In a previous post, I mentioned that the end of the college football season has created a sense of emptiness in my life. Fortunately for me, the NCAA has implemented a ridiculously large 32-game postseason.

Pat Forde, of ESPN.com, says it best:

There are 64 lucky teams still playing, which means just 55 didn’t make the cut. That also means many of you out there aren’t very good. If your team is 6-6 or 7-5, just smile and thank a system that rewards mediocrity.

Now most people would not pay attention to the first 20-25 of these bowl games, like last night’s San Diego County Credit Union Poinsettia Bowl.

(Yes, you read that bowl name correctly. Its sponsored by a frickin’ local bank. What does it take to sponsor a bowl game these days? Can I get one of my own? I wouldn’t mind taking the bottom two teams in the league (Florida International and Duke, both at 0-12), throwing them together on a high school football field, and seeing how long it takes before people just walk out and leave. We can call it the AmishShah.Com I’d Rather Actually Bowl.)

Anyway, as I was saying, most of these games are pretty boring, featuring teams and conferences which neither dominate college football, nor gather much national attention. But a buddy of mine introduced gambling to the bowl season, and put together a pool that makes even the most boring of games exciting.

The concept is simple. Each person picks the team for each game that will cover the spread.

(For non-gamblers, spread is the amount of points a team is ‘expected’ to win by. Usual sources for this info are the Vegas betting lines, where bookies do this for a living and thus have a pretty accurate knack for determining these kinds of things. In the Championship game, OSU is favored -’expected to win’- by 8.5 points. In order to get credit for the win in our pool, OSU would not only have to win the game, but would have to win by 9 or more points. On the flipside, someone picking Florida could get credit for the pick, even if Florida lost, as long as they lost by 8 or less points.)

After picking all 32 games, each person must also assign points to each game by ranking them based on the confidence of that pick. 32 goes to the highest confidence pick, and 1 goes to the lowest. Correct picks earn a person the amount of points they assigned.

The winner of the pool is not the person who picked the most games correctly, but the person who earns the most number of points. With 42 people in the pool, a few hundred bucks are up for grabs, and I now have a pretty vested interest in the outcome all 32 games over the next 20 days.

Like, for example, last night’s battle between the Texas Christian University Horned Frogs and the Northern Illinois Huskies. Or the International Bowl, a game which earns its ridiculousness from the fact that it takes place outside of the country.

My friend sent out a giant spreadsheet last night, detailing the picks and assigned rankings of all 42 entries in our pool. Because I am such a huge nerd (5 in AP Calc, yo!), I went all Will Hunting and analyzed the hell out of it. I’m sure there are flaws in the analysis I’ve done, but whatever, I only got a 4 in AP Statistics.

(Also, if you’re not a numbers person, I strongly urge you to stop reading this right now. If for some reason you decide to read this, I give you full permission to come over, give me a wedgie, dunk my head in a toilet, and steal my lunch money. I deserve it.)

Outcome
First I looked at the consensus in the pool on the outcome of the games, ignoring confidence levels. I calculated the total number of picks earned by each of the 64 teams playing to determine which teams had the highest and lowest number of people rooting for them (Obviously, both teams would come from the same game).

The Winner(s): The Texas Bowl, with Rutgers (-7.5) earning 34 of 42 picks and Kansas State scrapping together 8. I attribute this mostly to the fact that a majority of the people in the pool are (I think) originally from New Jersey, and if not, live close enough to it now. Too bad they won’t be able to watch the game.

Points of Note: Last night’s Poinsettia Bowl had the lowest consensus in the pool, with both teams earning exactly half of the pool’s picks. Across the board, the average gap in number of picks between two teams was 9.81.

Where do I fit in? I ran with the herd and also picked Rutgers to win. I (correctly) picked TCU in last night’s game. Both games are high point games for me (25 and 28 points, respectively).

Confidence
I then took a look at how the 32 teams fared in terms of confidence from the pool entrants.

Mean: The average confidence in each time gives an idea of which team our pool as a whole is most and least confident in. San Jose State earned the lowest confidence, with an average ranking of 8.19 (from 16 picks). True to form, Rutgers earned the most confidence, with an average of 23.29 (from 34 picks).

Standard Deviation: The standard deviation gives an idea of just how much our pools agrees with each other on confidence. The lower the standard deviation, the more we ‘agree.’ Texas Tech earned this honor, with a standard deviation of 6.13 (from 13 picks). South Florida seemed to cause the most disagreement (std dev of 10.71, from 29 picks).

Where do I fit in? My picks/rankings seemed fairly close to the pools as a whole. I actually picked the four schools above in each of their respective games, and assigned rankings to each that were within 1 or 2 points of the pool averages.

Games of Note
Obviously, the stats of same games are more important than the stats from other.

Rose Bowl: Michigan favored by 1.5. Michigan earned 24 picks from the 42, to USC’s 18. Surprisingly, both teams earned average confidence levels of 20.

Championship Game: OSU favored by 8.5. Like the Rose Bowl, this game also had a pick spread of 24-18, with the scale tipping in Ohio State’s favor. OSU and UF earned average confidence levels of 22 and 23, respectively.

The Other 3: The pool predicts Oklahoma, Louisville, and Notre Dame to win the other BCS games, with pick margins more or less equal to the other two games above.

Where do I fit in? Michigan, Oklahoma, Wake Forest, LSU, and Not-Gonna-Say.

Personal statistics
This is my third year doing this pool, and in the past, I’ve pretty much handed my money over. In 2004, I got LAST PLACE, and my improvement in 2005 was marginal at best. In a listing of All-Time Stats (compiled by the guy running the pool), my name appears twice. Once in the list of ‘Most Pts Lost - Single Season’, and once in ‘All Time Worst Winning %.”

I’m hoping that this year will be a little different. I put a little more effort into researching and making my picks, and have already opened strong, earning 28 points from the TCU win last night.

-Highest ranked picks: LSU (32), Oklahoma (31), Michigan (30)
-Lowest ranked picks: Southern Miss (3), Cincinnati (2), Tulsa (1)
-My picks agree with the overall pool consensus of the pool in 21 out of 32 games.
-On average, my ranking for a given team is 6 points away from the average ranking that team earned in the pool.

After pulling together all of this information, and plugging these stats into a nifty online calculator, I estimate my chance of winning to be approximately 54%.