Our Oksana Selekhmeteva vs Kamilla Rakhimova prediction for this US Open second rounder lands somewhere odd: T Money's fundamentals favor Rakhimova, the market has her at -161, and the pick is still a PASS. That is not a hedge. It means the case for Rakhimova is real, the case against her is also real, and neither one clears the bar to make this an official bet.
They play at 3:00 PM ET on September 2. Rakhimova is the side T Money's model prefers on paper. The reason this stays a PASS lives in a piece of recent history that the raw numbers cannot erase.
Why T Money likes Rakhimova
Strip away the head to head for a second and look at how these two are actually serving right now. Neither one has a real weapon on serve, both self rate as grinders, so the match gets decided on free points and who gives fewer of them away.
Selekhmeteva's double fault rate has been climbing across shrinking windows, from roughly 10% up to nearly 14%, with her first serve percentage sliding from the high 60s down into the high 50s and her second serve points won dropping into the low 30s. Rakhimova's line, by contrast, is flat: about 7% double faults and 44 to 46% on second serve win rate in every window T Money looked at. That gap is worth roughly a break a set if it holds.
Look at how each of them beat a tougher opponent in round one. Rakhimova needed seven match points to close out a seeded Krejcikova, but she did it by winning 72% of her first serve points, eleven points clear of her own baseline, and she held her nerve through the finish. Selekhmeteva's win over Ito was earned on break point conversion and second serve return, a genuinely good day, but it came inside a 51% total point margin behind a 6-2 6-4 scoreline, with a chunk of the swing coming off Ito's second serve collapsing rather than Selekhmeteva forcing it.
Read on percentile terms, the ceilings sit a tier apart. Rakhimova's better wins grade out much higher than Selekhmeteva's best result of the season. The read there is that Rakhimova tends to lose to players who are simply better than her, while Selekhmeteva tends to lose to her peers, often in deciders she does not close.
The history that keeps it off the card
Here is the problem. Selekhmeteva has beaten Rakhimova twice, and recently. She won 6-2 7-5 on hard court about six months ago, then beat her again on grass in June. Same surface as today's match, inside the last year, is exactly the kind of head to head evidence that should carry real weight, and it says Rakhimova has lost to this specific opponent in this specific way already.
Selekhmeteva also is not the same player who showed up to those matches with a leaking serve. Her US Open round one line, five double faults and an 89% hold rate, is her cleanest serving day in a month. And the version of her season the file initially missed matters here too: a hidden 125 level run in Philadelphia in late August that went to the semifinal, meaning her true form since late August is closer to 4-1 on hard than to a slump.
None of that flips the pick to Selekhmeteva. Schedule adjusted, her hard court edge and Rakhimova's hard court edge wash out to roughly the same level once you account for the quality of who they beat to get there. But it is enough documented recent risk, on the exact surface being played today, that T Money will not stake money on the side its own model technically prefers.
The number that matters
-161 is not an unreasonable price for the side with the better serve profile and the higher ceiling wins. It is under T Money's price ceiling for a play, so the number itself is not what is stopping this one. What stops it is that the required edge over that price is not there once a real, same-surface, recent head to head loss sits on the other side of the ledger. A price this short needs a cleaner case than a fundamentals read that a documented recent result already contradicts twice.
The verdict
This is a PASS. No official action, no side backed with money. T Money's serve and return numbers lean Rakhimova at -161, but Selekhmeteva has beaten her twice in the last several months, most recently on hard court, and that kind of recent same-surface history is not something the model overrides just because the underlying serve trends look better today. Watch it, do not bet it.