February 12, 2017

Double trouble

Which bowling pairs have enjoyed the most success in home Tests?

R Ashwin may be India's top Test bowler, but he owes a large part of his success to Ravindra Jadeja © Hindustan Times

Stats current to before the start of the Hyderabad Test

India beat England 4-0 in their recent Anthony D'Mello Trophy series in India. Normally such a scoreline would suggest Indian dominance with bat and ball. In Test cricket this usually means three high-quality bowlers and four or five high-quality batsmen. Duncan Fletcher's theory of a successful team involved a minimum of eight regulars who were successful and picked themselves. While Mohammed Shami (ten wickets at 25 apiece) was fit and available, it could be said that India had three outstanding specialist bowlers. But take the series as a whole and you find that R Ashwin and Ravindra Jadeja took 54 wickets between them at 28 runs apiece, while the other Indian bowlers combined for 36 wickets at 38 apiece. Amit Mishra's five wickets cost 55 runs each. It didn't help that Mishra played on the two most batsman-friendly wickets of the series, in Rajkot and Chennai.

Famous teams are marked by famous bowling trios and even quartets. But these are well known precisely because they are so rare. It is hard enough to find one Test-quality bowler for a given set of conditions, let alone three. For the most part, Test success is built on the backs of a pair of reliable, experienced bowlers. It is difficult to isolate such pairs. Even if we consider iconic combinations like McGrath and Warne, or Ambrose and Walsh, they tended to have above-average support bowling.

I identified all pairs of bowlers, bowling in home Tests, that satisfy the following two conditions. First, the pair should have played together in at least ten home Tests. Second, together they should have bowled the majority of their team's overs in all these Tests taken together. These conditions help to isolate pairs of bowlers who not only dominated the bowling for their team, but did so for a prolonged period in home conditions. A side with three strong bowlers would not have two out of the three dominating both the bowling and the wicket share for a prolonged period of time. A pair of bowlers who were not good at what they did would not survive as the main bowling pair of a Test team for a prolonged period of time.

There are 25 pairs of bowlers in Test history who qualify under the conditions set forth above. They are listed in the table below. The table also gives the number of balls they bowled per match, their combined average and strike rate in those Tests, match results, and the number of wickets they took in the match as a pair.

Leading bowling pairs in home Tests (min ten home Tests)
Bowlers Matches Wickets per match Balls bowled per match Average Strike rate Wins Losses Draws
McGrath-Warne 54 9.3 534 24.4 57.2 43 4 7
Murali-Vaas 51 10.5 560 21 53.5 27 10 14
Kumble-Harbhajan 34 10.5 656 27.2 62.6 14 6 14
Hadlee-Lance Cairns 23 8.2 495 23.5 60.6 9 3 11
Hadlee-Chatfield 21 7 492 27.4 70.3 7 3 11
Bedi-Chandrasekhar 20 9.6 637 25.1 66.7 7 6 7
Bedi-Prasanna 20 9.3 658 25.1 71.2 7 8 5
Qadir-Qasim 18 9.1 583 22.9 64.0 7 1 10
Bedi-Venkat 18 6.7 579 28.3 86.8 3 8 7
Ashwin-Jadeja 17 12.2 621 20.3 50.8 14 0 3
Murali-Dilhara Fernando 16 10.1 483 20.0 48.0 11 3 2
Malinga-Murali 15 10.7 443 20.0 41.3 11 1 3
Chandrasekhar-Venkat 15 7.7 589 30.5 76.1 3 5 7
Lock-Laker 14 10.6 452 11.7 42.5 9 0 5
Raju-Kumble 14 10.4 579 22.1 55.5 10 2 2
Gough-Caddick 13 8.9 393 25.2 44.4 6 5 3
Ojha-Ashwin 13 11.4 635 25.1 55.7 9 2 2
Qadir-Tauseef Ahmed 13 7.8 487 23.7 62.7 7 1 5
Mortaza-Rafique 13 6.4 494 36.1 77.4 1 10 2
Chandrasekhar-Prasanna 12 8.2 557 28.5 68.2 5 4 3
Shivlal Yadav-Maninder Singh 11 7.4 533 28.9 72.4 2 2 7
Qadir-Sarfraz 11 6.8 519 32.2 76.1 4 0 7
Murali-Herath 10 9.2 477 24.5 51.9 6 2 2
Gul-Kaneria 10 8.7 544 33.1 62.6 5 1 4
Goddard-Tayfield 10 7.6 773 27.1 101.7 2 5 3

Perhaps unsurprisingly, the list is not a long one. If the minimum match requirement is reduced to five home Tests, the list grows from 25 to 44 pairs. Perhaps equally unsurprisingly, the list is dominated by spin bowlers, since spinners tend to bowl more overs than the fast men. Twenty-two out of the 25 pairs include at least one spinner. The most prolific spinner of them all, Muttiah Muralitharan, appears on the list four times.

The Indian spin quartet's appearances are notable. India did systematically worse when S Venkataraghavan and one of the other three spinners combined to bowl the majority of India's overs than they did when any combination of the other three bowled together. Venkat and Erapalli Prasanna do not feature in this list as a pair. Was the quartet really a quartet?

The best bowling pair in home Tests: Murali and anyone © Getty Images

Abdul Qadir appears on this list as a part of three pairs. Iqbal Qasim was Qadir's most frequent spin-bowling partner. Imran Khan and Qadir do not appear as a pair. This is probably because Qadir bowled only about 25% of Pakistan's overs in Pakistan (the bowling share for a spinner is more commonly closer to 30% than 25%), while Imran bowled only about 21% of Pakistan's overs in Pakistan. Even though they played 22 Tests together in Pakistan, they accounted for 45% of Pakistan's overs in those matches and took 49.7% of Pakistan's wickets.

If we change the conditions for isolating pairs to consider the ones that took the majority of their team's wickets over 15 home Tests, the many great fast-bowling combinations in Test history make their appearance. The minimum here is increased to 15 Tests, as there are numerous pairs who have achieved this over ten. Further, if one considers strong attacks, then this high bar helps isolate the truly dominant performers in those attacks. For example, among the great West Indian speed quartets of the '70s and '80s, Malcolm Marshall, Joel Garner and Michael Holding feature on this list. West Indies never lost a Test at home when Marshall and Garner, or Marshall and Holding took the majority of the opposition wickets.

Leading bowling pairs in home Tests (qual: min 15 home Tests)
Bowlers Matches Wickets per match Balls bowled per match Average Strike rate Wins Losses Draws
Steyn-Ntini 20 9.3 413 25.4 44.7 11 7 2
Nel-Ntini 19 8.8 446 27.6 50.4 9 7 3
Waqar-Akram 24 9.5 392 20.1 41.1 11 4 9
Qadir-Qasim 18 9.1 583 22.9 64 7 1 10
Steyn-Philander 18 9.4 387 19.8 41.2 12 3 3
Kumble-Harbhajan 34 10.5 656 27.2 62.6 14 6 14
Zaheer-Harbhajan 30 8.5 531 28.9 62.3 17 3 10
McDermott-Merv Hughes 17 8.8 464 25.9 52.6 8 4 5
Trueman-Statham 19 9.2 499 22.4 54.4 9 5 5
Trueman-Lock 16 8.9 375 15.3 42 10 3 3
Bedi-Prasanna 20 9.3 658 25.1 71.2 7 8 5
Imran-Sarfraz 16 8.2 462 24.8 56.4 8 0 8
Murali-Vaas 51 10.5 560 21 53.5 27 10 14
Benaud-Davidson 17 9.1 643 25.1 71 7 4 6
Marshall-Garner 18 8.9 418 21.1 46.8 12 0 6
Pollock-Donald 28 8.6 407 20.8 47 17 3 8
Ashwin-Jadeja 17 12.2 621 20.3 50.8 14 0 3
Marshall-Holding 16 8.8 395 22 45.2 10 0 6
Murali-Dilhara Fernando 16 10.1 483 20 48 11 3 2
Hadlee-Lance Cairns 23 8.2 495 23.5 60.6 9 3 11
Bedi-Chandrasekhar 20 9.6 637 25.1 66.7 7 6 7

A number of pairs appear on both lists. These are pairs that bowled the majority of their team's overs and took the majority of their team's wickets, and played together in home Tests on at least 15 occasions. The standout pair from this select group is the Indian pair of Ashwin and Jadeja. This may come as a surprise to readers, but Jadeja now remarkably has 111 wickets in his 25 Tests. More importantly, his economy rate of 2.27 is the best among all bowlers in the 21st century who have at least 50 Test wickets. In India, Jadeja has taken 90 wickets in 17 Tests with an even better economy rate of 2.13. He bowls a maiden once every 3.7 overs. Jadeja is a captain's dream. He is nearly unhittable in India. At the other end, Ashwin has a significantly greater arsenal of variations and tricks, and can afford to experiment far more than he might have done without the miserly Jadeja as his partner.

India have won 14 out of the 17 Tests that Jadeja and Ashwin have played together at home. In these Tests, these two have accounted for just over 12 wickets per Test. Compare that with the returns of Marshall and Garner, or Ian Botham and Bob Willis, or any of the other celebrated fast-bowling pairs at home. The fast men take their wickets about an over quicker than Jadeja and Ashwin do, but they account for fewer than ten wickets per match as a pair. Over 17 Tests, the Indian duo have faced nearly all major contemporary Test opposition. By way of comparison, Anil Kumble and Harbhajan Singh bowled more per Test than Ashwin and Jadeja, and managed 10.5 wickets per Test as a pair in India.

Wasim and Waqar: a twosome to be feared © Getty Images

It is a truism in Test cricket that bowlers hunt in pairs. As with most things, this is probably out of necessity. One great bowler surrounded by modest colleagues is not sufficient to challenge most Test-quality opposition. Three top-quality bowlers are hard to find all at once. A pair is a nice compromise. A pair can also ensure that at least when the two are bowling together, the opposition batting faces a difficult examination.

If visiting teams are to do well in India in the near future, they will not only have to learn to read Ashwin's many variations and resist the temptation of being dragged off balance by his mesmerising drift, they will also have to find a way to score runs against Jadeja. For it is at Jadeja's end that India's dominance at home begins.

This article demonstrates an approach to studying pairs of bowlers in home Tests. It can be easily applied to other combinations of specialisations (middle-order batsmen, openers, lower-middle orders, spinners, fast men) as well. It tells a story that the more commonly studied statistical accounts of individual players does not allow us to see.

Stats current to before the start of the Hyderabad Test

Kartikeya Date writes at A Cricketing View. @cricketingview

Comments have now been closed for this article

  • Rajasundram on February 14, 2017, 13:20 GMT

    For a proper evaluation - pitch bias must be taken out - because Home teams prepare itches to suit their bowlers - either spin or seam. Thus a spin-seam ( eg Warne/McGrath) combination should be rated over seam/seam or spin/spin combinations. It will also be useful to finf the % of matches won at home by these combinations

  • Kartikeya on February 14, 2017, 10:34 GMT

    The point of limiting it to home tests is to look at pairs who made good Test teams. Away Test success is rare, with the exception of a few truly outstanding teams in Test history. But Test cricket is a representative sport (not a franchise sport), in which local conditions play a significant role. Competent Test teams are dominant at home. But this dominance is rarely due to 3 or more bowlers. Most commonly, its a pair of bowlers who do this damage. So, by setting a significant number of home Tests are a limit for a pair, the idea is to isolate periods of domination (similar to the one India have enjoyed in the last 18 months or so).

    The general point of using data to look at the history of the game is that (a) it is a game, and (b) its history is completely available in scorecards. This means that it is easy to test conventional wisdom about the game by looking at all the records. So, if a pair is absent here, its because they do not satisfy the cond. set here.

  • Cricmj Makes on February 14, 2017, 9:52 GMT

    Hi, interesting as this is I can't fathom the relevance limiting to home tests. Surely all tests would give a more balanced outcome and be more relevant. Honestly don't see an important point being made at all.

  • sreeji9928160 on February 14, 2017, 8:36 GMT

    wasim -waqar the fearsome jodi... the lightning of 90s.

  • Vidhyuth on February 14, 2017, 6:44 GMT

    The absence of an effective long time bowling partner for the legendary Kapil Dev is glaring from the list. For so long, the great man carried the burden all by himself - both as a striking force as well as a holding option. His record suffered both by the fact that he played majority of his test matches on unresponsive subcontinental pitches as well as by the absence of even decent support from the other end. Till the unbelievable workload he carried took its toll, Kapil's record matched the best in the world - 200 wickets in 50 test matches (majority of them in India & Pakistan), 300 in 81 test matches. The fact that Lillie and Hadlee played very few test matches in the subcontinent and Imran was brilliantly supported by a fine set of bowlers, should help us understand Kapil's brilliance in perspective.

  •   Venkatesh Venkatesh on February 14, 2017, 3:23 GMT

    Good article but why so many very good pair of bowlers are left out likes of Hall & Gibbs , Ambrose & Walsh is it deliberate to project only few , very difficult to understand

  • Master on February 13, 2017, 12:46 GMT

    The methodology in use for this article systematically favours teams who only picked 4 bowlers rather than those who picked 5. What reasoning is there for only looking at bowlers who bowled +50% of the overs or took +50% of the wickets? I would like to see this list based purely on Economy and Strike Rate while operating together never mind if they got 40% or 60% of the wickets.

  • Kartikeya on February 13, 2017, 10:22 GMT

    I notice lots of comments about why a pair isn't in the lists. So, some clarification.

    The lists are exhaustive. They are drawn from all Test matches. I did not look up pairs as I went along and add they to the list. I looked at all possible pairs and then filtered out all the ones which did not meet the following criteria:

    1. Bowled a majority of the team's overs for at least 10 home Tests (25 such pairs) 2. Took a majority of the team's wickets for at least 15 Tests. (44 such pairs)

    This was explained in the text, but many of the comments suggest that it was not read.

  • john on February 13, 2017, 9:41 GMT

    Why isn't sir ambrose and walsh on the list?

  • sam on February 13, 2017, 8:13 GMT

    Ashwin has the numbers certainly-at home that is. And certainly playing on our conducive home pitches has really been made best use of by him. But look at his o/s recd and you see the 'real' picture. His average 'inflates' to a healthy 50+. On the other hand,the likes of Mcgrath,Warne,Murali,Wasim,Lillee etc are one of a kind because they were just as lethal and match winners all across the world ,on a variety of pitches and conds and found a way to over come the most obstrusive/unfriendly of conds to bowlers/their kind of bowling to be successful,at least having a holld over the batting if not the outright threats they normally are. That's what makes them incomparable,rather than some random recds,'milestones' or anything else.

  • No featured comments at the moment.