Saw this absurdity today and just couldn't let it slide without a little mockery.
First of all, wow, commas! But we all know Berry's not a writer, so ok...
Since the chances of winning in overtime only happen after missing the last second shot, those multiply. Just a little thought would show you that. Imagine that you estimate OSU's chances of winning in overtime to be 60% instead of 52%, and our chances of scoring on a last second shot to be 40%. Just estimates, like Berry's, still reasonable. Looks like you could say we were guaranteed a win, or maybe even give us that coveted 105% chance of winning. 
If we do it correctly and use his numbers, our chances of winning given the ball, tie game, and a last second shot would be: 0.35 + (1 - 0.35)*0.52 = 68.8%. Doing the math for TCU gives you: 0.0 + (1 - 0.35) * 0.48 = 33.8%. Those conveniently add to 100%.
Tie game, late, you figure overtime is a 50-50 proposition, although in this case, OSU clearly was considered the better team, so maybe you give the Cowboys a 52-48 edge.
So whatever chance you give OSU of scoring on the last possession adds to the 52 percent... I'd estimate OSU's chances of scoring on a last-second shot somewhere between 30-40 percent... But add 35 percent to the 50/52 percent chance of overtime, and OSU was looking at an 85/87 percent chance of winning...

If we do it correctly and use his numbers, our chances of winning given the ball, tie game, and a last second shot would be: 0.35 + (1 - 0.35)*0.52 = 68.8%. Doing the math for TCU gives you: 0.0 + (1 - 0.35) * 0.48 = 33.8%. Those conveniently add to 100%.