By: Brian B

Brian B — Mon, 10 Mar 2014 22:05:47 +0000

To reply directly to Chase’s specific question, the variance concept says an underdog wants to bet as much of the game on as few plays as possible. Assuming OT would require more than 6 plays (1 for each 2-pt conversion), it would almost always be better for the underdog to prefer the 2-pt conversions to the OT. The only exception would be if one team or the other were prohibitively good or bad at short yardage/goal line situations.

This can be shown very persuasively using the Huygens corollary to the classic Gambler’s Ruin problem in statistics (not to be confused with the gambler’s fallacy). The GR problem proves that when the odds favor the house, a gambler should bet as much of ‘the pot’ as possible on as few ‘trials’ as possible to maximize his chances of going home a winner.

In response to the question of whether the 2-pt conversions are high enough leverage to outweigh the leverage of OT, I think that’s almost certain to be the case. In the 2-pt conversion game, you need to win only 4 out of 6 of them (just 1 more than 50%) to win the game. In the OT game, it would be rare (but not impossible, see Tebow, T. vs PIT) to win the game by just ‘winning’ one more play than an opponent.

By: Richie

Richie — Mon, 10 Mar 2014 19:32:33 +0000

In reply to Chase Stuart.

I agree with Dave.

The better team would rather play overtime. I think the extreme comparison would be baseball. A stronger baseball team will surely have a better record than a weaker team over 162 games. But the 7-game playoff series is much more random.

Unless the team is like your Panthers example above, that might have a particular skill at scoring and/or preventing 2-point plays.

Comments on: Would A “Mandatory Go For Two” Rule Lead to More Upsets?

By: Brian B

By: Richie