2009 CFB Preview: Notre Dame - Take 2

Last week, I did a post refuting Joe Schad's idea that Notre Dame will win ten games, but the numbers appeared off. I noticed the next night there was something wrong with the Monte Carlo calculator when I went to run the stats for two other teams. To fix the error, I re-downloaded the calculator from xlssports and performed the analysis again. The results look much better this time.

When you sum the individual probabilities in the median scenario, the expected value for Notre Dame's wins is 8.8, but last time the expected value for the record was 7-5, clearly wrong. At nine, the number appears to be correct now.

Offering Joe 2-1 is still slightly +EV, especially when considering that Notre Dame's preseason lines are always overinflated, but not quite to the degree I'd stated last week. Whatever, it's my dumb fault, like 5Dimes offering Alvaro Quiros -130 against the World's Greatest Golfer, so if Joe reads this or the tweet I send him, he can still have the bet.

I'm backing off the college football previews to one a day. It's too much work to bang out two of these a night, especially with my real job taking up more time over the next month or two.


adam said...

Woo hurricane season.

moneyline said...

Nit picking:

- ND -20 @ Purdue is off by almost a full TD. You're basically saying that Purdue is 16.5 points worse than Michigan and only 4 better than Wazzu. The Boilers are not quite that bad.

- You're short changing the Irish a good deal with ND -12 vs. Navy. OSU is -22 vs. Navy Week 1, and we know based off the USC at ND/OSU GOY lines that OSU is not 10 pts > ND.

The two "mistakes" pretty much cancel each other out I think.