Saturday, August 9, 2014

Catches Win Cricket Matches, But Probably Not Often

There is a cliché in cricket that ‘catches win matches’, as if catching was a major factor in determining who wins. I do not believe that - and you may not as well - but here are my personal reasons for thinking that quote is misleading.

I cannot find the exact numbers anywhere (assuming they exist) but I would guess that, at the highest levels of cricket, the difference between the best catching team and the worst catching team would be on average around one dropped catch per match. Or if you want to think of it the other way in terms of catching chances taken, the best catching team probably only snaps up one more catching chance per match than the worst catching team. Sure, we have seen teams that have dropped four or more catches in a match on occasion, but that is a relatively rare event.

How much does that one missed chance cost the fielding team on average? Well, in Test matches batsmen (including specialist bowlers) make about 30 runs on average before they are dismissed. How many Test matches are decided by 30 runs or less? Not many at all. Proponents of the ‘catches win matches’ idea might point out those occasions where teams have dropped batsmen, and those batsmen have gone on to make over 100 more runs, but what are not as easily remembered are those instances where batsmen have been dropped and it has not cost very many runs at all.

The effects of dropped catches might be a little different in one-day internationals and Twenty20 matches. Margins in these matches are generally much closer than Test matches. On the other hand batsmen make less runs on average in those matches, and there are less catching chances per match, hence limiting the effect of being the worst rather than the best catching team.

After going through that reasoning I had a look around the internet to see what others had found on this question. I found this 2013 article which argued that catching helps to win matches (which I don’t disagree with), but not as much as other aspects of the game.

The study analysed over 100 one-day international matches, and gave a value to each dismissal based on how many more runs would be expected to be scored given the wickets and number of balls remaining. It found that a batsman who is one standard deviation above the average batsman contributes 8 runs above par to his team’s total. However a fielder who is one standard deviation above average restricts the other team’s total by less than 2 runs compared with the average fielder. (Fielding is used synonymously with ability to catch here, ground fielding is ignored.)

That ‘2 runs’ seems quite a bit less than the figure I came up with through my rudimentary reasoning. But if you factor in the authors’ arguments that:

a)      most catches and run-outs in ODIs occur near the end of matches where they do not have much impact, and
b)      that they split the credit for a dismissal between the bowler and the fielder, and
c)      that their result is not the difference between the best and worst case scenario,
then their numbers are probably more comparable to my reasoning. (In any case I would go with their reasoning since they have thought through the effects a lot more.)
I am still not sure I agree though with their suggestion that the role of catching might be greater in Test cricket than in ODIs. I would agree that a fielder that is one standard deviation above average would probably restrict the other team’s total by more than the 2 runs compared with the average fielder that they restrict it in ODIs. But as noted above, given the larger margins in Test cricket, this may not necessarily translate into a higher effect on the probability of winning.
So catches don’t usually win matches. All of the above though will never take away from the magic of this moment.

No comments: