As 8.5 point underdogs, the Vikings managed to not only stay in the game with the Cowboys, but hold a lead with the ball nearing the end of regulation. Up 23-20, Christian Ponder faced a 4th-and-5 from the Cowboys 36-yard line with 3:04 remaining. While we are almost always a proponent of going for it in no-man's land, this presents an interesting decision point. Dallas had two timeouts remaining and with a conversion, Minnesota would be able to milk a good portion of the clock. With a made field goal from Blair Walsh - one of the top kickers in the league, despite his earlier missed extra point - the Vikings win probability actually falls to 75%. We have talked about this previously, being down six is sometimes better than being down three.
So, it really comes down to going for it or punting. 4th-and-5 converts at 49% league-wide and the Vikings are essentially a league-average offense. With a successful conversion, the Vikings would win the game 96% of the time. With a failure, giving the Cowboys great field position leads to only a 68% chance to win the game. A punt, league-wide, results in an 85% win probability due to an expected starting field position at the Dallas 12. While the Cowboys also have essentially a league-average offense, the Vikings defense, however, has allowed 73.5 points above expectation after adjusting for strength of opponents, No. 31 in the league. It's extremely hard to estimate how that would affect the Cowboys chances of driving downfield for the game-winning touchdown, but it takes a 3% discrepancy between punting and going for it and likely shrinks it to nothing - if not making going for it the better option so as to avoid putting their horrendous defense on the field.
So, let's play the what-if game. What if Blair hadn't missed the extra point and the Vikings still faced the same 4th-down decision? Here, the decision is much more clear, you certainly go for it assuming you can convert over 17% of the time. But, why are these expected win probabilities lower than when the team is up only three points?
It is for the same reason that being down six is sometimes better than being down three. Being down four forces the Cowboys to be aggressive and as a result, more efficient, rather than getting conservative and settling for the game-tying field goal.
Keith Goldner is the Chief Analyst at numberFire.com - The leading fantasy sports analytics platform - and creator of Drive-By Football. Follow him on twitter @drivebyfootball or check out numberFire on Facebook.
I'm not sure how the Vikings are a league-average offense; they're 30th in offensive efficiency, and the eyeball test seems to agree with that ranking. It's still an interesting breakdown, though.
Watching the vikings this year, punting on that down should have brought WP to 0. They have had the opportunity 4 times and failed to win late. They cannot stop anyone. Also I dont think ponder has ever gotten anyone to Jump offsides/
Thinking about this game made me aware of this whole problem of league wide strategy.
While league wide offensive or defensive efficiency make some kind of baseline sense, the coach on the other side is either aware of the realtiv FG/TD payoff or he isnt. it does not matter the slightest, weiter leaguewide 85% of coaches are not aware of the proper strategy or not, as long as you catch one of the other 4-5 then the proposed drop in WP for a 4-6 point lead just does not happen and you end up making the wrong decisions.
Also due to the highly dislinear correlation, between play efficiency and drive result, my guess is, that small differences in your estimate of the opposing teams quality, can have huge differences in the chances of the next drive reaching a field goal or TD. It would be nice to see the markov model used to see, how the WP behaves when the efficiency changes.
It doesn't matter what you think of the Vike's offense. The offense has a better chance of getting a first down than the defense has of stopping anyone. Going for it is the only option. You've won one game, your defense hasn't been able to hold a lead for the entire season. So why not go for the win? It kills me when bad teams made decisions like this.
Or, the Vikings did a great job of playing just competitively enough to keep fans interested without screwing up their draft order.