Jason Garrett Made a Massive Error in Overtime by Punting


Jason Garrett and Bill O'Brien engaged in a coaching clinic on Sunday Night Football, and Jason Garrett emerged as the winner … in the race to get on the hot seat. This despite O’Brien’s best efforts by kicking three field goals of 21 yards or less, and using a timeout on 3rd and 15 in regulation and then not running a play to get to the sticks.

Garrett’s final stroke of genius came when he punted on 4th and 1 from the Texans’ 43, on the first possession of overtime, with 5:40 remaining. I’ll note that this was NOT the same situation as Frank Reich faced when deciding not to play for the tie a week ago. The distance was shorter, and the ball was on the Texans side of the field.

By punting, Garrett was conceding a chance to win for fear of losing, but was putting his team in a situation where they could be sudden death losers. They were underdogs to win the game once he made that decision. Just do a mental experiment. What would the chances of converting that 4th and 1 have to have been to make it a bad move?

The answer is about 25%, given that field position and game situation. The actual conversion rate on 4th and 1 (non-red-zone situations) is 69.8% over the last three years.

Yes, Dallas is less likely to give up a score by punting than by giving Houston the ball at around the 43, but a score isn’t automatic. Based on past overtime possessions, the chances of a Houston score starting inside the 20 is about 27%, and it is 41% starting closer to the 43. Meanwhile, Dallas almost certainly scores if they convert that fourth down. Thirteen of 15 teams in the last decade to start an overtime possession between the opponent 45 and 35 scored (4 touchdowns, 9 field goals). Dallas could win outright, and even if they eventually settled for a field goal, they would be a solid favorite to win the game, as fewer than half the teams answer a field goal with a score, and if Houston answered with a field goal, Dallas would be a favorite again with possession and in sudden death.

I estimate, based on recent historical results in overtime, that the Dallas Cowboys were a 76% favorite to win if they convert the fourth down, a 31% favorite if they went and failed, and a 42% favorite if they punted. Yes, failing is worse than punting but the massive difference in chances of winning by going and getting the ball right on the edge of scoring range far offsets it.

Thus, using the 70% conversion figure on 4th and 1, Dallas overall chance of winning was 62% by going for it (without knowing the result) and 42% by punting.

That is a MASSIVE error in strategy. A 20% error. Yet, I doubt anyone else will call it the dumbest move this year, like folks were doing with Frank Reich last week (Reich’s decision may have been about a net -2 to -3% chance of winning (where a tie is a half-win).

That folks, is how you end up with two playoff appearances in what looks like now eight seasons in the NFL.