Thanks for joining me!
“You’d better listen to her, because the Pentagon does”
Top Gun (1986)
A bit about me before I get into the numbers:
It’s easy to have an opinion, and particularly easy to broadcast that view online. Filtering out the noise is a challenge.
So why should anyone care what I think about cricket?
Well, my cv for starters- Masters degree in Physics from Oxford (4th year was focused on simulations of Earth’s atmosphere), then qualified as an accountant, spent 2 years in Banking Front Office (where I cut my teeth on excel modelling), and after a further role in Banking Finance I’m now working for a FTSE-100 retailer, doing modelling and strategy.
It’s not quite the Pentagon, but you should listen to me, because some people at a FTSE-100 retailer do.
In 2011 I built a test match simulator – which could predict the outcome of an innings from a given starting point, based on ball by ball bowler vs batsman probabilities, and running the simulated innings enough times to get a reasonable sample (>1,000). This was mainly for gambling, and it works.
Later I expanded this to cover the two white ball formats, though the 50 over model has always received more attention than the 20-20 one – I don’t mind 20-20, but I struggle to love it.
With a full time job, and a young family, cricket data comes third on the list – and that means I will focus on red ball cricket. There’s a lot of professionals who have got further than me in 20-20, and I’m not going to stand out by splitting my efforts across 3 formats.
Let’s see if I can come up with some original thoughts, and some predictions which stand the test of time.
Ed Bayliss, Dec 2018.
Red Ball Data 