Here is his write up, it's very good (but long):
Strong Opinion
Missouri 35 Oklahoma (-16.5) 45 (at Kansas City)
05:00 PM Pacific, 06-Dec-08
I lost my Best Bet on Oklahoma State as a dog against Oklahoma last week, but I still think it was a pretty good bet. After all, Oklahoma State averaged 7.1 yards per play and scored almost every time they didn’t turn the ball over (they punted just twice, but turned the ball over twice while Oklahoma had zero turnovers) while Oklahoma averaged 6.8 yppl and would have averaged 5.9 yppl if not for a 73 yard TD pass that was initially broken up by the Cowboys before miraculously landing in the hands of an Oklahoma receiver 15 yards down the field with a clear path to the endzone. Now the Sooners are a 16 ½ point favorite over a good Missouri team, which is a line that simply can’t be justified. Oklahoma has an outstanding offense that has averaged 7.2 yppl and 53 points per game against Division 1A opponents and the Sooners rate at 2.2 yppl better than average with quarterback Sam Bradford under center. Missouri’s defense is 0.3 yppl better than average, which is the same as the average of teams that Oklahoma has faced this season, so the Sooners should average around their standard 7.3 yppl in this game. Oklahoma’s defense, however, has been susceptible to good offensive teams this season and Missouri’s offense is the best that they have seen. The Tigers have averaged 7.3 yppl in 11 games against 1A competition (against teams that would allow 5.4 yppl to an average team) and their +1.9 yppl rating is almost as good as Oklahoma’s rating. Oklahoma’s defense, like Missouri’s defense, is 0.3 yppl better than average and the Sooners have allowed an average of 34 points to good offensive teams Texas, Nebraska, Texas Tech, and Oklahoma State – and Missouri’s offense is better than all of those teams. The Sooners overall advantage from the line of scrimmage in this game is just 0.3 yppl – which works out to about 2 points with all else being equal. All else is not equal, however, as Oklahoma turns yards into points better than any team in the nation. Part of the reason for that is a +22 turnover margin, which is partly due to skill (Bradford has thrown just 6 interceptions this season) and partly due to luck, as the Soooners have only lost 2 fumbles all season while recovering 14 fumbles in 11 games against division 1 foes. Fumbles are 90% random in college football, so Oklahoma has been pretty fortunate to be +12 in fumble margin, which works out to about 4 points per game. Another reason for Oklahoma’s higher scoring efficiency is their incredibly good success rate on 3rd and 4th downs and the Sooners converted two 4th downs into touchdowns last week rather than settling for two field goals. One area where Oklahoma is not good is in special teams, as the Sooners have allowed 4 kick off return touchdowns this season, which could be a problem against Missouri’s Jeremy Maclin in this game. My math model projects Oklahoma with a 573 yards to 494 yards advantage and a +0.9 turnover advantage, which would result in about a 9 point margin even without factoring in special teams, which favor Missouri. My math model has a tendency to underestimate Oklahoma a bit since they score at a higher efficiency than a normal team would given the same yardage and turnover numbers. With that being the case I decided to run a model based on compensated points and adjusted for random events such as the Sooners’ +12 in fumble margin, and I got Oklahoma by 9 points with a total of 79 ½ points. Even if Oklahoma is their normal +1 in fumble margin that model would still favor the Sooners by only 12 points. Another way to look at the math using only points is to use margin of victory averages and schedule strengths. Oklahoma played a schedule of Division 1A teams that is 7.0 points better than average and they out-scored those teams by 26.1 points per game, which makes them +33.1 points better than average based purely on scoring. Missouri faced a schedule of 1A teams that is 4.1 points better than average and the Tigers out-scored those teams by an average of 17.8 points, which makes them +21.9 points better than average based purely on points. The difference in those point margin ratings is 11.2 points. But, that assumes that Oklahoma will be +2 in turnover margin in this game (since they average +1.9 in turnovers while Missouri averaged -0.2 in TO margin), which is very unlikely since past fumbles don’t correlate very highly with future fumbles and a big part of Oklahoma’s turnover margin is their +1.1 in fumbles. That prediction also assumes that the Sooners will continue to covert visits to the redzone into touchdowns at an 87% rate, which is also unlikely (Missouri has a great offense and they are at 72% touchdown rate in redzone visits, which is outstanding). So, Oklahoma should still only be favored by 11 points in this game even if the Sooners are as fortunate with turnovers and redzone scoring as they’ve been. The line on this game should be 9 points and 16 ½ points is simply way too high regardless of what kind of math you do, so Missouri is clearly the percentage side to take in this game. My only hesitation is a 14-0 ATS situation that I discovered when I was querying how teams that have scored 50 points or more in 3 or more consecutive games tend to do in their next game. The answer is as follows. Teams that have scored 50 points or more in 3 or more consecutive games are 14-0 ATS laying 21 points or less (or getting points) against a conference opponent that played the previous week. That is the only thing keeping me from making Missouri a Best Bet in this game and I’m still close to pulling the trigger on this game even with that trend favoring the Sooners. I’ll consider Missouri a Strong Opinion at +14 ½ points or more based on the line value.