2026-27 - Most Improved Player
Last Update: Nov. 11, 2026, 3:02 p.m.
Eligibility: 65 qualifying games (20+ minutes; up to two 15-20 minute games count). Ineligible players' predictions are set to zero.
Model Type: CatBoost Regression Model.
Target: Vote share scaled to sum to 1 by season.
Prediction: Vote share scaled to sum to 1.
| Name |
Qualifying Games |
Predicted Vote Share |
| Cooper Flagg |
10
|
0.187
|
| Dylan Harper |
10
|
0.152
|
| Ty Jerome |
9
|
0.105
|
| Luka Dončić |
10
|
0.084
|
| Davion Mitchell |
11
|
0.082
|
| Duncan Robinson |
10
|
0.073
|
| Victor Wembanyama |
10
|
0.065
|
| Ayo Dosunmu |
11
|
0.028
|
| Scottie Barnes |
8
|
0.025
|