What Use Are Stats in Predicting Cricket? | Cricket coaching, fitness and tips

What Use Are Stats in Predicting Cricket?

Recently, I had a great discussion with coach Ian Pont about the use of stats in predicting the future. Now I want to know what you think.

Can you use stats to help your cricket?


First, let’s go over the numbers.

In this case, England were playing India and scored 400 batting first. Someone looked at the stats for the ground and said the highest ever losing score was 347.

The implication was that England had the game wrapped up and the rest was already written. The truth was that India scored over 600 and bowled England out on wearing pitch to seal an innings victory.

The numbers had failed to predict the future.

Cricket astrology

As Ponty was quick to point out, this shows using historical numbers to make predictions as unreliable. History at the ground is interesting, but different players, wickets and conditions make a prediction the same as astrology. Useless.

Instead, the only way to make a reliable predicition is by deep knowledge of how people play. If you know these things - a task that takes great mastery and skill - you can be far more successful in your predictions.

And, at a playing level, if your predictions are better you are more likely to succeed.

That’s what we all want!

Whither stats?

So, does this mean we should forget about the numbers?

I don’t take as hard a line as Ian Pont on this one. I think there is room for both.

Of course, using experience, educated hunches and mastery of understanding of tactics is a powerful way to build success in cricket. If you can tap into these skills as well as people like Ian Pont (with decades of immersion in the game) then you should use this ability.

It’s proven to work, and work quickly.

But don’t throw out numbers because something else works better. You can still get insights that allow you to plan better. And that will also give you greater success.

Let me give you an example

Counting right

Last year at my club there was an obvious numerical trend across games: The first 10 overs. If the team scored more than 30 runs for less than two wickets they won the match.

However, the average score in the first 10 overs was 29–2.

So, the obvious solution to this is to plan to score more runs in the first 10 overs while playing safely. There are many ways you can do this, but the clear winner is picking up more singles; another stat we noticed on the way through the season (the higher the score the greater percentage of runs came in singles).

With this knowledge, the team can do a few things differently next year including building training around the top three improving strike rotation in safe areas. That’s preseason practice sorted for the top three.

Counting wrong

The arguement against this approach is to point out that it’s obvious that improving strike rotation will help your team. Who wouldn’t want to do that?

And that is correct, but with numbers to back things up, we know it to be the most effective option. Naturally, we still can’t predict the future and even changing the average first 10 over score to 40–1 might not do the job, but it’s still giving you focus.

Where you can go wrong here is throwing out the numbers altogether. They may only show you whats obvious, but now you have some facts to back up your argument. And, unless you are the grand lord of all you survey, you are going to need to do some convincing to make changes.

But the numbers can also go wrong too. You need to be careful.

To go back to my example, if we scored 31–1 in the first 10 we can’t relax thinking we have won. If you were to do that you are decreasing your chance of success.

All you know is that 31–1 is a decent start on average. Like 400 in the first innings, it’s improved your chance of success but offers no certainties. Trends can be bucked, records can be broken.

So, knowing your numbers is good but it’s like knowing your PB in a marathon: You might run a PB and still finish last. It depends on other factors that history can’t predict with certainty.

So, I’d argue it right to use both stats and skill to prepare and decide on tactics. Let’s not fight about it.


  • Predicting the future is hard and essentially a guess, but useful for cricketers.
  • Predictions can be based on stats or deep knowledge.
  • Neither is totally right, so use both to improve your chances of success.
  • Don’t let your predictions lull you into a false sense of security. Everyone is wrong sometimes.

Broadcast Your Cricket Matches!

Ever wanted your skills to be shown to the world? PV/MATCH is the revolutionary product for cricket clubs and schools to stream matches, upload HD highlights instantly to Twitter and Facebook and make you a hero!

PV/MATCH let's you score the game, record video of each ball, share it and use the outcomes to take to training and improve you further.

Click here for details.


I feel that people who RELY on stats have a real issue. As we saw with something as basic as 'predicting' the outcome of elections, statistics are ONLY a tool to potentially back up facts. And even then they should be come with a cautionary warning.

Cricket, like baseball, is a stats-based game. We use Run Rates to guide us, averages to help select players and previous performances to shape our thinking.

But statistics are simply a tool for any coach or player to utilise and not to make decisions on. We often hear a wonderful story about how a team has used 'Moneyball' tactics to win a tournament by being clever with statistics. For every one of those there are 100 that don't work, which we never hear about. Because the bottom line is player performance based on a series of dissimilar results against completely different sides, environments and in other circumstances cannot tell us the 'best' way to win a match. In T20 cricket, stats are not needed to know that boundaries and dot balls are important. So picking guys who can hit the fence regularly is a de facto plan. Stats would show us who can do that in the way they play.

On an individual basis it may well be highly useful to know a batsmen struggles against spin, Or that a certain bowler bowls better to left handed players. There might be something TECHNICALLY sound in observing those statistics - in that a player has a flaw we could expose in planning, to give us the best chance of success. I could pick a batsman who has a strike rate of 200 and hits 6s, over a player who has a lesser strike rate and is not a boundary hitter. Similarly, I could prefer a batsman who bats 'long' by occupying the crease in balls faced, over a batsman who simply swings at every ball he faces and doesn't bat past 25 balls.

But these are obvious. And I personally feel we are desperately being lead towards statistics as 'an answer' instead of

"Don't bowl short to Warner" might be an easy thing to work out without statistics. "Malinga bowls great yorkers" or "Starc swings in in to you" would be something we wouldn't need go back to work out on pitch maps. So hereby hangs the tail: cricket pundits on TV and Radio are now relying on statistics to make interesting points. "England has bowled 38% of it's deliveries at off stump, whereas India has bowled 67%". Did we really need that to tell us England should have bowled at off stump instead of 6th and 7th stump? Possibly, that's interesting for the armchair cricket watcher. And being more recent or 'live', that sort of statistic might be useful for certain coaches/captains in their team talk.

But as the statistic becomes older, more historic, more general - the less it has any relevance other than for notional entertainment. Test Match Special has a full-time statistician for instance who blurts out a variety of interesting facts. "This is the 4th lowest score at this ground against this opposition", "That's the third time in 8 innings this bowler has got out this batsman" or "In the last 10 matches a team scoring 175 for the first 2 wickets hasn't lost on a Friday".

My personal views are that individual statistics are more relevant than team ones. That's because teams change but individuals have trends due to their decision making, style and preferences. This is something that is potentially useful as mentioned above. But for team stuff, it is ALMOST an irrelevance.

To say statistics have a place in cricket is like saying water is wet. It's obvious - but only as a servant of coaching/playing if we are unsure or need to have confirmation of our 'hunches'. And as with sports science, it isn't there to tell us what to do.

I would love to see what stats there are on how accurate using stats to deliver results, actually are.

Great comments Ponty, and I think we are broadly in agreement. You can use stats for fun and profit, but be careful with them!