What JFas said.
Or look to do something in the realm of non-deterministic optimization. This will take a bit longer to converge (10 seconds+ depending on specific code and hardware) but alot of these techniques will give you more flexibility when you're not just trying to maximize mean points.
I also played in one of those formatted tourneys way back in the day. Don't think it ever got down to 1 on 1 though.
At least now I know why you're so messed up! Thanks for sharing Mr. Broken.
Next time you think you're alpha, remember that you'll never be this alpha:
http://espn.go.com/nba/story/_/id/12501628/nba-ja-adande-michael-jordan-1995-comeback-fax
I think I heard the NHLPA would piss and moan or something about OT being too long. I guess that makes sense from a labour standpoint.
From a pure entertainment standpoint, I'd much rather see 3 on 3 than another f***ing shootout.
Oh that was you that wrote that? I actually didn't know!
I liked your writing and explanation of what you did. It was very clear. Also you get a big +1 for ensuring that you tested out of sample.
I really don't like projections that basically just amount to running a regression. At least the inputs are reasonable and don't have a bunch of retarded exponents in them like SIERA.