I've been working on a rugby prediction algorithm for a while now. My basic premise is that the World Rugby rankings are a good indicator of past performance, but don't give the full picture when predicting future results. I've already visualised my predictions based on form alone, and the results look reasonable - so is there room for improvement?
My hypothesis is that a more accurate prediction can be made if we take into account short-term form and past performance against specific opponents.
With this in mind, my new model combines four key metrics:
- All-time performance - as shown by current world ranking
- Recent Form: tier 1 - how well have the team performed in their last 10 matches against tier 1 opposition.
- Recent Form: tier 2 - how well have the team performed in their last 10 matches against tier 2 opposition.
- History against opponent - how well have the team performed against the specific opponent we're predicting the result for?
There's obviously plenty more that I can add to this algorithm, but with the World Cup actually starting soon (tomorrow, at time of writing), I've got a hard deadline for "just shipping" whatever I've got.
- 1New Zealand418QF2
- 2South Africa315QF4
- 5United States00
Podcasts for Nerds
I often bore my friends by going on and on about great podcasts I've heard lately. But doing this one-on-one was getting a little stale, so I've launched a weekly podcast-recommendations newsletter so I can bug lots of people all at once!