Sunday, May 28, 2006

Batting first in the subcontinent

Responding to my previous post, kshah pointed out that teams batting last often still have to deal with difficult pitches in the subcontinent. I therefore spent a while at Statsguru to see what it told me.

The year brackets are slightly different to those of Z-Score, because I was entering the wrong years. Statsguru refuses to exclude Bangladesh matches, so they are included in the stats. I've only included matches in India, Pakistan and Sri Lanka, in which there was a home team. (I think this means that, eg, the Pakistan v Australia matches in Sri Lanka and Sharjah don't count.) Percentages in the following table are the percentage of victories that went to the team batting first.


There are lots of arguments you could have about various numbers in the above table. (For instance, Pakistan only lost two home Tests between 1969 and 1983 - did batting first or second make a difference?) Nevertheless, the figures for the past few years at least agree with the overall trend of an advantage to the side batting second, though it is much less pronounced than outside the subcontinent. I'm a little bit tempted to say that the figures in the last column for India and Sri Lanka are noise, given the 94-98 numbers.

I was surprised to see how team-batting-second-friendly Pakistan is.

Let us all stop criticising Ricky Ponting.

When you win the toss, ninety-nine times out of a hundred you should bat without hesitating. The other time, you should hesitate, and then decide to bat.

I've seen that piece of wisdom in various forms attributed to several Test captains from both hemispheres. However, it is no longer valid.

An analysis by Z-Score shows that it is much better, on average, to bowl first these days. The linked page may not display properly in your browser (it doesn't in my Firefox), so I'll summarise the key points.

Before WWII, 59.1% of Test wins went to the side batting first. From 1945-69, that figure was 54.0%. From 1970-84, it was 52.3%. From 1985-1994, 51.2%. From 1995-99, 43.2%. From 2000-05, 41.5%.

It's not clear precisely what has caused this gradual change. Z-Score mentions pitches, and this is probably the biggest factor. Before WWII, batting was usually easiest on the first day, and teams batting first averaged about 30 runs in the first innings more than the team batting second. Pitches would often be very deteriorated by day 5.

Unfortunately, I wasn't alive for most of the 20th century, so I don't know if pitches did start to get slowly better after 1945. It is certainly the case that many pitches today are extremely easy to bat on on day 5, so it makes sense that the team batting last should be having so much success.

But why should it be so favourable to the team batting second, rather than just being 50-50? Here Z-Score points out explicitly something that I've never seen written anywhere else, though in some way we are all aware of it. Teams chasing do just enough to win. Teams who set a target have to be more conservative and use up extra time to minimise the chance of losing. No-one until now seems to have noticed that this phenomenon translates into a sizable advantage for the team batting second. (Perhaps Steve Waugh realised it? He certainly won the toss and bowled frequently, but he also said that it didn't matter, since the better side would win.)

So, if a captain wins the toss and scores 400, and the team batting second makes 500, the captain should be criticised for not bowling first.

This page is powered by Blogger. Isn't yours?

Subscribe to Posts [Atom]