Suppose you have a very good team on your hands. On average, every starter on your offense will do his job well about 95% of the time.
At first glance, it would appear that you should expect most plays to be successful. They would have to, right? That rate means all the players are doing the right thing almost all of the time.
And yet, that’s not how probability works. The actual chances of every single guy succeeding at once is not much better than a coin flip.
In this scenario, a single player will execute well 95% of the time on one given play. Expand that to two players, and the chances of good execution are 95% raised to the power of two. That comes out to right around 90%. So far, the odds of moving the ball forward are still good.
For eleven players, the chances for proper execution are 95% to the power of eleven. That figure is about 57%. Again: your team’s chances of everyone doing enough to have a good play is not much better than 50-50.
Of course, not every play requires all eleven players to be perfect. The smallest number required to execute well might be three on something like a quick bubble screen. The quarterback will need to make a good throw. The receiver must make the catch and cut to the right area. One other receiver will need to block.
Using the 95% rate again, the chances of all three doing their jobs correctly is about 86%. We’re still in pretty good territory here.
Few plays are that simple. Maybe one or two players are mere decoys, but most plays require most of the guys on the field to succeed in their jobs.
Now, consider the guys up front. If everyone on the offensive line can execute at once, the chances of getting sufficiently good execution from the offense as a unit go dramatically upward. If you don’t have to worry about the five linemen, then you’re calculating probabilities of good execution for at most six players instead of eleven.
Employing the 95% rate one more time, the probability that six players will all execute at once is 73.5%. The odds of a good play have increased by 16.5 percentage points over the eleven-player scenario. If you can cut the important skill position players down to three — perhaps a quarterback, running back, and a blocking tight end — now you can run the ball with that 86% rate of proper execution. Ditching the tight end and running from a four-wide set will bump it up to 90%.
I could only guess at what percentage Florida’s starters will execute well. There also is the matter that sometimes defensive players make terrific plays. Your guy may do his job correctly, but the defender will win out by doing his job even better. This mode of thinking only goes so far, but it is useful in this context.
Fans and media spend a lot of time focusing on whether a quarterback makes a good throw, if a running back makes the right cut, or if a receiver gets open or catches passes. Yet, by rule there are five linemen on the field on every play. Wide receiver is the only position that ever matches that count of five, and only rarely does it do so in most offenses.
And while those five linemen are individuals, they do have to function as a single unit in a way that wideouts never do. In the world of finance, you’d say that the risk of failure with linemen is correlated risk. If one guy doesn’t do his job, it can cause a cascade of failures for his fellow trench warriors.
The right side of the line last year was a frequent example of these issues. Stewart Reese took a while to get used to playing with the guys around him since he only arrived in the summer. Sure, he’d played for Dan Mullen and John Hevesy before, but it had been years, and it’s vital to be in sync with the linemen around you. Also, Jean Delance has limited upside as a player and struggles with some common defensive techniques.
Say that each of them did the right thing about 80% of the time, which is probably in the neighborhood of right. It might even be low; the anger when they’d blow a play triggers availability bias and makes you think it happened more often than they did.
If we go with 80%, then there’s only a 64% chance that both would do the right thing on the same play. That’s not great in and of itself.
Remember, though, that lineman failure is correlated with failure of others. Meaning, if Reese or Delance can’t handle a something like stunt or blitz attacking the right side, it’s going to put the other in a bad spot. The one who was doing his job correctly based on the prescribed protection might then falter because of the other’s mistake.
So, if both do their jobs correctly 80% of the time, the correlated nature of their failures means that the success rate between the two of them will be less than 64%. This figure also doesn’t capture a broader failure cascade, where the center either makes a mistake or puts the right side in a bad spot while trying to make up for a problem on the left side of the line.
In three years, UF has had only two future NFL picks playing on the line in Jawaan Taylor and Stone Forsythe along with one UDFA who’s caught on. In Forsythe’s case, only in 2020 did he really play like a future draft pick. In Taylor’s case in 2018, it took the better part of the season for the line as a whole to get the new offense down pat. Florida didn’t get the full benefit of having a future 2nd round pick and consistent pro starter there. Fred Johnson was on that line too and he started almost half the Bengals’ games last year, but he had legitimate struggles that year. It’s a bit of a surprise he’s done as well as he has in the pro ranks.
It’s no wonder, then, that the line has been such a problem. The math here shows you why it’s such a big issue, and perhaps why Mullen has been so reluctant to try younger players.
Mullen knows the strengths and weaknesses of the older players best. They are what they are, and they’re probably not going to change much. He can try to scheme around their weaknesses and minimize the number of plays where he needs all five guys to get things right. After all, the hypothetical 80% or 95% rates are averages; the actual rates on different kinds of blocking schemes will be different.
I would’ve liked to see Forsythe and Richard Gouraige man the tackle spots last year with Delance on the bench, and we might’ve if Ethan White was healthy. But with him out of the picture, it would be easier to plan for a fourth-year player on the outside with Delance than a true freshman with no spring practice in the middle with Joshua Braun.
A lineup going Gouraige-Braun-Heggie-Reese-Forsythe might’ve been better in the abstract than the starting five we actually saw, but Mullen doesn’t coach in the abstract. Plus if Forsythe was head-and-shoulders better than Gouraige at protecting Kyle Trask’s blind side, then that alone might’ve been worth the tradeoff of having Delance at right tackle (since the coaches seem dead set on keeping Gouraige on the left side).
I trust Mullen enough based on his history that I don’t think he does things for no reason. There must be some reason why Delance started all last year and got the first crack at first team in the spring despite the clearly obvious problems. Stuff like this is the best I can come up with.