
Posted by < nate on 4/24/2009, 12:41 pm, in reply to "Re: Playoff odds"
68.3.44.X
I shudder to join the nerd masses (nerds? in Ultimate? Who knew?), but it seems to me that basing anything on a ten game sample with a slew of dynamic latent variables (attendance, injuries, etc.) is silly. I like the Nabity ratings as much as anybody, but whatever their merits, they've lacked predictive power this spring. They called the correct winner 22 of 32 times - you would expect 16 on chance alone - but were only right on the score 3 times. Sure, that's tough to nail, and you might want to give credit for being in the ballpark. But I'll just bust out my pocket protector and slide rule and note that in the other 29 predicted games, the predictions were off one way or the other by 172 total points. That's an average of over five per game overall! So if Nabity says you're going to win by two, you can reasonably expect anything from (on average) a win by 7 to a loss by 3. That's a huge swing in Ultimate, yah?
Don't get me wrong, I like the stats and simulations and all that. Woop Woop Woop. Maybe we should throw out some betting lines and get real Vegas-y with this operation. People could only bet on their own teams, of course - no Pete Roses here. I mean, it feels bad to get eliminated from pool play on double game point, but to get eliminated AND lose $500 'cause you didn't cover the spread? That's real pain. Does Volo still throw the behind the back hammer when he's got two weeks' pay on the line? Let's find out!
Anyhoo, don't believe the hype. I'd venture that this tourney is up for grabs, moreso than usual. Injuries, missing stars, and 20 mph winds on Saturday are gonna make things exciting. And give people built-in excuses as to why Champagne Room walked away with the trophy.
Message Thread:
![]()
« Back to thread