Coaching & Analytics: Newsletter #2
From the archives: V1 of the old Hoop Vision coaching newsletter
NOTE: This post was originally sent as an email to subscribers in June 2017; the Hoop Vision newsletter and operation was far different in that period…enjoy:
The Mid-Range Jumper
To make a slight overgeneralization, the leading concept from the basketball analytics revolution has been to avoid mid-range jumpers.
Spoiler Alert: it’s not a completely new concept.
The coaching axiom that foot-on-the-line jumpers are the worst shot in basketball has been around years before statisticians got involved. However, getting math involved can lead to a much more extreme stance on mid-range shot attempts.
The math is quite simple. Due to the extra point, shooting 33% on three-pointers will net you the same amount of points as shooting 50% on two-pointers. I don't want to under-emphasize the significance of that extra point. According to Synergy, the NCAA average on mid-range jumpers was about 36%. The NCAA average on three-point jumpers was about 35%.
To show the significance, consider a fake game between one team that only shot threes (at a 35% rate) and one team that only shot twos (at a 36% rate). The three-point shooting team would win by 26 points, assuming average pace and equal rebounding and turnovers.
So the math is pretty damning, but the point of this newsletter is not to continue to repeat the common arguments of “analytics guys” ad nauseam. Instead, it is to first recognize that there is an incredible amount of nuance in the debate - and in basketball offensive theory in general. Then it is to provide some actionable (and hopefully reasonable) advice to consider as a coach.
1. Play context matters
A good chunk of mid-range shots are taken by necessity. When the shot clock is running down, shot selection does (and should) change. Keeping the same standard for shot selection with five seconds on the shot clock as with 25 seconds on the shot clock would result in an increase in turnovers (and decrease in efficiency). This is also a big reason why mid-range shooting percentage is nearly equal to three-point percentage despite the shorter distance.
2. Not all players are equal
The list of players that can overcome the extra point and remain efficient in the mid-range isn't necessarily long, but it does exist. More generally speaking: Personnel matters. Blanket statements about mid-range shots are irresponsible to make when such a wide variety of player skill sets exist.
3. Defenses are smart too
Most of the best defenses are designed to take away the three-point line and protect the rim. This isn't football where a coach chooses "run" or "pass" before the play begins. Diversity is an important quality of an offense. Just like spreading the court with shooters opens up room to drive, pulling up for an occasional long two can keep the defense honest.
'Live and Die' is Dead
The Warriors have almost singlehandedly killed the "Live and Die by the Three" cliche. This point might be the least controversial in the entire newsletter. Coaches are now more than ever thinking about shot selection in terms of expected value - process over results. Here are quotes from two of the best:
From Gregg Popovich:
"We had good shots. We missed our threes and they go in or they won't. You know we can't really - we like them to go in but we don't really care."
And here’s Brad Stevens…
"We don't measure our performance on if shots go in. We measure it on -- I don't mean to be too analytical -- expected shot value."
In the NBA, it's gotten to the point where having bad shot selection is going to hurt a team far more than having good shot selection is going to help. That's because most teams have generally figured it out, by necessity. Shooting is more important in the NBA than any other league ("make or miss league"). The nature of the NCAA allows for a dominant offensive rebounding team like UNC or a dominant pressing team like West Virginia, but that's not the NBA model.
Consider a scenario where a coach takes over an AAU team for a weekend tournament. You have one practice before the tournament to get ready. I'd argue the most practical and valuable thing a coach could implement for immediate improvement would be shot selection. Even marginal improvements in expected value are going to improve offensive efficiency significantly.
So how do you improve shot selection? Usually the debate about mid-range jumpers is so combative that the conversation ends there. The rest of this newsletter goes deeper with ideas for actually coaching/teaching the skill.
Offensive Scheme Matters
Before you ever have a discussion with your players about shot selection, your offensive scheme is going to play a large role in shot selection (and ultimately in efficiency). Don't believe me that plays/sets determine shot selection? Let's take a look at maybe the most extreme offense still being run in D1 at varying degrees: The Flex.
Al Skinner and Gary Williams are the godfathers of the flex offense. Both became especially famous for the offense at high major jobs (BC and Maryland). The flex is continuity based, with the actions being repeated over and over. Skinner's version of the flex is the most common version used at different levels of basketball. The basic action is a screen and flex cut across the baseline followed by a down screen for the initial screener. If the ball is reversed, the same action happens on the other side. Gary Williams' version doesn't utilize the down screen. The initial screener instead ducks in for a post-up on the weak side block.
Regardless of the version, the flex offense is a direct contrast to the "pace and space" style of offense that has permeated high level basketball. Gary Williams retired in 2011, but Al Skinner just finished his second season at Kennesaw State. Skinner is still running a lot of flex of 2017. The compact spacing (often times all five players are a full step inside the three-point line) makes it almost look like Kennesaw State is playing a different sport entirely.
Skinner and Williams both have former assistants that are currently head coaches in the NCAA: Jimmy Patsos (Siena) and Ed Cooley (Providence). Neither runs the same offense as their former bosses, but at the very least incorporate some of it in conjunction with more modern play design.
Let’s take a look at all four of these coaches and their respective influences on shot selection in the simplest way possible: three-point attempts. For each coach, I identified the transition year that he took over or left a program. In other words, I created a ‘before-and-after’ snapshot by looking at how the team ranked nationally in three-point attempts (shot selection) with and without the flex offense coach.
Educating Players
The idea of shot selection as a product of team culture is the most polarizing aspect. Obviously the shot selection is ultimately decided by the player shooting the ball. A coach can put that player in the optimal position, but the rest is in the player’s control.
So conversations need to be had with players on shot selection, but how? A common argument against analytics-supported basketball strategy is: “Sure the two-for-one (for example) is mathematically correct, but it’s bad for culture.” When it comes to shot selection, the specific counter-argument is that it makes players play like robots.
In reality, there is so much more nuance than just simply declaring it “bad for culture.”
Consider the strategy of going two-for-one.
A team takes a quick (and potentially bad) shot instead of working for a quality shot to ensure that they get the ball back one more time to end the half/quarter. Coaches (most notably Jeff Van Gundy) have said that the marginal benefits gained by going two-for-one aren’t worth the selfish culture that it promotes. I think this all has to do with how you are framing the concept to your players. A “whatever it takes” mantra is often implemented by coaches to promote hustle and the “little things”. The two-for-one is no different, or at least it’s no different if you frame it that way. Taking a quick shot doesn’t need to be selfish, it actually becomes *unselfish* if thought of as sacrifice for the greater good of the team.
Getting back to shot selection and the mid-range jumper, it’s not much different to the example above. Tell a player/team to “never shoot a mid-range jumper” and it’s going to be viewed as oppressive. Instead, educate a player/team on the *why* behind shot selection and institute a culture where relentlessly pursuing the optimal decision (or in this specific case shot) is always the goal.