http://www.bayesff.com/bayesian101/
http://www.bayesff.com/historical/
I created this site for last year's fantasy football season and wanted to pass it along, mostly in the hope that you guys find it interesting, but also to see if anyone might have an interest in working to help improve the model for next season.
I don't really have any formal background in this type of analysis (to be honest I'm not 100% certain this falls under the purvey of 'machine learning', though I think it does), but I was frustrated with my in-season rankings and this seemed like a natural way to think about the problem.
I did the analysis in R/JAGS (and the data processing in Stata), but am working now on learning Python in the off-season.
Again, my expertise is likely lacking compared to most people on this sub, but I have a bunch of data, code for my first cut at the problem, and am excited about working it more/open to collaboraters, so please feel free to get in touch. I'm also very interested in coming up with tools to make the info more dynamic/easier for users to understand. I'm working on learning how to code some of those tools up myself, but this is a fun side-project to a (non ML/developer oriented full-time job), so it's a bit slow going.
Cheers,
Nate
[link][comment]