# Why captains should field first in day ODIs

S Rajesh talks about the bias towards the winner of the toss in day-night ODIs in this Numbers Game column. He says that this bias is less pronounced in day games. This masks a couple of underlying facts – there is a substantial bias towards the team batting second in day ODIs, and captains aren't as aware of it as they should be (though they may be learning). This bias is probably caused by the natural advantage the teams batting second have (in knowing precisely how many runs they need to win) and perhaps some help from early-morning moisture.

First of all, let's look at the overall results between top-eight sides since 2000. In day games, the side batting first has won 144 games and lost 204. In day-nighters, the corresponding figures are 202 and 167.

The probability that such imbalances would happen by chance are about 7% and 0.1% respectively – there's still some reasonable chance that day-nighters are fair, though I would agree with those who say that it depends on the ground. Clearly the night-time conditions are having a big effect (after all, the direction of the bias is reversed going from day to day-night), and it seems reasonable that the atmosphere in different cities would affect the white ball differently under lights.

It's interesting to break these results down further. Day-nighters, win the toss and bat: win 165, lose 123. Day-nighters, win the toss and field: win 44, lose 37. That latter result is not statistically significant, but it is at least suggestive that captains know what they're doing when they bowl first in a day-nighter.

Day games, win the toss and bat: win 55, lose 90. Day games, win the toss and field: win 114, lose 89. Clear message to captains: don't bat first in day ODIs.

Since 2000, captains batted first in about 42% of day ODIs, as opposed to 78% for day-nighters. Clearly they're aware that it's often better to bowl first in day games, but 42% is still far too high.

In 2008, though, there have been just four day ODIs where a captain has won the toss and batted, and seventeen instances of winning the toss and fielding. Are the captains learning or is it just a blip? We'll find out over the next couple of years.

I'd like to thank Pelham Barton for some discussions on this topic.

Comments have now been closed for this article

Sir Garfield Sobers is till date the ONLY GENUINE ALLOROUNDER.He can bat any where from 1 to 11,Open the Bowling,Bowl spin,Field close in or at deep. He is the only one GENUINE ALLROUNDER cricket has seen so far

The analysis already excludes them. "Top-eight sides": not Bangladesh, Zimbabwe, etc.

Can we have the same analysis without Bangladesh, Zimbabwe and non-Test playing teams? Or does the analysis already exclude them?

ShawnB, I'm not sure where you've got that impression from.

Teams batting first are 12 wins and 19 losses since 2000 (top-eight sides) in Sri Lanka, 2 and 3 at the Premadasa.

Not in Sri Lanka, especially at Premadasa Stadium!

Tejas, that's absurd. There is a very obvious advantage to batting second. The first batting side bats to maximise their average runs.* The second to bat bats to maximise their probability of reaching the first team's total. Everything else being equal this would lead to bowling first being the optimal decision. However, everything else isn't equal, which is why you use these stats. *It might be the case that this isn't the best strategy. A team who consistently make 210 might win less than a team who make 100 half the time and 300 the other half.

The day matches data and the day-night match data suggests one simple fact; bat in the afternoon and win! Can you dig further and find of how many of these matches were won batting in a hot summer afternoon; winters and in rainy seasons?

This arguement of batting first in day ODIs is particularly valid as it has been shown that most one day pitches in the sub continent(excluding ones in maybe Sri Lanka and certain parts of India) do not deteriorate much over the course of the game. This stat could also be skewed due to the tournaments we have had this year. The asia cup and the Kitply cup were huge indications that a total of even 280 or so which was almost a fortified score some 10 years ago can be chased down on a regular basis. This is proving that although incidents like THE Jo'burg ODI maybe scarce batsmen are getting used to chasing down convincing totals at over a run a ball due to the inflence of T20 cricket. This might not have anything to do with the toss but merely greater experience in continuing at a pace of more than a run a ball through the course of the innings.

The argument seems to have been built on the fallacy of causation vs correlation. "If Sachin Tendulkar averaged above 40 exactly 8 years before today, bat first." There might be a correlation, but no logic supporting the theory.

Aaron, it's interesting that you mention Australia. You're correct in saying that they usually bat first - my suspicion is that they do this because all their home games are day-nighters, where it is sensible to bat first. Since 2000 in day games against top-eight sides, they've batted first 23 times, with 16 wins and 6 losses (1 no-result). They've batted second 9 times, 7 wins and 2 losses.

Australia is particularly good at batting first in ODI's (the following stats are a few months old, but they should still be close enough). Batting first, day games, they win 73%. Second-best is Sri Lanka at 49%. That's a huge gap. Batting second, Australia wins 78%, second is South Africa at 71% - a much smaller gap.

The same is true in day-nighters, though to a lesser extent. The gap between Australia and #2 in batting first is 13 percentage points, and batting second it is only 7.

(Edited: I called myself Aaron when I posted this...?!)

Sir Garfield Sobers is till date the ONLY GENUINE ALLOROUNDER.He can bat any where from 1 to 11,Open the Bowling,Bowl spin,Field close in or at deep. He is the only one GENUINE ALLROUNDER cricket has seen so far

The analysis already excludes them. "Top-eight sides": not Bangladesh, Zimbabwe, etc.

Can we have the same analysis without Bangladesh, Zimbabwe and non-Test playing teams? Or does the analysis already exclude them?

ShawnB, I'm not sure where you've got that impression from.

Teams batting first are 12 wins and 19 losses since 2000 (top-eight sides) in Sri Lanka, 2 and 3 at the Premadasa.

Not in Sri Lanka, especially at Premadasa Stadium!

Tejas, that's absurd. There is a very obvious advantage to batting second. The first batting side bats to maximise their average runs.* The second to bat bats to maximise their probability of reaching the first team's total. Everything else being equal this would lead to bowling first being the optimal decision. However, everything else isn't equal, which is why you use these stats. *It might be the case that this isn't the best strategy. A team who consistently make 210 might win less than a team who make 100 half the time and 300 the other half.

The day matches data and the day-night match data suggests one simple fact; bat in the afternoon and win! Can you dig further and find of how many of these matches were won batting in a hot summer afternoon; winters and in rainy seasons?

This arguement of batting first in day ODIs is particularly valid as it has been shown that most one day pitches in the sub continent(excluding ones in maybe Sri Lanka and certain parts of India) do not deteriorate much over the course of the game. This stat could also be skewed due to the tournaments we have had this year. The asia cup and the Kitply cup were huge indications that a total of even 280 or so which was almost a fortified score some 10 years ago can be chased down on a regular basis. This is proving that although incidents like THE Jo'burg ODI maybe scarce batsmen are getting used to chasing down convincing totals at over a run a ball due to the inflence of T20 cricket. This might not have anything to do with the toss but merely greater experience in continuing at a pace of more than a run a ball through the course of the innings.

The argument seems to have been built on the fallacy of causation vs correlation. "If Sachin Tendulkar averaged above 40 exactly 8 years before today, bat first." There might be a correlation, but no logic supporting the theory.

Aaron, it's interesting that you mention Australia. You're correct in saying that they usually bat first - my suspicion is that they do this because all their home games are day-nighters, where it is sensible to bat first. Since 2000 in day games against top-eight sides, they've batted first 23 times, with 16 wins and 6 losses (1 no-result). They've batted second 9 times, 7 wins and 2 losses.

Australia is particularly good at batting first in ODI's (the following stats are a few months old, but they should still be close enough). Batting first, day games, they win 73%. Second-best is Sri Lanka at 49%. That's a huge gap. Batting second, Australia wins 78%, second is South Africa at 71% - a much smaller gap.

The same is true in day-nighters, though to a lesser extent. The gap between Australia and #2 in batting first is 13 percentage points, and batting second it is only 7.

(Edited: I called myself Aaron when I posted this...?!)

well stats are there for every one to see but certain things are not considered in this discussion and the most important of them is playing conditions especially when you compare subcontinent and English or Kiwi conditions.they have to be analyzed to say whether or not toss is dominating factor or rather it is the only bane of this game where unfair advantage is given to one side,

It is an interesting article when you consider that recently the Australian team almost always bats first after they win the toss (that seems to be the general impression). I wonder if you took out the results of a really good australian team, would that skew the analyis towards batting second even further?

No featured comments at the moment.

It is an interesting article when you consider that recently the Australian team almost always bats first after they win the toss (that seems to be the general impression). I wonder if you took out the results of a really good australian team, would that skew the analyis towards batting second even further?

well stats are there for every one to see but certain things are not considered in this discussion and the most important of them is playing conditions especially when you compare subcontinent and English or Kiwi conditions.they have to be analyzed to say whether or not toss is dominating factor or rather it is the only bane of this game where unfair advantage is given to one side,

Aaron, it's interesting that you mention Australia. You're correct in saying that they usually bat first - my suspicion is that they do this because all their home games are day-nighters, where it is sensible to bat first. Since 2000 in day games against top-eight sides, they've batted first 23 times, with 16 wins and 6 losses (1 no-result). They've batted second 9 times, 7 wins and 2 losses.

Australia is particularly good at batting first in ODI's (the following stats are a few months old, but they should still be close enough). Batting first, day games, they win 73%. Second-best is Sri Lanka at 49%. That's a huge gap. Batting second, Australia wins 78%, second is South Africa at 71% - a much smaller gap.

The same is true in day-nighters, though to a lesser extent. The gap between Australia and #2 in batting first is 13 percentage points, and batting second it is only 7.

(Edited: I called myself Aaron when I posted this...?!)

The argument seems to have been built on the fallacy of causation vs correlation. "If Sachin Tendulkar averaged above 40 exactly 8 years before today, bat first." There might be a correlation, but no logic supporting the theory.

This arguement of batting first in day ODIs is particularly valid as it has been shown that most one day pitches in the sub continent(excluding ones in maybe Sri Lanka and certain parts of India) do not deteriorate much over the course of the game. This stat could also be skewed due to the tournaments we have had this year. The asia cup and the Kitply cup were huge indications that a total of even 280 or so which was almost a fortified score some 10 years ago can be chased down on a regular basis. This is proving that although incidents like THE Jo'burg ODI maybe scarce batsmen are getting used to chasing down convincing totals at over a run a ball due to the inflence of T20 cricket. This might not have anything to do with the toss but merely greater experience in continuing at a pace of more than a run a ball through the course of the innings.

The day matches data and the day-night match data suggests one simple fact; bat in the afternoon and win! Can you dig further and find of how many of these matches were won batting in a hot summer afternoon; winters and in rainy seasons?

Tejas, that's absurd. There is a very obvious advantage to batting second. The first batting side bats to maximise their average runs.* The second to bat bats to maximise their probability of reaching the first team's total. Everything else being equal this would lead to bowling first being the optimal decision. However, everything else isn't equal, which is why you use these stats. *It might be the case that this isn't the best strategy. A team who consistently make 210 might win less than a team who make 100 half the time and 300 the other half.

Not in Sri Lanka, especially at Premadasa Stadium!

ShawnB, I'm not sure where you've got that impression from.

Teams batting first are 12 wins and 19 losses since 2000 (top-eight sides) in Sri Lanka, 2 and 3 at the Premadasa.

Can we have the same analysis without Bangladesh, Zimbabwe and non-Test playing teams? Or does the analysis already exclude them?