Show HN: Fantasy Sports Solved through Dynamic Programming (opens in new tab)

This looks really nice. I wrote something similar for personal use during the NFL season. Selfishly, I hope no one finds out about this.

your optimal lineup for today has danilo as a SF, projected to get 27+ points, but he hasn't played a game all season, and is out tonight... which makes me wonder...

what data are you using to project averages? 2012-2013 season?

Congrats on launching! Nice looking site, too. Was this a scratch-your-own-itch type of project? Did you do it on the side? I'd be interested to hear where it came from and where it's headed.

Nice work. My brother has been doing this manually for sometime... useful tool.

I made a python tool for doing this a couple of weeks ago. It solves the "knapsack problem with multiple constraints", using the OpenOpt knapsack module. Find the maximum sum of FPPG (fantasy points per game), with the constraints of $60k (fantasy) salary sum, and player positions.

It didn't work very well though, my generated NBA picks finished in the bottom 50% on FanDuel. For the next iteration, I'm going to try optimizing for MPG (minutes per game).

Fantasy betting has the benefit of a legal loophole, so it's growing much quicker than traditional sportsbook line betting.

Ha, I used to work at StarStreet - Rotogrinders has a similar tool, and you can bet people have intense spreadsheets to help them make their decisions; how has this tool been performing in the NBA season?