🏀Front Office2026-27
 
Atlanta Hawks
75
OVR

CJ McCollum

Atlanta Hawks · #3 · Rotation
PG6'3"190 lbage 3413 yrs expLehighCanton, OH
7th best of 15 on ATLworth +0.5 wins to the rotation
OFF79DEF67
High-level rim protector & scorer
ATL
Atlanta Hawks
Eastern · Southeast · proj. 41-41 (▲+1.7)
Most Improved Player

A polished veteran combo guard brought in to steady the backcourt, still good for 18.7 a night on 37.5% from three and a tidy 56.6% true shooting. He's a craft scorer and a willing setup man (3.9 assists, top-quartile ball security) who picks his spots out of the midrange. At 34 the limitations are real on the other end — he doesn't rebound and gets hunted on defense — but the shot-making and decision-making still matter.

18.7
PPG
3.3
RPG
3.9
APG
0.8
SPG
0.5
BPG
GP 76MPG 29.8FG% 45.53P% 37.5FT% 77.2TS% 56.6USG% 25.7PER 15.9BPM -0.3VORP 1EPM +0.2
2025-26 regular season · bars are percentile at his position

NBA player ratings

0–99 · 2025-26 form
Rim protection87
Scoring83
Playmaking76
Inside finishing75
Shooting74
Rebounding65
Perimeter defense60

Skill ratings are 2025-26 per-game and advanced stats (Basketball-Reference) ranked 0–99 within position (guards vs guards, bigs vs bigs). Overall is anchored to EPM impact on a non-linear curve, nudged by these skills; Offense and Defense come from box plus-minus. A player who missed the latest season is shown from his most recent one.

Career

2013-14 to 2025-26 · 13 seasons
19.5
PPG
3.6
RPG
3.8
APG
863
GAMES
Points 16,838Win Shares 54.9Starts 750Best season 2016-17 · 23.0 ppg
Games played, last three seasons81%
AMONG CURRENT PLAYERS
20th in career points

Regular season, 2003-04 onward (Basketball-Reference). Averages are weighted by games played; honours count awards won, not finishes. “Current players” means the 802 who played in the last three seasons.

Trajectory

2023-24 → 2025-26 · age 32 → 34
Declining

CJ, at 34, is trending downward — 6.4 → 3.0 wins of value over the window.

Age curve Behind curve
Falling off faster than age alone predicts.
2026-27 forecast 2.0 wins (-1.0)
Age takes about 1.3 wins off at 35. Range 0.04.6 wins.
Production (BPM)+3.1-0.3-3.426-27 forecast -0.8
His on-court production has slipped.
2023-24: +3.1 · 66 games2024-25: -1.4 · 56 games2025-26: -0.3 · 76 games2026-27: -0.8 (forecast)23-2425-2626-27
Value to team (wins)6.43.0-3.426-27 forecast 2.0
His win contribution is trending down.
2023-24: 6.4 · 66 games2024-25: 1.7 · 56 games2025-26: 3.0 · 76 games2026-27: 2.0 (forecast)23-2425-2626-27
Contract value2026-27: worth $4.6M vs $21.0M salary−$16.4M
He is paid more than his play is projected to be worth.
$21.0M salary2023-24: $32.2M · 66 games2024-25: $4.1M · 56 games2025-26: $6.9M · 76 games2026-27: $4.6M (forecast)23-2425-2626-27
Current deal: $21,000,000 cap hit · through '27. On the contract row the flat dashed line is this salary, so the gap to the 2026-27 forecast is the surplus.

Seasons from Basketball-Reference (2023-24 → 2025-26). Production is box plus-minus, a per-possession measure. Value is VORP-based wins, so it also counts games played. Contract value is what each season’s play would fetch on the open market. The dotted 2026-27 point is a forecast: his established level, a year older, assuming a normal season.

Contract

2026-27
$21,000,000Guaranteed
11% of the club’s payroll · 3rd of 15 priced deals
Runs through1 season
Est. market value$4,585,333 /yr
Surplus$16,414,667
ABOVE VALUEpaid above his on-court value

Market value is an estimate, not a reported figure: what his play is forecast to fetch on the open market in 2026-27 — his current rating, a year older. Cap thresholds are projected for 2026-27 until the league sets them.

2026-27 off-season

1 move

OFFICIAL means the league’s own records carry the move; AGREED means a trusted insider reported it and the league has not processed it yet. Terms and attribution come from the wire, which is the only source that carries them.

Plays like

same position

Nearest skill profiles across the seven rating axes — who he plays like, not who is as good. The rating beside each name answers that second question.