Impact of course variance on in-play model

There are obviously some courses that have higher variance than others, thus increasing win probability for chasers etc. Is this factored into the model?


Course-specific variance is in the model. This is something that is set pre-tournament however, and isn’t updated based on play during the event (although it could be).

1 Like

How do you determine it? Just historic results?

Yeah just historical scores. Definitely could incorporate course characteristics, but haven’t explored that.