We read every US Open qualifying opener: the first-set winner won 84% of matches — and exactly half the deciders

We read every first-round qualifying match the 2026 US Open had finished — all of them, both draws, not a highlight reel — and asked one question of each: did the player who won the first set win the match? As of the frozen capture at 21:08 UTC on 26 August 2026, ESPN's tournament scoreboard listed the qualifying first round as complete in both singles draws (the capture's mixed-doubles qualifying competitions are excluded) — 64 men's matches and 64 women's, 128 in all. Three ended in retirement (two men, one woman); we set those aside, leaving 125 that were played to a natural finish. That is the population, frozen to one capture so every figure below is re-derivable from it. In 105 of the 125, the player who won the opening set went on to win the match — 84.0%. Read alone, that number sells the first set as decisive. It is not. The edge lives entirely in straight-set matches. 85 of the 125 (68%) ended in two sets, and in a best-of-three you cannot win in straight sets without winning the first — so the first-set winner took all 85 by definition. Strip those out and look only at the 40 matches that reached a third set: the first-set winner won 20 and lost 20 — no net advantage in this selected subset. The split holds across both draws — men's deciders went 12-10 to the first-set winner, women's went 8-10 against, netting exactly 20-20. The shapes are ordinary. Grigor Dimitrov opened 6-0, 6-3 over Lorenzo Giustino and never gave the third set a chance. Cristian Garin took the first set, lost the second, and survived a third-set tiebreak 7-6 (5), 4-6, 7-6 (11-9) — a first-set winner who held on. Jordan Lee did the opposite to Mark Lajal, dropping the opener 4-6 before winning 6-3, 6-3; Genaro Olivieri beat Guy Den Ouden 5-7, 6-4, 7-5 the same way. Every one of those third sets was its own contest.

The following is analysis, not fact. The 84% figure and the 20-20 figure are not in tension — they are the same fact seen at two altitudes. Winning the first set correlates with winning the match because winning the first set easily is what a straight-set win looks like from the front ^mmodeled. The aggregate number is mostly counting straight-set wins. What it hides is that the moment an opponent answers — the moment a match becomes a three-setter — the first-set winner had no net match-win advantage: 20 wins, 20 losses, in a bucket of 40. This is the gap between how the first set feels and what it does. A broadcast frames the opening set as momentum, a lead banked. The complete read says the lead's edge is concentrated in the straight-set wins. Once the match is close enough to need a decider, the players have shown themselves near-equal, and the first set-winner held no net advantage among these 40 deciders ^mmodeled. The scoreboard remembers who won the set. That is a conditional-probability story, not a claim that momentum is a myth. The decomposition is mechanical: all 85 straight-set matches went to the first-set winner by definition, while among the 40 deciders the first-set winner went 20-20. A single televised set shows one of those outcomes; the whole round shows the split.

The honest counter is selection, and it is strong. The 40 deciders are not a random sample — they are precisely the matches close enough to reach a third set, which means we have already conditioned on two near-equal players ^mmodeled. Selection into three-set matches makes a more even result plausible; it does not require an exact 50-50, and it warns against generalizing the 20-20 beyond this subset. So the 20-20 may say more about which matches go three sets than about whether the first set "stops mattering," and a reader should resist reading it as momentum debunked ^mmodeled. The buckets are also small — 40 deciders, split 22 and 18 by draw — and this is one round of one tournament's qualifying, not the main draw and not a law of tennis. Qualifying fields are lower-ranked and arguably more even than a Grand Slam main draw, which could widen the coin-flip zone; we cannot say the same holds when Alcaraz opens against a wildcard ^sspeculative. We scoped every figure to exactly the 125 completed matches in the capture and set the 3 retirements aside, rather than sell a partial or padded count as the finding.

From one match, a first set won is a lead you trust. Across the whole round the mechanics are plain: all 85 straight-set matches went to the first-set winner by definition, and among the 40 that went the distance the first-set winner went 20-20 — no net advantage in this selected subset. Both readings describe the same capture.