The Belichick Debate

I'm the boss...need the info...I thought if there was ever a cause for unanimity in the sports world, it would be in the reaction to Belichick’s call on Sunday night. Everyone outside New England hates him and will jump at the chance to dog him, I thought, and New England fans would be so distraught over the game they’d be desperate for someone to blame.

But this is the Internet. And so a contrarian view began to spread yesterday, virally. A view I believe started with someone’s desire to take a fresh angle on the obvious story, but that’s neither here nor there.

Every one of these analyses seems rooted in a statistical argument encapsulated in a Deadspin post entitled “Bill Belichick Was Right.”

With 2:00 left and the Colts with only one timeout, a successful conversion wins the game for all practical purposes. A 4th and 2 conversion would be successful 60% of the time. Historically, in a situation with 2:00 left and needing a TD to either win or tie, teams get the TD 53% of the time from that field position. The total WP for the 4th down conversion attempt would therefore be:

(0.60 * 1) + (0.40 * (1-0.53)) = 0.79 WP [win probability]

A punt from the 28 typically nets 38 yards, starting the Colts at their own 34. Teams historically get the TD 30% of the time in that situation. So the punt gives the Pats about a 0.70 WP

I wasn’t going to harp on it anymore, because it’s ultimately a matter of opinion, not fact, and I can scream my conclusion till I’m blue in the face and it’s not going to make anyone agree with me, nor should it. But I’ve been irked to see this argument passed along uncritically, and spent so much time repeating myself on Facebook, on IMs, etc. having this same conversation that I just want to put my argument up and then be able to send someone a URL whenever I want to continue making the same points.

First off, this isn’t baseball. The sports are apples and oranges, to begin with. Also, football is far behind baseball in terms of the fine-grained statistics that are gathered (though places like Football Outsiders are doing a great job of changing that picture).

Thus I find especially suspect the reliance on the calculation that “Historically, in a situation with 2:00 left and needing a TD to either win or tie, teams get the TD 53% of the time from that field position.”

Not Peyton Manning, specifically. Not the Colts. This means that throughout all time, teams, which means everyone from the Sisters of the Poor to the top flight of the league, have a 53 percent chance of scoring. I think that statistic is debatable, to say the least.

To me it is not as debatable as it’s being made out to be that if you gave Peyton Manning the ball with 2 minutes on the opponent’s 30 yard line, given the other two insanely quick scoring drives that had just transpired and the way the Pats defense was playing late in the game, that he was going to score. I don’t care what the overall team’s red zone efficiency states, in that game, in that moment, with the Pats defense shortstaffed because of injury, clearly losing steam and also losing momentum, there’s no way IMO the Pats are going to make a goal-line stand and win the game.

There’s also no guarantee that a punt meant the defense would be able to stop him, either, but I find it really hard to believe that the odds of Manning scoring from his own 30 are lower higher than scoring from the New England 30. That’s the information that matters with 2 minutes left in the game. I’m finding it really hard to stomach ppl weighing in with overall historical fourth-down conversion stats to essentially argue that giving Manning the ball on the New England 30 was somehow preferable or at least neutral compared to giving him the ball on the Indianapolis 30. It’s like arguing in baseball that a runner on first has a better chance of scoring than a runner on third.

Similarly, it was brought up yesterday morning on WEEI that the Patriots have a 63% fourth-down conversion rate this season (also heard 60%, 73%…) But that statistic also doesn’t account for time to go and game situation. It’s not a stat that’s representative of situation the way baseball stats are — where you say with less than 2 outs or with a man in scoring position. This is just conversion rate each time they’ve faced a fourth down, not accounting for the yardage to go, score, time remaining, opponent…

D&C’s point was that you can’t go by that statistical calculation, because in many cases when a team goes for it on fourth down, they’re behind significantly and the other team is playing a softer containment defense, or the game is imbalanced in some other way. This wasn’t a decision you could make in a vacuum just looking at the overall conversion statistic.

In short, if you could tell me that on fourth and 2 in the final three minutes of a game against AFC opponents ranked 5th or higher overall, the Patriots, specifically, have a 63% conversion rate, that might be different.

I will concede the point that focusing overmuch on the 4th and 2 decision isn’t wise, either, since there were so many other factors that led up to that situation, and a series of decisions that followed it that were in some ways equally inexplicable. There was also Maroney’s fumble in the end zone (he’s lucky Belichick is drawing the amount of attention he is this week), and the notion, which I expressed at the time, that if they’re playing at Gillette, I wonder if the ref sees Faulk “bobble” the ball.

But if the percentages really line up in favor of Belichick’s decision, if it’s really been mathematically demonstrated it’s the right call to risk giving it up on downs deep in your own territory up by less than a score with two minutes to go in the game, why does the punt exist in football at all?

This entry was posted in Bill Belichick, Uncategorized and tagged , , , . Bookmark the permalink.

3 Responses to The Belichick Debate

  1. Harmy G says:

    “I find it really hard to believe that the odds of Manning scoring from his own 30 are lower than scoring from the New England 30.”

    You’re doing it wrong.

  2. Vern says:

    You missed the point a bit, so let me re-arrange the formula for you a bit. All it shows is the combined chance of winning in the go-for-it situation:

    convert the 4th + fail yet still stop manning from 30 yards out = 55 to 65% + 10 to 20% = 65 to 85% net total chance to win

    What you’re doing wrong is assuming the Colts win the game 100% of the time if the Pats fail to convert. It’s not that high. It’s more like 80% (Pats still win 20% of the time). And remember, this only happens the 40% or so of the time they do not convert the 4th.

    You have this thinking that the Colts are absolute world beaters from 30 yards out, but long shots from 70 yards out. The key is that the DIFFERENCE in their success rates from 30 yard or from 70 yards is not as high as you think. It’s big, but not SO big that punting is the only right thing to do.

    Example, from 70 yards away their success rate is maybe 35%, then from 30 it’s more like 70% NOT 100%. Think about it in terms of the number of plays they need to score. From 70 yards its maybe 10 plays, but from 30 it’s still about 5 or 6. The chance the D stops them is not going to go from 65% to ZERO due to 5 less plays, it’s going to go from like 65% stop to only 30% stop or so, but that’s enough to tip the odds in favor of going for it.

    Even if you use a 2-point conversion rate for 4th down, which is a 2 yard all-out-stop situation for the D, the average is STILL around 50% success and the odds still favor going for it. For the call to be wrong, you have to assume the Colts succeed 70%+ of the time or higher from the 30, but LESS than 35% of time from the 30.

    There is just not that much DIFFERENCE in scoring in that situation based on the 40 extra net yards a punt might give you.

    Punts earlier in the game make more sense because of the ability for the other team to get FGs even on a failed drive. Notice that Belichick pretty much goes for it any time he is at or past the 50, barring special situations and such.

    It’s one of those group think things at this point, but is well established that teams punt WAY too much. The reason is blame and history – the game used to be about field position and defense. Teams used to punt on 3rd down even back in the 30s. This created the sense that it’s the “safe” play. But in the last couple of decades, the rules changes favors the offense so much that punting (field position) is worth far less than it used to be. Punt today is done primarily because it shields the coach from blame. Punt and fail = defense fail. Go for it and fail = bad coach decision. The fall out from this game is all the reason you need to see why coaches do not go with the numbers that are as clear as day.

    Note that Belichick goes for it far more often than most because he doesn’t care about blame at this point in his career and is free to go with the numbers. Look back at all the other games where he’s gone for it on 4th, including the Miami game only a couple weeks back and you see the success.

  3. beth says:

    Hey Vern, thank you for a long and thoughtful comment. I think there has been a lot of good debate to come out of this game and ppl have made excellent points on both sides. I don’t know that we’ll ever come to agreement, though. :)

    Thanks again for commenting and reading. Hope you’ll continue.

Leave a Reply

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

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>