Go For it on 4th down? Cal-Berkley study says YES

Actual Paper (Got be REAL smart to read this)
www.econ.berkeley.edu/users/dromer/papers/PAPER_NFL_JULY05_FORWEB_CORRECTED.pdf

ok, now im just a jr high guy but we punted only 3 times in 4 years.... ITS A DOUBLE WING THING, YOU WOULDNT UNDERSTAND....there was a time or two that we should have punted but yeah, we got alot of em. i think (just my opinion) that far too many punt on 4th and 2 or 3.
_

Sounds like a typical Berkley plot to mess with the Establishment.

Now, if the study had come from a high academic school with some football history like USC, Michigan, Notre Dame, et al., then maybe one could buy it.

Is Cal running the double wing?!

I would much rather our opponents have the ball on their 20 then we have it on our twenty.

There was a study of college football done years ago that showed:

When the offense began a possession on their side of midfield, it scored once every 12 possessions. When the offense took over across the 50, it scored two out of every three possessions.

I'm a simple History teacher and that math stuff escapes me. I'll take your word for it that it makes sense. The DC in me, however, says, "By all means, go for it on 4th and 2 on your own 20".
Mr. Wizard doesn't get fired for losing a game after turning the ball over on the -20.

EXACTLY, SAME WITH KICKOFFS....TEAMS RUN IT BACK TO THE 40 ANYHOW, SO MIGHT AS WELL ONSIDE KICK IT.

Studies like that are great until it is a real situation. THen I think you have to go with what your gut tells you. Twice this year we had 4th and inches from our own 40, went for it, and got stuffed. The third time we got into a similar situation I elected to punt figuring I had better learn from the 1st two debacles. But there are so many other factors. How well is your D playing? Can you stop the other team? Where are you on the field? Own 20 I am going to punt. Thier 20 YL...go for it. BEtween the 40s? That's why they pay the HC more money right?

^{hal mumme when he was at valdosta and kentucky went for it on 4th down 55% of the time. he would line up in punt formation and fake or punt it. he would line up in a offensive formation and punt or fake it.}

I do think that if you plan on going for it a lot (and 55% is a lot at that level), you need to build something in to your system to create uncertainty for the opposition when it is 4th down. For us, that is a quick kick from a 2 TE formation.

tvt50, thanks for posting this... it's brought up some interesting discussion...

I don't really buy this argument. There is nothing "illogical" about the 5'10" kid whupping the bigger kid. If you watch the tape he's probably got better leverage, exerting more effort or force, or just more skilled. The only way it would defy logic would be is if some extremely low probability event happened, like the 270 lb'er broke his ankle in the middle of the play, and was thus out of position. It might be unexpected if we don't know anything about the 5'10" kid or the 6'4" kid ahead of time, but the unexpected result isn't defying physics or biology, but simply an example of how, ahead of time, we didn't have enough information to make an accurate guess of how their matchup would turn out.

Further, while I agree that one has to take all studies of this kind with a grain of salt, just like opponent tendencies that they play cover 2 40% of the time and cover 3 30% of the time on 3rd and 7+ on a particular part of the field are just guidelines, not prescriptions, as there is another human calling the plays against you. However, studies like this and our own scouting often help guide and shape our "on-the-fly" decisions, and often present surprising results. I've scouted teams who after watching tape I was convinced racked up huge yards with the veer, and then running the numbers I realized it was their Iso and sweep that they averaged 7+ a carry on and was what we needed to focus on. Studies like this are just an extension of those kinds of simple calculations. Crunch some numbers to make us rethink things. Coaches are often extra risk averse, maybe it is to our benefit to play the expected values a bit more and be confident in our defense to hold em if we miss it. That's a bit of Belichek's philosophy.



Isn't this an argument why 3rd down values are appropriate and at least reasonably estimate 4th down results? I think Romer's assumption is that it is a 1 shot play to get the first down. 6 yards on 3rd and 10 is considered a failure, just like 6 yards on 4th and 10 is considered a failure. On 4th and 10 defenses are probably even MORE likely to be fine with giving up 5 or 6 yards since there is no worry at all about them going for it on the next down with only 4 yards to go.



I'm sure it wouldn't be impossible, but it may or may not be helpful. I agree that I never quite know how helpful studies like this on NFL games are. Studying the NFL does have a few advantages, like that you can basically assume that all teams have roughly equal talent and similar percentages (they really do), and that no team goes for it at a hugely disproportionate amount. If you had a HS district where one team went for it 55% of the time like Mumme and other schools that went for it only 5% of the time, then your 4th down data would be dominated by that one school, and may not be accurate.

Not trying to attack you or anything, just some responses off the top of my head and this is a good discussion. I don't think you can take a study like this and throw out everything you've ever learned about coaching, but sometimes rethinking things is a good thing.

_{I don't really buy this argument. There is nothing "illogical" about the 5'10" kid whupping the bigger kid. If you watch the tape he's probably got better leverage, exerting more effort or force, or just more skilled. The only way it would defy logic would be is if some extremely low probability event happened, like the 270 lb'er broke his ankle in the middle of the play, and was thus out of position. It might be unexpected if we don't know anything about the 5'10" kid or the 6'4" kid ahead of time, but the unexpected result isn't defying physics or biology, but simply an example of how, ahead of time, we didn't have enough information to make an accurate guess of how their matchup would turn out.}.
I agree with you... however, most scientific studies would qualify this as "logical" because logic is a "philosophic system based on inference". Generally,myself being one of the smaller people on the field my whole career, I certainly heard enough of "the bigger guy should win" being infered... that's all I was saying. There is a human element that is non scientific. It was probably a bad analogy, but humans are not mathematical automations, some great events occurred in sports because we are not.

Even with the 3rd/4th down play calls- there are situations where, like Huey said- it wouldn't be wise to force anything. if it is 4th, an interception might be better than an incompletion. Or there are the 60s Packers teams that ran play action on 3rd and 1... if they didn't make it, they could always kick. 4th down failure has a much greater impact on momentum and field position therefore I don't see it as an accurate estimation.

Also, like I said, this study already encourages the (insane/brilliant) tendency we have to go for it on 4th down.

^{IMAO, the single most important dictum in football is "keep 'em guessing."}

I think we do this... we run right AND left. While I agree with that statement, there is still something to be said for being predictable and unstoppable. My favorite memories in games involve segments where we were running the same play 8 times in a row and they couldn't stop it.

This whole discussion/analysis on 4th down reminds me of a pre-game meeting one year- we meet by position for 10 min. before the game. One year, I didn't have a specific defensive responsibility (as the HC, I was the "wanderer"), so I sat in on the D-Line meeting. The DL coach had 10 min to discuss schemes, adjustments, etc... but this meeting took about 10 seconds- all he said was "line up and kick their ass". I'll never forget that- for the moment and that group of players it was perfect. Sometimes, we greatly over-analyze this game... I'm guilty of that too.

I suppose I do not want our players to think we go for it a lot on 4th because of some statistical analysis... I want them to think we do it because "we're tougher". I guess I've watched Braveheart, Rocky (I), Die-hard, Henry V and stuff like that too many times...

Also, I'm not a "math" guy...I was an English/PE double major... so I tend to quote people like e.e. cummings ("nothing measurable matters...), which probably explains why I have an aversion to studies like this. As I said, it is interesting though...

I've read most of these articles on statistical football.....I agree with everything written thus far.

The main issue involved is statistics....they can quantify a position, but they aren't necessarily the "truth".

There is also a statistical paper on the ratio between FIELD POSITION and scoring probability.

Your ability to "go for it" all depends on the defense you are facing, the punt coverage ability of your team and your opponent, and the capabilities of your punt unit.

Punting, IMO, is not a "bad" thing....sure you change possession, but you can dramatically alter the field position (if you're any good).

When you face good defenses, going for it on 4th can be a game-killer - sometimes you just have to manage all 4 quarters of the game, instead of thinking short-sighted.