was clearly the "worker" in both partnerships. Imran obviously had the younger man do the bulk of the bowling and for Waqar to fire on all cylinders, it was better for him to bowl in short spurts. The comments are observations based on the figures presented. Perhaps the word "stock" was misleading - Wasim was quite similar to his fellow lefthanders Davidson and Johnston just a bit more versatile. He was as attacking as any bowler,but he was not a hostile bowler. He had a most unique style, an artistry all his own, coupled with the charm associated with lefthanders. I will say that both he and Waqar benefitted immensely from their partnership !

  • Alex on October 23, 2012, 21:51 GMT

    @Ananth: Thanks. As I expected, R-H, L-M, and D-P top the category of % top order wkts with nearly 67% and do not have a large % of low order wkts (20%, 18%, 20%, respectively). In comparison, W&W have 57% top order and fully 27% bottom order ... incidentally, the figures are 62% & 24% for Imran-Wasim. [[ In the early part of this thread we had some useful discussions on late order wickets. WaspS made some telling comments on the technique of dismissing late order batsmen. I think we should not de-value the late orer wickets. If you notice the three Ws have a fairly high late order wickets %: maybe the reason for their successes. At least 5 Tests in the past year could have gone the other way if late order batsmen had been plucked early. Ananth: ]]

  • Alex on October 23, 2012, 17:01 GMT

    @Ananth:

    1. Avg is an excellent metric for bowlers but your stats on % of top order wkts and % bottom order wkts are also quite important. Your article came out last year, I think, but I am not able to locate it on Cricinfo --- can you pl post its URL? [[ I am not sure which article you are referring to. That information is available here itself. The complete data file is linked below. CLICK HERE. Ananth: ]] 2. I am requesting it because I think your article showed that W&W have a large share of bottom order wkts. The WI bowlers of '76-'92 may have benefited from quality support but lost out on these relatively easy pickings, which would have improved their wkt tally _AND_ average.

    3. In general (SR, RPO) gives much more useful info than just avg.

  • Alex on October 23, 2012, 14:07 GMT

    @Dr Talla:

    1. I said "don't go too much by stats". That is different from ignoring the stats, which you accuse me of. If Lillee's 23.92 avg makes him less of a bowler than Waqar (avg=23.66), no problem ... rate Sanga to be a better batsmen than Viv & Lara as well. [[ As I have already mentioned Bowling average is much better than Batting average. Ananth: ]] 2. After the breakdown in '94, Waqar's pace had dropped but, I think, he still touched 85 mph fairly consistently. His post '94 stats are: 54 tests, 183 wkts, avg=28.13; he averaged an excellent 26 in Pak but got pasted in Ind/NZ/WI/Oz. Lillee/Roberts/Holding also lost pace after their first 3-5 years but had better stats than this after losing their pace.

    3. Greenidge's age was 39 when he played Waqar. To appreciate what age 39 does, just watch the horror show that has been touring the world by the name Sachin Tendulkar for the last year.

  • Sarosh on October 23, 2012, 13:02 GMT

    Applying the “Big scores against” test, we find Scored Against:

    Marshall / Garner – Zero Double hundreds, and 4 150 + scores against them. Marshall/ Roberts – 1 Double and 1 150 + score! (Both Sunny Gavaskar!) Holding/ Roberts – 2 Doubles , 3 150 + scores Croft / Garner - Zero Doubles , 3 150 + scores Lillee/Thommo – Zero Doubles , 5 150 + scores Wasim/Waqar – One Double (Attapatu), 4 150+ scores Donald/Pollock - One Double (Blewett) , 4 150 + scores McGrath/Warne - 5 Doubles , 10 150+ scores.

    Not a single double hundred scored against Marshall/Garner , Croft/Garner , Lillee/Thommo.

    Among the toughest to score big off were the West Indian quicks , Lillee /Thommo , Wasim/Waqar and Donald/Pollock.

    Whereas if the better batsmen got stuck in against McGrath/Warne it was relatively easier to pile on some runs. [[ Could part of the reason be the fact that during the McGrath/Warne era, many Tests were played in India and Australia. Of the 15, how many would be by Indian batsmen. 281+241, 195/180/169/155/152/168 off the top of the head. But not all might be against Mcg/Warne. Ananth: ]]

  • Waspsting on October 23, 2012, 11:56 GMT

    "I have always believed that out of the 10 to 12 greatest fast men this game has produced you can never rate one on top of the other"

    @Dr.talha - i agree with this, wholeheartedly.

    If two guys are on the same "league" (so to speak) statistically, I tend to avoid strong phrasing of opinions (definitely, certainly, beyond doubt, etc.)

    Don't think there's much between any of the 23-20 average brigade. [[ And in my opinion, the Bowling average is the most stable and solid of all emasures. It is a ovely composite of the two bowler characteristics in equal measure. I need there are arguments on this. But I will go with this. Not one bowler in the top-25 is out of place. The Batting average is weak. The high NO % of players like Kallis, Andy Flower, Samaraweera, Steve Waugh, Alan Border, Chanderpaul (all above 15%) distorts the average quite a bit. Mentioned just in passing. Ananth: ]]

  • Waspsting on October 23, 2012, 11:49 GMT

    as for Imran-Wasim - the times they bowled together, Imran was old (but still a fine bowler), medium fast, accurate and swinging the ball.

    If wasim bowled more, its because Imran's stamina was down at that point in his career. That Imran still had a better strike rate than Wasim at his fastest shows Wasim was still a little green on skill (and Imran was very sharp, knew exactly what he was doing, despite drop in bowling quality from his prime)

    I don't see Wasim as the "stock bowling" partner, like say nigglers like Ambrose or Garner. He was basically a blaster out - just not as good at it as guys like Waqar (or Donald)

    Wasim took wickets with unplayable swinging balls, not by wearing batsmen out with a nagging persistence outside off stump

    (that's at the heart of what i'm thinking when i speak of "blasters" and "nigglers")

    Waqar-Steyn-Thommo vs Ambrose-McGrath-Garner ... all great bowlers, but very different in their GENERAL approach to bowling.

    Wasim's more in the 1st type

  • Waspsting on October 23, 2012, 11:40 GMT

    @Alex - you got me to thinking about combos of bowlers. I wonder what works best 1)blaster-blaster 2)niggler-niggler 3)blaster-niggler?

    McGrath, Garner, Ambrose, Statham, Walsh, S. Pollock -all niggler types

    Waqar, Wasim, Thommo, Trueman, Marshall, Donald - blaster types

    Walsh-Ambrose probably the most niggling of the pairs. Holding/Roberts and Lillee/Thommo both pretty on the blasting side.

    No conclusions - just a distinction worth considering.

    @Dale - "Wasim appears to have been the stock bowler in both partnerships but still managed to garner more wickets than his illustrious partners... his strike rate is inferior to both Imran and Waqar although his average is better than Waqar's. Evidently, his ability to bowl long spells and to keep the batsmen under pressure was of immense value to them and aided both to take wickets"

    I'm not interpreting this the way you seem to be. the stats indicate it was Wasim who was lost without Waqar, not the other way round (cont)

  • Waspsting on October 23, 2012, 11:29 GMT

    re: the speed competition in '78... the results were open to interpretation.

    Each bowler bowled 8 balls (6?) and the pace for each was measured. Thomson was comfortably first, but on average, Imran was second - ahead of Holding.

    Holding moved to 2nd because his last two balls were considerably quicker than Imrans.

    That seems fair enough - Holding had an extra gear to Imran at the time, but its well accepted that Imran wasn't at his top pace at the time (his action wasn't even the leaping one we've come to remember him by!). Assuming Holding was at about his quickest, and Imran gained in pace, it seems likely that Imran at his quickest was about as quick as Holding.

    as for bowling on flat tracks - Holding's 76 Oval performance is mind boggling, and he did very well the few times he played in Subcontinent. Holding's a handful - in any conditions

    ---

  • Dr.talha on October 23, 2012, 8:12 GMT

    Waqar & Lillie are two totally different bowlers and cannot be compared. The best comparison of Lillie would perhaps be his own countryman McGrath, or even Hadlee. I have immense respect for all the great fast bowlers this game has produced, whether Aussies,Pakistani or West indians. Their contribution to the game has been outstanding. And they have been the most lovley sight to watch when u are watching the match live in the stadium, or on the TV screen. I would end my argument by the words from the 2012 Olymic champion, "Usain Bolt", the Jamaican sprinter, who said "i started running fast when during my childhood i saw Waqar Younis bowl" Thats how fast Waqar ran to the wicket.

  • Dr.talha on October 23, 2012, 8:10 GMT

    I think we should also consider, that how many tests did a bowler got when he was at his peak. From Dec 1990 to June 1992,"for one and a half year", Waqar got the oppurtunity to play just 2 tests. And this was the peak time of his career,when he was at his most devastating best. Holding and even Marshal were lucky to get so many 5 to 6 test match series during their prime. [[ Peak period is difficult to define. As my Career-split article proved (the three-way split verion yet to be done) the peaks vary from bowler to bowler. If Waqar got to play only two Tests, how can that be defined conclusively to be his pek. Ananth: ]] @Alex. The ball has been quite strictly monitored since the 92 Pak-Eng series. But even after that Pak has been remarkable with the old ball. Infact reverse swing is not just about old ball. Once wasim said that it even includes skills with your wrist and the seam position. Pak bowlers have some unbelievable skills, which has transfered over generations. I remember in 99 Pak-Ind series wasim got the bowl to reverse in the 14th over of the match. What about Umar Gul doing it in T20 as well. He was brilliantly reversing it in 2009 T20 WC.

  • Dr.talha on October 23, 2012, 7:39 GMT

    @Alex. "Don't go too much by stats ... else, you will draw as ridiculous conclusions as Waqar being better than Lillee" [[ I question the use of the word "ridiculuous" especially as Lillee's bowling average at 23.96 is higher than Waqar's 23.56. And I would say that if one ignores stats altogether we will be only left with subjective conclusions, often coloured by personal observations. Ananth: ]] Stats can never be ignored the way u are doing,brother. After all its the stats that make Bradman the best. Nobody of us have seen him bat live. And who say Waqar cant be better than Lillie?? I have always believed that out of the 10 to 12 greatest fast men this game has produced you can never rate one on top of the other. If i am the skipper and i need to defend 100, i may throw the bowl to Lillie,if u play at Perth or MCG, but it would be a huge blunder if i opt for Lillie, playing in subcontinent. I would ceratinly look at Waqar,as my match winner. But we have seen that now even wickets in Aus & WI are not as bowling friendly as it used to be at the time of Holding & Lillie. Perhaps thats the reason why, apart from Steyn u see nobody who will end his career at an avearge of under 25.

  • Dr.talha on October 23, 2012, 7:18 GMT

    @Alex. U can NEVER compare bowling competitions to bowling fast in international matches. I remember in that competition Imran was number 3, and he said once that he was even faster a couple of years later. And i believe,Waqar at his peak was faster than Imran at his peak. Waqar in his peak years got 180 wickets in 28 consecutive tests (and 156 wickets in 85 ODI's) at an unbelievable strike rate. And that included 35 wickets against the best team of that era WI, which included some of the finest players of fast bowling. I never saw Greenidge struggle the way he did in front of Waqar. Lara in an interview said "in my debut test, Hooper was playing the ball in gully and wanted me to run singles,to get of strike, because of the pace at which W's were bowling" Can give u the name of the TV show if u want.

  • Dr. talha on October 23, 2012, 5:11 GMT

    @Alex.Holding just played a single series in the subcontinent. Though i strongly believe he could have repeated his performance.

    I believe SL & Zim were not worst of the batting line ups. Infact there have been periods in cricket when the batting of the major teams have been far worse than the minnows, of the game. Waqar got 11 wickets at candy in '94 and absolutely devastated SL. If u look at the SL line up in that series, it was as good as some of the Eng batting ups,against which Lillie or Holding played. [[ Zimbabwe, other than of recent times, have always been pretty good with the Flower brothers, Campbell and Streak. Ananth: ]]

  • dale on October 22, 2012, 23:51 GMT

    Warne/Macgill: Seems as if Warne brought out the best of MacGill 82 @22.11 but MacGill was much less effective without Warne 126 @33.5. Interesting to note that Warne also was better without MacGill. On the surface it is tempting to say that Australia would have been even more successful had they played together more often - very, high WPT - great combination for way too few tests.Anyway it is surprising that MacGill simply out performed Warne when they played together.Better ave,better SR and almost equal dominance. [[ Like Lara, MacGill retired couple of years too soon. However, unlike Lara, he made the decision himself. He would have been invaluable for Australia for a few Tests. Maybe more at home than away, but who knows what he would have done if he knew that he was the leader of the attack. Ananth: ]]

  • dale on October 22, 2012, 23:33 GMT

    The value of Wasim Akram in both partnerships with Imran and Waqar is even more evident when we examine the percentage of balls delivered and wickets taken. Wasim bowled 25.8% deliveries to Waqar's 21.7% and took 29% of the wickets compared to Waqar's 28.5%. With Imran, Wasim bowled 24.1% balls compared against Imran's 19.8% and took 26.2% of the wickets to Imran's 24.3%. Wasim appears to have been the stock bowler in both partnerships but still managed to garner more wickets than his illustrious partners. Naturally, his strike rate is inferior to both Imran and Waqar although his average is better than Waqar's. Evidently, his ability to bowl long spells and to keep the batsmen under pressure was of immense value to them and aided both to take wickets. It is a testament to his incomparable skills that he still was able to take wickets with such regularity. [[ Is there a single batsman who conquered Wasim? Ananth: ]]

  • Anil on October 22, 2012, 20:17 GMT

    Thanks, Ananth. I agree that we should not dig too much into the numerical aspects, and should instead see the beauty and enjoy. The beauty of your articles, apart from them being of very high quality and product of great efforts, is that they are associated with the writer's response to almost each reader's comments; that's unique that I do not remember seeing elsewhere. I wonder how you manage the workload. Hats off to your passion! I am not very sure of your profession; but you could make a very impressive and inspirational teacher. [[ Thanks, Anil, for the kind words. One reason why I emphasize always that I am neither a Statistician nor a Statistical Analyst but simply a Cricket Analyst. Just to give an example. I may define a new ODI Batting Index which is the Geometric mean of the RpI and S/R. This has no statistical backing but simply intuitive. The higher either value is the higher the index should be. Hence the product. The GM would be to narrow down the values. This requires nothing other than common sense to define and understand. I have taught a lot in IBM, way back before George Fernandes occured. Ananth: ]]

  • Deepanjan Datta on October 22, 2012, 17:26 GMT

    Thanks much for a wonderful analysis as always, Ananth! As you have mentioned - you have taken some liberties going beyond purely stats (and rightly so). One pair I often wanted to see more of was Warne-McGill (surprisingly missing from all user comments and discussions!). Given Australia's present travails it's a shame that McGill had to leave when he did, he'd probably walk into most international teams at his best. May we please have an analysis on their stats? [[ Even though they did not play many Tests together (16 Tests but nearly 10 WpT), I have done the work and uploaded the file because of what they achieved together. Maybe they could have played more often, at least 25 Tests. And because MacGill has out-performed Warne in a big big way. CLICK HERE. Ananth: ]]

  • Alex on October 22, 2012, 17:07 GMT

    @dr talla & @Wasp: It is also important whether the two bowlers bring in differing complementary factors. W&W's physical build was the same and their weapons were almost identical, just that Wasim was more skilled. The big difference was that Wasim was left-arm bowler.

    Here, G-M probably score over R-H, L-T, D-P, McGrath-Gillespie, Trueman-Statham, Ambrose-Walsh, and maybe even W&W. McGrath-Warne was, of course, a truly sensational pair in this regard.

  • Alex on October 22, 2012, 16:45 GMT

    @Dr Talla and @Wasp: Wasp's points are correct.

    1. As for the flat pitches, Holding's avg on India's flat pitches is as low as 22.10; Roberts too averaged barely 22 in the subcontinent (Ind-Pak). And they never bowled to Zim/BD/SL.

    2. Use the speed measurement and not the illusion. In the 1978 fast bowler's competition, after accounting for the speed adjustment, Holding's fastest was 95 mph; only Thommo was faster in that competition, comfortably #1 at 103 mph. Holding's fastest certainly was faster than Waqar's.

    3. Re Crowe's comment on Marshall, Marshall's pace had dropped by '86 itself. His "fastest bowler" mantle went to Patterson in '86 and then to Bishop in '88. Over '89-'91, Bishop probably edged Waqar for it before breaking down in '91 ... Waqar himself broke down in '94 and was never the same again.

    4. Don't go too much by stats ... else, you will draw as ridiculous conclusions as Waqar being better than Lillee.

  • Anil on October 22, 2012, 16:27 GMT

    Dear Ananth, Thank you for your response. I do understand that the ball-by-ball info is very limited, and thus detailed analyses like this are not feasible. However, I do not understand the merit of such extensive analysis--even as I do see statistically significant trend (like, A doing better in the company of B, and vice versa). What is the mechanistic explanation of that trend? If two bowlers took wickets in impressive manner in a match, even when they bowled 20s of overs apart, what do we read about them being a "pair"? Unlike the batting partnership, how does a bowler influence the game of the other? If there is no influence, why do we see a statistically significant trend? Besides, Ambrose offers a very different picture in association with Walsh. Did it mean that Walsh "ate into" the share of wickets Ambrose would anyway have taken? I am sorry that I do not understand the mechanistic relevance of this analysis. [[ Anil., why should every analysis have a mechanical relevance. Consider the analysis' theme as "Contribution to the team cause". The two bowlers bowling together captured 9.xx wickets per match at a combined average of 22.xx and had a winning % of 68 as against another pair ... Do not look for statistical relevance. Look for what they achieved together. Look at the comments of Alex, Talha and Waspsting, especially the last-named. There is a great insight into the West Indian, Australian and Pakistani pace bowling attacks. Let us have some fun. Ananth: ]]

  • Waspsting on October 22, 2012, 12:12 GMT

    @Dr. Talha - thats a strong argument in W's favor.

    Roberts and Holding have edges in bounce, bouncers, seam movement.

    But factor "bounce" is accounted for largely by pitches, factor "bouncers" accounted for by laws of the game.

    Seam movement falls under the broader heading of "sideways movement".

    Its Roberts-Holdings seam/cut vs Wasim/Waqar's swing - and i'd give the Ws an edge there, too.

    re: Waqar's pace - i vaguely recall the series your referring to, and Waqar topping Donald convincingly in it.

    Martin Crowe also named Waqar the fastest he faced, and he faced Marshall.

    Agree with Ananth too. Bouncers make the batsman appear more rushed than yorkers - it follows logically Holding appeared to hurry batsman maybe a bit more than Waqar.

    But the number of times Waqar crashed through the batsmen with pitched up balls - beating them for pace, makes me think he was as quick as they come.

    Not much in it between Waqar and Holding for pace, i think. both of the highest express type

  • Waspsting on October 22, 2012, 11:30 GMT

    against them (as stats confirm), is they're just barely outdoing their support bowling (says as much for strenght of support as it does them) - but how to call them the best pair then?

    It'd be like calling the Waugh brothers the best batting pair - considering the Aus line up that also featured some combo of Taylor, Slater, Hayden, Langer, Martyn, Ponting, Gilchrist

  • Waspsting on October 22, 2012, 11:24 GMT

    (possibly Croft-Garner when playing with R-H have a better partnership record than the lead pair - that's how good they were).

    M-G had many fine bowlers, but nothing quite at that level (esp. when they took to sharing the new ball)

    re: Thomson, were i sympathize with him is his injury was caused by an act of God run in with a fielder. I think this is VERY DIFFERENT from say Imran, Lillee or Bishop - whose injuries presumably came from their actions.

    That said, results are results, and R-H lasted much longer as a killer pair. Probably Thommo wouldn't have last as such a hot property relying on raw pace for that long anyway - and I fully agree that Holding was a better, more versatile bowling.

    (another point against Imran's opinion of Holding is that he thought Holding faster than Thommo - the only person who did, i think. Suggests some kind of personal preference, not very objective)

    I think the case for Roberts-Holding as the top pair rests heavily on longevity. main point (cont)

  • Waspsting on October 22, 2012, 11:17 GMT

    I was thinking of each pair as being the pair in the lead position.

    e.g Marshall/Garner played many matches together before Marshall was particularly good (let alone "great"). it all goes into the stats, but i'm thinking of the pair as the new ball pair. (sort of the way we don't think of Imran of '71 when we think of Imran Khan)

    I agree Holding/Roberts bowled longer than L-T and M-G as the "lead pair", and count it as a point in their favor.

    That said, seeing they were taking the new ball, they should edge the other pairs as strike forces (assuming equal ability, and others not using new ball together as a pair as often)

    They don't. Garner and Marshall have better averages, strike rates and economy rates than both overall (and i'd guess, even better than that with the new ball, particularly average and strike rate)

    Garner-Marshall had good support true - but i think R-H definitely had better. Croft and Garner were in their primes for much of R-H's partnership (cont)

  • Waspsting on October 22, 2012, 11:06 GMT

    @Alex - re: talent - i've heard this opinion from Imran, and frankly, never understood it. What makes a fast bowler "talented"?

    For a batsman, its easy to see - some players can do things that others just can't (e.g Lara, Viv Richards)

    For fast bowlers, you rarely hear people call them "talented" because it seems more like a learnt skill.

    I think Imran was referring to Holding's fitness(?), i've heard him say Holding was so naturally fit (unlike Imran himself and Lillee) that he didn't need to train that hard to stay in shape. Imran thought he could have continued playing test cricket for many more years if he'd worked harder on fitness

    Anyway, "talent" in a fast bowler doesn't mean much to me.

    re: support - Ws did NOT "often" have the bowlers you named. Imran and Qudair were gone very early in Ws partnership - though i agree that their support was reasonably good nonetheless (on top of Saqlain, Mushtaq Ahmed was pretty hot for awhile)

    ---

    Regarding partnership parameters (cont)

  • Sreevatsa on October 22, 2012, 6:33 GMT

    I was wondering how did the english pair of caddick-gough fared? can you please consider this request, hoping that they meet your qualifications. [[ They did quite well. The concerned file is attached. CLICK HERE. Ananth: ]]

  • Dr. talha on October 22, 2012, 5:50 GMT

    From Steve waugh to Brian Lara, and most of the players of that era said that Wasim was the best they ever faced. Wasim could bowl 6 different deliveries in an over, and i can bet that all 6 would beat the batsmen. @Ananth. Thanks for the Ambrose/Bishop figures. I told u that their figures would surprise a lot of people. Terrific bowling pair!! Cant understand how the Windies produced so many great bowling pairs. No wonder they dominated world cricket for so long.

  • Dr. talha on October 22, 2012, 5:44 GMT

    @Alex. "I do not go too much by averages for such comparisons" Its not only the avearge of W's that is better than Holding/Roberts, but even the Strike rate. So: - W's had better average - W's had better strike rates - W's were better in conventional swing - Reverse swing..No comparison!!! - Pace: Debatable - W's played on the most placid wickets - Had horrible fielding support

  • shrikanthk on October 22, 2012, 5:41 GMT

    Example looking at bowling averages, Warne with Mcgrath was better than Warne alone, and McGrath with Warne was better than McGrath alone. This would make a true pair

    This sounds nice. But not too practically useful. Because whether a bowler does well alone or does well with another bowler has got more to do with where they are in their respective careers than it has to do with any chemistry with the partner.

    Eg: Sobers and Gibbs were a better pair than Sobers and Ramadhin. That's purely because Sobers took a while to evolve as a bowler and became a potent weapon only in the early 60s by which time Ramadhin had quit. So a comparison of Sobers-alone versus Sobers with Gibbs tells us very little about their chemistry and more about Sobers' own development as a bowler which simply coincided with Gibbs' best years.

  • Dr. talha on October 22, 2012, 5:34 GMT

    @Alex. I dont know what made u say that Holding was faster then Waqar. This can only be told by a batsman who has faced both at their peak. Lara once said that Waqar was the fastest he faced in early 90's. I remember a tournament in 1993 (total cup)which had WI,SAF & Pak. Waqar showed that he was the fastest.The fastest ball Donald bowled was 144, while waqar constantly touched 150 Km. There may have been faster bowlers than Waqar and bowlers who swung the bowl more, but have u ever seen a bowler with waqar's pace, swinging the ball as much as he did???? [[ I get the feeling that some of Holding's deliveries might have seemed faster in view of the absence of protective gear. Ananth: ]]

  • shrikanthk on October 22, 2012, 5:25 GMT

    Pretty similar numbers for the Laker-Lock and Grimmett-O'Reilly pair, and fantastic numbers at that

    This is where one needs to look beyond the numbers. For all their similarity, there is absolutely no doubt on the superiority of the O'Reilly-Grimmett pair.

    For one thing, Grimmett and O'Reilly plied their trade against a much stronger batting opposition on MUCH FLATTER wickets.

    In contrast Laker and Lock often bowled on doctored dusty pitches against an often middling Aussie batting side.

    Secondly, Grimmett and O'Reilly seldom played when both were at their best. Grimmett was nearing his end when O'Reilly made his breakthrough into ths side. And O'Reilly's best was yet to come when Grimmett's career ended (Forced retirement influenced by the Don).

    Keeping these things in mind, there can be little doubt on how much greater Grimmett/O'Reilly were in comparison with Laker/Lock despite similar figures.

  • shrikanthk on October 22, 2012, 5:18 GMT

    Some thoughts on great test match pairs that might have been -

    1. Richardson and Lockwood : These two were the ORIGINAL pair to terrorize batsmen. Both bowled often in tandem for Surrey in county cricket. However in test matches, Lockwood's best years coincided with Richardson's worst and vice-versa. Hence, the pair never really set the world of cricket alight.

    2. VV Kumar and Chandrasekhar : Kumar was a terrific orthodox leg-spinner with an unbelievable FC record. Sadly because of competition and possibly politics, the guy never managed to play much for India. It would've been nice to see him pair up with the unorthodox Chandra.

    3. Anderson and Flintoff : Sadly Flintoff nearly faded from the scene before Anderson discovered his true potential. It would've been fun to see these two pair up in the mid 2000s. Imagine Anderson's conventional swing wreak havoc in combination with Flintoff's mix of hit-the-deck bowling and occasional reverse swing. [[ I would have liked a quality pace bowler like Kapil Dev to have bowled with the spinning trio of the 1960s. Ananth: ]]

  • shrikanthk on October 22, 2012, 5:12 GMT

    I am sure you must be aware that Lohmann/Barnes/Turner/Peel/Briggs/Blythe occupy the top six positions in the Bowling average table.

    This is a very diverse collection of bowlers. Lohman/Turner/Peel/Briggs are products of the 1880s - the pre-Golden Age era when cricketing techniques against overarm bowling were still evolving.

    Barnes and Blythe are products of the late-Golden Age from 1900s-1910s when cricket was more or less modern, pitches a lot firmer and batting a lot easier.

    One cannot club these two with those relics of the 1880s. [[ Shri At times you read too much into statements. I had only mentioned that the top-six positions in the Bowling average table was occupied by these Pre-WW1 tables. THAT IS ALL. This was in response to someone who mentioned that I should go only by numbers. Where do we get 1900-1910 and 1880s periods come in. Ananth: ]]

  • shrikanthk on October 22, 2012, 5:03 GMT

    Unless the wickets taking spin (usually it isn't), even good spinners are "stock bowling", basically giving the fast men a rest while keeping a check on the scoring

    Most wickets do take spin contrary to popular perception. Be it a test match at Durban, Edgbaston or Kolkata, pitches offer turn when you make the game last till the 5th day. This is precisely the reason why test matches are played over 5 days with 2 innings per side. Because the format brings spinners into the game as an offensive weapon. Votaries of 3-day/4-day cricket ought to bear this in mind.

    Also when we discuss great spinners, we reflexively mention only Warne, Murali, O'Reilly and Grimmett. But there are several others who are not far behind. Especially Verity, Underwood, Kumble and Laker. Underwood was not at all the wet-wicket wonder as he is often painted.

    The other great utility of spinners is that you get more overs out of them per match, and also their peaks last longer than the best fast men.

  • Anil on October 22, 2012, 2:39 GMT

    Great work, Ananth, as always--even as your choice of players/pairs is more imotional/nostalgia than science. Statustical analysis is science, and nostalgia should not creep in. Besides, I do not understand how the pair's performance is noteworthy, as analyzed here, if we do not consider their bowling in tandem? [[ 1. There is only one selection which is nostalgic. Chandra & Bedi over Kumble & Harbhajan. I think even the most die-hard of today's cricket followers would find it difficult to disagree with this selection because the spin trio/quartet represented India's first match-winning bowling attacks. Mankad/Gupte were lone exceptions. 2. If I present everything based on numbers, I would lose credibility. The bowling analyses will have Lohmann/Briggs and Barnes/Rhodes snd Spofforth/Turner at the top. Their figures were far superior. I am sure you must be aware that Lohmann/Barnes/Turner/Peel/Briggs/Blythe occupy the top six positions in the Bowling average table. 3. I have talked about the ball-by-ball data about 100 times. It is available only for the last few years and that too not freely. How do we know whether Lindwall & Miller bowled together or Miller & Johnston. 4. Bowling in tandem is a mirage and really does not mean anything. The bowlers have to do well, within an innings, when bowling together or separately. Ananth: ]]

  • rajwani on October 22, 2012, 1:43 GMT

    well...all i want to say is all above mentioned bowlers must have been super talented and champions in their own right...they hav won many matches for their sides which is important... i observed sumone here saying that 2w's depended on reverse swing,as a matter of fact,this shows the real depth in their fast bowling as they were grown up on those pitches like dustbowls..westindies australia southafrica had fast bouncy pitches hence they developed bowling that channel on off stump whereas wasim and waqar didnot hav that luxury of helpful pitches and to add to that a horrible fielding side..hence they also made sure of targetting the stumps and pads rather than lukin for edges... dunt wana go too much into numbers here...wasim and waqar got many wickets all around the world due to their skill...only marshal can be sed at par...

  • Alex on October 21, 2012, 15:52 GMT

    @Wasp & @Ananth: Imran himself rates Holding as the most talented bowler he ever saw.

    1. W&W (& Shoaib Akhtar) surpass almost everyone on yorkers & reverse swing.

    2. Mechanism to create reverse swing was viewed suspiciously way into 00's: I do not hold it against W&W but I prefer clean things.

    3. As Ananth said, Marshall-Garner (M-G) might be better than Roberts-Holding (R-H). But, R-H was a truer "pair" since they often shared the new ball 1976-82 ... M-G did that over '84-'86 only.

    4. M-G had phenomenal support: Holding/Walsh/Patterson/Gray. W&W, too, often had Imran/Qadir/Saqlain.

    Donald-Pollock was a bit like M-G. Lillee-Thommo was a mirror image of R-H ... but, effectively, lasted only 3 series before Thommo broke down. Here, Lillee was a bit better than Roberts while Holding was better & more versatile than Thommo. Likewise, Ambrose-Bishop combo was terrific but, effectively, lasted only 3 yrs since Bishop broke down at age 24!!

  • dr. amir on October 21, 2012, 14:23 GMT

    @ Alex . when i placed 2 w's ahead of the rest , i did it keeping in mind the nature of the wickets the mostly bowled at ( which bowlers from the rest of the world used to term as graveyards for fast bowling) , the little support from the rest of the bowling attack, the pathetic fielding , and usually the let down batting displays. also the rule changes for bouncers and the protection gears for the batsmen. the quality of batsmen they come up against and lastly their stats might suggest that they were more lethal to the lower order batsmen but anyone who saw them bowling acknowledge that even top order batsmen were not spared. lastly as stats suggest 10 times they took 15 wickets in a match together sums it up all. again i would like to thank Ananth for the hard work he did in providing us these stats. this article is a master piece.

  • Ananth on October 21, 2012, 13:19 GMT

    I have completed the following work and uploaded the Excel sheet. 1. Incorporated the data for all the 30 pairs selected in the Excel sheet. 2. Created a new Excel sub-sheet containing the balls bowled by the team, pair and bowlers and presented the % as well as strike rates. Lot of additional information has been presented. To download the complete table, the widest Excel sheet I have seen for a long time (75 columns!), please CLICK HERE. Ananth

  • Waspsting on October 21, 2012, 11:19 GMT

    "Roberts was arguably as fast & versatile as Wasim while Holding was certainly faster & better than Waqar. That settles the matter for me. "

    agree about Wasim and Roberts (and believe that much of Wasim's standing is hyped up, hence he'd probably usually get the edge in most people's comparison of these two)

    Holding faster than Waqar probably true, don't know about the better part. where Waqar has Holding is his ability to move the ball sharply. (Holding has Waqar in other areas too, of course)

    Holding was more a keep the seam up and be around off stump type - like Donald. with a lot of brilliant short stuff thrown in.

    No starting a foot outside off stump and hitting leg type stuff from Holding

    Very different types of bowlers - and both great - but i don't see a clear case for either being "certainly" better than the other. [[ The fact that there were four pace bowlers operating together leaving no breathing space for any batsman has to be considered. Lack of balance did not mean that limbs were not broken, especially with no protection gear. Ananth: ]]

  • Waspsting on October 21, 2012, 11:07 GMT

    but that got curtailed by Thommo's unfortunate accident (which i think is categorically different from an injury incurred due to playing - like Bishop's), though its doubtful Thommo could have stayed that hot for 6-7 years

    Marshall-Garner i think edge Roberts-Holding based on having less support.

    would be very glad to discuss this with you

  • Waspsting on October 21, 2012, 11:04 GMT

    True, but this is true for all the pairs other than Wasim-Waqar. I take your statement to be an acknowledgement that 2 Ws have a strong (cont) case for being #1? (just noticed your comment was a reply to Dr. Amir - explains it)

    Each pair had their own method, and i don't think placing one method over another has any merit. They had their reverse swing, Those guys had their bouncers (note that 2W's played mostly after limitations were placed on bouncers)

    I don't have any definite opinion on whose the best of the various pairs - and have found discussing matters with you is always interesting

    IOW this isn't an ego thing of trying to attack what you say - i'm keen to just explore the matter and see what comes out - but that involves questioning opinions. (want to be clear this isn't an attack, but a query)

    Why do you think Roberts-Holding > than say Lillee-Thomson or Donald-Pollock, as well as 2Ws?

    My gut feeling is that Lillee/Thommo were a hotter pair (cont)

  • Waspsting on October 21, 2012, 10:56 GMT

    @Alex - I'd be glad of a chance to talk to you about the various bowling pairs you've named (and I agree without undue stress to numbers works best in such a discussion).

    The difficulties i see with your argument in favor of Roberts-Holding ("Croft & Garner arrived in 1977 to claim a good share of wkts") is you aren't taking into account how Croft/Garner affected Roberts/Holding pair(?)

    Just as Roberts makes Holding more effective (and vice versa) - the heart of a good partnership (as opposed to two single bowlers), it seems to me that Garner-Croft might have made Roberts-Holding better?

    Statistically, since Garner-Croft came into the side, NOT ONCE did Roberts top the WI attack for a series in wickets, average or strike rate. (the other 3 have various tops in these fields)

    "Also, Roberts-Holding never needed reverse swing to take wkts."

    True, but this is true for all the pairs other than Wasim-Waqar. I take your statement to be an acknowledgement that 2 Ws have a strong (cont) [[ I also think the argument on W's lacking pace is facetious. The skill factor of the W's is awesome. Ananth: ]]

  • Waspsting on October 21, 2012, 10:35 GMT

    re: spin combos - I agree with both Dale that spin combos work best in tandem and Arjun on the limits of that idea as an over-generalized rule.

    Unless the wickets taking spin (usually it isn't), even good spinners are "stock bowling", basically giving the fast men a rest while keeping a check on the scoring (as opposed to taking wickets or threatening too)

    This is true even for "great" spinners - only guys like Murali and Warne (amongst those I've seen), possibly O'Reilly, maybe Grimmett - are exceptions, and wicket taking threats in most conditions.

    In a world 11, I pick both Murali and Warne, on the grounds that -if wickets turning, they become extra lethal

    -it its not, they're still pretty threatening (as in those two guys are this - can't just replace them with Laker and Qudair and keep the same dynamic)

    - with 3 paceman in the side as well, 2 spinners aren't detracting from the pace attack (benefits and all, it wouldn't be worth having the two if it were dampening pace power [[ If the third pace bowler is an all-rounder like Imran, 5 bowler-attack is a luxury the captains can afford. Anyhow the Best Test XV is an article not too far off in the future. Ananth: ]]

  • Waspsting on October 21, 2012, 10:27 GMT

    @Dale - re: bowling at tailenders

    The standard method of quality fast bowling is outside offstump looking for the outside edge (and compelling batsmen to play at those balls with the threat of in movement)

    This is often wasted on tail enders who simply miss such balls.

    It figures that Wasim, Imran, Waqar are better than most at removing the tail because they differ from the norm in their approach. Generally speaking, all 3 (esp. waqar) aimed to hit the stumps or the pads in front of the stumps. [[ It requires a special skill to get the late order without too much batting contributions from them. Especially with feisty late order like Harbhajan, Swann, Kulasekara et al who tend to attack than defend, the late order batsmen cannot be taken lightly. Ananth: ]] Some of the best fast bowlers ever were poor at removing the tail due to their inability to do this - Lillee, Marshall amongst them. They specialized in standard out-movement outside off stump methods

    And note how the best bowler today, Steyn, tends to bounce out tailenders rather than go with his outswingers to them.

    As for Qudair... leg spinners are always good against the tail, due to the confusing variety they usually have. [[ And Warne. Ananth: ]]

  • Mochu on October 21, 2012, 3:01 GMT

    Wow, what an in-depth analysis! Excellent job. I do have an un-related request...you may have already looked at this in the past. I have always been interested in learning who the true match winners are, statistically speaking. For example, a batsman's contribution in a test victory being weighted with respect to where (home/away) and who (test ranking at that time) and frequency of such contributions. Likewise for bowlers. From a statistical perspective, is Kumble a bigger match-winner compared to Tendulkar? Can one arrive at some such analysis or am I dreaming?

    Thanks again for starting such a wonderful discussion and debate across the world of cricket enthusiasts...it appears like the interest in test cricket is still alive! [[ Yes, it is possible once my Innings ratings work and Bowling Spell ratings work is completed. Since I always try for equivalence in Ratings work, it would be possible to measure the contributions. It would, however, be a slightly longer term project. Ananth: ]]

  • jitendra on October 20, 2012, 10:06 GMT

    I want to know about: Mcgrath-gilespee and warne-mcgill combination.Also Crains-vitori [[ I have only done Cairns-Vettori and uploaded the same. Wrne & MacGill have captured 70 wkts together and McGrath & Gillespie is not the premier combination. Warne was the leading one. CLICK HERE. Ananth: ]]

  • Charlie on October 20, 2012, 9:32 GMT

    Dear Ananth, firstly thank you for your masterful analyses,they always seem to shake out the outstanding from the merely great. For one who has watched cricket for over 70 years, from Lindwall & Miller to Steyn & Morkel, the all-time stars are the two Ws (especially Waqar) and Shane Warne. These are the only bowlers I have watched who one expected to get a wicket with every ball. Wagar completed the only clean-bowled hat-trick I ever saw, albeit in an ODI (the 2003 World Cup: Pak vs NZ). For me their best qualities were that they were compelling to watch, there never was a dull moment when they were performing. [[ I have watches Test criclet over 40+ years. To see that you have watched from Lindwall/Miller/Toshack onwards makes me feel like a novice. All the best sir. May your tribe flourish. Thank you. You have made my month. Ananth: ]]

  • Gerry_the_Merry on October 20, 2012, 6:50 GMT

    Hi Ananth, just saw this article, and I must read the article in detail. But could not find Ambrose and Bishop as a pair here. They would have taken around 180 wickets as a pair, so less than your cut-off. But in terms of simple averages in a 2005 article by S Rajesh, I remember them to be the top bowling pair by averages in history, and a couple of points clear of the next best. So wonder if your approach will validate this claim. You have put only the top 10 pairs in teh main XL file, so would be nice if you posted a consolidated XL file as the links to different files seem to be in different parts of the reader comments. In 1990, these two fairly trampled over each batting line up, and their numbers will be very very good. [[ No, they did a lot of damage together. In fact over 300 wickets. But not at that great an average: only around 22 (only ???). Bishop played only 4 Tests without Ambrose. CLICK HERE. I have onl;y the top-10 aand one selected pair being processed. Let me see whether I can make the XL a consolidated file. Program modifications required. Ananth: ]]

  • dale on October 20, 2012, 3:20 GMT

    Thanks for the "Waqar"correction - it is interesting that like Akram/Waqar, the Imran/Qadir combo also snared a high percentage (over 24) of late order wickets. I would imagine that in addition to Qadir's attacking mixture of deliveries,Imran's reverse swing (inswingers) and bowling directly at the stumps had a lot to do with this. Note on Australia:due to their rapid accumulation of runs from the first inning of batting, they pressured the opposition and also gave their bowlers more time to get at the batsmen. Plus,in addition to the immaculate skills of Warne,his presence ensured a fast over rate. By contrast, WI by having the four fast bowlers ,were always struggling to maintain the over rate. Finally, although WI had great batsmen, they never went after the bowling in the team's first inning and as a result never really dictated the pace of the game from the onset.

  • Alex on October 20, 2012, 3:13 GMT

    @dr amir: Holding-Roberts circa 1976-1982 were certainly more devastating than W&W. IMO, this was the most devastating pair over 5-6 year long period --- better than McGrath-Gillespie, Donald-Pollock, Lindwall-Miller, and Ambrose-Walsh. Their numbers do not seem that impressive only because (1) Packer series cut this sizable 7-year period down by more than 2 years and (2) Croft & Garner arrived in 1977 to claim a good share of wkts. Also, Roberts-Holding never needed reverse swing to take wkts.

    In general, I do not go too much by averages for such comparisons since the conditions & opponents can be significantly different. Roberts was arguably as fast & versatile as Wasim while Holding was certainly faster & better than Waqar. That settles the matter for me.

  • Ananth on October 20, 2012, 2:01 GMT

    Message to Mitul Gogri/Mitz4u (or whatever you are) Unfortunately this is the only way I can send you a message. I do not want to expose my mailid. You have persisted in sending obnoxious and malicious mails completely unrelated to the article, implying that I pull down a great player, who certainly does not need support of your ilk. Despite my requests you have not stopped this. And you have done so in this article also. So there is only one thing to be done. Your mails, the names, mailid and the computer reference will all be banned from this workspace. I will take steps with the publishing platform to junk these comments at the source itself. If I still receive your mails, I will trash those irrespective of the contents. This is a life ban and even when I do an article on that wonderful player, you will not have the opportunity to send any comments. You have lost that right irrevocably. My apologies to the other readers. Ananth

  • dr amir on October 19, 2012, 21:18 GMT

    great work and amazing stats. i would place wasim-waqar pair way ahead of the other not merely for the stats but also keeping in mind the subcontinental dead decks and the quality of batsmen they come up against.Pakistan's batting was vulnerable even in those days despite some top quality batsmen in their ranks and Pakistan's fielding has always been pathetic. i have bben following cricket since 80's but never had the same pleasant feelings of seeing those two masters destroying batting line ups with their lethal bowling. thanks Ananth for the hard work you put in this one. i loved it every bit. [[ I have already mentioned this. But all things considered, the W-twins would get my, and many other, votes as the most balanced, effective and devastating bowling pair ever. Ananth: ]]

  • Al on October 19, 2012, 21:03 GMT

    I really like your style. Also like the fact that you are not like a computer churning out stats, but use some cricketing logic as well.

    My biggest grouse is that when we are talking about pairs, we've to only consider the wickets/overs when the bowled in tandem. This stat maybe a bit difficult to obtain. But without that it fuzzes out the stats - we don't know how many overs Warne bowled opposite Lee or Gillespie or McGrath. [[ I have already completed the analysis on balls bowled by the pair, individual bowlers, strike rates, in relation to team balls etc. The problem is that it is very difficult to integrate this into the main article. So I am going to create a new Excel sheet and incotrporate it into the linked Excel document. You would have to download and see. By today evening, perhaps. Ananth: ]]

  • Gaurav Kapoor on October 19, 2012, 19:57 GMT

    Excellent work.

    Two things to consider:

    1. % wins have increased in modern era due to ODI influence so we can see more matches finishing to a result which favour Warne/McGrath. Can we factor the % win with overall % win of that era/decade and compare apples to apples? Still believe WI pairs of 1980s were more lethal. [[ My feeling is that the findings should not be diluted or otherwise in relation to the epecific era values. First this cannot be any predetermined era values. A bowling pair might have spanned two eras. It has to be a truly peer value: like averages. So it has to be highly customized. Probably not worth it. Second point is that Australia made their own success figures. They defied the odd, scored faster and did not allow the other teams to think of saving the Tests. Not every team thought of in this manner. So they deserved the credit. Let us also not forget that Ananth: ]] 2. Include innings where both bowlers bowled atleast 10 overs each or more to truly filter the "hunting in pairs" ... maybe too much to ask, sorry :) [[ No problems, but what do we do about 5.1-3-11-3 6.2-2-13-4 7.0-2-19-2 All Warne specials. Ananth: ]]

  • atul sondhi on October 19, 2012, 13:51 GMT

    great great great piece. Remarkable analysis [[ Thank you. A few more words would be nice. Even shortcomings. Ananth: ]]

  • Arjun on October 19, 2012, 11:55 GMT

    Ananth,

    "Spin bowling pairs(unlike fast bowling pairs) are heavily depended on each other". Not entirely true. I think selection on spin bowlers(more than one) in playing eleven is heavily depended on pitch conditions. If a team selects 2 spinners in the sqaud then pitch must be spin-friendly or opposition must be poor players of spin. That is why figures of spinners in pair are better than individually. [[ Yes, when the selection often is 4+0, normally 3+1, sometimes 2+2, rarely 1+3 and probably fewer than 10 times 4+0, the spinners do not normally enjoy the luxury of a top class spinning compatriot bowling at the other end. Ananth: ]]

  • Dr. talha on October 19, 2012, 10:08 GMT

    Ananth i guess you did the right thing by not considering those pairs who are currently playing. I think these kind of analysis should be done after a pair retires. What about Imran/Sarfaraz and Ambrose/Bishop?? I am sure Ambrose/Bishop will surprise a lot of people.

  • Uday on October 19, 2012, 10:05 GMT

    Pretty similar numbers for the Laker-Lock and Grimmett-O'Reilly pair, and fantastic numbers at that. Its interesting that in each of the top three spin pairs analyzed here (Bedi-Chandra being the third), the bowlers depended very heavily on each other and did not do nearly as well when the other was not playing. Wonder if this means that two spinners generally do much better in tandem than a pace-pace or pace-spin combo?

    However, the numbers for Bedi and Chandra are surprising, given that even if one or the other were not in the side, Prasanna and Venkat would have provided another really good spin partner [[ Tht means any two out o three, Chandra/Bedi/Pras might produce similar figures. Venkat was outside this trio. Ananth: ]]

  • Sreevatsa on October 19, 2012, 7:00 GMT

    Fantastic numbers for grimmet-o'reilly pair. over 11 wkts per test, excellent average for both home and away; these look like numbers for a fast bowling attack ( probably true with o'reilly's mindset to bowl as fast as he could). feeling bitter about the aussie mandarins and about the same for the Don. [[ Never realized that since I did the table in a hurry and uploaded. Over 160 wickets in 15 Tests together. That is something. Also see how much they drop off when they bowl separtely. Ananth: ]]

  • Meenakshi on October 19, 2012, 4:23 GMT

    Ananth, great, superb job on the bowlers!! Could you do a similar one for batsmen as well? Openers and big combinations like Sachin - Dravid or Dravid - Laxman (just examples). [[ Probably not worth it. I have already replied to this request. Pl go to the earlier responses. Ananth: ]]

  • Alex on October 19, 2012, 1:54 GMT

    @Ananth: This is one of your best and workmanlike articles. The text files contain a wealth of useful information. Congrats and thanks! [[ Don't forget to download the Excel sheet with 75 columns!!! Ananth: ]]

  • Madhu Iyengar on October 18, 2012, 20:48 GMT

    This is a nice article, and have to commend the amount of work to compile this.

    I just want to make a simple comment. There are some bowling pairs in this list, where bowler A complimented B (and vice versa). Example looking at bowling averages, Warne with Mcgrath was better than Warne alone, and McGrath with Warne was better than McGrath alone. This would make a true pair.

    However, take the case of Miller and Lindwall. Miller was much better bowling alone, compared to when he bowled with Lindwall. But, Lindwall did better with Miller.

    There's another case, Marshal vs Garner. Their ave didn't move much, though Marshal's was slightly better if he bowled by himself.

    Another, with Chandra and Bedi. They heavily depended on each other to do better. Their averages bowling by themselves, were pretty bad to say the least :-). [[ Yes, there are some pairs here who worked genuinely together as pairs and one was lost without the other. There are others who would have worked with any other good bowler. But the analysis is not "who jelled with who". It is rather which pair served its country very effectively and then on... Ananth: ]] There's a few more of these, but I'll stop here.

    Cheers.

  • dale on October 18, 2012, 19:57 GMT

    Re - Akram/Waquar :I think there are two main factors which contributed to their great success against the lower order. 1- Their incomparable skill bowling with the old ball and 2- In the case of Waquar,he bowled more directly at the stumps than most bowlers even when he was swinging the ball. They both were very intelligent bowlers who knew how to get tailenders out - especially Waquar. [[ When you play Scrabble you can only have words in which 'q' and 'u' go together. If proper names are allowed Waqar would be an excellent choice. You have given Waqar the Scrabble spelling!!! But points are well-made. Ananth: ]]

  • Jonathan Ellis on October 18, 2012, 17:58 GMT

    In the Imran/Qadir stats I see a very, very interesting discrepancy: Imran was a significantly better bowler with Qadir in the side, but Qadir was far better *without* Imran in the side. Does this say anything about Imran's ability to handle spinners, or more about the fact that Pakistan at home were more likely to prepare wickets that favoured pace if Imran was present, but spin if he wasn't?

    Also, did Imran and Akram work as a pair enough to be considered? [[ I think in all our discussions on Pakistani bowlers we tend to underestimate the importance of Abdul Qadir. Imran & Wasim did enough (both over 100 wkts) for me to get the summary and upload. CLICK HERE.

    Ananth: ]]

  • Ananth on October 18, 2012, 17:10 GMT

    WaspSting: Gregory & McDonald have only 86 wkts between them. Similarly Adcock & Heine have only 102 wickets between them. We have to maintain some levels. Hence I have not given these two. Click at link to view/download. Roberts-Holding: HERE. Laker-Lock: HERE. Grimmett-O'Reilly: HERE. Hall-Griffith: HERE. Ananth

  • dale on October 18, 2012, 16:31 GMT

    Could you post the Sobers/Gibbs figures ? I believe they bowled long spells together plus it would be interesting to see how they compare to Hall/Griffiths. Two stalwarts of South African cricket, Tayfied and Adcock were also a formidable pair and should make an interesting study. Re:Marshall/Garner - it is remarkable that their WPT 8.94 is the same at home or away. It is also significant that their support was also very consistent in their performances. Of all the great WI bowlers from their era ,Garner may not have been fully appreciated even though only Marshall can be securely rated above him. [[ Sobers-Gibbs uploaded. Click at link to view/download. Sobers-Gibbs:HERE. Ananth ]]

  • Waspsting on October 18, 2012, 11:50 GMT

    Roberts-Holding Lock-Laker O'Reilly-Grimmett Gregory-MacDonald Hall-Griffith Heine-Adcock

    could you put up the figures for the above combos please?

    I'm reminded of a most amusing comment from Mike Atherton in '96 regarding bowling pairs.

    India and then Pakistan were touring Eng that year, and after Eng won the series against India (and before the Pak series got underway), he said that Prasad-Srinath don't suffer in comparison to Waqar-Wasim and one of the Ind batsmen was of a higher calibre than any of the Pak players. [[ Will do by morning. Ananth: ]]

  • Waspsting on October 18, 2012, 11:42 GMT

    "The graph runs from 50 onwards to 20"

    ok. The second bar looks about 1/6 of the top one - and the numbers are 31 and 47. confused me.

    "for Donald and Pollock - Pollock was past his best when Donald left - don't know if its Donald's presence that accounts for the better record w/ or that..."

    The full thought was an attempt to explain some of the changes we see in the stats in matches with vs without partner.

    One way of looking at it is having the other guys around made the partner more effective.

    Taking X to bowler 1's "quality", Y to be bowler 2's "quality" and Xp and Yp to be the bowlers respective performances (statistically)- this way of looking at it suggests that keeping X constant but depending on if Y is present or absent will affect Xp.

    With Pollock and Donald, I think maybe the quality wasn't constant?

    In other words, by the time Donald left, Pollock wasn't as good a bowler as he had been when Donald was around. Hence the drop in stats, NOT Donald's support per se(?) [[ Probably applicable to all pairs. Wasim/Waqar a perfect example. Maybe the fact that the other top bowler is there is THE support factor. Ananth: ]]

  • Boll on October 18, 2012, 9:26 GMT

    @Waspsting. Yep, very fair. In my earlier comment re.bowling partnerships (as opposed to pairs) I was only trying to point out that almost any Sri Lankan bowler of Murali`s era would have bowled a lot(percentage wise) in partnership with him, simply because he bowled very lengthy spells from one end. I would be interested to know what percentage of overs he did bowl. 37.5% is my `snatch a figure from the air` guess. [[ Let me see whether I can do that work and post. "% of Team balls bowled". Ananth: ]] Vaas was clearly Sl`s second-most senior bowler for a long period and, while a bit outclassed by many of the names mentioned here, was an excellent international bowler for a long period of time. Loved watching him bowl as well.

  • dale on October 18, 2012, 0:23 GMT

    Thanks for featuring Bedi and Chandra. They were both great bowlers. As the only spin duo featured, I am impressed with their averages and their outstanding WPT of 8.76 along with the top order wickets which they snared (60.9%). Both measurements are on good terms with the other tandems. Could you post the stats for Ramadhin /Valentine ? Thanks again. [[ Ramadhin-Valentine: CLICK HERE. Almost identical figures from the spin-twins. Ananth: ]]

  • dale on October 17, 2012, 18:45 GMT

    Ananth, I believe this is perhaps your most comprehensive and detailed presentation. Great work as usual! Australia possessed one of the most balanced bowling attacks during the Lindwall/Miller period. They consistently played 5 frontline bowlers and this I think explains the low WPT for Lindwall/Miller. We are aware of the immense value of the great Bill Johnston but Australia also paraded the likes of Ian Johnson and to a lesser degree one or sometimes two of the following, a young Richie Benaud, Ron Archer,Doug Ring,the very effective Ernie Toshack,Colin McCool and a young Alan Davidson. I believe Keith Miller's all round ability plus his batting at either 4 or 5 in the order afforded them this luxury. [[ I feel that Linwall & Miller were as influential a pair as, say, Lillee & Thomson. They were primarrily responsible for the post-WW2 successes of Australia. Australia's problems started when their skills waned during mid-50s. Ananth: ]]

  • Tom on October 17, 2012, 16:57 GMT

    Ananth, I think this is the first bowling article I've seen from you where you've not mentioned SF Barnes once. I suppose he probably did not play enough Tests with another given bowler to get any meaningful statistics. Maybe Barnes and Rhodes?

    That aside, this was as interesting as ever. So many times I idly consider abstract statistical analyses that I know I have neither the time nor expertise to concoct, and find that you've already done them. [[ Excellent observation. I looked at Barnes only casually since I had an intuition that he was a loner. Now I did some serious pair work, feeling somewhat guilty by your sound statements. Barnes & Blythe had a completely overlapping career. But Blythe played between 1901 and 1910 during which time Barnes played very little. When Blythe retired in 1910, Barnes had played only 12 Tests and captured 67 wkts. Together they captured only 49 wickets (28 & 21). Not worth looking at. Rhodes & Barnes fared better. They played together in 24 matches. But the most lop-sided partnership one can think of. They captured 203 wkts but split as Rhodes 33 @ 32.21 and Barnes 170 @ 16.37. Partnership dominance 4-19. My intuition was correct. But you were also justified. Ananth: ]]

  • Dr. talha on October 17, 2012, 13:59 GMT

    I agree that Wasim got more effective when Waqar joined him. I have always believed when waqar got a back injury in 1995,that was the beginning of the downward slide of Pak. Pak won 13 of their 21 series from 1985-1995, and lost only 2. Before the emergence of waqar,Imran won the test matches for Pak. But after 95 Pak started losing more frequently, as Waqar was never the same bowler (barring 1996 Eng & 1998 SAF tour).Though from then onwards he missed a lot of test matches as well.

  • Vyasa on October 17, 2012, 13:55 GMT

    Thanks, Ananth. I'll rephrase. Is it possible to extend this pairwise interactions to about 50+ pairs? The top 50/100 pairs in history. That way, we could see the averaged out effect of two bowlers bowling in tandem. That would put to rest the debate about the merit of having illustrious team mates- whether it helps a bowler or eats into his tally. Can a general sense of numbers be got and then the analysis redone to see how specific cases work out?

    Vyasa. [[ This was done about four years back during Sep 2008. Ananth: ]]

  • Dr. talha on October 17, 2012, 13:52 GMT

    Agree with the readers that W's were probably the best. ne more factor adds to their brilliance..they never relied on the fielders. Pak has been a horrible fielding side. Can never forget 1995 Gabba test when Pak dropped at least 10 catches in a single inns. And there is a whole list of such performances. Ananth if u do an analysis of the highest number of bowled & LBW, 2W's will be again be at the top. Its great that you have mentioned the away performance. As expected Murali/Vaas,Bedi/Chandra & Botham/Willis has a dip in their performance. But surprised to see a dip even in Donald/Pollock efforts. [[ Why it is quite understandanle. They certainly got good assistance at home. Ananth: ]] I think you should have included the strike rates in the tables. [[ The average is composed of strike rate and rpb. As such it is a reflection of the strike rates indirectly. Already there is so much data that I felt completely swamped. Ananth: ]]

  • grant on October 17, 2012, 13:47 GMT

    Their best match performance has been an unprecedented 19 wickets (11 and 8) in the same Test. This is the only instance of one of these pairs capturing 19 wickets in a match.

    I don't know if I'm misunderstanding this comment. [[ The joke is on me. I never realized that I had had a typo. My apologies. Has since been corrected to 10 & 9. Thanks. Ananth: ]]

  • Nitin Gautam on October 17, 2012, 12:35 GMT

    "I feel Srinath left a year or two too soon." Cant say this for sure. I am a huge fan of his relentless pursuit of excellence during his heydays but during 02-04, his retirements & comeback episodes plus switching of formats (once it was ODIs only & later it was tests only) didnt help his cause rather it sort of ruined his chances. Barring WC-03 he lukd jaded & disinterested. That was the time India had a rich pool of promising pacers with zaheer had firmly found his feets & Nehra coming into his own with Agarkar & Pathan, Balaji, RP singh all knocking on doors (all seemed pretty well & gud future prospects)with lesser pacers like yohanan & Iqbal siddiqi were also in frame. I remember he had a chance of playing in some english county in probably 2004 due to which he made himself unavailable for Aus tour in 03-04 where Pathan,Balaji emerged heroes which continued in Pak tour too thus shutting all doors for Srinath. IMO It was a doom that he brought on himself by retiring & coming back [[ I agree that Srinath was average in the last 5 Tests. Until then he was quite effective. All his last 8 Tests were against West Indies, 5 away. But I was in touch with the Indian team at that time and I still think he should at least have gone to New Zealand. Ananth: ]]

  • Meer on October 17, 2012, 12:31 GMT

    Fantastic job Ananth. Cannot help but remember the two Ws with a sense of wonder. With them there were stretches where they completely dominated batting attacks, and batsmen just had to wait for periods when they could score with relief. And even as the match progressed there was a sense of expectation as you hoped to see that late movement signaling reverse swing as a factor. Glorious in their prime, with a fire that I just never found in other test bowlers. Maybe the West Indians, but watching the metronomical accuracy of McGrath just was not the same as the explosiveness of Wasim and Waqar. Thats all I have for now, going to go through the article in a lot more detail to see what i can glean from these great tables. Was just wondering if theres some measure that could detail, as in Laras article, the "form' of a bowler, considering Wasim always made a point how rhythm is so important to a fast bowler. Maybe consecutive innings with only one or less wicket, or a particularly poor ave [[ Tricky bit of analysis but a valid request. Also it should come in a nowler-specific article and not in a bowler-pairs article. But will keep in mind. Ananth: ]]

  • Sebastien on October 17, 2012, 11:45 GMT

    This is a superb article. Thanks for the time and effort that went in to it

  • Sagar on October 17, 2012, 11:45 GMT

    Dear Ananth, great work. 2 special requests if you could accomodate. 1. lillee-thommo 2. Zahir and Bhajji.. regards.. [[ Both uploaded.Click at link to view/download. Lillee-Thomson:HERE. Harbhajan-Zaheer:HERE. Ananth Ananth: ]]

  • Farrukh Hanif Awan on October 17, 2012, 11:42 GMT

    Rehman and ajmal maybe? [[ I will check. Probably too soon. Rahman has not even crossed 100 wickets. Ananth: ]]

  • Nitin Gautam on October 17, 2012, 11:42 GMT

    P11. Kumble 225 @ 27.94 & Srinath 187 @ 29.28 Total: 412 @ 28.55

    Thanks for this one. Even though this pair doesnt feature anywhere near the other greats yet from an Indian perspective probably this is the best pacer-spin combo we ever had. Special mention to Srinath who bowled in flat pitches/graveyards/dustbowls of India in 90s & took most of the burden, he probably bowled more than any other Indian bowlers in as many tests as he played. Standout fig. are WpT of 8.24 & 63% top order wickets. Together they took 55% of team with 26% better avg than team peers making it only 2 bowler team for most of the time. for a comparison, Botham-willis improved team peer avg by 17% only however K&S are nowhere close to 2 Ws or any other pair. 2 of srinath's performance are as gud as any..decimating SA in 1996 & 2nd inning against pak in probably 1999. not sure but perhaps 2nd one is best ever ina losing cause. IMO The best pair for India when bowling in tendom (In India) [[ I feel Srinath left a year or two too soon. Ananth: ]]

  • Waspsting on October 17, 2012, 11:32 GMT

    for Chandra/Bedi - is the graph for "away bowling vs (pair vs other bowlers)" drawn completely off-scale (?) [[ I am not sure what you mean. C&B average was 31.07 and the others' avge was 47.26. The graph runs from 50 onwards to 20. Ananth: ]] - Marshall and Garner, perhaps the best pair there's been. Supports my idea that Garner is vastly underrated, too. Those figures probably get even better than they are if you start counting from the time one or the other was taking the new ball.

  • Waspsting on October 17, 2012, 11:24 GMT

    things that caught my eye -

    Vaas' poor record without Murali. What to make of it? You'd think sans Murali, Vaas would be main threat so batsman would be more careful against him (with Murali in the side, he's the main threat, so batsmen more likely to be more attacking to Vaas relatively)

    Anyone, supports the obvious idea that this "partnership" is a most unbalanced one on the Murali side.

    - Tad confused with Wasim and Waqar's away from each other figures (i would have thought Waqar would have far fewer than 96 wickets independent of Wasim)

    I don't know if I'd conclude Wasim was dependent on Waqar though. more simply, his best years were all when Waqar was in the side (he'd probably have done almost as well if Waqar hadn't been, and wouldn't have done much better than he did if Waqar had been there when he wasn't) [[ Waqar played 12 matches after Wasim left and captured 40 wickets. So he has played 14 Tests durng their combined career without Wasim. Wasim played 25 Tests before Waqar made his debut. His haul was an extremely poor 76 wickets in these 25 Tests,. Ananth: ]] Similarly, for Donald and Pollock - Pollock was past his best when Donald left - don't know if its Donald's presence that accounts for the better record w/ or that (con) [[ Last comment incomplete. Ananth: ]]

  • Ananth on October 17, 2012, 11:16 GMT

    Further Special requests handled. Click at link to view/download. Steyn-Morkel: CLICK HERE. Anderson-Swann: CLICK HERE. Kumble-Srinath: CLICK HERE. McDermott-Hughes: CLICK HERE. Pl let me know if I have missed any request. Ananth

  • Waspsting on October 17, 2012, 11:11 GMT

    "To call Vaas 'x' is not fair"

    true, true. (space limit excuse - my fail safe when i've been too harsh)

    Put it this way. I don't think of Murali-Vaas as a great pair, I think of Murali as a great bowler, and Vaas as decent support.

    When I think of Wasim, 3-2 thoughts after the thought "Wasim" is the Wasim-Waqar partnership (and same for most of the pairs named)

    When I think of Murali, I do NOT think of the Murali-Vaas partnership at all (same for Hadlee and Cairns)

    But yes, dismissing Vaas as "x" is harsh on a decent bowler. [[ I agree with your views. Warne-McGrath/Garner-Marshall/W & W and even Lindwall-Miller were true bowling pairs. But Murali-Vaas was a 66.67-33.33 partnership. But let us give the plucky Sri Lankans their due. Ananth: ]]

  • grant on October 17, 2012, 10:37 GMT

    I see you count the first innings wickets as 11 and there is only 10 wickets available when counting the total number of wickets taken for some of the pairs. Like for donald and pollock you have it as 11 & 8 for a total of 19 where I would have it as 10 & 8 for a total of 18. [[ Who is missing what I do not know. Where have I shown 19 as 11+8. It is only 10+9. 6 for 53 & 4 for 16 plus 5 for 74 & 4 for 64. Surely there is a misunderstanding here. Pl point out whare I have gone wrong and I will correct it. Ananth: ]]

  • Farrukh Hanif Awan on October 17, 2012, 9:42 GMT

    This is a very interesting read again. For starters I am very happy to find the great W's again showing up in the top 5 again. I wonder though who is among the latest group of fast bowling pair or bowling pair a stand out? I mean we always had a pair when growing up. Ambrose and walsh. Wasim and waqar etc etc. The only three pairs i can actually really think about are Styen and Morkal, Swann with either Anderson or Broad and Ajmal with Rehman. That is it. Is it a tragedy or has it always been like this?

  • Anshu N Jain on October 17, 2012, 9:11 GMT

    Just looked them up: Srinath and Prasad took 168 wickets in 21 tests they played together. Their WpT is a round 8.00, on an even keel with Walsh & Ambrose, and marginally better than Willis & Botham! [[ My previous response was before reading this one. I think 168 is way too low. I can still do it if you insist. You cannot compare 95 and 60 matches with 21. Ananth: ]]

  • Anshu N Jain on October 17, 2012, 9:06 GMT

    I feel that 50% of team wickets in the innings is a tad low as the bar for measuring pair success. Teams go in with 4 bowlers, and usually need 3 to come good to win tests. If the best 2 bowlers only take 50% of the wickets and this is still qualified as a success, dont you think it leaves too much to ask of the other, obviously less-fancied, bowlers? [[ Don't forget that this is done by innings. And algorihmic. For odd numbers automatically this is pushed to the next number. 3 out of 5, 4 out of 7 and 5 out of 9. Surely these instances have to be treated as successes. For even numbers only the bar might be a bit lower. 3 out of 6, 4 out of 8 and 5 out of 10. If I go to the next higher integer, while 6 out of 10 looks fine, 4 out of 6 seems too much. Probably does not matter. Set the expectation % bar high. What is the alternative? Bring in top order batsmen. Ananth: ]]

  • Anshu N Jain on October 17, 2012, 8:59 GMT

    Ananth, since you are taking requests, could you do one for Srinath and Prasad? They must have taken about 200 odd wickets playing together if not more, and carried India's pace bowling fortunes in the late 90s. [[ Probably not worth it. As a pair, Srinath reaches exactly 100 wickets and Prasad only 68 wickets. I feel these numbers are too low. Ananth: ]]

  • West Indies Follower on October 17, 2012, 7:29 GMT

    First of all, a truly Hurculean effort with truly outstanding results. But then again, this is what we have come to expect of your analysis Ananth. I was wondering whether you could do an analysis by of the average of runs conceeded per innings/ wickets by the team of these pairs of bowlers, and then compare this with the average runs conceeded without one of the pair playing. Then you could do an analysis of the impact of these pairs in relation to the rest of their country's bowling resources. You could also do this comparason with bowlers who played after the retirement of these pairs. One issue with the above though could be that the sample sizes for when these pairs did not play is quite small.(e.g Curtly only 16 wickets), but I think it might have some statistical value [[ Not very clear. "per innings/wickets" seems confusing. Ananth: ]]

    Thanks!

  • Micko on October 17, 2012, 4:20 GMT

    Another great analysis as usal Ananth! It seems to give some weight to the theory that if you have a quality bowler at one end a wicket is more likely to fall at the other as well, due to the pressure being applied. The only pair where both had worse stats was marshall/garner, but their other bowling peers also were of the highest quality, unlike say murali/vass or chandra/bedi. [[ Yes, it is likely to happen when we have selected two out of four and not two out of two or three. Ananth: ]]

  • Harsh Thakor on October 17, 2012, 4:15 GMT

    My list in order of merit 1.Waqar-Wasim 2.Lillee-Thomson 3.Holding -Roberts 4.Garner-Marshall 5.Trueman -Statham 6.Donald-Pollock 7.Walsh-Ambrose 8.Lindwall-Miller 9.Botham -Willis 10.Imran Khan-Abdul Qadir

    Imran and Qadir together could be devastating.Donald-Pollock and Botham -Willis are superb combinations of out and out pace combined with intelligent medium pace.Alan Donald was arguably the quickest of his era while Shaun Pollock the most effective medium pace bowler.Hall and Griffith just miss out who were a brilliant pair.

  • Sarosh on October 17, 2012, 4:12 GMT

    Also did a check of Double hundreds scored against the top modern combinations:

    Warne/Mcgrath – 5 Doubles. Ambrose/Walsh – 3 Donald/Pollock - 1 Wasim/Waqar – 1 (By Attapatu, the ultimate converter of 100s to doubles).

    It was not easy scoring big against Donald/Pollock and Wasim/Waqar. [[ If you discount Atapattu, with no disrespect intended, you are really looking at no top batsman scoring 200. Who scored the 200 against Donald/Pollock. Blewett ??? So no top batsman against either pair. Ananth: ]]

  • Harsh Thakor on October 17, 2012, 4:08 GMT

    A great effort Ananth.Warne-Mcgrath being at the top shows that a bowling attack with the combination of great spin and pace is arguably more lethal than pure pace.Garner -Marshall attack was also lethal as it was a combination of the most accurate paceman of all in Joel Garner with one of the most hostile,quick and versatile fast bowlers of all in Malcolm Marshall.Thus pace and accuracy was brilliantly blended.However,the ultimate to me was Waqar and Wasim.Both brilliantly combined pace with control, swing and reverse-wing.Waqar was amongst the quickest of all time while Wasim was arguably the most versatile of all pacemen.Above all it was a left arm-right arm bowling combination which was lethal. [[ Three pairs will always stand out. Warne/McGrath, Garner/Marshall and Wasim/Waqar. I would say the W's get the edge for the reasons you have identified. If my life depended on a team being dismissed for below 100, that is the pair I would entrust my life with. Ananth: ]] Morally,Lillee-Thomson pair was the most lethal of all on fast tracks closely followed by Holding -Roberts.

  • Meety on October 17, 2012, 3:56 GMT

    Hi Ananth, This is one of those articles where I will have to keep coming back to, to make sure I absorb everything. Fascinating so far. Loathe to make a suggestion as you always seem to have more "tasks" after completing a project than at the beginning but.... Do you think there is anyway you could line up when the bowlers took the wickets when they were operating at the SAME TIME? (i.e genuinely in tandem). I think this would take SOME need out of looking at the influence of a third bowler. I am just thinking whilst it is true that McGrath may have taken 488 wickets in matches where Warne was playing, the truth is, he probably took a fair % of those wickets BEFORE Warne came on to bowl. If it's do-able something like Wickets in Tandem (WiT), & Wickets with Other (WwO) would be good classifications. I assume this would only be doable for matches in the cricinfo era. == == == Botham & Willis are countrymen? Learn something everyday! [[ What you ask for is only possible with complete ball-by-ball data which is not available to me. No point in making guesses here. Most of the top order wicklets for McGrath would have been taken when when he was operating with Gillespie or Lee. That is all. Nothing more. Ananth: ]]

  • Sarosh on October 17, 2012, 3:53 GMT

    Looks like I jumped the gun.A comment on the previous blog looks more appropriate here. In the Wisden Top 100 Test innings there are 5 against Mcgrath/Warne. If we include a later entry of Butcher at No.48. In fact, 5 out the Top 50 are against Warne/McGrath (All losses for Aus) As opposed to a solitary entry for Warne in the Best bowling list; no McGrath surprisingly (or Marshall). Am I right about this? Or have I missed something? [[ I am not getting into discussions on a 10-year old list, since updated about fifty times, now. That was during 2001. Ananth: ]] Other appearances in the Wisden batting list: 3 innings against Donald/Pollock. (1 loss, 1 draw, 1 win) 2 Ambrose/Walsh (2 losses) 1 Wasim/Waqar (loss)

  • Ananth on October 17, 2012, 3:28 GMT

    Abdullah I have completed the Top order Home and Away wkts splits and posted on the main tables. As also the Low order wickets. I will leave it to you to draw your own insights. In general, it seems like the bowling pairs from weaker teams tend to pick up a higher % of top order wickets away than home. Probably the absence of support bowlers. Ananth

  • Skeptic on October 16, 2012, 21:14 GMT

    Could a similar analysis be done on Batsmen as in Mahela/Sanga ? [[ There are normally 4/5 good batsmen in a side. A partnership analysis for batsmen makes less sense. Unless otherwise you are referring to actual batting partnerships which information is available easily. Ananth: ]]

  • Karim on October 16, 2012, 18:00 GMT

    Very interesting article which proves once again (if proof were needed!) that Indians have an amazing knack for figures. I am a Pakistani but cannot envisage a fellow countryman compiling this! [[ Thanks, but probably No. I have a few friends from your country, including some of the readers operating under pen-names who would make excellent analysts. It is just that they have other important responsibilities. Ananth: ]]

  • Ananth on October 16, 2012, 17:40 GMT

    Just a quick update. I did the % of low order wickets (9-11). The results are as you expected. Akram & Younis: 25.9% Warne & McGrath: 22.1% Garner & Marshall: 21.4% Others around 20% Lindwall & Miller: 18.3%. I will do the Home/Away split later. Ananth

  • Dinesh on October 16, 2012, 17:38 GMT

    [[ No apologies needed. At times what is obvious to one may not be to the other. Ananth: ]] I just did my own small analysis on Steyn and Morkel. They both took a combined 329 wickets. Morne's Presence or lack of it never affected Steyn. It shows Steyn Averaged 23.5 with him and 23.7 without him. And in Wins with Morkel in team, Steyn averaged an astonishing 17 and in lossess a ok ok average(by his standards) of 31. Morkel's average doesnt improve much with Steyn in team.He has taken hiss all but 10 wickets with Steyn's presence. So this shows that the Presence of each of them hasnt helped much. [[ I will do a Steyn & Morkel and post the table. Ananth: ]]

  • Pawan Mathur on October 16, 2012, 16:44 GMT

    1)Looking at these pair analysis, I really feel sorry for Kapil Dev. A great bowler, more so in the first ten years of his career, Kapil fought a lone hand in the bowling department for India in the 80s. Perhaps, even an average new ball bowler would have made things easier for Kapil. 2) Can you do a pair analysis of Mcdermott and Hughes ,not the most celebrated of pairs but still an integral part of the Australian revival 3) Among the current English bowling attack,is it appropriate to mention either of Anderson-Swann or Anderson-Broad or is it too early/they are overrated. [[ If I have to select one, i would select Anderson & Swann. Will do and post. Also McDermott & Hughes. Ananth: ]] P.s: I am really surprised to see readers comment on your "ignorance" of Botham and Willis

  • Arjun Nagarajan on October 16, 2012, 16:29 GMT

    Ananth,

    Is it possible to get data on how these pairs performed when they were bowling together(one at each end)? There is a fine difference between bowling pair/partnership and bowling team, just like the difference between Hayden and Ponting scoring together in partnership and Hayden and Gilly scoring runs in same innings. I suspect pairs like Pollock and Donald / Kumble and Harbhajan are more closer to the phrase 'hunting in pairs' than say Warne and McGrath [[ Where do you get this sort of information. Not from the scorecards. Ananth: ]]

  • Neocommunist on October 16, 2012, 14:59 GMT

    Can a list of most ineffective bowlers with >200 wickets be also made?maybe unlucky ishant sharma will be lucky there [[ 1. What is the definition of an ineffective bowler. 2. Ishant is nowhere near 200 wkts. 3. For what is worth I have given below the highest averages for 100+ wkt bowlers. Hooper 114 @ 49.43 Boje 100 @ 42.65 Shastri 151 @ 40.96. Ananth: ]]

  • Hassam on October 16, 2012, 14:00 GMT

    Ananth,

    Excellent stuff, all power to you.

    It is very interesting when numbers and statistics corroborate what us viewers have always felt or known. You mentioned it was a big surprise that Wasim and Waqar have a figure of 57.2%, the lowest amongst these pairs, for top order wickets. On the contrary it really wasn't a surprise. The pair were renowned for cleaning up the tail for nothing; absolutely lethal against the lower middle and tail, far more than any bowling pair before them or since. In fact it was rather late in their careers they figured out how effective the new ball could be, early on it was all about getting the ball old enough to reverse. The fact that 43% of their wickets came after number 6, the highest in the list, confirm their effectiveness with the old ball and eliminating the lower order. On the other side of the spectrum one always felt how great Pollock and Donald were with the new ball and again your factual analysis establishes this assertion. [[ Abdullah's add-ons would offer some nice insights. Ananth: ]]

  • Waspsting on October 16, 2012, 13:31 GMT

    Some confirmation for my suspicions re: Bill Johnston in the Lindwall-Miller dynamic.

    Miller without Johnston (irrespective of Lindwall) 54 wickets@24.2

    that's fractionally better than Miller with Lindwall, and lengthier than Miller without Lindwall.

    Lindwall without Johnston (irrespective of Miller) 79 wickets@24.67

    that's better than Lindwall without Miller.

    Matches with all 3 playing (36 matches) Lindwall 144 wickets @22.12 Johnston 140 wickets @23.80 Miller 107 wickets @22.98

    (no easy conclusions since these guys seem to do pretty standardly regardless of who else is in the side, but not quite clear if its a "Lindwall-Miller" analysis is fair, without particular note to the third bowler every bit as effective as either of them)

  • Boll on October 16, 2012, 13:25 GMT

    @Arjun. I`m not too sure if the analysis of bowling trios would add much that this article, and Ananth`s detailed analysis of overall bowling attacks hasn`t done.

    Warne/McGrath/Gillespie seems an obvious one in recent history. Anderson/Swann/Broad good as well - but doubt if Kumble/Raju/Chauhan(?? must admit the only Chauhan I`d heard of was Chetan - OK, RK Chetan 47 wickets at 40) would feature amongst the great trios though... [[ Boll, this would add value to Waspsting's points about the importane of the third quality bowler. His point that Johnston was a great third to M&L or B&D is valid. Gillespie or Lee or Holding also come into the mix. Also Prasanna, added to Chandra/Bedi is a wonderful idea. Ananth: ]]

  • Boll on October 16, 2012, 13:19 GMT

    Right - having satisfied myself with a weak attempt at humour I`m back in business. Great to see a fresh look at `bowling pairs` - I imagine a phrase chosen intentionally to differentiate from the more well-known `bowling partnerships` (usually confined to new-ball bowlers).

    It`s interesting to note that not many of the chosen duos are real new-ball pairings. Probably only Walsh/Ambrose, Wasim/Waqar and Pollock/Donald from the original 8 pairs really fit the bill. (Lindwall/Miller, Trueman/Statham redress the balance somewhat).

    Having said that I think that many of the others bowled in tandem for extended periods.

    Chandra/Bedi, pretty much any Sri Lankan bowler and Murali, Warne and McGrath all fit the bill.

    Right, as advised by the author I`m going to have a more focused read through it all before commenting. [[ Not really, Boll. Vaas was an excellent foil to Murali. He may not stand out when compared to Murali. But Murali/Fernando or Murali/Malinga would be ridiculuous. Vaas is like Pollock: grossly underrated. Ananth: ]]

  • Boll on October 16, 2012, 13:08 GMT

    I tend to trust, the usually very serious, Narayanan-san here. I sincerely believe that there is, at the very least, a better than even chance that he knows that Sir Ian Botham and Bob Willis are actually from the same country. I also think that he may have a good shot at guessing that that country is actually England. Furthermore, I deduce that he may understand that test cricketers often struggle to play tests against their own nation. I know I`ve been wrong before...just saying if I were a betting man I`d back him in. [[ If my single sentence was not understood, what chances are there that your collection of sentences would be. But thanks for the levity introduced. Ananth: ]]

  • Nitin Gautam on October 16, 2012, 12:55 GMT

    Luking at the last table, its very hard to go past 2Ws awesome record as a the best pair. close to 560 wickets in 55 tests at 3rd best avg speaks volume of quality. Can SR be added to the tables. I am sure 2Ws will trump ever other pair here. somewhere I read its 45.xx for 2 Ws. If only Anantha can prove it. 2nd their record in home matches 22 matches, 223 wickets @20 is beyond belief, regardless of few exceptions they bowled pretty much on dustbowls in 90s) & they pioneered reverse swing to deadly effect. Murli & Vaas comes close but that's primarily due to Murali's extremes but pacers with such record is extraordinary. Can we get pair records of Kumble & Srinath. they played a large part of their career together & scripted many special moments for India in 90s. [[ As far as possible let us stick to a bowler being a part of a single pair. Otehrwise next we will have Kumble & Prasad etc. And amongst the West ibdidain bowlers, virtualluy any combination. But will do K & S at least to see where they stand. Ananth: ]]

  • Abdullah on October 16, 2012, 12:23 GMT

    Hi Ananth Great article once again. Just when I thought, you cant do any better. I thought your article on Lara could only be matched by your article on the best batsman, but this is as good if not better. Being a financial analyst myself, its one thing to process large volume of data but even more difficult to present it in a simple and powerful manner as you have done. Hats off to you. [[ Thanks, Abdullah, for the kind words. Ananth: ]] I have a number of comments i would like to make, but the first one i would like to make is on the lower Top-6% for the Ws. I think it reinforces the point that they were both masters with the old ball and probably took more lower order wickets than the top order ones. It wouldnt have helped that the ball swings less and the pitches are non-condusive for swing bowling in the sub-continent where they played the majority of their matches. A split of their Top-6% by home and away might shed some more light on this. [[ Very good point. Off hand it was Shane Warne who was adept at cleaning up the tail. What I can do is to a 9-11 compilation also and post the results. Will do in a day. Similarl;y Top-6, home and away is fairly easy. Will do that also. Ananth: ]]

  • Arjun on October 16, 2012, 12:16 GMT

    Identifying trio won't be a problem. Just need to make sure that all the 3 bowlers have taken atleast 50/100 wkts individually and collectively that should be more than 400/500. [[ Yes, Arjun, it can be done. A clean two-stage process. I will do it after some time. Ananth: ]]

  • Waspsting on October 16, 2012, 12:15 GMT

    going through slowly as advised, couple of instant thoughts I had

    There's a potentially big confounding factor in assessing "pairs", that being the presence of a significant 3rd bowler.

    This is somewhat accounted for by stats on "other bowlers" - but if 1 bowler is on par with the pair in question, slotting him into "other bowlers" while retaining the "pair"... doesn't seem quite right to me(?)

    For Lindwall and Miller, there's Bill Johnston (but i see no big difference in the stats of any of the 3 away from each other)

    For Warne and McGrath, there's Gillispe maybe(?)

    ---

    I also think you've given a blind positive respect to the 1 from each side aspect of the selected pairs.

    Hadlee and Cairns? Murali and Vaas? As John Mcenroe's doubles partner once said, "the best doubles team in the world is John McEnroe and anyone else"

    I think most people would rather have seen Robert-Holding, Lock-Laker, Beanud-Davidson etc. then Murali/Hadlee and x/y

    get back to you in more detail soon, [[ Just one response. Ask about any pair as long as they complete the basic qualifications. In fact I had Benaud-Davidson for half the duration of my analysis. Then I changed. The key tables will be presented and you can draw your own conclusions. Frankly this is a bowling pairs analysis and is already the longest I have done. How can the impact of the third bowler be brought in. Only way to do is to do what Arjun has suggested. Look at Bowling trios. To call Vaas 'x' is not fair to a guy who captured more Test wickets than Donald, Willis, lee, Gibbs, Trueman, McDermott and Garner. Ananth: ]]

  • Arjun on October 16, 2012, 11:48 GMT

    Ananth,

    Great analysis. As we now know about great bowling pairs, can we have a table of Successful Bowling Trios ? England's recent success in Test matches has lot to do with Anderson, Board, Swann. Similarly, in early 90s, India had Kumble, Raju, Chauhan. [[ Very valid but tough request. I have to do some very special algorithms using a combination of bowler names/ids. The problam is how to identify the trios. Once that is done, the program will take over. But will keep in mind. Ananth: ]]

  • Ananth on October 16, 2012, 11:46 GMT

    Special requests handled. Imran-Qadir: CLICK HERE to view/download. Kumble-Harbhajan: CLICK HERE to view/download. Ananth

  • ruudraza on October 16, 2012, 11:42 GMT

    It wud be intresting to see how teams winning % changed when one of the two was missing. It wud give a fair idea of who was more important of the two. Intrestingly, Australia did better without Miller and Lindwall despite their careers almost completely overlaping each other.

  • Elvis on October 16, 2012, 10:51 GMT

    Great article once again. Does bowling in tandem from either end give a higher degree of co-relation. Commentators invariably point this out. There are a few variations to be considered: 1) Attack at both ends with no respite for the batsman. Marshall and Holding. 2) One bowler keeps the batsmen quiet at one end while the bowler at the other end reaps the reward. Laker and Lock, Bedi and Chandra, Venkat and the other three 3) A bowler takes no. 6 or 7 opening the floodgates for another bowler to mop up the tail. Probably ball by ball or over by over data to do this type of analysis is not available.

  • Uday on October 16, 2012, 10:41 GMT

    Also, when I think of Donald's bowling partner, its always De Villiers who comes to mind rather than Pollock. Is it possible to get their data? Im sure they didnt capture that many wickets together since De villiers retired soon after SA's readmission, but I bet their average is lower than Donald-Pollock, or even Pollock-Ntini [[ Too few matches and too few career wickets for de Villiers (18 and 85). It is certainly not a significant partnership. Ananth: ]]

  • Uday on October 16, 2012, 10:32 GMT

    Hey Ananth

    Great Analysis as always, and so much data! let me just make two quick observations upfront:

    (a) Personally, its very pleasing to see how well Ambrose did without Walsh. Big Curtly was always my favorite - to my mind, his greatness is never appreciated enough, and between him and Walsh, he was comfortably the deadlier of the two.

    (b) There seem to be three types of bowling pairs here - firstly, a case of two great bowlers who just happened to play at the same time (such as Warne-Mcgrath or Ambrose-Walsh). Secondly, a case where one bowler is clearly dominant, and the other got the benefit of playing alongside him (such as Murali- Vaas or Willis-Botham). Thirdly, where both bowlers depended on each other to get the best out of themselves (such as Wasim-Waqar/Bedi-Chandra). Its the third I would consider a true "partnership" - is it possible to organize a sequential table based on difference in averages when playing alone and together? [[ Very good analysis of the pair formation. Two tables , one a summarized one in the article and another mammoth one, which can be downloaded, are available for you yourself to analyze. Ananth: ]]

  • KR Sriram on October 16, 2012, 9:15 GMT

    Can we have the analysis for Kumble and Harbhajan? Statistically, they were the most successful Indian pair of bowlers. [[ Will do. Ananth: ]]

  • KR Sriram on October 16, 2012, 9:08 GMT

    Re your comment on Test # 1313, the Almanack Report says that Warne's toe was broken by an in-swinging yorker from Waqar Younis. [[ I have got the three volume Wisden Book of Test Match Cricket. There (this match no is 1312), the comment is "Despite being deprived of their star spinner...". That is all. Anyhow thanks. Ananth: ]]

  • Sumit on October 16, 2012, 8:50 GMT

    I dont think hadlee and cairns should have featured here...just on the basis of country...it would have been much better to have garner and holding instead of them. [[ Garner is already featured. How can be part of another pair. Ananth: ]]..

  • JAY on October 16, 2012, 8:49 GMT

    Wonderful article! Much insight into pairs. I agree with you that this is just the start! Much more analysis will come!

  • Vyasa on October 16, 2012, 8:41 GMT

    Excellent work, Ananth. It is nice to see this article. I know it would be quite difficult to expand this pairwise concept, but if you did, it would add a lot of value to who might be superior- bowler X or Y. Often, having great colleagues is termed as bad for someone's wicket taking (Warne vs Murali for instance). A bunch of articles will only shed more light on the issue. Vyasa. [[ Pl re-frame your request within the bowling framework. Ananth: ]]

  • HP on October 16, 2012, 7:58 GMT

    want to know about Kumble-Harbhajan pair. Also want to say how great Pollock is!! [[ Will do and post a link in a comment from me. Ananth: ]]

  • junaid on October 16, 2012, 7:36 GMT

    Superb stats n great info.

  • Nitin Gautam on October 16, 2012, 7:34 GMT

    Incredible is the word here.

    To envisage & compile something of this magnitude is truly genius.

    It took a gud 45 mins to read & understand the herculean effort you have put in this article still to post a comment matching the quality of the article, I have to say, It has to re-read & re-read again, doing which I cant not resist so will post comments later

    Just one thing that strike in 1st glance. you mentioned in start of article

    "The two other pairs are Akram and Younis, who reserved their best against West Indies and Willis & Botham, who, for some strange reason I cannot comprehend, failed to capture a single wicket against England, reserving their best for Australia".

    Willis & Botham both belonging to England cannot take wickets against England. Is it some typo or I missed something obvious here. [[ After being so kind to me with your words, you have completely missed the humour behind the way this was done. Two readers within the first hour !!! Ananth: ]]

  • ronnie on October 16, 2012, 7:27 GMT

    no Agakar and Balaji?

  • Dinesh on October 16, 2012, 7:20 GMT

    "Willis & Botham, who, for some strange reason I cannot comprehend, failed to capture a single wicket against England, reserving their best for Australia."

    Anath: was that Sarcasm? if yes then probably a stage has come where i cannot understand it or was it a typo mistake as Both Bob Willis and Botham were from England. So how could they take a wicket against England. [[ Dinesh, that was not sarcasm, but humour. Unfortunately it has come to a stage where my attempts at humour are taken seriously by readers. Do you SERIOUSLY think I did not know that Willis and Botham were from England. Then I probably should be in an institution for the mentally chllenged and not writing 20-page articles. Ananth: ]] This article will be a bouncer to Many readers who aren't frequent on this BlogoSpace. But let me take my hat of to you. Another Gem of an Article. Havent read it fully. Will post further on reading it deeply.

  • Saleem Basit on October 16, 2012, 7:08 GMT

    Hi. Great analysis Anantha. I always wondered how the 2 Ws compared to other great bowling pairs. Thanks for sharing.

    Just a thought for Wasim under-performing without Waqar. Could it be the lack of opportunities in the early phase of Wasim's career (1984-1989), due to Imran & Qadir dominating the wickets. May be Imran-Qadir stats shed some light on it.

    Thanks. [[ Will get Imran-Qadir stats, upload the file and post the link in these comemnts. You have to keep a watch here. But one thing cannot be said that Wasim did not have the support. He had good support. Possibly reductin in Imran's bowling effort meant that a larger pie was available for thes etwo. Ananth: ]]

  • No featured comments at the moment.

  • Saleem Basit on October 16, 2012, 7:08 GMT

    Hi. Great analysis Anantha. I always wondered how the 2 Ws compared to other great bowling pairs. Thanks for sharing.

    Just a thought for Wasim under-performing without Waqar. Could it be the lack of opportunities in the early phase of Wasim's career (1984-1989), due to Imran & Qadir dominating the wickets. May be Imran-Qadir stats shed some light on it.

    Thanks. [[ Will get Imran-Qadir stats, upload the file and post the link in these comemnts. You have to keep a watch here. But one thing cannot be said that Wasim did not have the support. He had good support. Possibly reductin in Imran's bowling effort meant that a larger pie was available for thes etwo. Ananth: ]]

  • Dinesh on October 16, 2012, 7:20 GMT

    "Willis & Botham, who, for some strange reason I cannot comprehend, failed to capture a single wicket against England, reserving their best for Australia."

    Anath: was that Sarcasm? if yes then probably a stage has come where i cannot understand it or was it a typo mistake as Both Bob Willis and Botham were from England. So how could they take a wicket against England. [[ Dinesh, that was not sarcasm, but humour. Unfortunately it has come to a stage where my attempts at humour are taken seriously by readers. Do you SERIOUSLY think I did not know that Willis and Botham were from England. Then I probably should be in an institution for the mentally chllenged and not writing 20-page articles. Ananth: ]] This article will be a bouncer to Many readers who aren't frequent on this BlogoSpace. But let me take my hat of to you. Another Gem of an Article. Havent read it fully. Will post further on reading it deeply.

  • ronnie on October 16, 2012, 7:27 GMT

    no Agakar and Balaji?

  • Nitin Gautam on October 16, 2012, 7:34 GMT

    Incredible is the word here.

    To envisage & compile something of this magnitude is truly genius.

    It took a gud 45 mins to read & understand the herculean effort you have put in this article still to post a comment matching the quality of the article, I have to say, It has to re-read & re-read again, doing which I cant not resist so will post comments later

    Just one thing that strike in 1st glance. you mentioned in start of article

    "The two other pairs are Akram and Younis, who reserved their best against West Indies and Willis & Botham, who, for some strange reason I cannot comprehend, failed to capture a single wicket against England, reserving their best for Australia".

    Willis & Botham both belonging to England cannot take wickets against England. Is it some typo or I missed something obvious here. [[ After being so kind to me with your words, you have completely missed the humour behind the way this was done. Two readers within the first hour !!! Ananth: ]]

  • junaid on October 16, 2012, 7:36 GMT

    Superb stats n great info.

  • HP on October 16, 2012, 7:58 GMT

    want to know about Kumble-Harbhajan pair. Also want to say how great Pollock is!! [[ Will do and post a link in a comment from me. Ananth: ]]

  • Vyasa on October 16, 2012, 8:41 GMT

    Excellent work, Ananth. It is nice to see this article. I know it would be quite difficult to expand this pairwise concept, but if you did, it would add a lot of value to who might be superior- bowler X or Y. Often, having great colleagues is termed as bad for someone's wicket taking (Warne vs Murali for instance). A bunch of articles will only shed more light on the issue. Vyasa. [[ Pl re-frame your request within the bowling framework. Ananth: ]]

  • JAY on October 16, 2012, 8:49 GMT

    Wonderful article! Much insight into pairs. I agree with you that this is just the start! Much more analysis will come!

  • Sumit on October 16, 2012, 8:50 GMT

    I dont think hadlee and cairns should have featured here...just on the basis of country...it would have been much better to have garner and holding instead of them. [[ Garner is already featured. How can be part of another pair. Ananth: ]]..

  • KR Sriram on October 16, 2012, 9:08 GMT

    Re your comment on Test # 1313, the Almanack Report says that Warne's toe was broken by an in-swinging yorker from Waqar Younis. [[ I have got the three volume Wisden Book of Test Match Cricket. There (this match no is 1312), the comment is "Despite being deprived of their star spinner...". That is all. Anyhow thanks. Ananth: ]]