Re: ***NBA FUTURES THREAD***
[ QUOTE ]
I don't know any thing about the NBA, but I want to get a little action down so I took pinny's line as the true line, assumed a Poisson distribution and then just looked for value at the two books I have right.
[/ QUOTE ]
Requisite ImStillBen response: "No"
Wong is just plain wrong about using Binomial distributions to model season wins in sports where they play a lot of games. For example, the 1995-96 Spurs won 59 games. With many returning players it would not be unreasonable to expect them to win 53 games in the 96-97 season so lets assume that that was the expected average number of games for a hypothetical Pinny line.
During the 96-97 season the Spurs won 20 games, for those keeping track at home, thats a 1 in 5 million event (to win 20 or fewer games). That event should only happen once in about 200,000 seasons.
So is it possible that we witnessed a 1 in 200,000 year event? No, not really - the truth is that the standard deviation is much higher than what Poisson gives you, and as a result RSW totals close to the posted lines occur less frequently than Wong's theory predicts.
|