5355Re: similarity scores
- Dec 6, 2004
> Using the (3FG * 3FG%)/MP is a way to rate players in theseThat's a good way of doing it. I considered it (I think) before
> categories that I hadn't considered; I'll have to put these into
> my formulae and see how the output shapes up.
deciding to go with something along the same lines but slightly
I'm not sure, however, that it makes sense to do the same for two-
point percentage, in that you don't have the kind of sample size
issues with twos you have with threes (i.e. Player X goes 2-3 --
that obviously isn't sufficient to conclude he's a 67% three-poitn
shooter) and in that two-point attempts are already contained
somewhat in possession rate.
I'd also suggest reporting your scores out of 100 instead of 1000,
since that's the format we're generally used to.
- << Previous post in topic Next post in topic >>