Spurs Lineup Data Complicates Fox-Castle-Harper Finals Debate

San Antonio's Finals guard debate was more complicated than simply choosing Dylan Harper and Stephon Castle over De'Aaron Fox.

A detailed Reddit lineup analysis argues that Castle, rather than Fox, was the bigger fourth-quarter problem. It also suggests Fox needed to play whenever Victor Wembanyama rested because the Spurs struggled to create offense without either of them.

The two-man lineup numbers favored Fox and Harper:

1. Fox-Castle: -3.3 net rating in 103 minutes, 1.44 assist-to-turnover ratio
2. Fox-Harper: +4.5 in 99 minutes, 4.5 assist-to-turnover ratio
3. Castle-Harper: -11.9 in 79 minutes, 1.0 assist-to-turnover ratio

Adding Wembanyama changed the picture:

1. Fox-Wembanyama-Castle: -1.4 in 100 minutes, 1.41 assist-to-turnover ratio
2. Fox-Wembanyama-Harper: -2.0 in 75 minutes, 2.24 assist-to-turnover ratio
3. Castle-Wembanyama-Harper: +3.6 in 62 minutes, 1.09 assist-to-turnover ratio

Castle, Harper and Wembanyama defended extremely well, but their offense fell sharply:

1. Fox-Wembanyama-Castle: 104.1 offensive rating, 105.3 defensive rating
2. Fox-Wembanyama-Harper: 102.0 offense, 104.0 defense
3. Castle-Wembanyama-Harper: 96.9 offense, 93.3 defense

The Castle-Harper pairing without Wembanyama posted a -15.5 net rating in 17 minutes. By contrast, the analysis found that most Fox lineups without Wembanyama finished with positive net ratings. Fox-Wembanyama minutes were less convincing.

The fourth-quarter Castle-Harper stretches were especially damaging:

• Game 5, 8:51-6:50: a 10-point Spurs lead with possession became a six-point lead with Knicks possession.
• Game 4, 8:43-6:24: a 15-point lead with possession became an 11-point lead with Knicks possession.
• Game 3, 9:18-7:31: a five-point lead with the ball remained a five-point lead with the ball.
• Game 3, 4:50-0:00: an eight-point lead became a four-point win, including two take-foul free throws.
• Game 2, 10:15-6:45: a nine-point deficit with Knicks possession became a 12-point deficit with Spurs possession.
• Game 1, 7:37-4:04: a tie with Knicks possession became a four-point deficit with Knicks possession.

Across those stretches, Castle and Harper scored 25 points in 18 minutes. San Antonio finished in a worse position every time except the first Game 3 segment, and at least 20% of the scoring came from free throws.

The samples are small and do not clear Fox. They challenge the idea that benching him for Castle and Harper would have fixed the late-game offense.

The contrast was clearest when Wembanyama was involved. Castle-Harper-Wembanyama produced the best defensive rating of the three featured groups at 93.3, but its 96.9 offensive rating was easily the lowest. Removing Fox improved the defense while taking away a needed source of creation. San Antonio could get stops, yet converting those stops into points became another problem.

The original breakdown also highlighted one extreme reserve configuration: Luke Kornet, Castle and Harper posted a -65.8 net rating. The tiny sample still illustrates how badly the non-Wembanyama offense could unravel.

Game 2 offered another useful detail. Castle sat for more than six minutes after the Castle-Harper stretch ended, and the Spurs recovered from a 12-point deficit during that period. He returned with 30.3 seconds remaining. That sequence does not prove one substitution decided the game, but it adds context to the criticism of San Antonio's late rotation choices.

The same Castle-Harper-Wembanyama trio had been more effective against Oklahoma City. That difference suggests the matchup with New York mattered. The Knicks could pressure San Antonio's young guards and force the Spurs to depend more heavily on Fox for half-court creation.