Did the New York Giants Risk Losing the Game to Cover the Spread?

Last Sunday, the New York Giants played the Chicago Bears in football. The point spread on the game favored the Giants by 1.5 points, meaning that if the Giants won by only one point, those who bet on them would lose.

Having watched the game myself, I was not at all surprised to receive the following e-mail from a fan who prefers to remain anonymous:

Yesterday was the usual Sunday fare, sitting on my couch watching football games that I had bet on while I worked on research. The day had gone in a most familiar way – lost hundred of dollars on the Chiefs +6 against the Chargers, but looked like I was about to win it back with the Raiders +3.5 against the Broncos.

My last game of the day was the Bears +1.5 against the New York Giants. The Bears grabbed an early lead, and things were going pretty well, with a 16-7 advantage midway through the fourth quarter. My 1.5 points looked good, since the worst case scenario looked like a 17-16 Giants win. In which case, I would still win my bet.

The Giants proceeded to score a touchdown with under 7 minutes left. No problem — the Bears will take the next kick-off, chew up several minutes with a few first downs, and, worst case scenario, punt. What actually happened was a three and out, and now the Bears punted to the Giants, but at least they started on their own 23 with less than 5 minutes remaining.

The Giants promptly drove down the field, and with 1:37 remaining found themselves at the Bears’ 2 yard line, at first and goal. I quickly computed in my head that this was okay, because in the process, the Bears had used 2 of their 3 timeouts. Thus, the Giants should just kneel a few times, get the ball to the middle of the field, and run the clock to a few seconds before kicking the game-winning field goal (which would amount to an extra point) as time expired. The fact that the Bears called a timeout to allow the Giants to think this through made me really confident that Coughlin would proceed in this direction. Even the announcer, Troy Aikman, recognized this and called for that strategy. The game would end 17-16, I would win my final bet of the afternoon, and maintain my cheerful holiday mood for our planned tree decorating later that night.

Alas, the ending was not as our hero had planned.

If the Giants had followed the obvious plan, they would have been left with a field goal roughly the length of an extra point to win the game. The success rate on extra points in the NFL is 96 percent. So the Giants were virtually assured of winning the game.

The only problem was that they would only win by one point. Which means they wouldn’t cover the spread.

Could that explain why instead, the Giants ran a sweep on first down, scoring a touchdown? This gave them a five point lead, but also gave the Bears the ball back with 1:33 on the clock and a timeout! It also required the Giants to kick off to Devin Hester, perhaps the most dangerous returner in the history of football. The Bears very nearly did come back to score a touchdown, although they ultimately failed.

So what do we think? Simply a case of bad decision making by an NFL coach (it wouldn’t be the first time according to economist David Romer)? Or something more devious?


Would you fools please forget about ex post outcomes and focus on ex ante probablities, as Professor Levitt is trying to stress.

Panem et Circanses

Football coaches' decisions are as bad as baseball managers' were, before writer Bill James and others showed them the error of their ways. When similar change happens in football, this sort of stuff will become rarer.


Viewing this from a probability standpoint:

1. Taking a knee, as mentioned above generally costs 2 yards of field position, and therefore is not a valid option - it takes you out of just needing to win the push battle at the line of scrimmage, and gives the defense the oppurtunity to cover the strong side, even when the running back starts going inside. A loss of two yards greatly reduces the Giants' probability of scoring a touchdown in that situation.

2. While, in your usual late game situation, you may find a higher probability for winning by running the clock out and then kicking a field goal, you cannot generalize and abstract the following conditions out of the equation:

A: The long snapper is a rookie, which increases the propbability of him muffing the snap. While DeOssie's probably not unfamiler with playing in bad weather (he played collegiantly at Brown), he's about to have to block players that are much bigger, stronger, and quicker than any of his Ivy league opponents.

B: Even if DeOssie does everything right, the ball is slippery as evidenced by Manning's earlier fumble - that could effect DeOssie's release and Feagle's catch.

C: There was a strong, variably gusting wind. The long snapper has to fling the ball 21 feet through the air, to a player who is kneeling, who has a limited area where he can catch the ball and still be able to get it down for the kicker, who is timing his approach off of the snap, not the holder getting the ball down. A couple of the long snaps on punts had been high. A high long snap on a field goal attempt turns it into a faked field goal play, and Feagles is not someone who is a threat to run the ball into the end zone.

D: Tynes is a below average place kicker, who is in what is arguably the highest pressure situation of his career. In addition, his coach has been riding him hard all game, which is probably doing wonders to his confidence and concentration.

E: To get the kick off, Tynes needs to be able to plant his non-kicking foot. The field is in horrible shape, and traction at the spot of the kick is far from guarunteed.

F: Assuming the kicker does his job and gets the kick off, it needs to travel 24 yards into that gusting wind (yes, 24, not 18, see below) - if the ball is kicked at all off center, the wind will push the ball wide.

Now, with those five factors added, I don't think the Giants had a 96% chance of converting the field goal as suggested by David, although he seems to be basing that on the Giants kicking from 17 yards out. Taking a knee 3 times (and running the clock down to 14 seconds - still enough time for a Hester kick off return, or enough time for the Bears to have 2 shots at getting into field goal range) would have cost the Giants 6 yards, and made the field goal a 24 yard field goal, not an 18 yard one.

The probabilty of Tynes missing such a kick under normal conditions is closer to 8%. This season, Tynes is 35 of 38 on kicks within 30 yards, or about 92%, but I disagree that Tynes had a 92% of converting in that specific circumstance with all of the other variables.



It's not so much as Giants coaching staff but Chicago also. Have anyone notice that Chicago just let Giants walk on in and hoping just like we all thought Devin Hester or Benard Berrian come to the recues like last week?? Also if you watch the gameYou can see that you can drive a dump truck through that's how wide is end zone on that play. Not all on Giants but on both side. You don't give I can't take. rule of thump


Last night's game illustrates the additionial elements that go into a field goal attempt - the Giants (having read this blog presumably) ran down the clock for their kicker to kick a game winning field goal as time expired.

The long snap was botched, throwing off the kicker's timing, and leading to a miss.


give me 16 games where im kicking a 20 yard field goal on 4th down to win the game as time expires u people are idiots


Remember the discussion of the value of punting versus going for it on 4th down? The theory of why coaches punt even though the odds favor going for it is that if they punt and the defense gives up a score, it's the defensive players' fault, but if they go for it and fail, it's the coach's fault. Same thing applies here. If they kneel to run out the clock and miss the field goal, it's the coach's fault for trying something fancy (versus trying two or three times to score a touchdown), while if they score right away and the Bears come back to win, it's the players' fault for giving up the touchdown.

a student of Economics

Scoring quickly is definitely a mistake, but the TD try might make sense.

Here's a potentially better strategy:

Kneel down twice, run the clock to 10 seconds and take a time out.
Then run up the middle (maybe from "kneel down" formation). If you don't score, take another time out and kick the field goal.
This gives you 2 chances to score the winning points, instead of just one, increasing your odds or winning from 96% to perhaps 97 or 98%.

Of course, you need to a) make sure no one fumbles on the running play and b) kick away from Devon (squib kick) on the ensuing K.O.


The move was probably made becaues of the Giants' kicker, Lawrence Tynes. He has missed two extra points this year, and his job is hanging on by a thread.

It might have been the wrong move, but that's because the Giants could have taken a knee to reduce the time left on the clock, and force the Bears to use another time out. I completely understand why Coughlin did it.


The Giants' kicker was having a reasonably bad game. Perhaps this was a factor in Coughlin's decision to try for the touchdown?


Nah. 1st down with more than a minute and a half left, why not try and score a TD? They just made it in on the first try.


Remember, we're dealing with the Bears' timekeeper. With that in mind, here's how the kneeling scenario might have played out, with what I would consider to be generous timing for the Giants...

1:37 1st Down: Kneel (call it a 3 second play)

1:34 TIMEOUT Chicago (#3 of 3)

1:34 2nd Down: Kneel (3 + 37 second play clock)

0:54 3rd Down: Kick Field Goal

Yes, they kick it on 3rd down. No football team in their right mind sets things up to kick on 4th down. They always kick it on 3rd down in case they experience the dreaded fumbled snap. In that case, they down the ball and line up for another try on 4th down. Two shots for one field goal.

The point is that no matter how you look at it, there would still be enough time on the clock (at least 45+ seconds in my scenario) for Hester to run the ball back far enough to lineup for a FG, if Hester didn't just run it back to the end zone. And if that's the case, I'd rather force Chicago and Rex Grossman (or Hester), to get into the end zone to beat me instead of just getting far enough for a FG.

And note that what the Giants did had the intended effect:

3-10-NYG 28 (:04) 8-R.Grossman pass incomplete deep left to 23-D.Hester (37-J.Butler).

A little quick math calculates that to be a 45-yard field goal. Quite makable for Gould, who is 11-for-12 from 40-49 this year. That 45-yard field goal would have won the game for the Bears. Instead, they had to keep going for the touchdown and lost.



Have you seen many Giants games this year? Their kicker has already missed two extra points. A smarter strategy was to run the ball right up the middle, not score, and make Chicago call a time out, and then score after that (whether on the ensuing play, or a later one).

(Of course, the smartest move would have been going for two instead of one, to go up by six points, but that's neither here nor there)


My first reaction is that despite the claims on the amount wagered, this is not a hardened gambler making this complaint. I have never in my LIFE thought about the possible outcomes of a wager and decided that in the "worst-case scenario" I am still guaranteed a win, especially once things started to turn against me. If he thought the Bears were guaranteed "a few first downs" then he hasn't been watching their offense the last few years. Of COURSE you expect the other team to go for a touchdown (and in my experience they usually do), and Vinatieri's missed chip-shot a few weeks ago was probably enough to scare an old-schooler like Coughlin into the obvious play first.


Todd, I'm sure this has already been pointed out, but the "fancy" play (and the one which would bring the most criticism on the coach) is actually going for the TD in this situation. If the players screw up a chip-shot field goal (basically an extra point), that is in no way the coach's fault. It's still an interesting point to bring up, but you've got it backwards.


I hope Giants don't care about the spread !

I think the market (bookmakers) clearly and correctly evaluated the odds, and that was close to it. the handicap was very well chosen

More to it
There is also a strong incentive for a team, (or a trainer, and in the long term for the league)to win with panache, people want some show and prefer TDs to punting.

That is a smart way to attract people and supporters in stadiums


This decision may not be quite as sub-optimal as it appears. The Football Commentary website helpfully offers a table of 2-minute drill probabilities (http://www.footballcommentary.com/twomindrillprobs.htm), and Chicago's situation (1:33 left, one timeout, down by >3 points) roughly corresponds with an 18% chance of getting a TD. Better than the 4% chance they have of winning by hoping that the Giants miss the FG, but then the Bears have a below-average offense and the Giants have an above-average defense. But then, as you point out, they have Hester. So a sub-optimal decision, yes. But one assumes as much from Tom Coughlin. A gambling-conspiracy allegation is a bit overblown.


Another possibility, is other incentives. Not to believe that the players went against the strategy the coach dictates, but even without betting on the game, players have incentive bonuses and just wish to have higher personal stats (especially if they "have themselves" in a fantasy league).

Suppose a coach is in a fantasy football league (as simple as on espn.com or yahoo), and has his own running back. It would be horrible from an ethical point of view, but he might want extra stats for his player, and adjust his strategy.

Or perhaps its just decision making. Aikman is obviously under less stress at the time when its occuring. And coaches make mistakes that they "should know" having been in the league that long (look at the consecutive time outs last week as well)


It may be questionable but it certainly is not a bad call.

If you go for the kick and you miss, you get 0 points and lose the game within feet of the end zone. If you try to score a touchdown and you fail [remember that it was first down, after all], you have two more downs to try to score before you give up and still have a shot at a close field goal on 4th down.


A fascinating post! While watching this game, I had exactly the same reaction: why not run down the clock and kick a field goal? Tynes is an experienced kicker--one of the best ever in the CFL. I agree that there might have been something fishy going on; after all, betting is widely perceived to be a huge problem in sports (the Rick Tocchet case a prime example in hockey). It would have been interesting if Aikman had offered up the betting explanation!