Serve Data on Hard Courts: Why Aces Don't Decide the Winner

On the centre court of a hard-court Grand Slam, the winning player sometimes hits only 12 aces while the loser hits 21. The stands remember the explosive serves, but my spreadsheet remembers a different column: second-serve points. Across nine years of watching professional tennis, from near-empty qualifying courts to packed finals, one thing keeps recurring: the flashiest serve is rarely the serve that decides the winner. The serve is the only shot a player fully controls. That is why analysts have spent decades measuring it: first-serve percentage, first-serve points won, ace count, double faults. Those traditional metrics hide more than they reveal. Once point-by-point data became widely available, through Hawkeye and advanced analytics platforms, we could finally measure what actually makes the difference: second-serve points and the shot immediately after the serve, what analysts call the "serve+1". A player can hit a monstrous first serve, but if he loses 40% of second-serve points, opponents only need to put the ball in play to take control. On medium-paced hard courts, the surface most common at the majors in Australia and the United States, the serve remains a weapon but no longer an absolute shield. The ball bounces true and higher than on grass, giving the returner more time. That factor turns second-serve points into a metric that separates levels. I track the serve data of leading players on Australian hard courts across many rounds. What catches my attention is not the average ace count, but the gap between first-serve points won and second-serve points won. The smaller the gap, the harder that player is to break in decisive games. Imagine two players both landing 65% of first serves. Player A wins 78% of first-serve points and 55% of second-serve points. Player B wins 74% of first-serve points but 62% of second-serve points. On aces and first serves, Player A looks the better server. But in hold probability, Player B edges ahead, especially in tense games, because he does not collapse when the first serve misses. Many fans, and some commentators, overlook this. They count aces like goals, when tennis does not score that way. An ace is not worth more than a second serve followed by a forehand into the corner; both are simply one point. The difference lies in how often it repeats and when it happens. The data reveals another paradox. Big servers often win a high share of tie-breaks, but they also get dragged into more tie-breaks, because they lean on the serve so heavily that they rarely create break chances. Conversely, a player with steady second-serve points and a strong return generates more break points, and that is the shorter path to winning a set. One more metric I always track is return points won. On hard courts, title-winning players usually win more than 30% of return points. That threshold separates the contenders from the rest of the draw more clearly than any ace count. Whenever rules, surfaces or playing conditions change, I rebuild a before-and-after model. That method helps me separate signal from noise, something tennis analysis often lacks. It is also why I distrust conclusions drawn from only a handful of matches. I still remember watching a five-set quarter-final where the higher-rated player hit 24 aces but converted only two break points. His opponent served more modestly yet held 64% of second-serve points and turned the match around. My spreadsheet showed what the stands had not yet seen: the man losing the ace column was winning the column that mattered. That observation can lead to a hasty conclusion: steady second serves equal victory. Not true. Data does not lie; it is the person reading it who makes excuses. A high second-serve points won rate does not automatically mean a title. A player can hold serve well because of a weak early-round opponent, then collapse against an elite returner in the quarter-finals. The denominator changes, and the pretty rate loses its meaning. Nine years ago I built a prediction model on historical data and believed in it. Reality taught me a lesson still worth holding: in 2026 I learned that a 95% probability still has a 5% that laughs. In tennis, that 5% lives in a missed second serve in a tie-break, a slip of the foot, a time violation. The first data rebellion was never meant to overthrow anyone, only to prove that numbers deserve to be heard. So when I read serve data, I always set two variables beside it: opponent quality and scoreboard context. An impressive second-serve rate against a player outside the top 80 says little next to a modest rate against a top-10 opponent. And a second-serve point won at 5-5 matters far more than the same rate in a game led 5-1. When the next hard-court swing begins, I will not count aces. I will track second-serve points won, break points created and break-point conversion, three metrics that reveal more about a player. If a big-serving star suddenly drops in the second-serve column while still firing aces, that is the earliest signal of a fall in the next round. Who will win the next title? The answer likely does not lie in the prettiest serves, but in the column the stands rarely look at.