Fourth-and-1 from the Pats’ 24-yard line

[maverick4]

Obviously everyone is happy that it worked out great, and it was a key point in the game.

However, no amount of pseudo-math or pseudo-stats can make a legit rational case why it was a smart move, based on reason or analysis, to prove that going for it on 4th down at your own 25 in the 3rd quarter in a close game, is a smart rational move. Using a simple payoff tree, the benefit/payoff is a 1st down near your own 25, the penalty/damage is a sure FG for the other team, or a decent chance at giving up a TD. No matter how much you manipulate the percentages of likelihood, it's not a smart call. Of course, it's still amazing that it worked and everyone is happy after the fact.

Even Belichick admitted it wasn't a smart rational move in his post game interview, he admitted he would have gotten roasted if it didn't work.


I have read those research papers advocating going for it on 4th down. I generally agree that especially past your own 40 yard line, there are circumstances where you should go for it more. However, some of their research claims and inputs are just crazy. Some economists advocate that once you are past the opponent's 35 yard line, you should always go for it, even if it is 4th and 10. These people are using odd inputs, assumptions, and illogical football knowledge to make their cases.

 

[maverick4]

SOURCE: TMQ: Punting the punt - ESPN Page 2

Did you even read that article? I have. And a few other papers advocating other scenarios.

Read that guy's inputs and assumptions about how he used the stats. They make no sense at all.

Going for it on 4th and 10 once you're in FG range, never kicking a FG, is what this guy is advocating.

Also, TMQ in general is a huge tool who isn't nearly as intelligent as he thinks he is.

 

[Absurdly Metro]

What complete loser of a Patriots player back then droped the 4th down pass that woulda made a statement? I wonder...

I don't remember. Who was it? Terry Glenn??

 

[fumbrunner]

It's only genius if you convert it. Anyone remember going for it on 4th and long in the last superbowl instead of trying a FG? There were alot of name calling after that call and I don't recall "genius" being one of them.

 

[JSn]

It's only genius if you convert it. Anyone remember going fo it on 4th and long in the last superbowl instead of trying a FG? There were alot of name calling after that call and I don't recall "genius" beig one of them.

Hey, thanks for mentioning that...

 

[emoney_33]

This is about the stupidest thing I've read on this site ever.

Its not just the one play that is affected by this outcome its the whole game. In football the W at the end of the game is so important. We were risking the W for a very small gain here. The situation is totally different to poker.

It was one play and it's not an all-in play and is very analogous to a hold-em tournament play. There is NO single make or break play in a football game. It's effect mixed in with all the other plays in the game leads to a W or L.

But yea I guess it was the stupidest thing you've ever read

 

[patriots pam]

When I was watching, I yelled "GO FOR IT!" at the TV. I guess BB heard me and didn't realize that I didn't really mean it. Imagine my surprise when the actually did. I was excited that they were going to do it, and was thinking it just might be the right idea. Then again, after a few beers or so, the list of things that I don't consider a good idea is pretty short.

They say it's a fine line between bravery and foolishness, but since it's BB I'm going to err on the side of bravery. Soooo glad it actually worked as planned. Whew!

 

[PatsWickedPissah]

I don't remember. Who was it? Terry Glenn??

No, the stump who droped the easy 4th down opening drive pass that woulda rocked the Broncos was not Terry Glenn. It was...

 

[ForThoseAboutToRock]

Did you even read that article? I have. And a few other papers advocating other scenarios.

Read that guy's inputs and assumptions about how he used the stats. They make no sense at all.

Going for it on 4th and 10 once you're in FG range, never kicking a FG, is what this guy is advocating.

Also, TMQ in general is a huge tool who isn't nearly as intelligent as he thinks he is.

I've read the Romer paper. To me, it's interesting, but not wildly practical due to the concessions he makes in order to try to establish as much stability as he can from a statistical variance standpoint (only counting 1st Q situations, measuring his values on 3rd down instead of 4th down (when intuition tells us that the yardage on 4th down should be shorter than on 3rd down on average)) etc.

Still, his claim that teams are hurting their chances of success by being overly risk averse, is true even when allowing for his margin of error. This is a nice model for academic theory, but demands caution when applied to real situations (same with the risk models devised by M.I.T. physicists and mathematicians that bankers used to 'prove' to themselves that all their subprime mortgages were within their risk tolerances... and we obviously know how that worked out).

I found TMQ's anti-punt guidelines pretty practical. He burned my edges with all the Spygate talk, but I like the anti-punt feature in his column, because I think he's right (and I feel that the stats out there proving him and others right are valid)... how I feel about how smart he is, or how smart he thinks he is, on other subjects doesn't really make much difference since that's not what I'm referring to in using his column for my post.

 

[Patspsycho]

No, the stump who droped the easy 4th down opening drive pass that woulda rocked the Broncos was not Terry Glenn. It was...

Yes that's right, it was....

 

[chris_in_sunnyvale]

No, the stump who droped the easy 4th down opening drive pass that woulda rocked the Broncos was not Terry Glenn. It was...

Bruschi? Didn't the Pats fake the punt and Bruschi was the intended receiver?

Regards,
Chris

 

[eagle eye]

It was one play and it's not an all-in play and is very analogous to a hold-em tournament play. There is NO single make or break play in a football game. It's effect mixed in with all the other plays in the game leads to a W or L.

But yea I guess it was the stupidest thing you've ever read

You don't get it do you?

I play poker and I understand ev and reverse odds and everything else about the game. This is not in any way similar to poker. There are too many variables that you do not have control over. Its incorrect to assess this using ev, well to assess it as a +ev play anyways.

The smartest play in that circumstance is to fake like you are going for it and try and get an offside call, but it they don't go for it you kick it away.

 

[emoney_33]

You don't get it do you?

I play poker and I understand ev and reverse odds and everything else about the game. This is not in any way similar to poker. There are too many variables that you do not have control over. Its incorrect to assess this using ev, well to assess it as a +ev play anyways.

The smartest play in that circumstance is to fake like you are going for it and try and get an offside call, but it they don't go for it you kick it away.

Maybe you should re-read the thread (and my point). I was pointing out that any ONE specific 4th down situation should not be determined by the overall statistical evidence that suggests going for it is the right play. I compared it to poker in the sense that just because a play is right against 90% of opponents doesn't mean you should ALWAYS make that play (for instance if you had a good read on your opponent that suggested it was the wrong move, you wouldn't make that move just because 90% of the time against the general poker population it would be right).

 

[Satchboogie3]

This is about the stupidest thing I've read on this site ever.

Its not just the one play that is affected by this outcome its the whole game. In football the W at the end of the game is so important. We were risking the W for a very small gain here. The situation is totally different to poker.

You are not reading what he wrote. He is saying that, like poker, going for it on 4th and 1 isn't just automatic, there are TONS of variables you need to consider. In yesterday's game, the variables were that it was in the 3rd quarter, at our own 25 when we were ahead by 6. That would seem to dictated that we just punt. HOWEVER, our offense was moving the chains at will, the running game was going well, and Belichick knew their interior defense was weak.

Then you do a risk/reward assessment. If we make it, it is a big blow to the Falcons, whom (I assume) got worked up at the opportunity to take over on offense. We keep the ball, continue the clock, and march down to likely more points scored. If we don't make it, we give the Falcons good field position. Now you have to weigh the odds of us converting the 4th down (which, according to all the variables, was incredibly high). You factor in the clock time we can kill, which detracts from the odds of the Falcons having time to come back, along with the odds of the offense building a bigger lead, minus the odds of the Falcons scoring if we give it up. Looking at this objectively, it appears that the Pats would very good +EV if they go for it. That swings this situation in favor of going for it, when otherwise, you would think to punt. THAT is why he was comparing this to poker.

Anyways, his overall point about making +EV (expected value) plays, no matter has slim, in the correct situations will lead to more wins/money. It's a matter of risk reward, and when the reward outweighs the risk, you have to take it, otherwise you are surrendering equity. I understand that there is an absolutely crap load of variables (in poker, such as tournament vs cash game, tournament structure - MTT - STT, being on the money bubble, etc), but the overall point stands.

To do a crude estimate, if you think that we would convert that 4th and inches 75% of the time given all those game factors, it would be a move that would be worth +1.19 points.
Breakdown: (.75*3.25) - (.25*5) = 2.44 - 1.25 = 1.19
That is the odds of making it multiplied by the points we score (on average, 8 drives, 6 scoring for a total of 26 points) minus the odds of us turning the ball over multiplied by the points surrendered (2 RZ trips, 10 points). +1.19 is the amount of points we expect to get in the long run. No doubt this a crude estimate that doesn't take into account many factors (including more pro Pats factors), but it's the method behind making the correct decision.

 

[AzPatsFan]

I thought the play all was gutsy and only something that a self-assured Coach would dare. But not nearly as dangerous as what some ostensible Pats fans have been calling for almost dailly, all-out blitzing.

Possible Gain +1 Yd and First down and retain possession; or Possible Loss: -1 Yd and 0 points and FD for opponent on your 25. No change in possession status as Punt concedes possession, but +40 yds on punt.

Compare that to what some Pats fans have been yelling for:
Possible Gain: +7 yds and sack; or Possible Loss -75 yds and -7 points

By comparison it is much safer to Go for it.

 

[eagle eye]

You are not reading what he wrote. He is saying that, like poker, going for it on 4th and 1 isn't just automatic, there are TONS of variables you need to consider. In yesterday's game, the variables were that it was in the 3rd quarter, at our own 25 when we were ahead by 6. That would seem to dictated that we just punt. HOWEVER, our offense was moving the chains at will, the running game was going well, and Belichick knew their interior defense was weak.

Then you do a risk/reward assessment. If we make it, it is a big blow to the Falcons, whom (I assume) got worked up at the opportunity to take over on offense. We keep the ball, continue the clock, and march down to likely more points scored. If we don't make it, we give the Falcons good field position. Now you have to weigh the odds of us converting the 4th down (which, according to all the variables, was incredibly high). You factor in the clock time we can kill, which detracts from the odds of the Falcons having time to come back, along with the odds of the offense building a bigger lead, minus the odds of the Falcons scoring if we give it up. Looking at this objectively, it appears that the Pats would very good +EV if they go for it. That swings this situation in favor of going for it, when otherwise, you would think to punt. THAT is why he was comparing this to poker.

Anyways, his overall point about making +EV (expected value) plays, no matter has slim, in the correct situations will lead to more wins/money. It's a matter of risk reward, and when the reward outweighs the risk, you have to take it, otherwise you are surrendering equity. I understand that there is an absolutely crap load of variables (in poker, such as tournament vs cash game, tournament structure - MTT - STT, being on the money bubble, etc), but the overall point stands.

To do a crude estimate, if you think that we would convert that 4th and inches 75% of the time given all those game factors, it would be a move that would be worth +1.19 points.
Breakdown: (.75*3.25) - (.25*5) = 2.44 - 1.25 = 1.19
That is the odds of making it multiplied by the points we score (on average, 8 drives, 6 scoring for a total of 26 points) minus the odds of us turning the ball over multiplied by the points surrendered (2 RZ trips, 10 points). +1.19 is the amount of points we expect to get in the long run. No doubt this a crude estimate that doesn't take into account many factors (including more pro Pats factors), but it's the method behind making the correct decision.

You see this is all absolute rubbish, the reason being that you never have enough of a sample size to make an ev decision.
You cannot use previous seasons due to personnel changes on both teams, conditions of play on the day of a game. So you never have an adequate sample to make a decision like this mathematically.

 

[maverick4]

I compared it to poker in the sense that just because a play is right against 90% of opponents doesn't mean you should ALWAYS make that play

What emoney is saying (I think/hope) with his poker analogy, is that sometimes weird stupid ass plays win, even if the percentages aren't with you.

An example of what he's talking about, is going all in with a pair of fives early on in a tournament match.

Of course it's not the best mathematical play, but you could win based on a hunch or a feeling.

That is basically what Belichick did yesterday. There is no payoff-tree, or stat, or math that can justify trying to gain a 1st down on your own 25, in exchange for possibly giving up a sure FG or possibly TD.

But, it still works sometimes, and it did for Belichick yesterday.

 

[PatsFaninAZ]

I've argued in this forum that these so called "studies" about going for it on 4th down are severely flawed. They simply do not take into account other bad scenarios that can occur to a team while on offense that are negated with a chance to punt. They also assume an unwarranted likelihood of success in the remainder of the drive.

The best examples are actually inside the opponents' end of the field. Assume fourth and inches on the 10. These scenarios -- at least the ones that I've seen -- assume that the primary downside of going for it is the chance of not converting. They do not calculate the fact that you can make the yardage but then throw a pick or fumble, thus depriving yourself of the 90percent chance you had at 3 points. Nor do they calculate the possibility of penalties that make a field goal relatively more difficult. They also do not take into account time management concepts, which would mainly apply when the team is losing, although they might apply in other circumstances too.

I simply do not believe in any way shape or form that the decision Belichick made yesterday can be justified as better than a 50/50 proposition in terms of risk reward, going by the game situation and the advantage gained by making it versues the advantage gained by punting. No way. My guess is that even BB would agree. But this is not to say it was the wrong call. There may have been other reasons to do it -- based on the Head Coach's attitude about motivation. If anything, the fact that it is not a better than 50/50 proposition is perhaps a reason for a coach to try it as something unusual to give his team a kick start. Perhaps he viewed the benefits to making it to be greater than just having a first down on the 25 yard line, and perhaps he thought there was some benefit the team would derive mentally in trying but failing. All of those are things you want your HC to be in touch with. So long as he is not going with his "gut". If he understands the risk/reward, and chooses to take a lower proposition for reasons that he understands because he has the pulse of the team -- and doesn't make a habit out of it -- I'm all for it. Obviously, it worked yesterday.

 

[jbb9s]

Bruschi? Didn't the Pats fake the punt and Bruschi was the intended receiver?

Regards,
Chris

I think so. Because the memorable TV color man's call "Bruschi surely would have caught it if it was a brew-ski". Am I making this up?

 

[crypto]

It was a horrifically bad call. As for the "It boosted the defense" garbage, the defense sure as **** was playing well before that play and to say that it needed a boost is utter crap.