Change vs. Regression: Field Goal Percentage

One of the most common ways for a player to be overrated or underrated in their season-long stats is field goal percentage, and one of the biggest mistakes that people make when it comes to that stat is not correctly evaluating the reasoning behind an over or under-performance in a player's shooting. While most of you guys will probably know to look for regression from absurdly high percentages, just leaning on that as a blanket explanation will cause you to miss out on real changes. To truly get better at evaluation, it's important to understand the instances when the most likely answer is not the correct one.

It's easy to fall into the trap of rigid expectations. "In this world, nothing can be expected to be certain, except death, taxes and Khris Davis hitting .247." Unfortunately, not all players will function with the same meme-worthy consistency as Khris Davis, and being able to adjust the goal posts with new information and change expectations based off of real information is a key level-up moment when it comes to player evaluation.

It's really easy to just hand-wave any hot shooting streak and say "he's a .420 shooter, he can't keep this up" or "he has a 50% field goal percentage, that will go up", and a lot of the time, you will be right. In fact, maybe 70-80% of the time you will be right. But the times when you are not right end up hurting really badly, as you cut bait on a guy that ends up being the next breakout or trot out a guy that guts your efficiency and costs you a category week in and week out when he really isn't a great fit for your team, and that's when the importance of knowing what to look for comes to the forefront.

Shot Distribution Can Change

The biggest thing to look for, and most common explanation for most permanent changes in field goal percentage is shot distribution, and this should be something that is relatively easy to spot and figure out as an explanation for a new normal. A big who starts to shoot 3s will see their FG% drop down while a guard who starts attacking the rim more will see a bump. It's easiest to use some examples from the past few years to illustrate this effect.

Modernizing Your Big

The easiest example of a big adding 3s to their game and the subsequent drop in FG% they offered was Marc Gasol, as he went from attempting 3 three-pointers in 2015-2016 to attempting 265 in 2016-2017. Gasol hovered in the high .400s the previous few years, with his FG% dropping to .464 and .459 the next 2 seasons before cratering at .420 last year. Over that span, he still remained a very valuable player but his value was distributed in a different way, and those who valued him incorrectly in that first year shooting 3s might have ended up with a player who didn't fit their team as well as they thought.

It's not just adding a 3-pointer that can change shot distribution though, a change of scheme either from a new coach or moving to a new team can change shot distribution drastically. Serge Ibaka this year is a great example of this, moving from almost a pure stretch 4 to a small ball 5 with the Raptors' coaching change and personnel shakeup. This is leading to a 43% share of shots at the rim, his highest since 2011-2012, which has fueled his 57% field goal percentage. While 57% percent is still a bit high, it won't drop as much as you think if the usage pattern continues.

Accepting Your Fate

I call this the Rajon Rondo effect - when a bad shooter just decides to take fewer shots and live almost exclusively off getting to the rim. This leads to guys who are anecdotally bad shooters becoming surprising positives in FG% despite a lot of people just simply not noticing the trend.

This year, Elfrid Payton is the best example of the Rondo effect, and with him finally being on a team that isn't terrible, he's upped his attempts at the rim up to a career-high 65% while cutting down his share of mid-range jumpers to a career-low 22%. His 50% FG% so far this season, while a small sample, isn't buoyed by unsustainable percentages in shooting, but instead, by an extreme Rondo effect which makes me believe that his 50% field goal percentage is more believable than it looks, once you set the reputation aside.

Signs Of Regression

On the flip side, the way to really look regression candidates is by taking a look at percentages after breaking it down into shot types, and looking for something that is way off when compared to career norms. A great example last year was Aaron Gordon shooting over 40% in 3s for the first few months before completely falling off a cliff with regards to his three-point shooting.

An example of a regression candidate this year would be Kemba Walker, who's shot distribution hasn't changed too much, but is shooting a blistering 65% from long mid-range, compared to career averages in the low 40s. That jump is fueling his surprisingly high field goal percentage, and you can expect a drop as his long mid-range 2s fall closer to career norms.

Putting It Together

I'll finish this with doing a quick analysis on a player and the expectations I'll set for him. I'll choose Nikola Vucevic for this exercise.

It's easy to write off Vucevic's start of the season to an unsustainable 53% from 3, but that doesn't tell the whole story. Vooch is attempting half as many 3s as last season (11% down from 23%) and attempting 37% of his shots at the rim, his highest total since 2014-2015. While regression on that 53% number is coming for sure, it won't have as big of an impact as you'd think due to the lower volume of 3s being taken.

Vooch's mid-range numbers are in line with his career averages, and his 77% mark at the rim, while very high, is only a bit higher than the 73% he put up last season in a fairly large 57 game sample. This is a 66 game sample of ~75% conversion rate at the rim, which is a big enough sample for me to be confident that it won't drop down to the ~60% numbers from before in his career, but I still would anticipate a small drop into something close to the 70% range. Here's a breakdown and quick calculation:

With all this together, while I don't think Vooch will be able to keep up his 56% rate, I don't agree with calling for regression all the way down to the 47-48% range we saw the last 2 seasons, and I expect him to settle in at around the 50-52% range for the season barring a major change in the trends we are seeing so far this year.

Hopefully, this piece shed a new light on the FG% category for you!