Saturday, August 21, 2010

You Don't Necessarily Need 20 Wickets to Win a Test

It irritates me no end when experts talk about a side needing 20 wickets to win a test. At the risk of sounding pedantic, I'm going to point out that there have been 131 tests in which a team lost despite losing less than 20 wickets. Ofcourse most of the teams that lost in such a manner did so because one or more of their batsmen were either retired hurt or absent hurt.

However, there have been numerous instances in which a team has declared either or both its innings and gone on to lose. South Africa lost at the SCG in 2006 because Graeme Smith was ready to risk losing the series 2-0 if it gave him a chance to level it at 1-1. He declared both innings and Australia chased down the target on the back of a second hundred in the game from captain Ricky Ponting. Less than a year later Shane Warne masterminded the collapse that broke England's spirit for the remainder of the series. Warne and his mates had looked insipid in the first innings when England piled on a score of 551/6 before declaring their innings. India's historic win at Chennai came after Kevin Pietersen had declared England's second innings with one wicket remaining.

Sure these occurrences are rare (Statsguru doesn't allow me to sort the declared innings from the others), but they do happen. More often than a team winning despite scoring less runs than the opposition. How many times has a team done that? Never. Why? Because in test cricket, just like in any other sport, you cannot win if you don't score more than the opposition. And yet experts keep talking about how bowlers win you matches, how you're not a world class side if you don't have a strong bowling attack, how you can't win if your bowlers don't have the ability to take 20 wickets.

Saying you need 20 wickets to win a test is the biggest cliche in cricket. You can say it because it's not like you can take more than 20 wickets. But can anyone predict the number of runs to win a test before the start of the match? Obviously not, because in order to win you need atleast one more run than the opposition although you could win by scoring a lot more as well. On the other hand, you could score a lot more runs than the opposition and yet not win! So while scoring more than the opposition is a necessary condition to win a test, it may not be sufficient. But taking 20 wickets is neither a necessary nor a sufficient condition to win.

If you took a snap poll among cricketers on who has the better bowling attack between India and Pakistan the verdict would be overwhelmingly in favour of the latter. A similar poll on the batting strength of the two sides would result in India getting the vote over their neighbours. And yet it's the Indian team that hasn't lost a series in two years (They have won 5 during this period, 4 if you exclude Bangladesh), while Pakistan have lost two series and not won any during the same period.

The only two teams both have faced during this period are Sri Lanka (Home and Away) and New Zealand (Away). India beat Sri Lanka at home while Pakistan drew against Sri Lanka (The first test was drawn and the second one was cut short after the attack on the Sri Lankan team bus). Pakistan lost to Sri Lanka 2-0 away from home, failing to chase a sub-200 target in one of the tests while India drew their series in Sri Lanka 1-1, chasing down a 200-plus target in the final test. Coming to New Zealand, India beat the Kiwis 1-0 and it might well have been 2-0 had it not rained on Day 5 with India needing just two more wickets. Pakistan, on the other hand, drew their series 1-1. You can say Bond didn't play against India and New Zealand won the only test he played against Pakistan, but Pakistan would have lost had it not rained on Day 5 of the final test at Napier.

This isn't an exercise in proving India is better than Pakistan when it comes to test cricket. I've used these two teams because they have forever played contrasting styles of cricket. A few years ago Pakistan were clearly the better test side because they had batsmen like Saeed Anwar, Javed Miandad, Salim Malik, Inzamam-ul-Haq to go along with bowlers like Wasim, Waqar, Imran, Shoaib, Saqlain. Compare that lineup to the current lineup that has only one batsman averaging over 40 and you know why Pakistan aren't among the top sides in the world. In contrast, India have figured out ways to win in all conditions with Zaheer Khan being the only consistent performer. Each of their top 4 averages over 50 and the number 5 batsman averages almost 60 over the last 4 years, while the captain averages over 60 when he's leading the side.

The two greatest test sides were so good because they were extremely strong in both disciplines. It's not like Marshall, Holding, Garner and Roberts had a bunch of bunnies batting in their side. Greenidge, Haynes, Richards, Lloyd is as strong a top four as you'll get. The Australian juggernaut had Hayden, Langer, Ponting, Waughs, and Gilchrist. They have combined for over 150 test hundreds!

The point I'm making is there are different ways to be a quality test side. While being a strong batting side doesn't guarantee it, being a great bowling side doesn't either. Remember, you can't win unless you score more than the opposition but you can win even if you take less wickets than them.

Edit: David Barry just showed me how to filter out finished innings so that we only have innings which were declared.


  1. Statsguru does let you filter to allow only declared innings - link.

    I don't really get why you're annoyed by the cliche. A team that consistently takes less than 20 wickets a Test is going to win very few games.

  2. Thanks for sharing the link, David. I hope you don't mind me adding it at the end of my post.

    I agree that a team that consistently takes less than 20 wickets a test is going to win very few games. However, is there a similar number as far as batting sides are concerned? The way commentators talk about it would make one think a good bowling attack is more than enough to win matches, as if the batsmen are only as good as the bowlers allow them to be.

    Besides, the ability of a side to take 20 wickets depends on the conditions as well as the quality of batsmen fronting up, just like how many runs a side scores varies depending on the conditions and the bowling attack up against it. However, you might lose a test even though you take 20 wickets but you will never lose a test in which you score more runs than the opposition.

    PS: It would have been interesting to see an imaginary series in England featuring teams comprising Indian batsmen and Pakistani bowlers against a team comprising Sri Lankan batsmen and English bowlers. At the same time, how would a series in Sri Lanka that featured Pakistani batsmen with Sri Lankan bowlers against English batsmen with Indian bowlers.

  3. Well, 28 out of nearly 2000 Tests, and 1286 results, prove the opposite of your point. In all but 2.17% of results, 20 wickets have been essential for victory. So clearly, being a quality side in almost 98% of the case involves having the ability to take 20 wickets on all types of wickets.

  4. Hey Mahek,
    As Kartikeya shows,it is only in 2% of the results where 20 wickets weren't required.Besides,
    these days captains are too defensive and usually don't take the risk.
    The Chennai and Graeme Smith incidents you mentioned are just one off incidences.

  5. Kartikeya, how many of these 2000 tests were won by teams scoring less than the opposition? How many times did a side lose inspite of taking 20 wickets?

    Clearly, taking 20 wickets isn't a sufficient condition to be a quality test side. In fact, whether a side can take 20 wickets in a test often depends on the conditions and the quality of opposition batsmen. Sure you have to take 20 wickets more often than not if you want to win consistently, but is there an equivalent figure for the number of runs a side needs to score to win most of its matches?

    Freehit, the three examples I've mentioned are from the past 4 years. If captains were too defensive these days we wouldn't have seen it happen.

  6. Mahek, I would say that your ire is better directed to another cliche: bowlers win matches, batsmen save them.

  7. Yeah you articulated it a lot better than I could. I think that's what commentators imply when they talk about needing a wicket-taking fast bowler or a unit that consistently takes 20 wickets, at the same time failing to mention the batting requirements that define cricketing dynasties.

  8. Mahek, you've got a point about the batting being as necessary as the bowling, although of course any comparisons like the one you make between India and Pakistan are more meaningful if you talk about how much better one batting/bowler order is than the other. And then there's the conditions/opposition as you mention.

    However, I don't think you're hearing the right emphasis in the "need to take 20 wickets talk". On the one hand, we've seen a period of relative lack of good bowling attacks. IMHO the Australian batsmen of 5-10 years ago had a relatively easy time, while the bowling attack has only since then started to receive a proper amount of credit for just how much difference they made. It has also meant that when stronger attacks started appearing more recently, many batting orders have been unprepared. I think the use of the cliche reflects the fact that the bowling was almost missing for a while - pointing out the extra factor when the batting was already being talked up.

    You might disagree with that, but I think there is a more fundamental point to the cliche. You speak of not being able to win without scoring more runs, but while bad batting leads to a loss without good bowling, good batting without the bowling leads to a draw. A match now and then with a risky, sporting or even mistaken declaration doesn't take much away from how bowling well enough to take 20 opposition wickets in the relevant conditions is built into the structure of the game.

    Yes, the 20 wickets doesn't correspond to runs threshold for the batsmen, but that's just part of how the game is, too.

  9. Mahek, an interesting exercise for you: covert cricket scores to football scores with the following formula: less than 350 runs in an innings = 1 goal conceded, 270-190=2, 190-120=3, 120 and below = 4. The results of the cricket games match, rather precisely, what you'd expect football games producing those scores to produce.

    Ergo, batting is defence; it takes discipline and concentration above all else. Now, the standard football cliche is that a team needs a good defence to win. Pakistan are all joga bonito, scoring goals with ease and playing no defence, but India are the better side. Attacking batting is a bit like a forward press: it makes you more likely to concede goals, but it gives you more time to score them too.

  10. Jonathan, you might be right about me not hearing the right emphasis. However, can we really say there has been a period during which there aren't enough good bowlers around world cricket? Ambrose and Walsh, Wasim, Waqar, Shoaib, Asif, Saqlain for Pakistan, Warne, McGrath, Lee, Gillespie for Australia, Donald, Pollock, Ntini and Steyn for South Africa, Murali and Vaas for Sri Lanka, Srinath, Zaheer, Harbhajan and Kumble for India, Bond and Vettori for New Zealand, Hoggard, Flintoff, Harmison for England. All of them have played in the past two decades. Surely we've had quality bowlers throughout. Even now there's Amir, Asif, Anderson, Swann, Zaheer, Steyn, Morkel. I could add a few more names like Harbhajan, Bollinger, Johnson, Taylor, Roach, Edwards but their status as quality bowlers may be arguable.

    You're quite right about the effects of bad batting as opposed to bad bowling, but shouldn't it also mean teams need good batting to ensure they're always in the game? Also, good batting can give mediocre bowlers enough time to take 20 wickets. I read a stat somewhere that in India's last 25 tests they took 20 wickets in 11 tests and won each of those matches, they had one more victory which came after England declared with 9 wickets down. I guess there are always two ways at looking at situations.

    I think the Aussie bowlers got a lot more credit than their batsmen. Warne and McGrath were the stars of the Australian dynasty, Ponting and Gilchrist to a lesser extent. Hayden was the cavalier while Langer was the foot soldier who always stood up in tough situations.

    Damn, that was a REALLY long response. Hope I didn't ramble on for too long.

  11. Russ, I'd be fine with that theory expect Spain are proving it's possible to defend with a strong attack. They keep possession of the ball, keep it up the pitch and win it back before it gets into their defensive third. Also, scoring big has to go alongwith scoring quickly if a side wants to win consistently. A score of 340 in 80 overs isn't the same as a score of 340 in 120 overs. However, having seen your website I'm quite sure you've done enough research into your theory and you're right about such a parallel between the two sports.

  12. Mahek, you are too kind. Spain's possession based game is the equivalent of bowling tight lines - it keeps the opposition from having a chance to bowl you out, without actually scoring. Their pressing game with aggressive batting - scoring quickly to maximize bowling time. It's a unique quirk of cricket that both the defending and attacking teams get to score, and that the score that ultimately counts is the defending one, but true, nevertheless.

    (The true football scoring equivalent would be one team keeping the ball until they've scored 10 goals, and giving the defending team a point for every turn-over they force. Which doesn't make cricket sound very exciting, to be honest.)

  13. Mahek, I'm certainly not going to complain about anyone else rambling on! Looking at the debut/maturing and retirement/decline dates of quite a few in your list explains why I said 5-10 years ago. Throw in the lack of cricket played by Pakistan in that time, and Australia's 3-4 man strong attack really stands out. (My perception is that apart from Warne the credit to them came more after retirement.) England improved towards the end of that period, India's results more or less reflected the strengths of their attack in different conditions. "A lack of quality bowling attacks" might by too crude, but in general, a lot of series featured fairly evenly balanced batting lineups and were decided by the differences in bowling. This continued in 04/05 and onwards even as IMHO the general bowling quality improved.

    But yes, you’re absolutely right that the batting is just as important. I just think the reason for
    the special mentions is not to play the batting down, but because the bowling is noteworthy through differences - between teams, between limited overs/longer games, in having a fairly black/white target of 20 wickets.