Saturday night, TCU stormed back against Oklahoma, rallying from a 30-13 deficit to trail by a single point, 30-29. The Horned Frogs needed a win to have any hope of winning the Big 12 or making the College Football Playoff, and Gary Patterson decided his team would try for it right then and there with a two-point conversion.
It didn't work.
Third-string QB Bram Kohlhausen scooted around and drew defenders toward him, which opened up the receivers they were covering. He lofted a pass to his receiver ... and it was juuuuuuust swatted down by an Oklahoma defender. Game more or less over.
Old-school football thinking tells us that the decision to go for two here was "risky." But really, the decision isn't about "risk."
NFL teams make about 50 percent of their two-point conversions. College teams make about 50 percent of their two-point conversions. Going for two with a win or loss on the line generally means you have about a 50 percent chance of winning and a 50 percent chance of losing. If you're good at offense, you probably have a better than 50 percent chance of winning. If the other team is good at defense, you probably have a worse than 50 percent chance. But still, about 50 percent.
This decision is called "risky" because it opens up the opportunity for a loss, which rarely happens if you attempt a PAT. But it also opens up the opportunity for a win, which never happens if you attempt a PAT. It shifts the decisive climax of the game from the overtime to a single play.
You should go for two if you feel that 50-50 shot gives you a better chance at winning than a sustained overtime session. Are your chances at beating a team on a single play better than your chance of beating a team over the course of at least two full series? If the answer is "yes," you should go for two.
Gary Patterson had plenty of reasons to think that was the case. His team was an 18-point underdog on the road at Oklahoma. He was playing his team's third-string QB, and the team's world-beating WR was out, too. The longer the game went, the more likely those disparities would be a problem than on a single play.
But Patterson also had plenty of reasons to play for overtime. Oklahoma also was playing with a backup QB, Trevor Knight, and the Sooners offense had been absolutely stagnant. Their last six drives were five punts and an interception. Knight finished the game 5-of-16 with 76 yards and a pick. And Kohlhausen had sparked TCU's offense to 16 unanswered points. The Horned Frogs were outperforming the Sooners on both sides of the ball.
It's hard to fault Patterson either way, but fans will. TCU missed. It was a reasonable decision and the play almost worked. The Horned Frogs were a slightly loftier pass away from a victory and a shot at the Big 12 title next week against Baylor. Instead, nah. So Patterson's call seems bad.
But when Bret Bielema did this a few weeks ago against Ole Miss and Arkansas got it, it seemed like a good idea. It's a 50-50 call that leads to 20-20 hindsight.
I really have just one problem with TCU's two-point play:
Oklahoma ran a play with four down linemen and zero linebackers in the box. TCU spread the field with five wide receivers, and Oklahoma responded by spreading the field and covering the hell out of those wide receivers. But to do that, the Sooners left the entire middle of the field open.
It's not just like this was a pre-snap thing — here's what the field looked like a few seconds into the play:
TCU only had to get three yards up the middle, and there were zero defenders in a position to stop them. A QB draw or scramble up the middle probably wins the game. Instead, Kohlhausen got into a footrace to the sideline, where there were even more defenders.
Against Trevone Boykin, Oklahoma never would've done this. The Sooners would've needed to put somebody in the middle to prevent a walk into the end zone, which would've left receivers open somewhere. I guess they didn't feel Kohlhausen was a threat to run, and TCU rewarded them by not running.