Tennis and the Data Gap Behind Confident Analyses

“I was wrong about school football data, and that was the most accurate finding I have ever had.” I still open my analyses with that line, even when I write about tennis. It reminds me that a conclusion only deserves trust when data stands behind it. Yet look at how tennis is described every week, and I see the opposite: the story comes first, and the data — if it exists at all — arrives later, usually only to serve a story written in advance. A player wins four matches in a row at a Masters 1000, and the media instantly builds an “heir” narrative. But his first-serve points won can still sit below the average of the top twenty. Nobody checks. Because the story is already compelling enough not to need checking. Men's tennis is going through the clearest generational transition in two decades. Roger Federer stopped at 20 Grand Slams, Rafael Nadal at 22, Novak Djokovic at 24. A power vacuum was left behind, and the media needs a clean story to fill it. Carlos Alcaraz won his first Grand Slam at 19. Jannik Sinner rose as a steady force. The two names were quickly framed as a “new era.” But the ball does not move according to a media script. The ATP ranking uses a 52-week window and counts only the best 18 results. That means a player can hold a high position on points banked six months ago while his actual form fell long ago. A “ranking position” does not measure current form; it measures recent history. Once you understand this, many headlines peel away from their glossy coat. This is where process data separates from outcome data. Outcomes are wins, titles, rankings. Process is first-serve points won, return points won, break-point conversion, unforced errors. A player can win a title on a few impulsive points while the process numbers say he played below standard all tournament. When I cross-reference data across matches, a pattern often appears. The “comebacks” highlighted in the press come with an unusually low break-point conversion from the opponent. In other words, the win did not come from the winner's brilliance but from the loser's collapse. The media calls it nerve. Data calls it conditional luck. Based on my experience tracking matches, a tennis analysis only deserves trust when it states which process metrics it relies on. No metrics, no analysis — only a feeling presented as a firm conclusion. Here the counterintuitive view appears. Most people believe a bad analysis fails at the model — wrong algorithm, wrong calculation. My experience says otherwise: the failure is at the input. A perfect model running on empty data still returns empty results, except it presents them with a confident face. I once sat in front of such a table: full of frames, sections, and headings, but not a single number. The frightening part is that it still read smoothly like a real report. I believe in data, but I believe more in the mistakes data cannot measure. A stat sheet can tell you a player won 78% of first-serve points. It cannot tell you what state he served in: a nagging shoulder injury, an endorsement deal about to expire, or a coaching split three weeks earlier. Those variables are not in the sheet, but they decide its contents. The gap between market expectation and objective reality is where process data is most valuable. The market prices a player on reputation. Process metrics price him on what he actually does on court. And the story does not stop at content. It moves money. A player pushed by the media as “the heir” gets more wildcards, more sponsorship deals, and most importantly a higher seeding at big events. There, avoiding a strong opponent in the first round is worth several rounds. Ranking is not just a number on a board; it is money, schedule, opportunity. When a wrong analysis distorts how a player is seen, that distortion flows straight into his commercial value. The tennis industry sells both the matches and the stories about them. Tennis analysis often uses a lack of data as a crutch for certainty. The less information there is, the easier it becomes to assert. In my “debate room,” I saw it repeat: the strongest opinions always came from the people who read the fewest numbers. Tennis media does not lack experts. It lacks people willing to say “I do not have enough data to conclude.” For fans, the filter is simple. When you read a take about your favourite player, ask which process metrics it relies on. If the answer is “form,” “nerve,” or “class” with no number attached, you are reading a story labelled as analysis. Next season there will again be a young player called “the heir.” I will again open his process numbers first, check his return points won, and only then decide whether to believe. And if I am wrong, I will write it again, as always.