Vegas Knows Best: How The Media Can Make You A Better Football Team

The hits just keep on commin': Penn State opened as a ten point dog against USC Sunday night.  No respect, I tell ya.

While the BCS match-ups were officially announced this weekend, we've known the Rose Bowl pairing for a while now.  Vegas has had a lot of time to figure this one out.

The interesting thing, then, was trying to guess how the experts would handle a very high profile game between two very high profile schools. Unlike the past two years, Pete Carroll is actually being matched up with a Big Ten champion...although you wouldn't know it based on some of the stories we've been linking here at BSD:

"Penn State sucks.*"

[*not an actual quote, but you get the idea]

This puts Vegas in an interesting position. Frankly, you have no reason to believe that USC is substantially better than Penn State. Yes, they've been dominating on defense, but they haven't exactly been facing offensive juggernauts on a weekly basis. Besides, Penn State has been quietly playing some decent defense of their own, and have one of the most productive offenses in the nation (points/game and yards/game are better than any other Big Ten team, top 15 nationally in both categories). Then, to add, you've got the often misguided but interesting transitive property in which we use Oregon State as the middle man.

Several of the gambling blogs I check in on from time to time made guesses last night at all of the bowl game odds. The Money Line Journal and You Guessed It, Frank Stallone came up with -6 and -7, respectively, for the Rose Bowl.

These aren't totally taken out of thin air, of course. Besides being familiar with their trade based on an entire season of looking at lines, there are also third part tools available. Vegas Insiders runs a power ranking specifically designed to predict spreads, and they have the two teams separated by just 3 points (with a home field adjustment, USC would be a 6 point favorite based on their model, click here for a look at why this is probably appropriate). Sagarin does a similar exercise with his PREDICTOR formula; he has Penn State and USC separated by 5.2 points (8.2 with a home field adjustment).

The other BCS bowls:

So not all of the real lines were precisely predicted by these models, but I think you can explain many of the differences away. Alabama is going to be playing in a Sugar Bowl that will probably be 90% Tide fans, that is probably good for at least three points. Besides, non-BCS schools have always been difficult to project in these formulas because their schedule is set-up so different. Texas is considered an elite team hosed by the BCS, and so there is obviously some public perception built into that Fiesta Bowl line that might otherwise not exist. The Orange Bowl and MNC are very close to the predicted value. 

So why is it that the public seems to be so off on Penn State?  They aren't in the Mountain West, have just one loss and have been blowing out most of the teams on their schedule.

Sports Investments might have found an answer:

So why does a ‘true' line of 6.5 turn into 10? Reasons below:
-USC dominates the rose bowl every year.
[...]
-USC has the greatest defense in the history of college football.

More cliches are listed there, in jest of course, click though for a good review of the premise most LA Times articles will be based on for the next couple of weeks.

Update: So the lines for the rest of the bowl games have been released and, um, brace yourselves (per Covers.com):

Wisconsin vs Florida State (-5)
# 20 Northwestern vs #23 Missouri (-13.5)
Minnesota vs Kansas (-10)
Iowa (-3) vs South Carolina
#18 Michigan State vs #17 Georgia (-7.5)
#10 Ohio State vs #3 Texas (-10)
#6 Penn State vs #4 Southern California (-10)

That's a total spread of Big Ten +53. Spread out over seven bowls it averages to +7.5 per game.