Monday, January 05, 2009

Look Before You Leap

I began tonight's Fiesta Bowl without a rooting interest, but I ended it with one. That's one consequence of always rooting against stupid decisions.

Ohio State scored a touchdown with 2:05 left in tonight's game, putting them up 21-17. The two teams had a combined five timeouts left, so the game was far from over. Jim Tressel then called for a two-point conversion.

I assume you know what happened in the game, so there's not much point in hiding the results from you. Still, why go for two here?

The goal, of course, is to extend the lead to six in the hopes that the opponents will miss their PAT (after scoring a touchdown) and the game will be tied. Assuming they score, this scenario will happen maybe 1% of the time. (Texas was 71-71 on extra points this year at the time of the decision.) Then, if OSU fails to retaliate with a score, we head to overtime instead of seeing a Mack Brown Gatorade shower.

Meanwhile, what happens when the Buckeyes miss--as they did--and Texas scores a TD? Now if OSU responds with a field goal, this will merely tie the game instead of winning it.

So, we have two possible scenarios where the decision to go for two will influence the outcome. We will assume Texas scores a touchdown, since the point is moot otherwise:

- If OSU succeeds AND Texas misses the extra point AND OSU fails to score in response, this will change a Texas win into an overtime game.
- If OSU fails AND Texas makes the extra point AND OSU kicks a field goal in response, this will change an OSU win into an overtime game.

My quick-and-dirty estimates say that the Buckeyes make the 2-point conversion 45% of the time, the Texas PAT is good 99% of the time, and Ohio State wins 45% of the overtime contests. Using these figures, going for two is only a good play if you believe that OSU's chances of kicking a game-winning/tying field goal at the end of regulation are less than...

Drumroll please...


As in, one time in 140.

Jim Tressel apparently doesn't have much faith in his two-minute offense. Or, like so many coaches, he doesn't think ahead.

I could have better enjoyed the game's conclusion had Ohio State kicked a field goal at the end of regulation and gone on to lose in overtime, but seeing them lose at all was reward enough.

EDIT: I somehow forgot that Texas will go for two in this scenario. Assuming they make it 45% of the time, that changes the break-even percentage to 1.3%, which I still think is way too low to make Tressel's decision correct.


ben said...

tressel knew that i had horns -8/-2 and was helping me out.

it is worth noting that there is a breakeven point in terms of time/timeouts that would make it correct to go for 2, but ur analysis is obv right in this situation.

Garrett W. said...

I think your estimates are flawed. Using season PAT estimates for this scenario is gonna result in a higher percentage than reality since there would be pressure, whereas there is usually virtually no pressure.