Foul Trouble & The "Auto-Bench"
The math and nuance behind benching a player in foul trouble.
Welcome back to the Hoop Vision Weekly!
While we are on a different schedule in the offseason, the extra time allows us more time to crunch the numbers and dive into basketball talking points and conventional wisdom in a way that we simply cannot during the season.
Today, that leads us to a breakdown of coaching decisions based around foul trouble.
We take a look at the decisions coaches must make when players pick up early fouls, and analyze the consequences that unfold from a conventional “auto-bench” move, versus the aggressive move to let players play through foul trouble.
For one reason or another, certain viewpoints tend to become associated with the analytics movement — even in the absence of concrete supporting data.
Perhaps the best example of this is the “foul or defend” debate. For whatever reason, fouling while up three is perceived as the analytically-savvy move — despite the fact that studies are mostly indifferent on the decision.
The “auto-bench” — choosing whether or not to sit a player in foul trouble — is another one of those topics.
According to Ken Pomeroy’s 2-foul participation data this past season, when starters picked up two fouls in the first half, they played only 23% of available minutes during the rest of the half.
In other words, coaches are generally conservative when it comes to allowing players to play through foul trouble.
Not only are there a large group of people on social media critical of how conservative coaches are in benching players, but some coaches have even begun to speak out against the conventional wisdom. Most notably: San Francisco head coach Todd Golden.
In a Q&A with Brian Bennett of The Athletic, Golden discussed his strategy:
A lot of coaches don’t play guys when they get two fouls in the first half, but we ignored that and continued to play guys. Not all of it was perfect, but overall I think it helped our program.
The main logic behind the “let them play” strategy is relatively straightforward.
The auto-bench is essentially fouling your own players out
Consider a player, for example, who picks up their second foul five minutes into the game, right around the 15 minute mark of the first half.
By using the “auto-bench” strategy for the rest of the half, the most possible minutes that player can now play in the game is 25.
By allowing the player to stay on the floor, there is now a much wider range of possibilities, but the expected value of minutes is likely to be higher than the auto-bench.
Leading us to point number two…
The current benching strategy for (many) coaches is arbitrary
As it currently stands for some coaches, 19 minutes and 59 seconds into a game is being treated completely differently than 20 minutes — simply because that’s when the first half ends.
Although only one second has elapsed between the two moments, the halftime whistle seemingly acts as a magical forcefield against perceived foul trouble.
In fact, I’d predict that if an NCAA rule change lengthened each half to 21 minutes, many coaches’ definition of foul trouble would magically change accordingly.
Using the halftime intermission as an arbitrary endpoint is a simple, low-stress way to handle foul trouble, but it’s likely leading to suboptimal decision making.
To be fair to conservative coaches though, the decision to keep a player in the game is complex.
The biggest reason why?
Player behavior changes when playing through foul trouble
The basketball world has pretty concrete evidence that a player’s defensive performance changes when in foul trouble.
In Ken Pomeroy’s study from 2016, he found that:
Starters with zero fouls in the first half commit 2.66 fouls per 40 minutes
Starters with one foul commit 2.34 fouls per 40 minutes
Starters with two fouls commit 1.71 fouls per 40 minutes.
In order to avoid further foul trouble, players become more conservative on the defensive end.
Sometimes, that new behavior might even be at the direction of the head coach. A player in foul trouble, for example, might be instructed to “let them go” when defending a fast break or a shot at the rim.
It’s important to note that the opposing team is also acutely aware of this information. When a player in the game is in foul trouble, the opposing team may be instructed to attack that player.
So even if playing through foul trouble leads to more minutes, there’s an argument to be made for quality over quantity. Getting a player from 25 minutes to 30 minutes might not be worth it if the player was forced to play timid defense throughout those extra five minutes.
On top of that, there’s another argument to be made for trusting your bench…
An individual player’s value isn’t as high as we often think
Generally speaking, individual players aren’t worth as many points over their replacements as conventional wisdom might expect. (This was one of the topics in an old episode of Solving Basketball with Bart Torvik).
The easiest way to see that is by looking at betting markets. When a player is ruled out from a game, the Vegas lines don’t move nearly as much as one might think.
With that in mind, you can build out an argument in favor of sitting players in foul trouble. The logic would be as follows…
A starter is only worth a couple points per game more than their replacement
A starter in foul trouble is worth even less, because they are forced to play extra carefully on defense
With those two factors in mind: you’re better off playing the bench player
But it’s not as simple as “bench or play”
For a coach, the decision to bench or play isn’t a completely binary decision — particularly if you choose to let a player in foul trouble stay on the floor.
For example, a coach might choose to bring back a star player, but make adjustments to the individual defensive matchups. Hiding the player in foul trouble on the weakest offensive player on the court.
Or a coach may choose to make a schematic change, deciding not to hard hedge on ball screens — which risks a foul if the big gets too aggressive on the hedge — in favor of a more passive ball screen coverage.
In these cases, the coach would be aiming to strategically reduce the probability of the star player picking up another foul, all while sacrificing defensive performance as little as possible.
Perhaps the most extreme way to avoid fouling is by switching defenses entirely from man-to-man to zone.
Which leads us to another point…
Zone coaches are more likely to allow players to stay on the court in foul trouble
Per Synergy Sports, eight teams played zone on over 80% of their defensive plays last season.
Those eight teams finished with an average 2-Foul Participation of 51% — more than double the NCAA average of 23%.
The number in parenthesis is each team’s DI ranking in 2-Foul Participation. As you can see, all eight zone teams finished in the top 50 of that figure.
Meanwhile, strict man-to-man coaches who rely on tough on-ball defense — like Tony Bennett, Porter Moser, and Tom Izzo — are consistently at the bottom of 2-Foul Participation.
Personal opinion: I believe this can be seen as evidence that coaches are actually acting more rationally than we may think.
Coaches who rely on tough and disciplined on-ball defense are choosing to bring in bench players instead of sacrifice defensive performance. Coaches who rely on zone-based schemes are choosing to stick with their best players.
In the end, I think it’s hard — if not impossible — to create blanket rules for handling foul trouble.
The three main questions I would ask are:
How foul prone is the player in foul trouble?
How big of a drop off is the replacement player?
What type of defensive role will the player be taking on?
It can be frustrating when a coach ignores question number one — choosing to auto-bench a player that rarely commits fouls. But on the other hand, I suspect the critics often ignore the challenges associated with question number three.
As always…basketball is a game of trade-offs.
Links from around the internet
I was a guest on the And 1 podcast. We discussed analytics and the similarities and differences between different levels of basketball
Tony Bennett made a (relatively) rare podcast appearance on Hardwood Hustle
CJ Moore did a feature on Mark Adams for The Athletic, including more background on the no-middle (aka “side”) defense
Among current DI head coaches who played basketball collegiately, 27.8% transferred at least once during their playing career
If you’d like to receive access to exclusive content and support Hoop Vision, please consider subscribing to Hoop Vision PLUS. For $10/month or $100/year, your subscription unlocks access to exclusive offseason content and helps Hoop Vision stay independent and ad-free.