Our Alexandra Eala vs Tatiana Prozorova prediction points to Eala on the court, but the tier is a PASS. This is an official-card PASS, so there is no official action from us, and we are not backing it at -667. The Singapore WTA opener is set for Thursday, September 24, 2026 at 6:30 AM ET.
The fundamentals lean toward Eala, and the reasons are worth laying out. So are the ways this match could get awkward for her.
Why the model likes Eala on the court
The clearest gap is on hard courts. Over 395 service games against stronger opposition, Eala holds 66% and wins 39% of return games. Prozorova sits at 62% and 34% across 133 games against weaker opposition. That schedule difference matters: Eala earned her numbers against the tougher field.
The serve and return clash runs her way too. Prozorova's biggest leak is against mid-level servers, where she holds 62% and faces 0.95 break points per service game. She wins only 45.6% of second-serve points and double-faults on 3.5% of service points. Eala's best channel is the return of second serves, where she wins 57% of the points, and she wins 45% of return points overall. Her lefty delivery going into a two-handed backhand return also helps, since that return wins only 35% of points off first serves.
Prozorova's own return is the weakest number on the table: 41% of return points won, with 33% of return games broken and a 1-2 count against grinder-style servers.
What each player's losses tell us
Every hard-court defeat in Eala's 12-month table came against a top-tier ball-striker: Jovic, Anisimova, Bencic, Muchova and Noskova. Prozorova's table shows what happens to her against real tour level: Stearns 6-1 6-2, Krejcikova 6-1 6-2, Korneeva 6-1 6-3 and Sasnovich 6-4 6-1. Her best wins top out around the 19th-ranked opponent tier, while Eala's floor sits above that ceiling.
Eala's recent baseline is strong as well. Her three-month window is 13-4 with her best serve numbers of the year, and August brought the Washington WTA 500 title over Osaka, Svitolina and Pegula. Her last match was the three-set US Open loss to Jovic in round three.
The case against: rust, the roof and Prozorova's fight
Eala has played no match since the US Open, about 18 days, and has never hit a ball on this indoor court. She says she feels healthy, and the layoff was reportedly planned rest, but first-match rust is real. This is also the first match of her event.
Her hold is modest. She holds 66% on hard, faces 0.83 break points per service game, and is 8-9 against mid-level servers. Prozorova's raw delivery is bigger, with a 3.5% ace rate to Eala's 1.9%. If an indoor surface rewards that serve and Eala is broken early, the underdog has just shown she will stay in a fight: she beat Sofia Costoulas 5-7 7-6(2) 6-2 in 2 hours 51 minutes on Tuesday, coming back from a set and 4-2 down.
That win is not a clean upgrade, though. It came against a player ranked No. 152, and the match stats behind it are unavailable. Prozorova also carries that court time into a one-day turnaround after roughly seven weeks with one other competitive singles match.
The number that matters
The market has Eala at -667. Nothing in the fundamentals reads as a fair fight, but rust, an unfamiliar indoor court and a modest hold keep this from being a match we want to put our name on. We are not making this bet, and the record we grade in public will not include it as a play.
The verdict
Tier: PASS. This is an official-card PASS with no official action, and it is not a bet we are making. Eala is the better player on the numbers we have, but we are not backing her at -667 on Thursday, September 24, 2026.
- Event · Singapore · WTA (2026) · Thursday, September 24, 2026 · 6:30 AM ET
- Market · Alexandra Eala -667 — recorded when the pick was posted
- Model · T Money, Botcappers' AI Tennis model — how it is built and what it does not know
- Record · every pick is timestamped before the event and graded in public, win or lose
- Statistics · the model's own match/game database, current to the posting time above. Figures a reader cannot open here are internal to that database, not published third-party data.

