I went 1-3.
But I could have easily gone 4-0.
No, no. Hear me out. Alabama was putting a beat down on Houston, but the Tide not only let the Cougars back in, they let them fool around with their wife. Oklahoma missed a throat-cutting field goal inside the last minute and I saw that line close to as low as 9 before kickoff (Hey, we have to get this going on Wednesday). And Kansas State? I just refused to believe they lost to Kansas.
I didn’t say they were good excuses. I think I completely shat the bed on Gabe Kaplan’s picks. If you factor those in, I go 1-7. Fortunately, I don’t count those. Still, I’m 2-6 on the year. And I admittedly have no idea what I am talking about.
I am going to say I am getting killed by turnovers. I bet I’m averaging minus-3 in turnover margin in my losses. I’m too lazy to look it up right now. But seriously, somebody hold onto the ball like your scholarship depends on it.
But guess what? Vegas is just as bad as I am. I think. I looked at the results of 52 games factoring in the spread to the final score. Only 22 of those games were within one score (I used 8 points—a TD and a 2-point conversion) when the final whistle blew. That’s just over 40% of the games.
Granted that’s a little misleading because coaches don’t call games with the spread in mind. Coaches not named Belichick anyway. As an example. Duke lost to Wake Forest 41-36. Factoring in that Wake was a 7-point favorite, the line missed the number by minus-2 (or (41-36)-7=-2). But suppose Duke was sitting on the Wake 6 yard line with 3 seconds to play. They aren’t kicking a field goal to cover. They are going for the TD and the win. If they get that, the difference is now a minus-9.
So these numbers aren’t a perfect barometer, but they are interesting.
First just look at the average “miss.” For this just take the final score, subtract the losers score from the winner to get the difference, then subtract the number if the favorite won or add the number if the underdog won. Take the absolute value of the difference. Then average them out. For the 52 games, Vegas averaged being off by almost 11.8 points.
That’s more than two scores (TD and 2-point, and a FG). That’s not good, is it?
I don’t know. It’s tough to make sense of these numbers in a vacuum. For all I know this is a typical week for Vegas. It’s probably not (four upsets in the Top 10) but it’ll take a few weeks of this to see if this is the norm or not.
You might be tempted to think one of the greatest upsets in college football history—A 40-point Stanford underdog winning at USC—is skewing that average up. But it’s probably not pulling it too much as there were still 15 games where Vegas were off by more than 17 points.
In any instance it might not matter too much for a couple of reasons. 1) Vegas doesn’t care that they are off. They only care that you think they are right enough to where they get even money on each side of the line. 2) Even if this was statistically an odd week, big deal. You can’t know that ahead of time so you are fairly powerless to exploit it (I think… Still playing with this).
I’d give you the distribution in a dot plot, but Excel, being the piece of shit that it is, can’t even do a dot plot without having to trick it through about seven steps. If you are a little more stat-nerdy, the data has a mean of 8.2 and a standard deviation of 13.1. Eyeballing it though, it doesn’t look normal (I’ll post the spreadsheet with the data on it later).