Friday, November 03, 2006

Packer Final Season Record Prediction

Ask an individual to predict the outcome of a sporting event and he can be pretty stupid. Ask a whole bunch of people and the group prediction can be amazingly accurate. This is proven in Vegas all the time. In that vain, I wanted people to give me the estimated probabilities for the Packers winning each of their final 9 games:

@ Buffalo (2-5)
@ Minnesota (4-3)
New England (6-1)
NY Jets (4-4)
@ Seattle (4-3)
@ San Francisco (2-5)
Detroit (1-6)
Minnesota (4-3)
@ Chicago (7-0)

I asked the fine folks at Lambeauleapforum.net to do just that and received 20, probably biased results. See, for this to completely "work", I would have needed to ask that question to non-biased football fans. Quite honestly, that's something a Packer fan would never be accused of. Packer fans seem to fall into two camps these days; the unrelenting optimists and perpetual pessimists. Despite that, I think those opposing forces will have a tendency to balance itself out in the long run. The resulting predictions seem to support that:



While some of those individual predictions might be unrealistic, the average of those predictions seem pretty reasonable. Let's see what final season records result from those probabilities:



Cumulatively, Lambeauleap figures about a one in two chance of the Packers finishing with 6 or 7 wins, and about a one in four chance of them finishing at .500. Again, that sounds very reasonable to me. After all, even bad teams have some chance of winning every game. While the Packers might not have the talent of a .500 team, the luck of the draw might allow them to finish with a .500 record.

The 3% chance of finishing 10-6 (and probably making the playoffs) unfortunately sounds reasonable as well. The Packers simply don't have anywhere near the talent of a 10-6 team. All the balls would have to bounce the Packer's way for them to make the post season, which only means that the Packers would most likely be extremely overmatched by whomever they would play there. Not a pretty picture.

I guess I've been right to focus on the goal of the Packers finishing at .500 this year. Hey, it's something...

2 Comments:

Anonymous Rob said...

Ugh. That's misusing the idea of the "wisdom of crowds". All it states is that market mechanisms can provide the best estimator of future values since it reflects all information. Even if we believe in a full efficient market hypothesis, which is very unlikely to occur, it still doesn't guarantee anything that would be called a good estimate. And to make this really work, you'd need everyone to bet say $20 on their beliefs to get to the right incentives.

Nothing personal, just sick of people having no idea what they are talking about going on about market mechanisms for information dispresion.

10:07 PM  
Blogger rluzinski said...

I think you are taking this post much too seriously. I don't claim to understand the economic theory behind "wisdom of the crowds". Heck, I didn’t even mention that phrase. I simply stated that Vegas does an excellent job in predicting the outcome of games and wanted to do something similar. Now, 20 biased Packer fans don't even begin to compare with thousands of mostly unbiased betters with a financial incentive to be correct but I thought I had pretty much conceded that fact.

I carried on with the exercise because I wanted to accomplish two things. First, I wanted to show how misleading it can be to estimate a team’s final record by predicting the outcome of each game separately and then adding up the results. Second, I wanted to see how reasonable the consensus final Packer record prediction would even be.

Is sports-betting and the concept of “wisdom of the crowds” related?

9:47 AM  

Post a Comment

Links to this post:

Create a Link

<< Home