Lu Dort — Two-Way Deep Dive (3 Seasons)

Analysis date: 2026-04-17 Source: BBall Index (bball-index.com) — rerun with the Basketball Analytics Expert Sage active.

The Question

Three-year look at Dort: what's the true-talent baseline on each side of the ball, how much of 2024-25 was skill growth versus role expansion, and how much of 2025-26's regression is signal versus noise? The framing matters — single-season composites swing, and the right read is usually a weighted multi-year composite with priors named.

Part 1: The Big Picture

Dort's 2024-25 was an above-prior season driven by role expansion. LEBRON went from -1.15 to 0.51 — net-positive for the first time — then back to -0.69 this year. A 1.2-point swing season-over-season is meaningful, but LEBRON is a talent-in-context composite (box prior × adjusted plus-minus × role/stabilization layer), and single-season composites typically move around a multi-year baseline by more than people assume. Weighting all three years toward the current one, the read is a slightly-below-replacement role player, not a player who "was good and is now bad."

The split tells the real story: D-LEBRON is steady-positive across all three years (0.17 → 0.78 → 0.60), O-LEBRON is steady-negative with one above-prior spike (-1.32 → -0.27 → -1.30). The defensive value is the stable trait. The offensive swing is the seasonal variable.

Metric	2023-24	2024-25	2025-26	Read
LEBRON	-1.15	0.51	-0.69	One above-prior year on a negative baseline
O-LEBRON	-1.32	-0.27	-1.30	24-25 was the outlier
D-LEBRON	0.17	0.78	0.60	Defense is the stable positive
LEBRON WAR	1.84	3.86	2.04	Volume (minutes) shaped by role
Offensive Talent	-0.48	-0.61	-1.45	Talent read declining with role change
Minutes Per Game	28.4	29.2	26.8	Contracted this year

Part 2: The Offense

Shooting efficiency — a prior question, not a slump question

True Shooting % dropped 6.6 points to 52.5% and Effective FG% hit 50.8%. Before reading that as collapse, name the prior. His multi-year Stable FG3% is 38.5 → 39.3 → 35.0, which suggests his 3PT true-talent prior sits around 37-38%. The 41.2% in 2024-25 was above-prior; the 34.4% this year is below-prior. The multi-year expectation is closer to 37% than to either single-year reading.

That said, the drop is not only variance. Overall Shot Quality has declined from 1.12 to 0.47 — the shots available to him are genuinely worse this year, which isn't noise, it's context (teammate roster, role change, defensive attention distribution). Overall Shot Making is -0.75, so he's also converting below expectation on those worse looks. Both the terrain and the execution moved against him simultaneously. The prior-naming says the shot-making -0.75 is probably partly variance, the shot-quality 0.47 is probably partly structural.

Metric	2023-24	2024-25	2025-26	Read
True Shooting %	59.1%	58.7%	52.5%	Below a multi-year ~56% prior
Effective FG%	55.7%	57.8%	50.8%	Below prior, both layers moving
Overall Shooting Talent	-0.87	-0.70	-1.16	Worst single-year read, prior likely ~-0.9
Overall Shot Making	-0.42	-0.16	-0.75	Execution below expectation
Overall Shot Quality	0.96	1.12	0.47	Worse looks — structural, not variance
Overall Shot Creation	-1.07	-1.25	-1.18	Stable and limited

Three-point shooting — the load-bearing offensive question

This is where the analysis should concentrate, because the 3PT volume is high enough for the signal to stabilize and the archetype depends on this skill.

On volume: 3PT FGA Per 75 at 7.38 and 3PT Attempt Rate at 71% — the offense is asking him to be a shooter even though the Stable FG3% prior says he's a 37-38% shooter, not the 41% he posted in 2024-25 nor the 34% he's posting this year.

On the terrain: 3PT Shot Quality collapsed from 1.74 to 0.29. That's the single most important offensive data point. In 2024-25 Dort was getting elite-quality looks (likely: more off-ball screens, better spacing, defenses sagging off him because they were correctly treating him as a secondary threat behind Shai and Jalen). In 2025-26 the quality has reset to near league-average. Part of this is opponents respecting him less off-ball, part is internal scheme/usage shifts.

On execution: 3PT Shot Making flipped from +0.38 to -0.38. On the lower-quality looks he now gets, he's also under-converting. That two-layer move — worse looks AND below-expected execution — is the real offensive issue.

Cross-check via C&S 3PT Shooting Talent: 0.76 → 1.12 → 0.43. The catch-and-shoot strength that made 2024-25 work has softened, though not collapsed to negative. If that recovers to ~0.9-1.0 on stable volume, his offense stabilizes closer to the 2023-24 baseline than to the 2025-26 trough.

Metric	2023-24	2024-25	2025-26	Read
3PT %	39.4%	41.2%	34.4%	Bracketing a ~37% prior
Stable FG3%	38.5%	39.3%	35.0%	Prior ~37-38%
3PT FGA Per 75	6.36	7.17	7.38	Volume trustworthy for the read
3PT Attempt Rate	60.3%	69.7%	71.0%	Role is more 3-dependent
3PT Shooting Talent	0.07	0.31	-0.41	24-25 above prior; prior ~0
3PT Shot Making	0.23	0.38	-0.38	Executing below expectation now
3PT Shot Quality	1.05	1.74	0.29	Terrain reset — the structural issue
C&S 3PT Shooting Talent	0.76	1.12	0.43	The skill that powered 24-25 has softened

Self-creation and midrange

Isolation Shooting Talent (-0.89 → -0.50) and One on One Shooting Talent (-0.94 → -0.51) have improved each year, which is a quieter signal than headline readers will notice. But his Usage Rate is 14.1% — low enough that these numbers describe a small fraction of possessions and should be weighted accordingly. The improvement is real but role-limited in its impact.

The surrounding creation picture is consistent with the archetype: 3PT Pull Up Talent worse (-0.64), Midrange Talent flat-negative (-0.76), Floater Talent stable-negative (-0.72). Dort is not a self-creator and the metrics have never suggested he would become one.

Metric	2023-24	2024-25	2025-26	Read
3PT Pull Up Talent	-0.48	-0.43	-0.64	Off-the-dribble 3 is not a threat
Isolation Shooting Talent	-0.89	-0.70	-0.50	Quiet upward trend — small sample
One on One Shooting Talent	-0.94	-0.64	-0.51	Same: low-volume signal
Midrange Talent	-0.70	-0.76	-0.76	Not his game, not changing
Floater Talent	-0.42	-0.72	-0.72	Stable-low
Self-Created Shot Making	-0.62	-0.47	-0.72	Fits the archetype

Finishing at the rim — selectivity, not skill

Rim FG% jumped to 65.1% on half the volume — Rim Shot Attempts Per 75 went from 2.55 to 1.25. Finishing Talent has nudged from -1.60 to -1.36 across three years, which is the true skill read; it has not moved meaningfully. The improved rim percentage is almost entirely selection-driven: he's avoiding contested looks and only finishing when the read is easy. Useful for possession-by-possession efficiency; not a talent story.

Drives Per 75 at 2.98 (down from 5.32) reinforces this — he's not getting to the rim less because defenses are stopping him, he's choosing not to put the ball on the floor. Drive Foul Drawn Rate at 6.0% is a reasonable guess at his prior; the 2.97% in 2024-25 was probably small-sample variance.

Metric	2023-24	2024-25	2025-26	Read
Finishing Talent	-1.60	-1.45	-1.36	Talent prior hasn't moved
Rim FG%	57.9%	57.5%	65.1%	Selection, not skill
Rim Shot Attempts Per 75	2.55	1.63	1.25	Driving less, not stopped less
Drives Per 75	5.32	3.51	2.98	Role contraction
Contact Finish Rate	19.4%	32.0%	28.0%	Noisy at this sample
Drive Foul Drawn Rate	6.65%	2.97%	6.0%	24-25 was low-sample variance

Playmaking, usage, and gravity

Off-Ball Gravity is the downstream telltale: 1.20 → 0.49. Defenses are sagging off him more because the shot-making isn't keeping them attached, which costs OKC spacing on every possession Dort is on the floor. On-Ball Gravity is still negative (-0.35), consistent with the archetype — he's not commanding help with the ball.

Box Creation dropping from 1.39 to 0.79 and Passing Creation Quality flipping negative (-0.18) are inside the "role-contracted" story: less on-ball responsibility, less creation output. These are opportunity-driven more than skill-driven.

Metric	2023-24	2024-25	2025-26	Read
Playmaking Talent	-0.66	-0.66	-0.83	Prior ~-0.7, noisy movement
Box Creation	1.39	1.21	0.79	Opportunity-driven decline
Passing Creation Quality	0.09	0.06	-0.18	Marginal — not meaningful
Usage Rate	15.4%	13.6%	14.1%	Consistent role (low)
True Usage	21.7%	20.0%	19.8%	Decline in ball time
Ball Dominance %	8.9%	7.6%	9.6%	Noisy around ~9%
On-Ball Gravity	-0.48	-0.52	-0.35	Stable-negative
Off-Ball Gravity	1.16	1.20	0.49	Spacing cost — real issue

Part 3: The Defense

Perimeter defense: the stable trait

Perimeter Isolation Defense at 3.42 is a three-year high in a metric where he's been elite every year (3.16 → 3.11 → 3.42). This is the stable positive-talent read. Single-year movement inside that band is less meaningful than the fact that all three years sit in the top tier of isolation defenders on-ball. Matchup Difficulty at 3.03 confirms he keeps drawing tough assignments — the number is consistent across all three seasons.

What actually shifted this year is assignment type, not assignment difficulty. Guarding Shot Creators went from 34% → 49% while Primary Ball Handlers dropped from 22% → 14%. Same difficulty, different job. OKC is using him as a wing stopper against shot creators rather than a point-of-attack defender against lead guards.

Metric	2023-24	2024-25	2025-26	Read
Perimeter Isolation Defense	3.16	3.11	3.42	Stable elite — the trait
Matchup Difficulty	3.22	2.81	3.03	Consistently hard assignments
% Guarding Shot Creators	34.3%	44.3%	48.8%	New role: wing stopper
% Guarding Primary Ball Handlers	21.9%	16.9%	14.0%	Less point-of-attack

Defensive metrics with wider noise bands

Here is where the analysis must be most careful. Ball Screen Navigation went -1.28 → 2.96 → 1.34 and Passing Lane Defense went 2.60 → 3.50 → 0.80. Those look like breakout-and-regression stories, but defensive metrics derived partially from plus-minus carry wide single-season error bands — the methodology documentation is explicit that defensive responsibility diffuses across possessions in ways that introduce noise. The 2024-25 spikes may have been partly skill (scheme change, real improvement) and partly the same kind of positive variance that makes single-season plus-minus derivatives overshoot.

The correct read is the multi-year composite:

Ball Screen Navigation: 3-year average ~1.0, with one spike. The 2024-25 "breakout" was likely part-skill, part-noise; the 2025-26 "regression" is likely part-correction, part-noise.
Passing Lane Defense: 3-year average ~2.3, which is still solid-positive. The 0.80 this year is below prior; the 3.50 last year was above it.

Neither "fell off a cliff" and neither "broke out." Both are somewhere in between across all three years.

Metric	2023-24	2024-25	2025-26	Read
Ball Screen Navigation	-1.28	2.96	1.34	Multi-year prior ~1.0, wide noise
Off-Ball Chaser Defense	1.36	1.57	0.77	Declined, prior ~1.2
Passing Lane Defense	2.60	3.50	0.80	Prior ~2.3, 25-26 below
Pickpocket Rating	0.64	0.66	0.43	Inside noise
Steals Per 75	1.13	1.37	1.24	Stable around 1.2

Effort, coverage, and role

Defensive Miles Per 75 dropped from 1.31 to 1.01 and Coverage Aggression is down. These are real changes, but paired with Defensive Positional Versatility at a three-year high (83.6), the pattern is scheme, not effort — OKC is asking him to cover more positions with less ground-covering aggression. That's a defensive scheme shift consistent with moving him off point-of-attack and onto wings.

Help Defense Talent at -0.68 is the worst single-year reading, but his prior on help defense has always been negative (-0.52, -0.33, -0.68). He is and has always been a narrow-role defender: elite on-ball against perimeter creators, below-average as a help defender. The 2025-26 reading doesn't change the archetype; it moves inside a band he's been in for three years.

Rim Protection at 0.51 is stable-positive and matters more than readers usually give it credit for in a 6'3" guard — he helps enough not to be a liability.

Metric	2023-24	2024-25	2025-26	Read
Defensive Miles Per 75	1.31	1.29	1.01	Scheme shift, not effort
Coverage Aggression	52.1	52.6	49.7	Same
Defensive Positional Versatility	81.3	78.9	83.6	Guarding more positions
Help Defense Talent	-0.52	-0.33	-0.68	Narrow prior, always negative
Help Defensive Activity	-0.42	-0.42	-0.59	Consistent with the archetype
Rim Protection	0.54	0.37	0.51	Stable-positive for his size
Defensive Role Versatility	46.1	39.1	35.2	Role narrowing — by design

Part 4: The Complete Picture

The multi-year prior

Dort's three-year weighted composite is a high-usage-for-role perimeter defender with a below-replacement offensive baseline and one above-prior offensive season. Call that prior a ~-0.5 LEBRON player — net-slightly-negative, most of the defensive value offset by most of the offensive cost, with heavy role-conditionality.

What is actually moving

Signal (real, role-conditional):

Perimeter Isolation Defense at a three-year high is the most stable positive trait he has
Overall Shot Quality genuinely declined — that's a structural/scheme reality, not noise
Off-Ball Gravity collapse (1.20 → 0.49) is a downstream spacing problem for OKC on every possession
Ball Screen Navigation volatility is partly real improvement followed by partial regression; multi-year ~1.0 is the honest read
Assignment type shifted — shot creators and small forwards up, primary ball handlers down

Noise / prior-adjusted:

3PT % swings from 41% to 34% are bracketing a ~37% prior. Neither extreme is "who he is."
Passing Lane Defense 3.50 → 0.80 is wider than the single-year skill move plausibly accounts for — multi-year composite ~2.3 is the honest read
Rim FG% at 65.1% is selection-driven and doesn't change Finishing Talent (still -1.36)
Single-year LEBRON swings of ~1.2 points are inside the typical noise envelope for single-season composites on role-contracted minutes

The role-conditional verdict

Dort is a defensive-specialist wing with a narrow, context-dependent offensive role. His 2024-25 was a role-expanded, above-prior season on multiple dimensions (shot quality, ball-screen navigation, overall impact); his 2025-26 is a role-contracted return toward baseline with the same defensive specialist archetype intact. The correct forward-looking read for OKC is neither "2024-25 Dort" nor "2025-26 Dort" but a blended, multi-year composite — a defensive-plus, offensive-negative role player whose total value depends on how aggressively the offense leans on him to shoot and whether his 3PT% lands near its ~37% prior or continues to underperform it.

The actionable question isn't "is 2025-26 Dort the real Dort"; it's "what does OKC need Dort to do in the playoffs." If the answer is "elite perimeter isolation defense against the other team's best wing scorer for 25-28 minutes," the three-year record says he does that job reliably. If the answer is "and also hit 38% of his threes at 7+ per 75," the prior says that's above-expectation — it happened once and it would need to happen again.