A Team That Never Punts

[strohs]

If teams would kick it out of bounds inside the 20, then I would advocate punting from the 40.
But if you are at the 40 and you are just going to kick it into the endzone anyway, I say go for it.
I just dont understand is the obsession with always trying to pin the opposing team inside the 5 when it rarely works...

 

[martinss01]

According to Kelley's data (much of which came from a documentary he saw), when a team punts from that deep, the opponents will take possession inside the 40-yard line and will then score a touchdown 77% of the time.

Ask him to defend his methods, and he revs up his Dell laptop and refers to his statistics.

so he has the documentary on his laptop then?

ya know, theres only 1 reason why punting the ball from your own 1 on 4th and 20 is the same as punting the ball and the opposing team getting it on your 40 carries essentially the same risk. YOUR DEFENSE SUCKS DONKEY BALLS. seriously, if your opponent getting the ball at your 1 is even remotely close to the same risk as them getting it at your 40... your defense must be HORRIBLE!

 

[HoldenesqueAlex]

The Berkeley economist David Romer wrote a paper about the merits of going for it on 4th down, available here for you nerds: http://elsa.berkeley.edu/users/dromer/papers/nber9024.pdf

The conclusion is that coaches very often do not make the rational decision on 4th down. (That is, rational in the sense of win-probability maximizing.) One of the striking results was that it made sense to go for it on 4th and goal within the 5. This is because it is (on average) worth points just to pin your opponent that deep in their own territory. It's not just that you get more points for the TD, if you don't make it then you've still pinned them within their five. Using all of the data from the NFL, starting a drive at your 1 yard line is worth about -1.5 points to you (See page 38 of paper for graph of field position values). So, even if you assume a 100% success rate on field goals, you don't need a huge probability of converting the 4th down for it to be worth it to go for it on an expected value basis.

Romer used only data from the NFL in the first quarter (to eliminate end-of-game/half issues) to make the field position value graph and conversion probabilities. The results may vary for college football.

The larger question is: what are the coaches' true goal? Maximizing win-probability, or keeping their job? Obviously, winning is good for the latter goal also, but it looks very bad when you make an aggressive play that doesn't work, even if it was the logical choice. Athletic directors likely haven't read the Romer paper either.

So, should Tressel have gone for it on that 4th down against USC? (Left as exercise for the reader.)

 

[martinss01]

The Berkeley economist David Romer wrote a paper about the merits of going for it on 4th down, available here for you nerds: http://elsa.berkeley.edu/users/dromer/papers/nber9024.pdf

The conclusion is that coaches very often do not make the rational decision on 4th down. (That is, rational in the sense of win-probability maximizing.) One of the striking results was that it made sense to go for it on 4th and goal within the 5. This is because it is (on average) worth points just to pin your opponent that deep in their own territory. It's not just that you get more points for the TD, if you don't make it then you've still pinned them within their five. Using all of the data from the NFL, starting a drive at your 1 yard line is worth about -1.5 points to you (See page 38 of paper for graph of field position values). So, even if you assume a 100% success rate on field goals, you don't need a huge probability of converting the 4th down for it to be worth it to go for it on an expected value basis.

Romer used only data from the NFL in the first quarter (to eliminate end-of-game/half issues) to make the field position value graph and conversion probabilities. The results may vary for college football.

The larger question is: what are the coaches' true goal? Maximizing win-probability, or keeping their job? Obviously, winning is good for the latter goal also, but it looks very bad when you make an aggressive play that doesn't work, even if it was the logical choice. Athletic directors likely haven't read the Romer paper either.

So, should Tressel have gone for it on that 4th down against USC? (Left as exercise for the reader.)

you of course are assuming that the instances in which you didn't go for it would have had the exact same outcome as the instances in which you did. i suspect talent levels, score, records going into the game, and momentum are not taken into account in this little study.

for example, texas tech just had the ball inside the houston 1 yrd line and went for it on 4th and inches. while the resulting houston drive ended in an int. texas tech not only lost the almost sure 3 points but they also lost the field position as they now have the ball on their own 34. not sure how your math works but imo 3 points on the board is better than having the ball on your own 34. expecially when you realize the field goal would have forced houston to either score a td and a 2pt conversion to tie and score twice to win.

edit: last season tOSU went for it on 4th and long with a fake punt from around the 50 against northwestern late in the 4th with a big lead as time was running out. i "suspect" that play was more likely to be successful than going for it on 4th and goal from the 2 against usc in the 1st quarter. but then i don't have a documentary on my laptop so what do i know?

edit #2: texas tech just lost to houston specifically because they went for it on 4th and goal from inside the houston 1 rather than kicking a field goal. this made it a 5 point game rather than an 8 point game. houston drove down the field and scored a td. they then failed to complete the 2pt conversion. had texas tech kicked the field goal rather than going for it they potentially would have won the game 38-36. -1.5 points indeed....

 

[HoldenesqueAlex]

you of course are assuming that the instances in which you didn't go for it would have had the exact same outcome as the instances in which you did. i suspect talent levels, score, records going into the game, and momentum are not taken into account in this little study.

for example, texas tech just had the ball inside the houston 1 yrd line and went for it on 4th and inches. while the resulting houston drive ended in an int. texas tech not only lost the almost sure 3 points but they also lost the field position as they now have the ball on their own 34. not sure how your math works but imo 3 points on the board is better than having the ball on your own 34. expecially when you realize the field goal would have forced houston to either score a td and a 2pt conversion to tie and score twice to win.

edit: last season tOSU went for it on 4th and long with a fake punt from around the 50 against northwestern late in the 4th with a big lead as time was running out. i "suspect" that play was more likely to be successful than going for it on 4th and goal from the 2 against usc in the 1st quarter. but then i don't have a documentary on my laptop so what do i know?

edit #2: texas tech just lost to houston specifically because they went for it on 4th and goal from inside the houston 1 rather than kicking a field goal. this made it a 5 point game rather than an 8 point game. houston drove down the field and scored a td. they then failed to complete the 2pt conversion. had texas tech kicked the field goal rather than going for it they potentially would have won the game 38-36. -1.5 points indeed....

-1.5 points is an average. Sometimes the team will make a 99 yard drive. Sometimes they'll give up a safety. Sometimes they'll go 3 and out and give the other team great field position. There are many possibilities, but on average (in NFL 1st quarters...) it's worth -1.5 points. Romer's analysis isn't immediately applicable to the Texas Tech situation because it had end-of-game strategy affecting it, but I'm still not sure it was the wrong decision.

It is possible that Texas Tech could have converted their 4th down. Then they would have had a two possession lead (rather valuable I'd say). Would you be critical of the coaching then? We only get to see one of the possible outcomes. Even if that outcome is undesirable, it doesn't necessarily mean it was the wrong decision.

Let's play this game: You get to choose between doors A & B. Door A has $1,000,000 behind it 99% of the time, nothing 1% of the time. Door B has $100 behind it 100% of the time. You choose Door A, get nothing. Did you make the wrong decision? After all, you could have had a sure $100.

 

[Jaxbuck]

Trying to use economics to find inefficiencies in the system is always a good idea. No one is going to just go for it on every 4th down if they like their job but I am all for anything that makes people think and opens their eyes to a different(possibly better) way of doing things.

I have long been a proponent of going for it more on the plus side of the field but it always comes down to game situation and each teams relative strengths and weaknesses.

 

[matcar]

Trying to use economics to find inefficiencies in the system is always a good idea. No one is going to just go for it on every 4th down if they like their job but I am all for anything that makes people think and opens their eyes to a different(possibly better) way of doing things.

I have long been a proponent of going for it more on the plus side of the field but it always comes down to game situation and each teams relative strengths and weaknesses.

I live in Little Rock and hence I see this team a fair amount. My son's school is in the same division and it's amazing to watch this team go for it on 4th down, inside their 10, with 15 yards to go. I'm not suggesting that this would work at the college level (much as the spread doesn't seem to work at the pro level) but at the high school level, it seems to work quite well. It's a whole different level of pressure on the defense when you know that there are 4 downs that the offense will use and likewise, it reduces the pressure somewhat on the offense as you always have an additional play to get the first down.

Now, it also creates additional pressure on your own defense as you often put them in difficult positions where the opponent inherits the ball in your territory, but it works quite well for the Bruins as they won the class 5A championship last year. Even as an opponent, it's a marvel to watch how "ballsy" they are sometimes.

 

[Jaxbuck]

It's a whole different level of pressure on the defense when you know that there are 4 downs that the offense will use and likewise, it reduces the pressure somewhat on the offense as you always have an additional play to get the first down.

Just looking at it on the surface that is exactly the thing I was thinking about. I have always thought it was deflating to a defense to make a stop just shy on 3rd down and instead of getting off the field seeing the offense line up to go get their 1/2 yard or whatever.

Almost feels like the defense is thinking "hey wait, the book says you should punt now we should be out of danger, thats what we do when we are on offense" while the other teams offense is saying "fuck you and your book, you are still very much in danger if you can't stop us".

I'm not saying blindly go for it on every 4th down but instead of always thinking about the pressure applied if you don't make it I think the pressure applied by going for it needs to be given consideration.

 

[BB73]

The type of game factors into the decision for me.

If teams are both having a tough time moving the ball, field position is more valuable. If the game's a total shootout, giving up the ball is basically just a lost opportunity for points.

I call the latter type of game a 'points per possession' game, as opposed to a field position game. Going for it on 4th and short between the -45 and +35 makes more sense in a 'points per possession' game.

tOSU doesn't usually play in very many games like that, and for me, that explains why JT plays it so conservatively. Overall, I agree with JT most of the time in those situations, only occasionally thinking that the punt should be foregone for a 4th down attempt.

 

[Jaxbuck]

The type of game factors into the decision for me.

If teams are both having a tough time moving the ball, field position is more valuable. If the game's a total shootout, giving up the ball is basically just a lost opportunity for points.

I call the latter type of game a 'points per possession' game, as opposed to a field position game. Going for it on 4th and short between the -45 and +35 makes more sense in a 'points per possession' game.

tOSU doesn't usually play in very many games like that, and for me, that explains why JT plays it so conservatively. Overall, I agree with JT most of the time in those situations, only occasionally thinking that the punt should be foregone for a 4th down attempt.

One instance that always comes to mind was The Game 2002. scUM wasn't scoring but they were converting 3rd downs like mad and moving the ball. Tress punted on 4th and inches somewhere near the scUM 40 and they proceeded to go down and kick the FG that made it 9-7. Everyone around where I was seemed to instinctively know that we should have gone for it and tried to keep their offense off the field. The next time down in almost the exact situation Krenzel went nuts and they went for it, got it and it ended up being the game winning TD drive iirc. In fact that was the point where I started really paying attention to and questioning the auto punt on 4th down strategy. It almost cost us The Game.

People can always come up with specific examples of why it works or doesn't work but I think anyone who studies it closely should be able to see an opportunity for improvement.

 

[martinss01]

The type of game factors into the decision for me.

If teams are both having a tough time moving the ball, field position is more valuable. If the game's a total shootout, giving up the ball is basically just a lost opportunity for points.

I call the latter type of game a 'points per possession' game, as opposed to a field position game. Going for it on 4th and short between the -45 and +35 makes more sense in a 'points per possession' game.

tOSU doesn't usually play in very many games like that, and for me, that explains why JT plays it so conservatively. Overall, I agree with JT most of the time in those situations, only occasionally thinking that the punt should be foregone for a 4th down attempt.

exactly! the problem with making your decisions purely based on statistics is that excel doesn't take the situation at hand into account when it spits out #'s.

if tOSU plays miami (oh), toledo, then ysu going 9 for 10 on 4th down conversions of 4 yrds or less. do you go into a game against the gators believing you will be 90% successful on 4th and 4?

to the people who think a metrics is some type of super wisdom, do you honestly tell me you go for it on your first drive from your own 6 against florida's d on 4th and 4? seriously?

jaxbuck, there is a HUGE difference between going for it on 4th and inches anywhere on the field and going for it on 4th and 15. i played lb and i NEVER thought the offense was going against some mystical rule book when they went for it on 4th and inches. in fact, anytime my team put the opposing team into 4th and inches that told me that we failed but were being given the chance to man up and make it right. only time i don't go for it on 4th and inches is when the opposing line is blowing mine off the ball.

if you as a head coach believe the opposing team having the ball at your 40 is just as likely to end in 7 points for the opposing team as giving it to them on your own 1. your defense is horrible. no, more than that. your defensive strategy sucks and you are a terrible coach as you are clearly failing your players.

 

[martinss01]

It is possible that Texas Tech could have converted their 4th down. Then they would have had a two possession lead (rather valuable I'd say). Would you be critical of the coaching then? We only get to see one of the possible outcomes. Even if that outcome is undesirable, it doesn't necessarily mean it was the wrong decision.

of course it means its the wrong decision. go up 8 which forces your opponent to score a td AND a 2 pt conversion just to tie with 6 minutes left in the 4th? my god man, thats what coaching wet dreams are made of! with luck the opposing team has to drive 80 yrds for 6. assuming they can even accomplish that they have 1 shot from the 3 to do it again. THEN, even if they can pull both of those off. i still get the ball back with time on the clock to win the game with nothing more than a field goal. tie game, time on the clock and the ball in my hands.... what more could a coach ask for in any game?

thats the situation every offensive player in the history of the sport has dreamed about.

 

[fourteenandoh]

I read this after the USC game
Deconstructing: The grisly demise of 'Tressel Ball' - Dr. Saturday - NCAAF - Yahoo! Sports

which linked to this (a commentary on the Romer study already noted by Holden, good post btw)
Advanced NFL Stats: Romer 4th Down Study Underestimates Itself


the part that sticks out to me is the expected point curve graph. you should check it out martin because it does not think going for 4th and 15 on your own 1 is a good idea. The expected value does not cross into positive territory until the offense is on their own 29 yd line.

 

[xcrunner]

The problem with these 'on average' curves and whatnot is that there is very rarely an 'on average' curve for teams with your team's exact offensive and defensive capacity.

By that, I mean-- obviously, there are some teams that should go for it on 4th more often than not. Namely, teams with great offenses that are likely to convert the 4th, and bad defenses that are likely to give up the 7 no matter where the opposing team gets the ball. Conversely, teams with bad offenses and great defenses (a trend of the Tressel era) are going to want to punt it more.

For each team, there's some magic number, the magic spot range, etc. Of course, this also all depends on that team's opponent. So these statistics tend to be pretty flawed when trying to guide a gameplan. But they can almost certainly be at least of a little use.

 

[DaddyBigBucks]

I think there is something to be gained by tracking the data and using them as a tool to help guide decisions and strategy.

But as has been mentioned several times already, there is a lot more that goes into it. I won't rehash the whole thread here; but to base every 4th down decision on a graph is to misunderstand what numbers can tell you and what they most certainly cannot.