The Sinner–Alcaraz Race and the Discipline of Data Verification at Tennis's Summit

In the fourth set of a Masters 1000 semifinal this season, I sat close enough to the hard court to hear the ball compress against the surface. The eventual winner landed only 58% of his first serves — a figure that, skimmed off a stat sheet, would instantly suggest he was playing below par. But when I isolated the 42 most important points — break points, tie-breaks and games at 4-4 or beyond — his first-serve points-won rate climbed to nearly 76%. The same line of data, two opposing stories. That is why, after twenty-eight years of watching tennis through data, I have learned that the most dangerous thing is not a shortage of numbers, but trusting a single one. The context of this annual season makes the data question more urgent than ever. After the Big Three era closed with Novak Djokovic's final campaigns, men's tennis entered a phase dominated by two names in almost every discussion: Jannik Sinner and Carlos Alcaraz. The pressure of a two-man race turns even minor matches into historical tests. Media need stories, sponsors need characters, fans need a summit to compare against Roger Federer, Rafael Nadal and Djokovic. In that atmosphere, numbers get stretched to serve a pre-existing conclusion rather than to lead one. Having once built my own metric for penalty shootouts after a hard lesson, I watch this season with caution. Each week my tracking system collects hundreds of variables: first-serve points won, second-serve points won, return points won, break-point conversion and a dynamic Elo rating. But the thing I watch most closely is not in that table. It is the question: which conclusion are these numbers serving, and would that conclusion still stand if one link were removed? The first thing I do with any analysis of Sinner is separate his serve data from his return data, then split it further by surface. On hard courts his serve pattern is fairly stable: a high first-serve points-won rate, but more notable is the pace of the shot immediately after the serve. When that pace exceeds his own average, his probability of winning the game rises sharply. This is a familiar finding, but it restates a principle I always follow: conclusions must come after layers of evidence are stacked, not after one flashy metric. With Alcaraz the picture is more complex. He is the kind of player conventional data under-describes. His winner and unforced-error counts swing widely between tournaments, partly because he accepts high risk at moments when most players choose the safe option. Looking only at his unforced-error rate, one might conclude he plays carelessly. But set that number beside his return position, and you see he often stands deeper and shifted to the left compared with the norm — a tactical choice that forces more movement in exchange for a wider return angle. Here the unforced errors are the price of an advantage, not a sign of decline. The annual season also carries a variable media usually ignore: ranking points to defend. Each player enters a tournament with a points window to protect, and that pressure differs by position. A top-5 player defends thousands of points at major events, while a world No. 20 needs only a few hundred to avoid sliding. When I watch a match between two players from these groups, I always ask: is the result reflecting form, or reflecting the pressure they carry onto court? In most cases I examine, the answer is both — and no single metric fully separates the two. Injury is the next variable. At thirty, a player recovers far more slowly than at twenty, and this shows most clearly in the fourth and fifth sets. I have compared one player's first-serve points-won rate across the first two sets and the last two across many tournaments. The average gap is only a few percentage points, but it appears consistently enough for me to treat it as signal, not noise. Still, that signal is only trustworthy when cross-checked against the schedule: if the player contested a three-set marathon two days earlier, the late-match dip is no longer evidence of poor fitness but a consequence of scheduling. Attributing it to a single cause would be intellectual laziness. The truth lies deep beneath the stat sheet, where headlines never reach. With electronic line-calling in tennis, I take the same stance. One wrongly called point cannot explain an entire defeat, but neither should it be dismissed as harmless. What interests me is how data is disclosed: when a decision is controversial, fans in the stadium are often given too little context to understand why it was made. Transparency, in many cases, is merely a slogan repeated at press conferences. An empty stadium does not make the result wrong; it only strips away our illusion that we understood what was happening. The most counterintuitive thing this season's data has shown me is the relationship between ace counts and win rates. We are often told a big serve is the decisive weapon. But when I track the ace distribution of top-10 players across tournaments, the correlation between average aces per match and win rate is weaker than many assume. Some players serve enormously yet win at modest rates, because their second-serve points-won rate is low. An ace does not build an empire. It is one point, inside a far more complex chain. Even using Elo for predictions has limits. Elo describes relative strength based on past results, but it does not know whether a player has a sore wrist today. So when I offer a judgment, I always attach a probability range rather than a declaration. 'Roughly a 70% chance this player wins' is a sentence I can defend. 'This player will certainly win' is a sentence I will not write, however clear the match may look. This is where the lesson from my own data mishap more than two decades ago still holds. That year, I rushed a conclusion about a player based on a defensive metric, entirely ignoring the new tactical context his coaching staff imposed on him. The metric was not wrong. My interpretation of it was. Since then, every analysis of mine reserves a section for the role variable — the tactical system, how the player is used, and the limits of the data I am holding. Looking ahead to the rest of the annual season, the signal I track most is not in the rankings, but in the consistency between how a player wins and how he loses. When a player wins only on serve, that streak is usually more fragile than it looks. Fans watch with their eyes; I watch with a probability distribution — and the distribution always reminds me that the right answer, in tennis as in data, tends to lie where we have not yet looked.