Our Mirra Andreeva vs Aliaksandra Sasnovich prediction for Wednesday, September 23, 2026 starts with a gap in class and ends with a gap in price. Andreeva is off a US Open quarterfinal run and Sasnovich is one match into Singapore after getting into the draw as a replacement, so the résumés point one direction. The market agrees, and agrees hard.
Andreeva is the pick here. The market has her at -1000 on the moneyline. That is the whole story in two sentences: right side, wrong number. This is a PASS, an official-card no-play, and we are not backing it. Here is how the model got to Andreeva anyway, and why the price is the problem.
Why the model is on Andreeva
Strip out the layoff for a second and look at the twelve-month hard-court baselines. Andreeva is 23-12 on hard against a tougher field, holding serve 75% of the time and breaking 42% of the time. Sasnovich is 21-16 on hard against a softer, mostly qualifying-level schedule, holding 67% and breaking 39%. Both numbers run in Andreeva's favor.
The specific lever is the second serve. Andreeva's return game has been winning roughly 58% of points against second serves in recent tracking, and Sasnovich's second serve is the weak link in her game: 45% of those points won over the last twelve months, 43% over the last three. That is the matchup the model leans on, a strong return against a serve that has been slipping, not just a name on the draw sheet.
The case against
Andreeva has not played a match since a three-set US Open quarterfinal loss to Coco Gauff on September 9, a 6-2 6-7(7) 2-6 defeat where she held two match points in the second-set tiebreak before it slipped away. That is fourteen days of rust walking into Wednesday, and the loss table has a pattern worth naming: Andreeva has dropped deciding sets on hard courts after winning the first, against Zhu Lin, Siniakova, Kostyuk and Gauff. A slow start against Sasnovich would not be shocking.
Sasnovich, meanwhile, just turned in her cleanest match in a while. She beat Daria Kasatkina 7-5 6-0, closing it out by winning the last eight games, and her own serve looked sharp doing it: zero double faults and 73% of first-serve points won, both well above her recent form. She got only one day off before facing Andreeva, but she arrives with real momentum, not just an easy draw.
There is also a channel that runs against Andreeva specifically. Her second serve has been a soft spot of its own, opponents have been winning the majority of those points, with her double-fault rate climbing to 6.9% of service points over the last month. Sasnovich's return on second serves sits at 55% won over twelve months. That is a live path for Sasnovich to stay in games even while Andreeva is dictating from the ad court.
The number that matters
None of that changes who the better player is here. It changes what this is worth as a bet. The market has Andreeva at -1000, and that price is the reason this stays a PASS. We are not naming which threshold it tripped, only that the fundamentals case and the executable price are two different questions, and on this one they land in different places.
A PASS is not a hedge and not a soft version of a play. It means no side goes on the official card for this match, full stop.
The verdict
Andreeva is the sharper player walking in, built on a better hard-court baseline and a return game aimed straight at Sasnovich's second serve. But fourteen days of rust, a history of blowing deciding sets after winning the first, and a red-hot Sasnovich coming off a clean win over Kasatkina all keep this from being automatic. Combine that with where the number sits, and this is a PASS. Andreeva wins the fundamentals argument. She does not win a spot on the card.
- Event · Singapore · WTA (2026) · Wednesday, September 23, 2026 · 6:30 AM ET
- Market · Mirra Andreeva -1000 — 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.

