Going for it on fourth down

[SeattleUte]

The fake punt, by the way, came on 4th and six on Utah's 39-- 61 yards out. Graph says to punt.

Yeah, our resident scientist's brother can't read a simple graph.

[CardiacCoug]

Yeah, our resident scientist's brother can't read a simple graph.

The psychology behind the decision whether or not to go for it on fourth down is really pretty fascinating.

All available data pretty clearly demonstrate that coaches should be much MORE aggressive in going for it on fourth down than they are currently. But tradition and convention has been to adopt a very conservative fourth down strategy. So fans like SU have come to expect a conservative strategy and are under the mistaken impression that a conservative fourth down strategy is "good coaching" when it is actually bad coaching and more likely to lead to the team losing.

Teams should almost NEVER punt on 4th and 1 on the 50 yard line, for example. It's a demonstrably horrible coaching decision, no matter what anybody intuitively thinks should be done based on years of coaching, playing, or watching football.

But it's a catch-22 for coaches. Do they make aggressive fourth down decisions that actually are more likely to help their team win but leave them more open to criticism from fans (after all, it won't always work out)? Or do they stay conservative, which makes it more likely their team will lose but may be perceived as better coaching by dumb fans?

[CardiacCoug]

Just thinking about this some more I think going for it on 4th and 1 from the 50 yard line is kind of like this risk aversion example:

Someone gives you a choice between taking a coin flip that will either give you $100 or $0 OR you can just take a guaranteed $40.

Going for it on 4th and 1 from the 50 is choosing the coin flip (high risk, high reward). Punting it is taking the guaranteed $40.

Sure if you go for it a lot on 4th down you will end up with $0 a bunch of times, but over the course of the season it will pay off to take the coin flip by aggressively go for it on 4th down (per the chart).

Obviously there are specific variables for each team involving the FG kicker, punter, game situation, and offensive/defensive strengths and weaknesses. Bottom line though is that coaches have traditionally been way too conservative because of human psychology related to risk aversion.

[pelagius]

Just thinking about this some more I think going for it on 4th and 1 from the 50 yard line is kind of like this risk aversion example:

Someone gives you a choice between taking a coin flip that will either give you $100 or $0 OR you can just take a guaranteed $40.

Going for it on 4th and 1 from the 50 is choosing the coin flip (high risk, high reward). Punting it is taking the guaranteed $40.

Sure if you go for it a lot on 4th down you will end up with $0 a bunch of times, but over the course of the season it will pay off to take the coin flip by aggressively go for it on 4th down (per the chart).

Obviously there are specific variables for each team involving the FG kicker, punter, game situation, and offensive/defensive strengths and weaknesses. Bottom line though is that coaches have traditionally been way too conservative because of human psychology related to risk aversion.

Cardiac,

I absolutely agree that coaches are affected by something like risk aversion in these cases; probably best thought of as risk aversion related to losing their job (or really maximizing the present value of wealth or other things that affect coaches happiness: power, presitge). Deviations, as you point out, from the basic heuristics of conventional wisdom can be an issue here. Totally agree.

The reason why risk aversion shouldn't matter in terms of going for on fourth down (at least in the sense we use it in economics and purely in terms of on the field decision making) is because all decisions in football collapse to a single parameter is terms of optimality: you pick the option the maximizes the probability of winning (if margin of victory matters like it can in college football then this isn't true and things get more complicated and you may end up with risk loving behavior being optimal for some decisions). That probability defines both the mean and the variance (unlike with something like maximizing wealth) because each decision is a draw from bernoulli distribution. Thus, unlike in finance, where two parameters matter only one matters here.

Now solving for what is optimal can be quite difficult. It is basically a long recursive dynamic chain, but how it collapses into a single parameter distribution is fairly straight forward, I think.

[Mormon Red Death]

Chuck Klosterman on going for it on 4th down

2.I like when teams go for it on fourth down, and I support any reasoning for doing so. I believe it's (probably) a winning strategy. But there's something about the overwhelming mathematical logic of going for it on fourth-and-4 at midfield that doesn't seem complete to me. I'm assuming someone in the comments will explain to me why I'm wrong about this (and I won't disagree if you do), but — as of yet — no one has adequately convinced me of why my counterargument is flawed. Here's my confusion: The reasoning behind going for it on fourth down is built on the base rate of success, which can be calculated here. I do not doubt these calculations. However, isn't part of the reason the numbers suggest going for it on fourth down at least partially because almost no one regularly does so? Statistics aren't predictive; they can only show us what happened in the past. So if going for it on fourth-and-4 at midfield is still a relative rarity, isn't the available data for its rate of success questionable? And isn't it buoyed by the specific situations in which it occurs? I mean, what kind of team tends to go for it on fourth-and-4 from midfield? It generally seems like it's teams who are desperate (and sometimes facing a prevent defense) or teams who feel confident that they have the personnel and the play-calling acumen to succeed (most notably the Patriots). But let's say every team started doing this, all the time (which appears to be what the stat-heads want). Won't the base rate drastically change in potentially unexpected ways? Or let's say only three NFL teams suddenly decide they're going to apply this theory wholesale, but those three teams are the Browns, the Chiefs, and the Jaguars. Wouldn't this wreck the metric? Or would it somehow prove it? Why do I find myself suspecting that — if absolutely everyone started going for it on fourth down — advanced statisticians would respond by telling teams they should consider punting?

[Donuthole]

Chuck Klosterman on going for it on 4th down

I'm guessing there's a large enough sample size of going for it on 4th down that the stats wouldn't change too dramatically. Why would the fact that a team is desperate change anything statistically? Why would a team's belief in its personnel change anything statistically?

I love Klosterman, but his reasoning here assumes a lot of immeasurable stuff, and that years and years of measurable data is inaccurate based on those immeasurables.

[Indy Coug]

If football had 162 game seasons and averaged 20+ scoring plays a game, then going for it on 4th down would almost certainly be a winning strategy. However, since seasons are really 10 to 20 games long and have 6 to 12 scoring plays a game, the variability with that 4th down strategy is high enough that it can seriously bite you in the butt.

[Scorcho]

Seven PAC-12 Schools had a 50% or better success rate on 4th down conversions this year.

Oregon was an incredible 19 of 27 (71%).

[Indy Coug]

BYU was 20 for 28 (71%), which was the 8th best conversion rate this season.

[HuskyFreeNorthwest]

Seven PAC-12 Schools had a 50% or better success rate on 4th down conversions this year.

Oregon was an incredible 19 of 27 (71%).

The failed attempt against Stanford from inside the 10 yard line cost them a trip to Miami. No biggie.

[mUUser]

I've spoken to a bunch of people about this successful HS coach in Little Rock who never punts and always kicks onsides. I understand he has a ton of onside kick formations he uses. My goal is to get to LR and watch a game next season. His philosophy is based on studies by a Harvard professor.


http://www.straitpinkie.com/sports/c...-kevin-kelley/

[Scorcho]

BYU was 20 for 28 (71%), which was the 8th best conversion rate this season.

Lark would have gone 30 for 28

Just sayin'

[The_Douger]

You've got that right.

[RC Vikings]

Football just needs to eliminate the field goal and point after. It would be a much better game.

[SeattleUte]

The failed attempt against Stanford from inside the 10 yard line cost them a trip to Miami. No biggie.

Wasu beat the Huskies because Leach was smart enough to kick a field goal for the tie near end of regulation rather than go for it on fourth and short. Whatever the lying stats say, that seems to happen every time,doesn't it.