Stats, science and sport. Sexy.

This week's Grade Cricketer podcast (one of my favourites) featured an interview with Ben Jones, editor at CricVis. Those who've watched any cricket this season would be familiar with the WinViz pie charts constantly being displayed, which is essentially a real-time result predictor.

I've always been a bit of a nerd for stats, especially when they're anything to do with sport. I used to even fill out imaginary cricket scorebooks, complete with ball-by-ball bowling data to ensure realism. Whilst I'm sure activities like these stifled my social development, I still loved it.

Anyway, Ben was on TGC to have a chat with the boys about this new age of data in cricket, and how uncovering new metrics are giving stakeholders fresh insights into the game we all love (to hate).

One of the biggest "eureka" moments in listening to Ben was how they've introduced relativity to traditional statistics in an effort to give them more context. For the longest time, we've simply been concerned with averages, strike rates and run rates, but we've only ever looked at them in the most simplistic means, as opposed to comparing figures with more meaningful benchmarks.

Bradman averaged 99.94. Mike Hussey averaged 51.52. To a cricket fan, these both look good. To those unfamiliar with the game, Huss's average might look pretty poor when stacked up against the Don. Forgive me for I don't have numbers to support the following statements, but I find it interesting that CricVis are able to measure the feats of batsmen like these two in a different context.

For example, Bradman batted at 3, Huss batted at 6. What are the value of those averages in context with runs scored by numbers 3 and 6 respectively? I don't know, but CricVis does, and this is valuable information!

CricVis dives deep into cricket. Really deep. Measuring and comparing swing deviations for bowlers day-to-day, and even time-of-day to time-of-day. A team is scoring at 4.16 runs per over. In the simple sense, that sounds good. But is it? Is that a fair scoring rate given the context of how the bowlers are performing versus how playable the pitch is? Again, I don't know - but you bet your arse CricVis does.

This might not seem like I learned a lot, because I still don't necessarily process the CricVis methods all that clearly, but having my eyes opened to this sort of modern data approach to a sport I love is of great interest to me. I'd love for my professional life to revolve around sport. I also love analysing statistics and data. A match made in cricket heaven? Maybe.

So, today I learned a little more about the value of data analysis in sport. And I also learned that when applying relevant metrics to Don Bradman's untouchable average, it may just be a little inflated. Not to take anything away from our Sir Don.