Cricket analytics

Nobody asked me to write this. I went through every completed T20 match in the Brampton Cricket League's 2025 season — 319 games across six divisions — and compared the six division champions against the other 42 teams to find what winners were actually doing differently.

Some of what came out of it:

• The median first-innings score in a winning game was 185. Score under 160 batting first and you are below a coin flip.
• Champions did not blast the powerplay. They scored about 30% of their runs in the first six overs, then built from there.
• Across a 14-game season, division winners bowled roughly 104 more dot balls than their rivals. Seven or eight a match, and the least glamorous number in the report.
• Winners hit about five more boundaries a game, worth around 24 runs. Put that together with the dot balls and you have the entire gap between a team scoring 150 and a team scoring 185.
• Depth beat star power. Champion sides had about the same number of players capable of a fifty, but scored nearly twice as many of them.

The practical use is recruitment. If you are picking for next season, this says take the player with three fifties over the one with a century.

Read the full report  ·  PDF version

The league read it and hired me. I now produce advanced analytical stats for BCL through the season, published on the league's own channels. That is the part I am most pleased about — the analysis goes to the people playing, not into a folder.

I have worked with two teams directly, building a dashboard that shows them where they currently stand and where they are losing ground — the gaps that are invisible when you only have the scorecard and a feeling about how the season went.

The other half of that work is match by match. Before a game I put together a prep sheet: what the venue and the toss are likely to do, how the opposition scores and where they are vulnerable, which of their bowlers to attack, and what our own side should not change. It ends in three or four instructions the captain can use on the day.

Read a redacted prep sheet — real work from a completed season, with names and identities removed.

Most club cricket decisions get made on reputation and memory. A surprising amount of that falls apart the moment you put numbers next to it, which is the useful bit.

I am helping as product manager on the Data Premier League, an international cricket analytics hackathon run out of the UK and India, with finals in Bengaluru.

My job is not to run this year's event. It is to turn it into something that can run every year — a product rather than an occasion. There is a real gap for a serious, accessible entry point into cricket analytics, and DPL can fill it if it is built to last.

So far that has meant mapping the whole thing end to end: how people find it, register and form teams; what onboarding and the hackathon experience should feel like; what happens to participants after the final; and the metrics that tell us whether any of it worked. That map is what next year gets planned against.