February 6, 2012

Win Probabilities

The Harvard College Sports Analysis Collective argue that Bill Belichick made the right call allowing the Giants to score with a minute left in the game.

But no one knew that his score would decide the game. Before he ran the ball in, the Giants had 0.94 win probability (per Advanced NFL Stats). After the play, the Giants’ win probability dropped to 0.85. Had he instead taken a Brian Westbrook or Maurice Jones-Drew-esque knee on the goal line, the Giants would have had a 0.96 win probability. Assuming the Patriots used their final time out, the Giants would have had 3rd and Goal from the 1-yard line with around 1:04 left to play. At this point, the Giants could either attempt to score a touchdown or take a knee. Assuming the touchdown try was unsuccessful or that Eli Manning kneeled, the Giants could have let the clock run all the way down to 0:25 before using the Giants’ final time out. With 4th and Goal from the 2 with 25 seconds left to play, the Giants would have a 0.92 win probability, 0.07 higher than after Bradshaw scored the touchdown of his life.

I respectfully disagree. First, I assume those probabilities are based on every 20-yard field goal kicked, most of which are probably made in no pressure situations. With the championship on the line, the Patriots would have to assume that three players would hold up under pressure; the snapper, the holder and the kicker. We saw a bad snap by the Giants earlier in the game, but NY recovered and just made the field goal.

Second, the analysis doesn’t take into account defending the touchdown play. If the Pats went after Bradshaw, he might have fumbled. The Giants might have fumbled a snap when trying to take a knee. Given that Eli could not manage the play clock, there could have been a penalty that pushed the ball back and made the field goal more difficult.

Finally, I would rather defend a lead and force the other team to beat me. When the Patriots allowed the Giants to score, they put all the pressure on themselves in a low probability win scenario. I’m sorry, Belichick gave this game away. He should have had the Pats defense fight to the end.

8 thoughts on “Win Probabilities

  1. Ed

    Football is more complicated than baseball!

    I think what the analysis doesn’t make clear is that with a two point Patriots lead, the Giants can win by kicking a field goal and holding on until the end of the game, or scoring a touchdown and holding on until the end of the game. At that point whether they score via field goal or by touchdown makes little difference, except in how much time this leaves the Patriots to score, and also in that with a field goal, the Patriots themselves only need a field goal too.

    The Giants had gotten to the point where they could score a field goal easily, but opted not too, I assume mainly to leave less time on the clock at the end for the Patriots to come back.

    But did the analysis take into account the possibility of the Patriots defense recovering a fumble or getting an interception? There is also the possibility of a game changing penalty call. Rooting for the Giants, I was more nervous about this happening than I was about the Patriots scoring on the last drive. There are more things that can go wrong for the offense on a touchdown drive than on a field goal attempt.

    ReplyReply
  2. rbj

    So super genius Belichick got hoisted on his own petard. hee hee! Still can’t beat the Tom & Eli show in the Big Game.

    ReplyReply
  3. Plank

    I understand the logic, but I think it’s a bad idea to give up the lead with so little time left. There is no guarantee the Giants score there. They made it a guarantee that the Giants would lead with a minute left in the game. Seems like a bad move to me.

    ReplyReply
  4. mcsnide

    So you think that it’s easier to prevent what is basically a PAT than it is for Tom Brady to drive for a TD with a minute and 1 TO? Neither is easy, but I’d think that, even with the pressure of the situation, you’d have to put the chance of making that kick at over 90%. I’d certainly think that Brady and the Pats would score a TD in that situation more than 10% of the time.

    To me, the odd part of the decision was that they did it on 2nd down. Why wouldn’t you do it on 1st and save the seconds and the TO?

    ReplyReply
  5. Subrata Sircar

    Belichick gave the game away only in that he didn’t assemble a defense capable of stopping the Giants from crossing midfield at will.

    The fact is that pressure and turnovers are all extremely low-probability outcomes, particularly when one team is focused on avoiding them. The chances that they can make Bradshaw fumble on a play where he’s locked up the ball with both hands – and recover the fumble – are considerably worse than 10%, which is their gain via this strategy.

    Say Belichick doesn’t let them score. So they stuff him at the 5 and call timeout with about a minute to go. The Giants come out and kneel on 3rd and goal – nothing the Patriots can do about that – and let the clock run down to about 25 seconds. They then kick the field goal on 4th down. NFL kickers are about 99% from this distance. Even assuming a generous estimate for failure due to increased pressure (which btw, no one has ever been able to prove – that is, there are clutch performances in all sports, but not clutch players, no matter what your lyin’ eyes tell you), that’s still about 95%. Now they kick off – a squib, and the Patriots have no real home-run threat on the kickoff, and the Giants are good at coverage – so Brady takes over on his own 30 with 10 seconds to go. If he can complete a 40 yard pass in 5 seconds out-of-bounds the Patriots can try a 47-yard field goal for the win – and NFL kickers are well under 50% from that distance. Otherwise it’s a Hail Mary.

    Not to mention that what he actually did was trust his 2nd-ranked offense to score with a minute to go rather than his 31st-ranked defense to get a turnover on 1 play. Put that way the choice seems obvious to me.

    Put another way, the same analysis that leads to the poor results on intentional walks or sacrifice bunts shows this to be a good decision. Just because the result “felt” bad – and I agree, it should never be a good idea to let a team score – doesn’t make it wrong.

    ReplyReply
  6. Capybara

    I would be very surprised if the win probability analysis does not include the possibility of a fumble of some sort.

    ReplyReply
  7. pft

    Patriots defense has been the achilles heel all year. Better off seeing if Brady can score in the last minute like he has done so many times before.

    I doubt too many Pats fans were 2nd guessing it at the time, and were glad to see the ball in Brady’s hands. Just did not happen.

    ReplyReply
  8. Sal

    you are incorrect in that the analysis does include the chances of a turnover. It was the right decision. NY would’ve bled the clock down to less than 5 seconds and kicked a FG that was essentially an extra point. The accuracy of NFL kickers on extra points likely exceeds 99%. Even accouting for the pressure of the situation, they probably still make the kick something between 95 and 99% of the time (if not more).

    You simply can’t let them take the clock down to nothing and kick a short FG to win. You have to find a way to get the ball back with time on the clock. The Pats did. And they had a reasonable shot at driving the field that took a major hit with Tuck’s sack that burned their last timeout.

    ReplyReply

Leave a Reply

Your email address will not be published. Required fields are marked *