I made predictions on 47 games last week: 2 of them were too close to the spread for me to make a call and 1 wound up in a push (Central Florida was favored by 14 over SMU and won 31-17). That left 44 games to measure, and my method (VHR) went 29-15, or 65.9%. For those who don't gamble, break even against the spread is 52.3% (or 11/21). 29-15 is a +28% return (if you bet $1000 total on my picks, you'd have $1280 back).

I don't expect to go 29-15 every week.

The best picks of the week:

Predicted: Texas 38 - Colorado 13

Actual: Texas 38 - Colorado 14

Predicted: Oklahoma 50 - Baylor 17

Actual: Oklahoma 49 - Baylor 17

There were some bad ones in there, but I'd rather focus on the good ones.

Before I get to the coming week, one housekeeping note. Last week I linked USA Today's spreads and used a column labeled "Opening". Someone who actually does bet tells me that's not appropriate, because it's hard to actual place a wager against the opening line. So this week, I'll placate that concern, I hope, by using the line marked Logans. I have no idea who the Logans are, but their spread is in effect for Week 7.

So on to Week 7 predictions.

And you'll note one very amusing one: Texas 23 - Oklahoma 23, and it declares Oklahoma the winner. That's because it's actually Texas 22.8 - Oklahoma 23.1.

Remember, these do not factor in HFA, not in the data that is used to calculate the score nor as an adjustment at the end.  Picks with the greatest difference between my score and the spread are at the top, and the closest are at the bottom, including a surprising 6 games where my pick is <1 pt from the spread.