Awesome work! I’m curious why a poisson is used here instead of a negative binomial? I’d love to see an article on your points model if you ever get a chance.
Thanks for the comment. 100% negative binomial is more appropriate here since there is a lot more variance than my Poisson model captures. In fact, my more recent models all use a negative binomial distribution instead of Poisson.
Awesome work! I’m curious why a poisson is used here instead of a negative binomial? I’d love to see an article on your points model if you ever get a chance.
Thanks for the comment. 100% negative binomial is more appropriate here since there is a lot more variance than my Poisson model captures. In fact, my more recent models all use a negative binomial distribution instead of Poisson.
I really would like to know what Morey would say about Harden. Like, 'yeah...we know.'