Analytics have changed the way we look at basketball.
They play a huge role in the way coaches look at the game, though it’s influence can have various levels of affect. Some coaches use statistical information for huge choices like what players they should recruit or not and whether they should play primarily man or zone defense others use it with a bit more subtlety and nuance.
If you read my work at Gator Country you’ll know I have been critical of a few Florida coaching decisions that I didn’t feel were grounded in a sound statistical basis. Many of these were rather large elements of the game, namely rotations, shot selection, and timeout usage.
While I get great satisfaction out of discussing these key elements of the game of basketball and what information we can glean about it from the numbers, today is a bit different. We’ll be talking about something on the extreme subtle side, and show why tiny decisions can have a huge impact on whether or not a shot is going to go through the hoop. This will give you something to look for when college basketball comes back, whenever that day is, and will be an example of how numbers can help inform small decisions that can have huge returns. Let’s get into it.
Mike White is fascinated with the three-point shot. As a coach in modern basketball, you have to be. He has been open about how he wants his team to get up a lot of threes offensively, while eliminating opponent three-point attempts on the other side of the floor.
This is backed up with results. The Gators have been in the top half of the country in three-point attempts in four of five seasons (this past year being the first time they haven’t made the top half) and defensively they have accomplished their goal finishing in the top half of limiting opponent threes.
Over the five seasons of the Mike White era they have averaged 133rd in the country in three-point attempts offensively and 84th in attempts defensively. While the team has been relatively successful in both of their three-point goals their defense is a bit ahead of their offense right now and moving forward they’ll be looking to get that offensive three-point number ratcheted up a bit.
One of the common ways to get shooters open for a three is to set screens for them. You see this more at the NBA than the college level but as NCAA teams continue to follow the pro’s lead it’s likely you’ll start to see more off-ball screen actions for shooters.
The average NCAA team took 4.9% of their shots off screens this year, usually as an auxiliary option to their offense.
The Gators only took 2.9% of their shots off screens this year.
Was that because they struggled at shooting off of screens? Here is our first opportunity to use some numbers. No, they did not struggle at shooting off screens. In fact, it was quite the opposite. The Gators were actually 14th in the country at offensive efficiency off screens, going for a gaudy 1.1 points per possession on those plays. That made it more efficient for them than their pick and roll offense, their post-up derived offense, and their transition offense.
This would suggest to me that they weren’t using screens to free up shooters enough. It’s likely that their efficiency would dip a bit as they started to take more attempts, but if you’re coming from a place of 1.1 points per possession and 14th in the country you can take a dip and still have it be one of your most effective offensive options.
Another reason why utilizing screens more often could be wise is that the Gators have an absolute flamethrower in Noah Locke, someone who is going to enter his junior season as arguably the best shooter in college basketball. Considering he came off a season where in conference play he shot 48% from deep on 6 attempts per game, there aren’t going to be too many arguments for other players to have that crown.
Locke shot 44% off screens last year, good for 1.2 points per possession making it an incredible shot for him and the Gators. Whenever any player executes any particular play type at that type of clip, it’s safe to say they should be utilized in that role more often. Keyontae Johnson and Scottie Lewis both touched the 1 point per possession mark as well, making it a great shot for them as well.
Locke was incredible coming off screens, this we know. But, what if he could be made even better, and what if there was an easy way to do it? Here is where analytics can inform decision making.
When it comes to shooting off screens the player is running and either turning to their left or to their right to receive the pass before pulling up. This is, for obvious reasons, dictated by which side of the court a particular play is set up on. The shooter, Noah Locke in most cases, is positioned somewhere on the floor when a screen or two comes, laying out a path for him to sprint through before getting open and receiving the ball.
Rhetorical pop quiz–do you think Locke shot a relatively similar percentage after going to his left or to his right off screens?
There is something you should know about nearly every basketball player. They don’t shoot an even percent all around the court, even if it’s identical distances from the rim. Nearly every player has a side of the floor they shoot better from (side note–it’s usually the left side of the floor for right handed players) and often the difference between a player’s preferred side and the one that isn’t is rather large, upwards of 5%.
Knowing that, I think you can guess where I’m headed with this.
Locke was not equally as good going off screens to either side.
In fact, the difference is quite large. Coming off screens to his right, Locke shot an otherworldly 58%, or 1.75 points per possession. That number is comically large, just an absolutely ridiculous mark that could only come from one of the best shooters in the country.
Now…coming off the left, that’s a different story. He shot 30% (0.79 points per possession) on these plays.
That is a monumental difference. 28% and nearly 1 point per possession of difference between these attempts is hard to fathom, but it showed how different it is shooting off screens to the right vs screens to the left.
Think about the mechanics of Locke going to his right versus going to his left. Since he’s a right-handed shooter, think about the more natural way his feet are aligned when catching and shooting from his right than when he has to go to his left. This makes it two very different shots and as you can see, there are two very different results.
Since he shot so much better going to his right than to his left, he must have gotten more opportunities to his right then…..right?
Unfortunately, no.
The breakdown of Locke’s attempts off of screens: 54% to his left, and only 46% to his much more dominant right side.
This is where analytics can help coaches make decisions that help put their players in the best positions to score. When running a play for Locke to come off a screen, the decision of whether he’s going to the right or to his left is a 28% decision, one that changes the expected shot value by nearly 1 point (I still can’t get over how ridiculous that number is).
As I mentioned earlier, Locke was at 1.2 points per possession coming off screens on the season, and that was with a majority of his attempts going to his weaker side. Let’s say instead of that breakdown, 60% of his attempts off a screen came to his dominant right side. Using his stats from last year, that would suggest he would be up to 1.4 points per possession off screens.
That is a major boost in efficiency, and it would be done with extreme ease. It’s not drastically changing the Gators’ style of play, it’s not changing what personnel you have on the floor, and it’s not changing the pace you play it. It’s simply running the same plays you already run for Locke, but using the numbers to realize what side he is largely more effective from and running it to that side.
What about the rest of the Gators, did they have a preferred way to shoot off screens as a whole?
Yes.
As a team the Gators shot 32% going to their left and 55% off screens to their right. Another gigantic disparity, and one that should show the coaching staff moving forward which way they should design plays that have screens set for shooters.
Basketball is all about finding incremental advantages throughout a 40 minute game. You’ve got about 68 offensive and 68 defensive possessions (depending on the game, obviously) to try to best your opponent in any way possible and at the end of it the team who had the most small wins in efficiency are going to get the win.
Would knowing how much better Locke (and the rest of the Gators) shoot off one side of screens from the other make the Gators win a bunch more games? No, but it very realistically could help them get on the right side of one or two of them. The Gators played in a lot of close games last year and were winless in one-possession games. Now, look again at the efficiency difference between Locke shooting off screens to his right and to his left again. Would more of the drastically more efficient shots in those games changed the outcomes? There is definitely a chance.
Gaining incremental advantages is what differentiates great teams from good teams and if the Gators want to take the leap everyone wants to see from them next season using analytics to gain those advantages could really help.