**Metric: Aces per match. Category: court Surface.** I chose these as the most decision-relevant pair. Surface is the main classifier for deciding where a player performs best, and it cleanly separates serve outcomes. Aces per match was the clearest serve metric. This is a judgement call on my part, not something the data ranks. The rendered result comes from the gold table `raw_kaggle_by_Surface`. **How aces vary by surface (average per player-match):** - **Carpet:** about 7.94 aces (median 7), from 2,477 player-matches. - **Grass:** about 7.57 aces (median 6), from 9,479. - **Hard:** about 5.88 aces (median 5), from 117,571. - **Clay:** about 3.38 aces (median 3), from 107,650. - **Fast versus slow surfaces:** Carpet and grass produce more than twice as many aces as clay (roughly 7.6–7.9 versus 3.4). Hard sits in between. Clay suppresses serve dominance, and grass and carpet reward it. - **Double faults are much flatter:** They average 2.81 on clay, 2.97 on hard, 3.11 on carpet and 3.17 on grass. Surface changes how often serves win outright far more than how often they fail. - **Opponent aces (`vA`) follow the same pattern:** Opponents average about 8.13 on carpet, 7.59 on grass, 5.74 on hard and 3.21 on clay. This is consistent with the ace gap coming from the surface and not just from the players. **Caveats:** - **Uneven sample sizes:** Hard and clay account for most of the data. Grass and carpet are small (9,479 and 2,477 rows), so their averages are less stable. - **Player mix differs by surface:** The average ranking (`Rk`) is about 151 on grass, 311 on hard, 361 on clay and 517 on carpet. Part of the difference may come from who plays on each surface, not only from the surface itself. - **Blank surface rows:** 28 rows have no surface label and are excluded from the comparison.