Geeks rule! Why Browns Kevin Stefanski is going for it on fourth down

Has Browns coach Kevin Stefanski lost his mind by going for it so frequently on fourth down, or is he just listening to the analytics experts in the front office?

Cleveland Browns coach Kevin Stefanski appears to be relying on the advice of analytics geeks who are telling him it’s smart to go for it on fourth down. Here we provide the best geekological analysis possible for how he arrived at the conclusion that it makes sense to go for it on fourth-and-medium situations that old-school coaches used to avoid like the plague.

This is part of a recent trend that NFL teams are going for it on fourth down more than ever before. The implication is that there are computer geeks that have figured out that this is a good idea. The Browns are supposed to be industry pacesetters, so the presumption is that the stat supports what the Browns are doing.

This author is a bit of a math geek himself, and while he doesn’t have a massive computer program of his own to fall back on, he can discuss some of the issues that the geeks have disagreed with over the years.

The best way to illustrate this is with a concrete example that violates old-school principles. Why, for example, did the Browns elect to go for it on fourth and three from the Arizona 13 in Week 6?

A complete answer would require a computer analysis tailored to the specific personnel of both teams, encompassing hundreds if not thousands of factors. But we can outline how the calculation is made with some representative data, understanding that the Browns coaching staff and front office have access to better and much more thorough stats.

So this is not going to be a mathematical proof positive, but it is going to be a sanity check on the coach.

The field goal expectation is not that difficult. We can compare that to data for extra points. We really want to know the probability of the Scottish Nailer, Chase McLaughlin, making a 13-yard field goal.

That might be hard to estimate based on his limited career stats, but possibly it is a little higher than the league average for the probability of making an extra point from the 15-yard line. That is known to be historically around 94 percent. Perhaps two yards closer distance bumps up the accuracy to 95 percent.

So the expectation for points is going to be 3*(0.95) = 2.85 points. But afterward, there will be a kickoff, and the ball will be put back in play on the 25 or returned, usually for a slight gain. If the field goal is missed (5% of the time), the ball is placed seven yards back, on the 20 in this case. On average, let’s estimate that Arizona would get the ball back on the 26-yard line. Thus, the Browns are essentially penalized 13 yards if they elect to go for the field goal.

If they elect to go for the first down and ultimately try to score the touchdown, we need data on fourth and three. Brian Ritchie and Cory Slusser of Kent State (advised by Jun Li) compiled data in 2016 on success probability as a function of down and distance. According to their data, the fourth and three play would have had a 45 percent chance of success from the 13-yard line.

This calculation is based on NFL-wide information, so if your team (e.g., the Cleveland Browns) is believed to be better than average, or if Arizona’s defense is believed to be weaker than average, we might bump up the estimate several points higher. The coach would probably rather trust the likes of Hunt, Beckham, and Mayfield over a young placekicker, but that is a subjective judgment.

Let’s guess that Stefanski and his geek cadre believed that they were going to succeed 57 percent or in other words 12 percent higher than the league average. A Browns first down would have resulted in a goal-to-go from inside the 10-yard line.

Football Outsiders compiled some data back in 2013 that claimed that a first down in the red zone, and specifically from inside the 10 resulted in a touchdown about 60 percent of the time. The data is old and should be replaced by a more current analysis, but guessing here is that the numbers have not changed that much in eight years. The game has not been revolutionized that much. After all, Tom Brady still goes to the Super Bowl every year, so how different can it be?

Note that if the Browns make the first down in the red zone, and then are stopped dead in their tracks for their next set of downs, facing a fourth and long, they still have an opportunity to kick the field goal. They have some control of this situation. As a guess, this could happen about half the time. The probability tree looks like follows:

I. GO FOR IT:

• 4th & 3→ 57% chance CLE makes 1st down

  • → 60% chance of TD @ 6.94 points = 2.37 points, AZ ball @ 26 yd line

  • → 20% chance of FG @ 2.95 points = 0.34 points, AZ ball @ 26 yd line

  •  → 20% chance of 0 points, turnover on downs, AZ ball @ ~3yd line

• 4th & 3 → 43% chance CLE fails to make the 1st down → AZ ball @ ~13yd line

•  4th & 3 → ~ 0% chance (neglected) of CLE pick 6, other disaster.

Total expectation 2.71 points, Arizona’s average field position 16.7-yard line (83.3 yards to go for a TD).

II. ATTEMPT A FIELD GOAL:

• 4th & 3 → 95% chance CLE makes FG → 2.95 points, AZ ball @ 26-yd line.

•  4th & 3 –> 5% chance CLE misses the FG–> 0 points, AZ ball @ 20-yd line.

•  4th &3 CLE suffers blocked kick for TD, other disasters (neglected).

Net expectation 2.95 points, Arizona’s average field position 25.7-yard line (74.3 yards to go for a TD).

It looks like the field goal is a better idea, coach. But the coach folds his arms, shakes his head and asks stubbornly, “Did you calculate the value of nine yards of field position?”

Uh-oh, we better try to calculate that.

A possible answer is provided by a blog called PhD Football that has compiled NFL data on scoring versus field position.

According to the graphs (and drawing a line through his data points by hand, which may not be terribly accurate) the probability of a drive culminating in a touchdown is about 17.7 percent from the 25.7-yard line versus 15.0 percent from the 16.7-yard line.

The probability of a drive ending in a field goal is about 12.1 percent from the 25.7-yard line and 10.7 percent from the 16.7-yard line. The total points expectation is better by (0.177-0.150)*6.95+(0.121-0.107)*2.95 = 0.23 points. That is, if the Browns go for the TD, it is probably going to help out the defense because of better field position worth an average of 0.23 points.

There’s yet another factor that needs to be accounted for. If Arizona is forced to punt, the Browns have a chance to inherit some of that field position advantage on offense. Their chance of scoring on offense will be higher than average.

If Arizona’s total chance of scoring a touchdown or a field goal from the 16.7-yard line is (15.0% + 10.7% = 25.7%), that means there is a 74.3 percent chance that the Browns’ defense will hold them, and some or perhaps all of that field position is thus inherited by the offense. That will be worth additional fractions of a point.

This tends to tip the scales in favor of going for it on fourth down on the original decision point above.

Still, there is enough uncertainty in the calculation that it could go either way. This fan is not claiming actual *proof* that Stefanski is making the right call by going for it on fourth down more than previous generations of conservative coaches. However, by looking at relevant calculations that others have done, it’s at least very plausible that the coach knows what he’s doing.

The geeks, then, probably know what they are doing in recommending that the Browns go for it on fourth down. If the coach believes in his red-zone offense versus the opponent’s defense (and perhaps does not believe in his kicker completely) then go for it on fourth down and medium yardage. It also matters not only how good he is at field goals, but how reliable he is at putting the ball in the end zone on kickoffs.

Conversely, a coach like Mike Tomlin, with a superior kicker in Chris Boswell, but the lowest-scoring offense in the division, might not make the same fourth-down decisions as the Browns, based on the way his team is now constituted.

These days, some Cleveland fans want Kevin Stefanski to give up play-calling duties and let offensive coordinator Alex Van Pelt have more responsibility. This fan disagrees. The offense is not broken. Stefanski — and Van Pelt — have done a fine job, and the fourth-down play calls simply show that the Browns coaches think further into the future than other coaches. Field position matters.

This fan will have to see something far crazier than going for it on fourth down and three in the red zone before he gets upset at the coach’s play-calling.

Yeah, it ain’t the way the old school used to do it. But the old school is probably wrong.