Ben Shelton's semifinal contract closed at 11:03 p.m. ET on Friday, and Kalshi listed the US Open men's final, Alexander Zverev against Shelton, at 11:25 p.m., 22 minutes later. By 2:53 a.m. ET on September 12, 2026, the match market had traded 494,000 contracts, and 430,171 of them, 87 percent, were on the Shelton side. Since midnight the board has moved about a cent on each leg, toward the volume, which is the point: the biggest buying spree in this market's short life walked Shelton's offer up two cents and moved his bid not at all. Zverev, the top seed and reigning French Open champion, is 56¢ bid, 57¢ ask, a 56 to 57 percent favorite read as a probability. Shelton, the No. 8 seed, is 43¢ bid, 45¢ ask. Sunday's final is at 2 p.m. ET, per the ATP Tour's schedule note, on Arthur Ashe Stadium, and the two men have never played at a major.
So the price says Zverev and the volume says Shelton, and my read before the evidence is that both are telling the truth about different things. The board is pricing the record: Zverev has played Shelton five times and won five, ten of the 11 sets, all four tiebreaks. The crowd is pricing the story. Shelton is the first American man in this final since Taylor Fritz in 2024, chasing the first American men's title here since Andy Roddick in 2003. He is the first Black American man in a US Open final since Arthur Ashe in 1972, and he was 0-3 against Carlos Alcaraz until the early hours of Wednesday. Whether a record should outweigh a fortnight is the argument, and the tape has the two seconds, at 1:27 and 1:43 a.m., when the story bought in size and the record did not blink. There is also a one-cent gap between this market and the tournament board that a Shelton buyer should know about, and I will get to it.
How Friday Made The Final
The two semifinals were played back to back on Ashe, and each one told the market something different about Sunday.
Zverev beat Karen Khachanov 6-3, 7-6(7), 7-6(6) in the afternoon, and the scoreline flatters him. Khachanov led the second-set tiebreak 5-3, and Zverev took six of the last eight points to win it 9-7, the last of them a 34-shot rally. In the third-set tiebreak Khachanov led 5-2 and held two set points at 6-4; both ended in forehand errors, and Zverev won it 8-6. Khachanov won more points than Zverev in the second set, 44 to 43, and lost the set. Zverev's serve carried him: 15 aces, 79 percent of points won behind the first serve, 37 winners against 34 unforced errors. "Straight-set win, but anything but straightforward in a way," he said. "He played incredibly well." It is his sixth major final and his third in a row, after the French Open title in June and the Wimbledon final, and ESPN has him at 24-2 in Grand Slam matches this season, 52 tour-level wins by its count, two ahead of the season table below, and 19 hours and 38 minutes on court to reach Sunday. Kalshi's market on the match had watched all of it: Zverev's contract printed its Friday low, 78¢, at 5:03 p.m., and closed at 99¢ when the market shut at 6:17. It is also his first US Open final since 2020, which is a match this page will come back to.
Shelton's night was the one the crowd came for. He beat Frances Tiafoe 4-6, 6-3, 6-3, 7-5, from a first set in which he made 14 unforced errors to Tiafoe's three, and won 14 of the next 20 games. The adjustment was on the return: after Tiafoe won nine of 13 second-serve points in the first set, he won 11 of 39 the rest of the way. Shelton finished with 20 aces and won 77 percent of the points behind his first serve; one stretch of second-set serves went 125, 144, 139 and 140 mph. The match turned on the last game of the third set, a 15-minute game that ended on a Tiafoe forehand error, and the fourth ended, per Yahoo's live blog, on a Tiafoe double fault. Tennis Majors counted 47 winners to 27, and five breaks from 11 chances against two from seven. It also noted a serve at 2-1 in the third set that the stadium gun read at 158 mph and Hawk-Eye read at 144, which leaves the number disputed rather than a record. "Ben played well tonight," Tiafoe said. "Served really well tonight, which made it really tough." The match market told the same story in three prints: Shelton's contract, 72¢ to 73¢ all afternoon, printed its Friday low of 49¢ at 8:27 p.m., was 90¢ by 10 p.m., and closed at 99¢ at 11:03.
The records are real, and they are the story the crowd is buying. Shelton is the first Black American man in this final since Ashe in 1972, the first in any major final since MaliVai Washington at Wimbledon in 1996, the youngest American man in a major final since Roddick in 2005, and the first American man in a US Open final since Fritz two years ago. No American man has won a major since Roddick here in 2003. "I'm going for it here, and I'm going to leave it all on the court," Shelton told the stadium. "Let's do it for the USA." Zverev, asked about it afterward, was dry: "Hopefully, I can not make the American dream come true on Sunday."
Two wins, one survived and one earned. The market watched both and priced the final within half an hour. Whether it priced it right starts with the record it has to answer for.
Five Matches, Eleven Sets, Ten To Zverev
Zverev leads the head-to-head 5-0, a count ESPN and the ATP Tour's own record both carry, all five since August 2024, and the shape of the five matters more than the count. Here they are.
| Date | Event | Round | Surface | Winner | Score |
|---|---|---|---|---|---|
| Aug. 2024 | Cincinnati | QF | Hard | Zverev | 3-6, 7-6, 7-5 |
| Apr. 2025 | Munich | Final | Clay | Zverev | 6-2, 6-4 |
| Jun. 2025 | Stuttgart | SF | Grass | Zverev | 7-6, 7-6 |
| Aug. 2025 | Cincinnati | QF | Hard | Zverev | 6-2, 6-2 |
| Nov. 2025 | ATP Finals | Round robin | Indoor hard | Zverev | 6-3, 7-6(6) |
The five tour-level meetings, per Sports Betting Dime's head-to-head record; every one was best-of-three. They have not played in 2026 and have never played at a major.
The row to sit with is the first one, because it is the only set Shelton has ever taken from Zverev and the only match that reached a third set, and it was on a hard court in August. Everything after it was straight sets. Three of the five were on hard courts, Shelton is 0-3 on them, and four of the 11 sets went to a tiebreak, all four to Zverev; where that leads is the next section. The counter, and it is the whole underdog case in one line, is that Shelton was 0-3 against Alcaraz before he beat him here in five sets, at 3:33 a.m. on Wednesday, the latest finish in US Open history. A record against a player is a record until the night it is not.
The 2020 final is the other piece of history the price has to carry. Zverev led Dominic Thiem by two sets in that match and lost it 6-2, 6-4, 4-6, 3-6, 6-7(6), and until June he was 0-3 in major finals. He is 1-4 in them now, and the one win came in Paris, on clay, in five sets against Flavio Cobolli. Shelton is 0-0. One man has a record of losing these, the other has no record at all, and the board has to decide which of those is worth more.
Ten sets out of 11 says the record is the whole case. The serve numbers say why it might not travel.
The Serve Numbers, And Why The Sets End In Tiebreaks
Both men hold, which is why the four tiebreaks matter, and the season figures make the serving closer than the head-to-head does. The numbers are from tennisstats.com's 2026 season tables for each player.
| 2026 Season | Ben Shelton | Alexander Zverev |
|---|---|---|
| Aces Per Match | 9.63 | 8.74 |
| First Serves In | 67.3% | 70.3% |
| Service Points Won | 71% | 71% |
| Service Games Held | 91% | 88% |
| Match Record | 38-13 | 50-12 |
Season-to-date figures across all surfaces, as posted on tennisstats.com at the time of writing; the site is a third-party aggregator, not the ATP's official stats feed, and its match records lag the tour's by a few days.
Look at the third and fourth lines. The two men win the same share of their service points, 71 percent each, and Shelton turns that into slightly more holds, 91 percent to 88. Two servers who hold nine games in ten do not break each other often, so their sets get settled by one break or by a tiebreak, and that is exactly what the head-to-head has looked like: four of the 11 sets went to a tiebreak, Zverev won all four, at 7-6 in Cincinnati in 2024, 7-6 twice in Stuttgart, and 7-6(6) at the ATP Finals, and on Friday he won two more against Khachanov, from 5-3 down and from 5-2 down. Shelton's own tiebreak week is split: he lost the first-set tiebreak to Alcaraz on Tuesday night and won the match tiebreak that ended it, 10-7. Friday's serving was a wash, Zverev with 15 aces and 79 percent of first-serve points won, Shelton with 20 and 77 percent. If Sunday's sets are as tight as those numbers say they should be, the record that matters is not the 5-0. It is the 4-0 in tiebreaks, and it runs the same way.
Zverev's own read of the man across the net was short. "Ben is just a big, powerful guy," he said. "Massive serve, massive groundstrokes." That is the honest description of both the 57 and the 43: the serve gets Shelton to the tiebreaks, and the record says what happens there. What the tape shows is how much money it took to decide that the record does not matter.
What Kalshi Is Pricing, And The Two Seconds At 1:27 And 1:43 A.M.
On Kalshi, each man is his own yes-or-no contract under the event ticker KXATPMATCH-26SEP13ZVESHE, paying $1.00 if he wins the final and nothing if he does not, with the price in cents standing in for probability; the plain-English guide to prediction markets covers the mechanics and the guide to betting sports on Kalshi covers placing an order. One sentence on the rules, because they matter for a final: the market settles only "after a ball has been played," a pre-match withdrawal resolves the contracts "to a fair price," and a postponement keeps the market open until the rescheduled match is done.
The Shelton leg opened at 11:26:40 p.m. with 969 contracts at 45¢, traded as high as 47¢ in its first ten minutes, and was 43¢ to 44¢ by 11:40 p.m. Through 1:26 a.m. it had done 53,786 contracts across 410 fills, steady, unremarkable, at 43¢ and 44¢. Then, at 1:27:36 a.m., in a single second, eight fills went through for 188,694 contracts: one of exactly 100,000 at 43¢, one of 70,311 at 43¢, 8,757 at 44¢, and one last 2,585 at 43¢, about $81,000 of Shelton in one print of the tape. Sixteen seconds later another 10,877 went at 44¢. At 1:43:43 a.m., in one more second, 26 fills took 151,967 contracts: 7,252 at 43¢, 127,949 at 44¢, two of them exactly 50,000, and 16,766 at 45¢, about $67,000 more. Seventeen minutes, 358,776 contracts, 83 percent of everything the Shelton contract has ever traded. Since 1:44 a.m. it has done 17,609. Of the 430,171 contracts on the leg, takers bought 429,500 and sold 671. Now the price. The bid was 42¢ to 43¢ before the sweeps and it is 43¢ now. The offers moved: the sweeps ate the 43s, then the 44s, then a slice of the 45s, and the makers reloaded a cent higher each time. Roughly $150,000 of Shelton, bought in two seconds, bought a two-cent walk up the ask and not one cent on the bid.
The Zverev leg is a different market. It opened at 11:27:37 p.m. at 60¢, and the first 3,139 contracts at that price were bought over nine minutes before sellers took it down: 59¢ by 11:39 p.m., then 58¢, where 35,712 of its 63,951 contracts have traded, and mostly 57¢ since about 1:45 a.m., where the biggest fill on the leg, 8,516 contracts, went through at 1:46:48 a.m. Buyers took 93 percent of the volume here too, on a leg that has averaged 57.7¢, but they paid 60, then 59, then 58, then 57; the sellers set the price and kept lowering it. The two legs say the same thing from opposite ends: the crowd wanted Shelton at any price the makers would quote, and it wanted Zverev only at a discount.
Here is the board at 2:53 a.m. ET on Saturday, the final's two legs beside the title contracts they now mirror.
| Market (Kalshi Event) | Contract | Bid | Ask | Contracts traded | Open interest |
|---|---|---|---|---|---|
| Zverev Vs Shelton, Final (Sun. 2 P.m. ET) | Alexander Zverev wins | 56¢ | 57¢ | 63,951 | 62,679 |
| Zverev Vs Shelton, Final (Sun. 2 P.m. ET) | Ben Shelton wins | 43¢ | 45¢ | 430,171 | 429,997 |
| US Open Men's Champion (KXATP-26USO) | Alexander Zverev | 56¢ | 57¢ | 2,447,250 | 1,559,744 |
| US Open Men's Champion (KXATP-26USO) | Ben Shelton | 43¢ | 44¢ | 3,291,897 | 2,236,429 |
Bid and ask are the live quotes on each yes contract as of September 12, 2026, at 2:53 a.m. ET; contracts traded and open interest are lifetime for each contract, all from one pull of Kalshi's public market data. With two names left, the two title midpoints sum to 100.
More live US Open boards: Sabalenka vs Rybakina, Saturday's women's final (Sabalenka 57¢ bid) · Shelton vs Tiafoe, Friday's semifinal (settled, Shelton in four)
The row that explains the crowd is Shelton's title line, because 62.8 percent of its 3.29 million lifetime contracts changed hands at a dime or less. On Tuesday night, half an hour before he played Alcaraz, its biggest fill of the tournament to that point took 46,096 contracts at 10¢; the biggest ever, 50,000 at 36¢, came Thursday morning. Everyone who bought at a dime is sitting on 43¢, and it is the same bet they made on Tuesday, with one round left to run.
The title board also shows what the market thought a Zverev final was worth before it knew the opponent, and it is the piece of Friday's tape I find most useful. Zverev's title contract was 49¢ to 50¢ at noon, before he played. It was 55¢ to 58¢ in the 4 p.m. half hour, 60¢ by 6:30 p.m., after his match market closed at 6:17. Then the second semifinal started, and while Tiafoe was ahead in it, Zverev's title line went up: 30,000 contracts bought at 61¢ at 8:26 p.m., 29,979 at 66¢ at 8:46 p.m., a high of 67¢ in that half hour. Shelton's semifinal contract printed its Friday low, 49¢, at 8:27 p.m. As Shelton took the match over, Zverev's line came back: 58¢ to 61¢ in the 9:30 half hour, with 17,736 contracts sold at 58¢ at 9:44 p.m., and at 11:01 p.m., two minutes before the semifinal closed, 15,934 sold at 57¢ and 7,942 at 56¢. Since midnight it has traded between 56¢ and 58¢. The tape records prices, not points, so I will not map fills to games, and a print in the middle of a semifinal is a blend over both possible opponents, weighted by whatever the score was that minute. But the direction is not in doubt: with the second semifinal in the balance, buyers paid up to 67¢ for a Zverev final, and once the opponent was Shelton the price was 57. The market thought Shelton was the harder final, and the crowd that has bought 430,000 contracts of him since the market opened Friday night agrees; they only disagree about how much harder.
Is 57¢ The Right Price?
The short version: the two-cent spread on Shelton is the tell, and the honest question is whether a man who is 0-5 against his opponent is a 45. The long version starts with what each side actually pays. On the Zverev leg the 56¢ bid holds 636 contracts, then 4,749 at 55¢ and 8,252 at 54¢; the 57¢ ask holds 115,347, with 224,503 more at 58¢. On the Shelton leg the 43¢ bid holds 7,524, then 82,215 at 42¢ and 82,961 at 41¢; there is nothing at 44¢ on either side, and the 45¢ ask holds 41,454 with 86,916 at 46¢. So the quote a Shelton buyer actually gets is 45, not 44, and a Shelton seller gets 43 for the first 7,500 and 42 after that. The two asks add to 102, which is the market's rent, and the shape of the books is the market-maker signature from every other page in this series: five- and six-figure size one cent either side of the touch on both legs, a thin 636 on Zverev's bid and an empty 44¢ on Shelton, where anything bigger than a small order slips a cent.
Then the fee. On 100 contracts of Zverev at 57¢, Kalshi's trading fee is 0.07 × 100 × 0.57 × 0.43, which is $1.72, so the position breaks even if he wins a little under 59 percent of the time. On 100 contracts of Shelton at 45¢, the ask that will actually fill, the fee is $1.74 and the break-even is a little under 47 percent. The fee schedule has the full formula. A reader who believes 5-0 makes Zverev a 60 is barely paid to say so at 57; a reader who believes the serve makes this a coin flip is getting Shelton at 45 and needs him to be a 47.
The read in one line: the quoted 43/45 on Shelton is 45 to buy in any size, 43 to sell for the first 7,500 contracts and 42 after that; the title board sells him a cent cheaper at 44; and the fee means a buyer at 45 needs him to win about 47 percent of the time to break even. Four hundred 30 thousand contracts of buying moved the offer two cents and the bid none.
Now the one-cent gap I promised. With two names left, the title contract and the match contract pay on the same event, and at 2:53 a.m. they were quoted at the same price on Zverev, 56/57. On Shelton they were not: 43/45 on the match market, 43/44 on the title board. A Shelton buyer gets a cent cheaper on the tournament contract, at 44¢, though only 9,391 contracts are offered there before the price is 45¢ again, with 106,583 behind it. There is one real difference between the two contracts, and it is the walkover rule. If the final never starts, the match contract resolves "to a fair price"; the title contract's own rules say that a participant who "forfeits, withdraws from consideration, or takes any official action that removes them from being able to win" resolves to No, which leaves the other man's contract to pay as the champion's. For anyone trading size, the title board is also where the depth lives, 1.6 million to 2.2 million contracts of open interest per name against 63,000 and 430,000 here.
The sportsbooks agree with the price and, for once, disagree with the crowd. ESPN's odds page, updated at 12:36 a.m. ET Saturday, reports that DK Sports' match line opened Friday night with Zverev at an implied 57 percent and Shelton at 48, margin included, and that its championship table, now a two-man market, had Zverev at an implied 60 percent and Shelton at 45 by 12:36 a.m. Those are two markets on the same question, and with the margin stripped out they put Shelton at about 46 and about 43; Kalshi's buyers were paying 45. The same page says Caesars and DraftKings will both lose on the tournament-winner market if Shelton wins; Zverev, it adds, "has not been an overly popular pick in the futures markets." The American money, in other words, is at the books too, and neither the book nor the exchange has moved its price toward it. The books' match lines are up and will thicken through Sunday morning; as OddsShopper's screen picks up the match, the live odds screen is where to compare their number across the major books against Kalshi's 57.
That leaves the question, and here is what 57 has to be wrong about for Shelton to be the buy. It has to be wrong that 5-0 travels: three of those matches were best-of-three on a hard court in August or in an indoor arena in November, none was at a major, and the man who compiled the 0-5 was also 0-3 against the defending champion until he beat him from a set down in a fifth-set tiebreak. And it has to be wrong that Zverev's tiebreak record is a skill and not a sample, because Friday's two, from 5-3 and 5-2 down, are the sort of thing that either happens to a player or gets done to him, and the market cannot tell which from six tiebreaks.
And here is what 45 has to be wrong about for the favorite to be the buy. It has to be wrong that the 2020 final is history, because a man who led that one by two sets and lost it is the same man who has since played four more major finals, won one, and is 24-2 at the majors this year. And it has to be wrong that Shelton's serve gets him to tiebreaks against a player who has never lost one to him, because if the sets are close, the record says who wins them.
If I had to name the one number I keep coming back to, it is the eleventh set: Shelton has taken one set from Zverev in two years, and the crowd has bought 430,000 contracts on him taking three of five in an afternoon. That is not a crowd that has misread the record. That is a crowd that has decided Tuesday night mattered more, and 45¢ is the price of finding out.
More on this: Shelton Vs Tiafoe Odds: 33,000 Contracts On The Underdog, Again · Zverev Vs Van De Zandschulp Odds: The Man Who Beat Alcaraz Here Is 7-To-1 · Khachanov Vs Blockx Odds: The Rib The Market Is Pricing
Where The Real Question Sits
The two seconds at 1:27 and 1:43 a.m. bought 340,661 contracts of Shelton for about $148,000. On Sunday afternoon that position pays $340,661 or it pays nothing, and there is no exit at cost in between: the 43¢ bid holds 7,524 contracts before the price is 42, so whoever owns those sweeps is not getting out at what they paid. That is the size of the bet the crowd has made on the story over the record, and the board has answered it by moving the offer two cents and the bid none. The only price on this page worth arguing with is the 45¢ a Shelton buyer really pays, and the case for it is the same case the 10¢ buyers made on Tuesday: that the serve is real, the records are old, and the man who beat Alcaraz at 3:33 a.m. is not the one who lost 6-2, 6-2 in Cincinnati. Sunday's final settles it the only way it can be settled, for $5.5 million to the winner and $2.8 million to the runner-up, and for the first American men's champion here in 23 years or for the man who stops him.
If pricing a single final is not your game, OddsShopper's free expert picks are the conventional route. For the wider board, the US Open odds comparison across Kalshi, Polymarket and the sportsbooks maps how the exchanges and the books priced the whole field, the Shelton-Tiafoe semifinal page and the Zverev-van de Zandschulp quarterfinal page show what the market thought earlier in the week, and the daily Kalshi picks hub grades the panel's calls in public.
Stay ahead of the markets.
Daily insights and expert picks on Kalshi, Polymarket, and what's moving markets.
Free forever. Unsubscribe anytime.
These are market prices and my own reading of them, not model estimates, not predictions of fact and not financial advice. Kalshi is a CFTC-regulated event-contract exchange (18+; availability varies by state). Prices quoted are the snapshot named above and will have moved by the time you read this.
