If only...

Another post at the Slate/Deadspin rountable. 

Any of these five plays could easily have gone the other way. In total they represent 0.73 WP—nearly the entire difference between winning and losing. Win probability is obviously an abstract concept, but it helps put a concrete number on what we already intuitively understand. Numbers like this underscore how razor-thin the difference between winning and losing is, especially when the two opponents are evenly matched. This year's Super Bowl was like most other NFL games—it hinged on a handful of critical and unusual events. The conference championship games were no different: The Giants won thanks to two bungled punt returns by the 49ers, and the Patriots won in large part thanks to a missed 32-yard field goal attempt by the Ravens.

If you missed my first one from Sunday night, here's the link. It covered the end-game strategy decisions of allowing the final TD.

The smartest play of all would've been for Belichick to have allowed the touchdown even earlier. The Patriots certainly could have done so on the play prior to Bradshaw's touchdown run, when he was stopped for a one-yard gain, forcing New England to burn its second timeout. In fact, they probably should have allowed a touchdown as early as the two-minute warning. That’s the point at which the Win Probability of receiving a kickoff down by four or six points (0.23) exceeds the Win Probability of trying to stop the Giants from bleeding the clock dry (0.2). The Patriots would have had almost two minutes, two timeouts, and all four downs available to get a touchdown and steal the win.

Ring Probability Added

...by Joe Harris over at the Community Site.

More Super Bowl Analysis

Comparison to Broncos – Packers in SB 32

Many people are pointing out the similarity between last night's final touchdown and the situation in Super Bowl 32 between the Packers and Broncos. I’ve previously written about that game, and I agree that the right thing to do was to allow Denver to score. With about 1:47 to play and the score tied at 24, the Packers allowed Tyrell Davis to walk into the end zone for the go-ahead score. The Packers got back possession but were unable to move the ball on their first series of the drive. The Broncos went on to win.

So why should Belichick’s decision to allow the touchdown be considered gutsier than Holmgren’s? The primary reason is that in Belichick’s case, his team had the lead. The Patriots strategically forfeited a Super Bowl lead. Had they forced a field goal attempt and it was missed or blocked, the Patriots would have won the game. In Holmgren’s case, the best case scenario for the Packers was a tie. Had they held the Broncos scoreless at that point, they still would have likely needed overtime to win.

Although the ultimate effect was the same in the two situations, the notion of intentionally forfeiting a lead is qualitatively distinct from allowing a team to break a tie. It’s instinctively more difficult, and therefore I believe required ‘more guts.’ (Or in Deadspin parlance, it was "ballsiest".)

The second reason was the game situation. The Broncos had a second and goal at the 1 with about 1:47 left on the clock, and a touchdown was a very likely outcome anyway. In the Patriots’ case, a Giants touchdown was far less assured. Belichick’s decision also relied on the faith that the Giants would not take a knee at the one as they should have. From the 1, there was no risk that the Broncos could get any closer, making the FG any shorter.

Was the Giants’ 12th man penalty intentional?

Super Bowl 46 Analysis

This became a really interesting game in the final two minutes.

First off, as I wrote for Slate and Deadspin, the big thing from a strategy perspective was the final go-ahead touchdown of the game. Belichick may have made the gutsiest call in Super Bowl history by apparently instructing his defense to allow the score with a minute to play. Had the Giants run out the clock in favor of a chip shot field goal, it would have extinguished any chance the Patriots had. Instead, they had a minute and one timeout to respond. It wasn't clear if Bradshaw was tippy-toeing trying to go down at the one, or if he was just burning a couple extra seconds prior to scoring like video-gamers are know to do. Either way, he should have taken a knee.

Beyond that, Eli Manning was a defensible choice as MVP, with 0.70 WPA and 12.3 EPA. But I thought the real difference in the game was the battle in the trenches. The NYG defensive front won the battle, handing the NE offensive line -0.24 WPA and -0.4 EPA. They had only 2 sacks, but more importantly they put Tom Brady on the ground 7 times. They added four tackles for losses and held the NE running game to a lower than typical success rate.

The big stars on the NYG defense were Justin Tuck and Jason Pierre-Paul. Tuck had 0.24 +WPA, 4.7 +EPA, and a Success Count (SC) of 4. Pierre-Paul notched 0.20 +WPA, 5.1 +EPA and a 6 success plays. Blackburn made the one interception of the game, but it was so deep it didn't move the needle very much. He made a bigger impact with his tackles, earning a 1.14 Tackle Factor for the game.

WP Forfeited

It's 4th and 2 from the opponent's 42. The score is tied at 17. There's 1 minute and 17 seconds left on the clock. What would you do? If you're Ken Wisenhunt, you send in Dave Zastudil to punt.

One thing I've learned about human nature is that to help convince someone of something, I should frame the issue in terms of fearing a potential loss. That's usually a stronger motivation than expecting a potential gain. For example, I wouldn't suggest to a coach that he could improve his chances of winning by going for it on 4th down more often. Instead, I'd tell him he's forfeiting a significant chance of winning by not going for it. Nobody likes forfeiting stuff. My wife suggested using the phrase "leaving points on the table." Coach Wisenhunt forfeited 13% chance of winning that game.


By far, the most common question I get from reporters is whether teams are going for it more often. My answer is almost always "it's complicated, but I think so." The difficulty in measuring 4th down aggressiveness is that it's so dependent on the situational variables. You can't just count how often teams go for it rather than kicking. To-go distance, score, and time all weigh heavily on the decisions, and there are just too many possible combinations to compare rates from year to year or even decade to decade.

If we constrain the analysis to certain parameters--inside opponent territory and when the score difference is within reason, for example--we'd get an incomplete picture. We've learned over the last few years that there can be many situations outside traditional 'go-for-it' limits in which it can be beneficial to go for the conversion rather than kick or punt. And each situation can have a drastically different magnitude of effect on a team's chances of winning. Also, why would we reward a coach if he goes for it on 4th and whatever when he's down by 5 with a minute left to play? Coaches always do that.

Here's my stab at the problem. With every 4th down situation in which it would usually make sense to go for it but a coach decides to kick, he forfeits some amount of win probability. We can total all the WP forfeited to measure the degree to which teams are erring on the side of conservatism.