The League

David Aldridge
Sports Reporter

David Aldridge

A nationally recognized sports journalist.

Smart move, Bill


Bill Belichick, according to the almost-unanimous opinion of the football punditry, is an idiot for going for it on fourth and two at his own 28, up six, with 2:03 left in the fourth quarter last night against the Colts. Rodney Harrison, Tony Dungy, Trent Dilfer -- all possessors of Super Bowl championship rings -- said so in the aftermath of the Colts' 35-34 last-second victory.

This is conventional, safe football thinking. But it's wrong. (Yes, the guy who lasted 10 days playing JV in high school, is calling the guys with a lifetime in pads, meeting rooms and games wrong.)

How many times do we have to see Parcells and Belichick, whether with the Giants or Patriots, going for it in big games, with championships on the line, before we realize that going for it is the superior football strategy?

Forget, for a minute, that the logic of the game pointed toward going for it; Tom Brady had shredded the Colts' decimated secondary all night. There was no reason not to think that Brady, from the gun, in a spread, wouldn't find somebody open.

This is not my idea, nor have I tested it. But others have. Gregg Easterbrook, of the Tuesday Morning Quarterback column on, is a strong proponent of going for it almost from everywhere on the field. He's cited statistical studies from Brian Burke at Advanced Football Stats, and commissioned his own comprehensive study of 10,000 simulated games from the 2006 season by Accuscore, a statistical website that crunches the numbers to find "win probabilities" (I know, this is a little wonky) in games.

Accuscore found in its simulated games from the '06 season that a team that avoided punts altogether added, on average, one point to its per-game scoring, without adding any points to the opponents' average scoring. Teams avoiding punting were five percent more likely to win, which adds up to almost one extra win per season. And one more win, as you know, is often the difference between making the playoffs and not making the playoffs.

Easterbrook concluded, based on the statistical evidence, that from your own 21-yard-line to your own 35, you should go for it on fourth and two or less if you want to win. Which is exactly what Bill Belichick did.

And there is precedent for a coach going for it, on fourth down, inside his own 30, on the road, late in the fourth quarter. This coach's team didn't make it, either, and went on to lose the game. Everyone thought they were postseason toast. But they went on to win the Super Bowl.
And, yes, I have just compared Bill Belichick, favorably, with Barry Switzer. I'm sure he'll sleep better now. Belichick, I mean.

By David Aldridge  |  November 16, 2009; 8:29 AM ET  | Category:  Indianapolis Colts , New England Patriots Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati  
Previous: What do you think of Belichick's call? | Next: Belichick's Blunder


Please email us to report offensive comments.

I am almost certain that this analysis is incorrect.

The Easterbrook/Advanced Football Stats analysis is based on going for it on fourth down ALL THE TIME, FOR THE ENTIRE SEASON. That does not in any way mean that if you go for it only some of the time in some situations in some games that it is always the correct decision. Huge difference.

Posted by: KeithW2 | November 16, 2009 2:51 PM

Of course conventional wisdom tells you to punt in this situation, but is it the correct decision? To evaluate this, you have to do the following:

1) Evaluate the probability of getting a first down in this situation.
2) Evaluate the probability of the Colts scoring a TD from the 29 Yard line with 2 minutes left.
3) Evaluate the probability of the Colts scoring a TD from the 67 Yard line with 2 minutes left.

Given the relative strength of the teams and the scoring patterns in this specific game, I'd estimate these as follows:

1) 50%
2) 40% (or 60% that they won't score a TD)
3) 25% (or 75% that they won't score a TD)

This is not based on any particular stats but simply on what I would estimate the probabilities to be in a game like this. If these numbers are true, by going for it, Belichick had a chance of 80% (50% + (50%*60%))of winning the game and a 75% of winning had he punted.

I don't think that Belichick's decision was as stupid as most people would make it seem. I think that most NFL coaches are extremely risk adverse by training. Belichick is different and smarter (as a Jets fan it hurts me to say that) then the rest of the bunch and that's why he is so successful. I wonder what most people would say today if the play had succeeded?

Posted by: mburix | November 16, 2009 3:35 PM

I too believe that the statistical data are being taken out of context to make this point. If true, the basic premise could then be extended to all play calls during the game and we could eliminate play-callers altogether, but we can't. The reason is because IN ADDITION to the statistical likelihood of a ball being thrown accurately or being successfully caught you have to incorporate other factors such as crowd noise, fatigue, momentum, injuries, and others too numerous to list here into every decision. I do think the basic information should have been considered in the decision, but so should have the fact that the 3rd down throw was almost intercepted and returned for a touchdown, that it was a home game for the Colts, that Manning had moved his team down the field effortsly for two scores, etc. Looking at the likelihood that Manning could move his team 80 yards in 2 minutes vs 30 yards also has statistical merit. I would have taken my chances that more could have gone wrong in 80 yards than in 30 and punted.

Posted by: zalumark | November 16, 2009 4:11 PM

I always say that if you can't get 2 yards when it counts, you don't deserve to win. Teams punt the ball away on a regular basis. Forgive me if I'm wrong but don't half of them lose anyway?

Show me some stats that say that going for it on 4th down with 2 yards or less inside your own 50 with less than a 7 point advantage means that you will lose the lead, more than half the time. You still have to consider the dynamics of *that game*. Just think of it, the odds are fairly decent that the Colts score on the next play and the Pats get the ball on the ensuing kickoff (which is just as likely to be an onside kick) and there's a fumble and they lose outright anyway. You have to make this decision in the moment, stick with it, do your best with it, and take what happens.

People forget what probability means. It means that sometimes you're wrong. Regardless of how smart you are and how stacked you think the odds are in your favor. If they had punted the ball and still lost what would you say then, "good coaching"? I doubt it.

Posted by: dubya1938 | November 16, 2009 4:46 PM

Have to agree with KeithW2. The first part of your argument is flawed. Those stats you presented assume a team always forgo punting for an entire season. It's like going to Vegas and trying to beat the house by counting cards at blackjack. The minute you deviate from it you put the odds back in the house's favor.

Posted by: spagball | November 16, 2009 4:52 PM

The main thing is that in a 35-27 game you can't deny that he had almost 100% expectation that they would get two frigging yards on an offensive play. It was only in doubt in the first place because it was 4th down. So what are you going to do, punt, hope *that* goes well, hope it doesn't get snapped over the kickers head or short and recovered by the Colts anyway, or blocked & run back for a TD, hope that there isn't a penalty before the kick that puts you 5 yards farther back and have to repeat 4th down with the fake kick or a straight play now that much less likely to succeed, hope that the punt isn't run back for a TD, hope the punt is a decent 45 yard punt with no hang-time or penalties in the return, hope that you actually stop the return for no gain, and hope that your defense suddenly is going to stop them from marching 70 yards in 2 minutes? Against Peyton Manning & the Colts?

How smart is *that*? That's 10 chances to lose the game outright, right there. Vs getting 2 measly yards (maybe more) and running out the clock, vs a possible turnover on downs, interception, sack or fumble? Gee I can't imagine which one I'd take.

Posted by: dubya1938 | November 16, 2009 4:58 PM

i watched the game and i absolutely did not want the patriots to win... and i dont like their team given that i root for another team in the afc east

that being said i think he did the right thing in going for it...

the bottom line is that the colts had all the momentum at the stage in the game, which you cannot discount, and if the pats had punted, with the two minute warning and the colts having a couple of timeouts, with peyton manning playing well, there's not much doubt in my mind that they wouldnt have scored...

it's a gutsy call but i think they did the right thing...

Posted by: m01lew33 | November 16, 2009 5:10 PM

A case example of hubristic "snatching defeat from the jaws of victory."

Posted by: washpost16 | November 16, 2009 7:36 PM

Heloooooo!?!?! You can make all the statistical analysis you want, but who needs it. We have a real life game that tells us if you go for it on fourth and two with the ball on your own 28 guess what lose the game.

I think even the great Bellichick would throw out your statistical anyalysis now.

Posted by: PASmitty | November 16, 2009 10:17 PM

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2011 The Washington Post Company