OT: 4th and 2 in your own territory near the end of the game

Submitted by michelin on
This blog previously considered the question whether or not one should go for it on 4th and goal from the 2. I recall the analysis, among other advantages, showing that, from NFL stats, the chance of making a touchdown (7pts) was about .6, yielding .6*7=4.2 pts on average, which exceeds the score from a made field goal. Now Bill Belichick is being fried by the talking heads in New England for making a similar decision to go for...but to go for it not on the goal line but in his own territory (30 yd line) on 4th and 2 with 2 minutes left in the game vs IND, who is down by 6pts. Brady's pass was complete but marked inches short when it was bobbled, enabling IND to have a short field (30 yards) to score the winning TD. I know that Belichick's decision to go for it violates intuition and almost universal coaching practice. But was the decision really a mistake, as so many claim? According to my VERY ROUGH and QUICK calculations and the graphical link below, the chance of NE winning was .68 with the decision to go for it vs. .5 if he decided to punt. So, I think Bill made the right choice. I'd be interested to know if anyone has different opinions about the chances of winning based on the assumptions, I have made below. Note: Remember that, if NE succeeds on 4th and 2, the game is virtually over as they can run out the clock. I've assumed a 40 yard punt and no unforeseen turnovers. I considered that Peyton Manning had a hot hand, the NE defense was tired, and that IND had a higher than usual chance of making a TD, whether it was from 30 yards out (after NE fails on 4th and 2)(Pr=.8) or 70 yards out (ie after a punt)(Pr=.5). Summary: NE had a greater chance winning by going for it on 4th and 2 (.68 vs. .5) Analysis of the 2 possible decisions: 1. NE goes for it P(make 4th and two )=.6 =Chance PATS win immediately Pr(don’t make 4th and two) = .4 p(Ind TD with 30 yard drive)=.8, Pr(IND does not make TD with 30 yard drive)=.2 Pr(IND fails to make TD on 30 yard drive and Pats fail on 4th and 2)=.2*.4=0.08 TOTAL CHANCE PATS WIN=.6+.08=.68 2. NE punts Pr(IND TD with 70 yard drive)=.5* Pr(IND fails to score on 70 yard drive)=.5 TOTAL CHANCE PATS WIN=.5 http://cid-4bf9d75c782b05b1.skydrive.live.com/self.aspx/ne%204th%20and%… Here's an article on the game decision.http://www.nj.com/sports/ledger/politi/index.ssf/2009/11/new_england_pa…

Big Shot

November 16th, 2009 at 8:31 PM ^

Belichick made the right decision given the situation. The only reason he's getting fried by the talking heads is because he made an unconventional decision. I'm too lazy to look at the details of your statistical analysis right now, but your final probability figures are close to what I expected.

michelin

November 16th, 2009 at 9:08 PM ^

The formula is correct, but I don't see any numbers applied to prove that going for it was actually better. Also, the numbers I've suggested are, in part, based on actual published NFL stats on the chances of making 4th and 2. In any case, thanks for directing me to the discussion...the board changes so fast that I don't keep up with all of it.

Plegerize

November 16th, 2009 at 8:37 PM ^

Indeed. The Pats were playing to win, not to not lose. Had they converted that play, Belicheck would've been considered a genius for finishing the game and taking the risk.

mfan_in_ohio

November 16th, 2009 at 8:37 PM ^

IIRC, the Colts still had the two minute warning and a TO left, so if the Pats make it, they would probably still have to go run-run-run-punt, and the Colts get it with 30 seconds left in their own territory with no TO's. Still unlikely they win, but not impossible. My issue with Belichick was the play calling. If he was going to go for it on 4th and 2, he should have run it on 3rd and 2. Maybe he gets the first, maybe not, but he might get it close enough for a sneak, plus Indy has to either use a TO or let it go down to the two minute warning, in which case NE gets a free TO to plan its 4th down call. Plus, now NE gets to keep its TO so Indy can't run the clock all the way down before they score. Either way, Kirk Ferentz is stupider.

UMWest22

November 16th, 2009 at 8:47 PM ^

That Monday Night Countdown just spent pretty much the entire hour and a half discussing this issue. This might be the best example of killing a (non)issue I've ever seen, esp. since, as someone said earlier, if they would have converted, he would have been hailed as the greatest tactician of all time.

The FannMan

November 16th, 2009 at 8:55 PM ^

Same facts (minus the 2 minnute warning), except it is Saturday. We are somehow winning, it is fourth and two on our 28. You are Rich Rod. Do you punt or do you go for it? Given our defense, I would try to get the two yards.

michelin

November 16th, 2009 at 9:14 PM ^

If you mean that you are asking whether to go for it, when it is not the end of the game, though, it would be a lot more complicated to figure out. Maybe you could do simulations. If it was worthwhile to go for it on 4th and 2, in general, that could affect outcomes much more, since the narrow end-of-game strategies are not frequent enough to matter a lot.

remdog

November 16th, 2009 at 8:59 PM ^

But a bit too complicated for me. Apparently New England has had a 75% success rated on fourth down against Indy in recent years. If the chance of stopping Manning and the Colts from a TD from 70 yards out is less than 75% it appears to be a mathematically good decision. NE stopped Indy short of a TD 6/8 drives in the first half (75%) but only 3/5 (60%) drives in the second half prior to the last drive. The last two TD drives Indy scored on 79 yards out in 1:59 and 1:44. In between, Manning threw a pick. Manning would have had about 2 minutes and probably 70 yards to go on the last drive after a punt. So, it appears that NE probably had a slightly better chance of winning by picking up a first down then stopping Manning and the Colts after a punt. So, based on the numbers, it was not a bad call. But I still think NE should have punted. The numbers are deceiving. NE had just failed to pick up a first down on 3rd and 2. Also, if not for a bad spot, NE actually should have had a first down. So it is arguable that NE made a first down but got robbed. So they not only had to get a first down - but by a decent margin. This makes their 75% chance of a first down a bit lower. It's also important to have some faith in your defense's ability to keep Indy from going 70 yards in 2 minutes. If you can't do this at least once in the fourth quarter, you don't deserve to win. Belicheck made a couple other glaring errors: 1. Throwing on third down when he apparently planned to go for it on 4th down. Indy was sitting on the short passes on the last few plays. A run would have likely gained some yardage and made a 4th down attempt that much easier. A run would have burned some clock and/or an Indy timeout. 2. Burning his own timeouts.

Swayze Howell Sheen

November 16th, 2009 at 9:24 PM ^

simple: public relations. if you lose after going for it, you get fried. if you lose after punting, people don't think twice about it. as the old political saying goes, "if you're explaining, you're losing." bill had to spend a lot of time explaining his decision, and hence...

michelin

November 16th, 2009 at 9:38 PM ^

You make a valid point, in general, about why coaches don't do what they should do here. The regret--and negative press-- over a choice that is unconventional is much greater than over a conventional one. If RR were in Bill's shoes, you'd never hear the end of it. However, Bill is an economist and knows what he's doing when he makes choices. His boss appreciates that. So, I think that there is no possibility whatever that Bill gets fired ever for such decisions, given the number of wins and SB's he's racked up. That's probably why he felt free to make the decision, IMO.

wishitwas97

November 16th, 2009 at 9:55 PM ^

to go for it. The odd of him converting 4th down and stopping Peyton Manning from scoring from 30 yards are better than punting the ball and have Peyton drive 70 yards to score and then try to get a game winning FG. Even with out that odd, chances of converting 4th and 2 are better than preventing Peyton from driving down the field(70 yards) to score a TD.

MadtownMaize

November 16th, 2009 at 10:03 PM ^

for arguments sake. Many Michigan fans often had a problem with Carr because he always put the game in the hands of his defense, even when his offense was the best unit on the team. Belichick believed that Tom Brady getting 2 yards in one try gave his team the best chance to win the game. I am sure that if they were not playing Peyton Manning he very likely would have punted. I have no problem with the call. I do have a problem with Kirk Ferentz's call. (FAIL)

jmblue

November 16th, 2009 at 10:50 PM ^

I thought this was going to be about the Purdue game. I wanted us to go for it then, but then, it's easy for fans like me to be riverboat gamblers - we don't have to live with the consequences.

AMazinBlue

November 16th, 2009 at 11:00 PM ^

are stats based on a team going for 4th and 2 from their own 30 yd line appx on the road with the game on the line. To simply use going for it on 4th down or even 4th and 2 means relatively nothing. Without the extenuating circumstances duplicated, all the statisical evidence shows are the chances of success in a generic 4th down situation. Whether Belichick was right or wrong is debatable only in that we know the outcome and are able to debate it. Had it succeeded, this is a non-issue, or better yet a "brilliant" call. Logic says punt and make them go 70 yards, roughly. Apparently Manning had not been 100% successful from that distance. Obviously Manning is more likely to score from 30 yards than 70 insode of 2 minutes. Had this been Jim Zorn or a rookie HC, he would be villified, maybe fired if it was Zorn. But Belichick should get a pass because of his proven success. Either way I have yet to see any evidence with the specific scenario described. Without, all this is pure conjecture and opinion.