Game Analysis

Strong Opinion – UNDER (58 ½) – Washington (-7 ½)  28   Colorado  20

I don’t see any value playing the side in this game but the total is too high based on my math. Let’s start by breaking down the offense versus defense matchups.

Colorado’s offense is only 0.3 yards per play better than average for the season (5.8 yppl against teams that would allow 5.5 yppl) but the Buffaloes are efficient because quarterback Sefo Liufau has only thrown 3 interceptions on 277 passes and only 9 going back to the beginning of last season. The Buffaloes may not be likely to turn the ball over but they’re also not likely to move the ball well against an elite Washington defense whose starters have yielded just 4.6 yppl to teams that would combine to average 6.1 yppl against an average defensive team. Colorado has a below average rushing attack (5.0 yprp against teams that would allow 5.2 yprp) and the Buffaloes certainly won’t be able to run the ball unless Liufau has some big runs from out of the pocket. The Buffs’ aerial attack also doesn’t figure to have much consistent success and is projected for just 5.0 yards per pass play against a Washington pass defense that has allowed just 4.8 yppp when their starters have been in the game (against quarterbacks that would average 6.7 yppp against an average defense). Colorado did perform relatively well (4.9 yards per play) against Michigan’s elite defense earlier this season but they’ve averaged only 19 points against the 4 best defensive teams that they’ve faced (Michigan, USC, Stanford, and UCLA) and my model projects 20.2 points for the Buffaloes in this game.

Washington’s offense has averaged 44.8 points per game on 488 yards at 7.5 yards per play while facing teams that would combine to allow 6.1 yppl to an average attack. However, the Huskies managed just 13 points on 286 yards at 4.5 yppl against USC – the only really good defense they’ve faced. Yes, Stanford does have a good defense and Washington averaged 6.8 yppl and scored 44 points against the Cardinal. However, that was one of the games in which both of Stanford’s starting cornerbacks were out. Stanford’s pass defense was horrible in the two games without both starting corners (they also gave up 42 points and 7.7 yards per pass play to Washington State, who would average 6.8 yppp on the road against an average defense, so Washington didn’t face a good Stanford defense – they faced the weakened version that couldn’t stop the pass with a collection of backup defensive backs.

Colorado has the best defense that Washington has faced and the Buffaloes’ defensive rating is very similar to Washington’s defensive rating, as the Buffs have allowed just 5.0 yppl and 19.8 points per game to teams that would combine to average 6.3 yppl against an average stop unit. As good as Washington’s offense has been (1.4 yppl better than average), they are only 0.1 yppl better than the Colorado defense and my math model projects a modest 5.8 yards per play in this game for the Huskies. Colorado has held their last 8 opponents to 24 points of fewer, including giving up just 21 points to a very potent USC attack (with Sam Darnold at quarterback), and the 45 points they gave up to Michigan is deceiving given that the Wolverines averaged only 5.6 yppl in that game. Michigan blocked two punts and ran back a punt for a TD in that game but Colorado did keep the very good Wolverines’ offense in check. The only bad game that Colorado’s defense has played was against Oregon (6.5 yppl and 38 points back in week 4) but that was against a spread offense and the Buffs’ are well suited to defend a pro-style attack. My math model projects 27.7 points for Washington in this game.

My math model projects just 48 total points in this game based on the projected statistics and that’s a reasonable prediction given that the compensated points model comes up with the same answer. Colorado games have averaged 52.6 total points per game against a schedule of teams that combine to average 60.6 total points per game, so the Buffaloes are 8.0 total points lower scoring than average. Washington’s games have totaled 62.6 points per game but the Huskies have faced teams that combine to average 63.2 total points per game, so Washington games have been 0.6 points lower scoring that average. That’s a combined 8.6 points lower scoring and the national average for points is 56.6 points, which includes overtime. The simple compensated points model (there are more advanced models that predict the same thing but this was easier to explain) would project 48.0 points also (56.6 – 8.0 – 0.6 = 48.0). The total is high because Washington’s offense has been so good, but they were shut down by the only elite (and healthy) defense they’ve faced and Colorado has averaged only 19 points against really good defensive teams they’ve faced. This game should go under the total and I’ll consider the Under a Strong Opinion at 56 points or higher.