Stats, analysis, commentary, and a performance ekg
The story of Deni Avdija’s season goes something like this: he improved throughout the season, performed well with greater offensive responsibility, and emerged as at least a significant part of the Wizards’ future. Some of that is even true.
Let’s start here — Avdija definitely improved this season, as he has in each NBA season. This year was the first that rated above average in my Player Production Average (PPA). (PPA is my overall production metric that rewards players for things that help teams win and dings them for things that hurt the cause of winning — each in proper proportion. PPA is pace neutral and has an accounting for defense and role. In PPA, 100 is average, higher is better, and 45 is replacement level.)
Here’s Avdija’s PPA progression by season:
- 2020-21 — 48
- 2021-22 — 69
- 2022-23 — 76
- 2023-24 — 111
This was the best shooting season of his career by miles. He had a 56.0% effective field goal percentage, shot 37.4% from three, and 73.7% on at-rim attempts.
The biggest change was his willingness to attack the paint off the dribble. A career high 33.6% of his field goal attempts came at-rim (within three feet of the basket, according to Basketball-Reference) where he shot that career best 73.7%.
As an aside, the idea that Avdija was a poor finisher was really true only in his second season when he converted on just 63.3% of his at-rim attempts. In each of his other three seasons, his at-rim conversion rate has been above average and above 70%.
The increased drives show in the numbers. First, while he had his best three-point shooting season, his attempts dropped to a career low 4.9 per 100 team possessions. His previous low was his third season at 5.6 attempts.
He exchanged those three-point attempts for drives. This season, he drove 13.9 times per 48 minutes, according to NBA tracking data. Here’s a quick look at his per 48-minute drives season-by-season:
- 2020-21 — 3.2
- 2021-22 — 7.7
- 2022-23 — 8.7
- 2023-24 — 13.9
Here’s where things start getting interesting. His first two seasons, it was probably smart to limit his attempts to drive. He struggled to score on drives (sub 50% shooting each season), didn’t get to the free throw line with much regularity, wasn’t an effective playmaker for teammates, and had turnover problems. In each of his first two seasons, his assist-to-turnover ratio was below one — in other words, more turnovers than assists.
That all actually changed in 2022-23, Avdija’s third season. He did a better job converting drives into points, his assists went up, and his turnovers fell. The ast/tov ratio was still pretty meh (1.35), but the improvement was there.
This season was largely a repeat in driving effectiveness (slight dips in some categories) but with dramatically increased volume. For example, he shot 54.5% on field goal attempts that came on drives this season — down from 56.0% last season. He produced points on 57.7% of drives this season vs. 62.8% last year. His points per drive fell from 0.63 in 2022-23 to 0.58 this season. His free throw rate also fell. His ast/tov rate was nearly identical (1.38 this year to 1.35 last). His rates of assists and turnovers were nearly identical.
One thing he did a lot better: knocking down free throws. This season, he shot 77.3% on free throws that resulted from drives vs. 68.8% last year.
As makes sense given that 2023-24 was the best season of his career, Avdija posted career highs in several key categories, including:
- Points — 23.0
- Assists — 6.0
- Usage — 20.4%
- Offensive rating — 114
Those numbers are from my spreadsheet and differ slightly from Basketball-Reference tallies. I won’t make you read why.
His rebounding was about the same as it was the previous season, as were stats like blocks and steals. His turnovers and assists rose, and his ast/tov ratio was below average. One bright spot: he fouled a lot less — 3.9 per 100 team possessions after hovering around 5.0 for his first three seasons.
I suggest a minor tap of the brakes on Avdija’s offensive “improvement” over the final 30 or so games of the season. The per game numbers went up because his minutes and usage rose. His offensive efficiency remained a bit below average during that span.
A couple reasons for just a tap of the brakes instead of a stomp.
- His efficiency was okay and maintained at about the same level even when his usage rose. That’s encouraging because players often lose efficiency as they try to do more.
- There are obvious ways he can improve his skills in the offseason and boost his efficiency in future seasons. Chief among these is developing the ability to confidently finish with his left hand.
Below is a chart I call a “Performance EKG” for Avdija this season. The ekg tracks a player’s production game-by-game throughout the season. The red line is the player’s full season PPA after each to that point in the season. The gray line is a five-game moving average, the blue is a 10-game moving average, and the pink is a 20-game average.
A few observations:
- The numbers don’t really support the idea of significant in-season improvement. As mentioned previously, the per game glory stats improved because Avdija got more minutes and increased usage. The quality of his play was in the “about the same” range.
- The chart indicates that Avdija was at his best early in the season, went through a rough patch and then stabilized.
- Avdija’s full-season PPA fell from 126 after 23 games to 110 just three games later. After that, his full-season PPA got as high as 117 (after the 43-point outburst against New Orleans) but never lower than that 110. He finished at 111.
- Avdija scored a 78 in my consistency index metric, which I dusted off and added to the ekgs this season. A score of zero would mean perfectly consistent, and a higher score means less consistent. I’ll probably come back to this metric later in the offseason after I’ve had a chance to run it on more players and get a read on what the numbers mean in today’s environment.
Overall, this was an encouraging and promising season from Avdija. He got stronger, played tougher, and produced the best season of his career. He has a good platform to continue improving and become even more useful to the Wizards (or another NBA team) in the future.
Another aside to wrap things up — part of my preseason forecasting process includes what I call the MILK method (name inspired by Wages of Wins author David Berri, who wrote that NBA players age like milk). It’s a simple approach: take a player’s performance in one season at any age and then apply average aging effects to his future production.
Here’s Avdija’s MILK forecast starting from his rookie year and his actual PPA scores through the end of his contract extension:
- Age 20 season: 48 — actual rookie PPA
- 21: 70 — 69
- 22: 90 — 76
- 23: 104 — 111
- 24: 111
- 25: 115
- 26: 116
- 27: 115
Now, here’s what it looks like over the same span including his actual score for this season:
- Age 23 season: 111 — PPA this season
- 24: 118
- 25: 122
- 26: 123
- 27: 122
Altogether, this approach suggests Avdija is likely to achieve relatively modest improvement over the next couple seasons. The MILK method is built around the fact that most players make their biggest improvements early in their careers.
That said, there are plenty of exceptions, late bloomers, and abrupt surges. Bradley Beal, for example, rated about average in each of his first four seasons before blossoming into an All-Star after he signed his first max contract.
Chauncey Billups didn’t become a consistently above average player until he was 25 — his fifth NBA season and fourth team.
Kyle Korver rated below average for his first eight seasons (he did nudge to a little above average in year two but the rest were well below the average line) before hitting his career peak at age 33.
In other words, while the odds are that what Avdija showed this season is most likely the kind and overall level of production he’ll provide over the next 6-8 seasons, there’s still reason to think he could get significantly better. A lot will depend on how hard and smart he works.
And, even if the improvement from this point is closer to the relatively modest level predicted by the MILK method, that’s still a decent offensive player who defends, rebounds and competes. Guys like that are useful to good teams and tend to stick in the league for a long time.