Lakers 2026 offseason · a fan-built fit tool
Luka Fit Index
LeBron is gone. Ayton got traded. Eight new Lakers arrived. It's Luka's team now — and this tool asks the only question that matters about every single move: does it actually fit the way Luka plays?
Player performance: 2025–26 · Roster & market updated Aug. 30, 2026
No blended scores, ever
The story so far
The Lakers won 53 games, took the 4 seed, and got swept out of the playoffs by Oklahoma City. Then the real season started. LeBron James declined to re-sign after 8 seasons in Los Angeles. Deandre Ayton was traded to Washington. The first five arrivals made the July board; Kevon Looney, Matisse Thybulle, and Ziaire Williams changed the roster math after it.
Every bet rides on one player: Luka Dončić, who led the NBA in 2025–26 usage and scoring. This Aug. 30 snapshot scores eight arrivals, the returning rotation, and only the affordable names still on the late-summer board. New-Laker scores remain projections; returning-Laker scores retain the stronger evidence of shared floor time.
The reference point
Luka's profile isn't an opinion. Every axis derives from these numbers.
25-26 record
53-29 4th West
Season ended
0-4 swept by OKC
The rules we scored by
Decided before the scores, so the scores couldn't bend them. The tool keeps a public record of correcting itself.
WHY USAGE?The whole thesis
Usage rate is the share of team possessions a player ends — a shot, free throws, or a turnover. It's the cleanest measurable answer to "who has the ball?" Luka ended a higher share of team possessions than any NBA player in 2025–26, so every teammate's fit starts with one question: can you be great without it?
NEVERNo blended score
One composite number implies a verdict the data can't back. Four axes stay four axes. If you want a single number, this is the wrong tool on purpose.
AXESFour measurable needs
Spacing (3P% + volume), play finishing (TS% without the ball), defensive cover (who guards so Luka doesn't), and ball-need — inverted: lower usage scores higher next to the league's highest-usage player.
MUSTConfidence tags are first-class
Sample size, shooting volume, team context, role, and injury status sit next to every score, not in a footnote. A bar without its context tag is an incomplete claim.
FOUNDThe Luka rest problem
Season usage blends minutes next to Luka with minutes while he rests — opposite signals. Fix: with/without splits. Applying it softened the tool's own loudest take (see Reaves).
UPGRADEDDefense, second pass
Cover scores originally leaned on steals, blocks, and reputation. They now start with individual defensive rating — points allowed per 100 possessions on court. It remains box-derived and team-flavored, so reputation never overrides the number and team context stays visible.
LENSFit and need never merge
Fit asks "does his game complement Luka?" Need asks "does this current roster need his role?" Looney, Thybulle, and Williams materially changed the need answer without changing the fit axes. The two lenses stay separate.
RUBRICEditorial, not algorithmic
The bars are disclosed editorial ratings informed by the displayed statistic, role, and volume: 0 means the skill is absent; 1 poor; 2 below average; 3 workable; 4 strong; 5 elite for a Luka complement. Ball-need reverses usage. No arithmetic combines the axes, and context tags can lower confidence without secretly changing a bar.
How to read
more gold = stronger fit vs Luka's needs (0–5)
purple = fit flag
Ball-need inverted: lower usage = more gold
Ch. 1 · Who came in
Eight new Lakers scored on 2025–26 prior-team context (Kessler uses 2024–25 because his 2025–26 sample was five games). Zero shared possessions with Luka; every score here is a projection.
WK
Walker Kessler
C · scored on 24-25 (58 GP, Utah)
Defensive cover
114.6 DRtg
Best rim-running profile on the board. Lowest confidence among the arrivals: the sample is a year old and the shoulder is surgically repaired; Looney now keeps that risk from being the entire center plan.
QG
Quentin Grimes
SG · 75 GP, Philadelphia
Defensive cover
118.2 DRtg
Signed as a 3&D wing coming off the worst shooting season of his career. The fit verdict rides entirely on which Grimes shows up — and the defense was quieter than the rep too (118.2 DRtg, below team average). Cover score dropped from 4 to 2 in the data rescore.
CS
Collin Sexton
G · 68 GP, Charlotte/Chicago
Defensive cover
117.1 DRtg
Elite shooting season, real talent. But 25.9 usage in 23.7 bench minutes is a ball-dominant profile next to a 38.0 usage franchise player. Regular-season staggering hides it; playoffs won't.
SM
Sandro Mamukelashvili
PF/C · 80 GP, Toronto
Defensive cover
112.0 DRtg
The quiet win of the offseason by these axes: stretch big, career year, huge sample, low ball-need. Defensive numbers ride Toronto's 5th-ranked defense — context flag.
JH
Jaden Hardy
G · via the Ayton trade, Jul 3 · 57 GP (WAS, 17-65)
Defensive cover
124.1 DRtg
The 39.7 3P% looks like a spacing win until you see it took 24.5 usage on the league's worst team to produce. Same archetype as Sexton, worse context — and like every arrival, zero shared possessions with Luka.
KL
Kevon Looney
C · 70 GP, Golden State · low-minute veteran role
Defensive cover
113.8 DRtg
Not a spacer, but a low-usage screener, finisher, rebounder, and credible center body. Looney changes the roster-need conclusion more than the fit thesis: Kessler's repaired shoulder is no longer backed only by small-ball options.
MT
Matisse Thybulle
G/F · 61 GP, Portland · perimeter specialist
Defensive cover
112.9 DRtg
The measured defensive rating supports the reputation, but does not make the offense disappear. Thybulle supplies the point-of-attack option the July roster lacked; his low-volume shot means lineup spacing still needs protection.
ZW
Ziaire Williams
F · 72 GP, Brooklyn · wing-size bet
Defensive cover
118.0 DRtg
The size and shooting shape fit better than the defensive number. Williams eases the raw wing-count problem, but 118.0 DRtg does not support calling him defensive cover; that job still belongs to Thybulle and Vanderbilt.
Verdict: the second wave fixed roles, not every fit. Looney insures the center spot, Thybulle supplies real perimeter cover, and Williams adds wing size. Mamukelashvili remains the clean two-way statistical fit; Sexton and Hardy still duplicate ball-heavy guard creation.
Ch. 2 · Who stayed
The only scores built on real shared-floor context with Luka. Higher-relevance data than any signing can offer.
AR
Austin Reaves
G · co-star · 51 GP · 23.3 ppg, 64.1 TS%, +182
Defensive cover
117.8 DRtg
Ball-need fit
25.2 w/ Luka
Austin Reaves, 2025–26 season and Luka availability splits
| Split | GP | USG% | PPG | APG | +/- |
|---|
| Season | 51 | 26.6 | 23.3 | 5.5 | +182 |
|---|
| With Luka | 41 | 25.2 | 22.0 | 4.9 | +184 |
|---|
| No Luka | 10 | 34.3 | 28.6 | 8.0 | -2 |
|---|
The tool's best self-correction. His season usage was inflated by 10 solo games at a 34.3 rate and 67.5 TS%. That's not a fit problem — that's your co-star proving he can carry the Luka-off minutes.
JV
Jarred Vanderbilt
F · 65 GP · 4.4 ppg, 4.5 rpg, 17.4 mpg
Defensive cover
113.4 DRtg
Ball-need fit
11.7 w/ Luka
A returning low-usage glue player with real shared-floor evidence. The spacing hole means he cannot share the floor with too many of the roster's new non-shooters.
JL
Jake LaRavia
F · 82 GP · 8.2 ppg, 25.1 mpg
Defensive cover
115.7 DRtg
Ball-need fit
14.4 w/ Luka
The most role-stable returning Laker: his usage is identical with or without Luka on the floor (14.4 vs 14.3). Played all 82 nights; the new low-usage arrivals make that connector role less isolated.
DK
Dalton Knecht
G/F · 54 GP · 10.2 mpg, 553 total minutes
Defensive cover
118.8 DRtg
Ball-need fit
16.5 w/ Luka
The shooting profile fits on paper. 553 minutes isn't enough evidence to bet a rotation spot on.
!
Pattern found
Roster construction, not one player
The high-usage guard overlap remains: Reaves, Sexton, and Hardy all need creation possessions. But the low-usage connector group is no longer just Vanderbilt and LaRavia. Looney and Thybulle add defense without demanding the ball, while Williams adds real wing size. The tradeoff moved from pure scarcity to lineup compatibility: Looney, Thybulle, and Vanderbilt cannot all solve defense at once without squeezing spacing.
Verdict: the glue room grew; the guard overlap stayed. Thybulle changes the perimeter-defense conclusion, Looney changes the center-insurance conclusion, and Williams changes the wing-count conclusion. None removes the need to stagger Luka, Reaves, Sexton, and Hardy.
Development roster (not scored — small n by design): Bronny James, Adou Thiero, Cameron Carr.
Two-way: Chris Manon, AK Okereke, Peter Suder.
Roster snapshot: 15 standard contracts and three two-way contracts on Aug. 30.
Departed: LeBron James (declined to re-sign), Deandre Ayton (traded to Washington July 3).
Ch. 3 · Who to chase
The Lakers reached 15 standard contracts by Aug. 30, so this is no longer an open-roster-spot shopping list. It is a contingency board for a later transaction, restricted to two late-August names whose roles still make basketball sense.
Two lenses, never merged
Luka fit asks whether the player can space, finish, cover defensively, and work without the ball. Roster need asks what remains after Looney added center insurance, Thybulle added point-of-attack defense, and Williams added wing size. Those additions lowered urgency; they did not turn fit and need into one number.
No. 1 · best contingency fit
BB
Bruce Brown
G/F · 82 GP, Denver · unsigned in late-August review
Defensive cover
117.1 DRtg
The low-usage connector archetype still fits Luka, but the need case softened once Thybulle and Williams arrived. His 117.1 defensive rating is workable, not reputation-driven proof of stopper value.
No. 2 · spacing-forward contingency
NB
Nicolas Batum
F · 74 GP, Clippers · listed in late-August market review
Defensive cover
116.8 DRtg
The best pure spacing-and-ball-need profile left on the old board. Williams reduced the wing emergency, but Batum's shooting would make low-spacing defensive lineups easier to build. Age 37 keeps the context flag attached.
Verdict: hold, unless another move creates the opening. Brown is the connector contingency; Batum is the spacing-forward contingency. With the standard roster full in this snapshot, neither is presented as an immediately signable sixteenth contract.
Off the board since our last update: Gary Trent Jr. re-signed with Milwaukee on July 16; Gary Payton II returned to Golden State on Aug. 1; Nick Richards signed with Miami on Aug. 27. The July rankings remain part of the page's history, but they are no longer active recommendations.
The data
Every displayed statistic traces to the sources below. Ratings are human-reviewed editorial judgments, not a hidden formula, and context labels identify where a clean number would overstate certainty.
STATSStatMuse, 2025–26
Player totals, shooting, true shooting, usage, defensive rating, and the Reaves availability splits come from StatMuse. Kessler alone uses his 2024–25 full-season line because the following season ended after five games.
NEW ARRIVALSSame prior-team standard
Looney, Thybulle, and Williams use the same 2025–26 StatMuse categories as the July arrivals: 3P% with volume context, TS%, individual DRtg, USG%, games, role, and prior team. Their Lakers fit remains a projection until shared-floor data exists.
MARKETLate-August transaction check
The active board was checked against the NBA free-agent tracker and team transaction reporting. Signed former targets remain visible with dates so the original July conclusion is corrected rather than erased.
METHODThe 0–5 bars
The bars are editorial ratings, using the disclosed 0 absent / 1 poor / 2 below average / 3 workable / 4 strong / 5 elite rubric. Defensive cover begins with individual DRtg but retains a team-context warning. Ball-need reverses usage and uses with-Luka splits for returnees, season usage for arrivals. The four bars are never averaged.
About this project
The Luka Fit Index is a for-fun analytics experiment built in one brainstorming session: a bias-visibility tool disguised as a sports dashboard. The interesting part was never the scores — it's that the assumptions are printed on the page, and the tool corrected itself twice when the data disagreed with its takes.
Built PRD-first with a multi-LLM workflow. Deterministic structure, labeled interpretation, receipts kept in a spike file the whole way.
AI assisted. Human reviewed. Powered by NLP.
See more of my work →