Basketball Bytes: The Thunder Roll, Despite An Offensive Cold Front
OKC is an elite team with an elite record — but why are they struggling to score? Plus, is free throw defense varying more than usual this year?
Welcome to Basketball Bytes — a ✨brand-new spin-off✨ of my baseball column, in which I point out several byte-sized pieces of information that jumped out to me from my various basketball spreadsheets. If you’ve noticed a Basketball Byte of your own, email me and I’ll feature it in a future column!
🏀 OKC’s Offensive Turbulence
A year ago, the Oklahoma City Thunder went into the NBA season as the league’s Team of the Future. Then they just became the Team of the Present — winning 57 games and capturing the No. 1 seed in the Western Conference during the regular season. While they fell short in the second round to a Dallas Mavericks team that punched above their weight in the playoffs, it was easy to predict that OKC would continue their ascent in 2024-25, particularly with a year of much-needed postseason experience under their belts.
And indeed, that’s exactly what has happened at the start of this year’s schedule. While they suffered their first two losses of the year in the past week, the Thunder still went into Monday’s action ranked second in the West with an 8-2 record, checked in at No. 2 in my composite forecast’s title odds (15.5 percent, trailing only the powerhouse Boston Celtics) and were tied with the surprising Cleveland Cavaliers for the best net rating in the league (+12.5 points per 100 possessions).
All of this was supposed to happen for Oklahoma City, particularly after having one of the better offseasons of any NBA team — they ranked third in the league in Net Wins Above Replacement (WAR) added minus subtracted following the additions of Isaiah Hartenstein and Alex Caruso, while losing little of note aside from Josh Giddey and a retiring Gordon Hayward.
But the way OKC has been winning early in 2024-25 is more than a bit surprising.
The Thunder rank No. 1 by an absolute mile in defensive efficiency, allowing just 100.3 points per 100 possessions. The second-best defensive team, the Orlando Magic, are at 107.5 points allowed per 100; the difference between No. 1 OKC and No. 2 Orlando is roughly the same as the difference between No. 2 Orlando and No. 18 Chicago Bulls (115.3 defensive rating).
That in and of itself represents a big leap from last year’s OKC defense, which was good (No. 4) but far from historically elite. The flip-side, though, is that OKC’s offense — which ranked third a year ago — is now 0.3 points per 100 below average, checking in at No. 16 in the league. An up-and-coming title contender notable for its balance has become a defensive juggernaut with a weak offense.
Why? The simplest explanation is shooting. The Thunder had the NBA’s sixth-best FG% on 2-pointers and the very best FG% on 3-pointers last season; this year, those rankings are down to 19th and 15th, respectively. Layer on the fact that they’re also shooting worse from the line (79.8% versus 82.5%) and getting there a lot less as a share of FG attempts — they’re down to second-worst in the league in FTA/FGA — and it makes sense that this offense has not been efficient this time around.
One other interesting note is in the split between shot quality and shot-making. Based on where they took their shots from the field, we’d expect OKC to have an Effective FG% of 53.9 this season if they had league-average shot-makers, which ranks 14th and barely scrapes above the NBA mean. But that’s not really different from last season, when OKC’s 54.8% mark ranked 13th and was only fractionally above average.
Instead, the main difference between the two seasons has come in terms of shot-making: A year ago, OKC outshot its expected eFG% by 2.5 points, the fourth-best rate in the league, while this year they are undershooting expected eFG% by 0.4 points, which ranks 18th.
Brilliant as Shai Gilgeous-Alexander is, his eFG% has dipped relative to the league — in part because his 3P% has fallen from 35.3% to 26.4% — and Cason Wallace and Ousmane Dieng’s shooting number are down considerably as well.
But among Thunder regulars (double-digit MPG), nobody is faring worse than Caruso.
The former L.A. Laker and Chicago Bull was traded for Giddey because he was supposed to help OKC at both ends of the court. While he’s known primarily as a defender (earning DPOY votes in each of the past two seasons), Caruso also shot 40.8% from downtown and boasted a +0.3 offensive Estimated RAPTOR last season to go with his strong +2.3 mark on defense.
But although Caruso has undoubtedly contributed to the Thunder’s dominant D so far this year, he’s provided whatever the opposite of a spark is on offense. He’s shooting 27.1% from the floor and 18.5% from 3, has a -4.4 offensive Estimated RAPTOR and OKC scores 9.8 fewer points per 100 with him on the court versus off. Perhaps not coincidentally — and granting that he’s also playing for a much better team now — Caruso’s minutes per game have dipped from a career-high 28.7 in Chicago last season to just 19.5 so far this year in OKC.
The good news for OKC (and Caruso specifically) is that shot-making is prone to huge random swings in a small sample. Given their wealth of offensive weapons, from trigger-men SGA and Jalen Williams to shooters like Wallace, Lu Dort, Isaiah Joe and Aaron Wiggins, the Thunder should be making a lot more 3s — and shots in general — than they are right now. While we may look back on Caruso’s 2023-24 season as a career year in terms of offensive effectiveness,1 he has a history as an above-average deep shooter.
And when the Thunder get their offensive issues straightened out, this team could potentially be even scarier than they already have been to start the season.
🏀 Free Throw Defense Is All Over the Map
This week, we have a mailbag question from reader Adam, who writes:
“I'm surprised by the spread of opponent FT% in the NBA. The "best" team - the Knicks (who were also the best last year) have been holding opponents to 73.3%, whereas the "worst" team, the Bucks, allows 81.9%. Obviously the range will narrow by the end of the season, but in past years there's been a 4-6% difference. I would expect something lower. Any idea what's going on?”
In addition to being a good question, it’s also a thematic one, because the Thunder now actually have the lowest opponent FT% allowed in the league this season. (This changed since Adam’s question, but the Knicks still rank sixth-best.)
Typically, we think of teams having little control over opponent accuracy from the stripe, aside from small factors at the margins, which suggests even more strongly that OKC needs to sort out its offense before that FT% defense regresses!
But back to Adam’s point: The spread of free throw defense seems to be much wider than usual so far this year. And indeed, the standard deviation of opponent FT% in 2024-25 is 2.5%, well higher than last year’s 1.2% mark. In fact, if this season continued at that pace, we would be staring at the biggest spread in opponent FT shooting since at least the NBA-ABA merger in 1976 by a full percentage point over 1982-83, the previous record-holder for widest spread of FT% defense in a single season.
As a general rule, NBA stats tend to stabilize earlier in a season than in other sports. But even within basketball, this varies a lot based on the statistic. As I mentioned last week, player tendency stats such as Usage Rate — or even shot locations/shot quality — tend to tell us something real fairly quickly. Something more random, though, such as 3-point percentage (or the even more unstable opponent 3P%) can take far longer into a season before stabilizing.
And it’s tough to find a stat more luck-driven than opponent FT%, which (relative to league average) has a year-over-year correlation of just 0.15 for NBA teams since the merger.
That’s why, if instead of comparing the start of 2024-25 with full historical seasons, we look at just the first 10 games per team of each season, this year’s leaguewide standard deviation of opponent FT% stops standing out:
Our 2.5% figure from above, which would have towered over any full season in recent history, actually comes in less spread-out than the typical season since 1983-84 when we compare it to just the first 10 games per team each year. (By comparison, last season’s opponent FT% numbers had a standard deviation of 3.5% through this same point in the schedule.)
Mainly, this just showcases how much noise there still is in opponent FT% early in the season. The differences between the top and bottom teams look wide now, but they are destined to get a lot more compressed by the end of the season.
Filed under: NBA, Basketball, Basketball Bytes
It was the only time he ever recorded a positive Estimated RAPTOR at that end of the court.
I think to properly assess Free Throw Defense, we'd actually have to look at the difference relative to expectation. Teams in the Atlantic division (outside of Philly) play Embiid more often than other teams and he takes more FTs and hits at a higher rate than most other players (and they had him and Harden simultaneously!). Surely that accounts for some of the variance.
I also wonder if some teams are smarter about fouling bad free throw shooters than others? Though this may have been filtered out more recently with some rule changes discouraging intentional fouls.