The Empty Cell: The Discipline of Saying “Insufficient Data” in Tennis Analysis

The Empty Cell At row seven of the press tribune, I keep an old habit from my years in the stats room: printing the match data sheet on paper. That evening the match ended after nineteen minutes. A nineteen-year-old retired in the fourth game with an ankle injury, the racket still wrapped in fresh grip. The sheet in my hand carried six stat lines, and five of them read “N/A”. First-serve percentage: insufficient data. Points won on first serve: insufficient data. Net points won, unforced errors, break-point conversion: all blank. One line had a number, and that line was match duration. The man beside me, a male sports editor, looked at the sheet and asked: “So what do we write?” Forty minutes later his headline read: “The fighting spirit of the nineteen-year-old boy.” I do not blame him. I understand the pressure: the feeling that an empty cell is a professional failure, that silence is laziness, that if you write nothing someone else will write it in your place, and that person will bend it further. Tennis is measured more thoroughly than almost any other sport. Every serve carries its own clock: twenty-five seconds across the ATP circuit. Every ball is logged to the millimetre, and from the 2026 season Wimbledon replaced its entire line-judging crew with electronic line calling. Every result converts into points along an architecture so precise it turns arid: a Grand Slam title is worth 2,000 points, the runner-up 1,300, a semi-final 780, a quarter-final 430. A Masters 1000 title is worth 1,000 points; an ATP 250 title is worth 250. The rankings run on a rolling 52-week window. Every point earned dissolves automatically exactly one year later, quietly, without notice, without farewell. That structure breeds a second industry layered on top of the competitive one: the interpretation industry. And that industry runs on one unwritten rule — an empty cell must be filled before the deadline. I entered the trade from the smallest role: fact-checker at an American sports magazine. The job taught me that most of the time a fact-checker does exactly one thing: says “I don’t know”, and says it firmly enough that it cannot be bent into a headline. Professional sports analysis runs on two tiers. Tier one extracts facts: who, when, how many, on which surface, for how long. Tier two is analysis proper: tactics, form, risk, outlook. Tier two depends entirely on tier one. When tier one returns a null payload, the correct output of tier two can only be a null conclusion: insufficient information to assess. Any other conclusion is built out of air. A null result is not one thing. There are at least three kinds, and they carry three different meanings. The first is absence: the match has not been played, the player has not walked on court, the data does not yet exist. The second is thin sampling: the match happened but was too short, or the player has too few matches on that surface for any comparison to carry weight. The third is contradiction: two sources disagree and nothing is strong enough to adjudicate. Collapsing all three into a single “N/A” is the fastest way to lose information. Take thin sampling. A young player wins eight of his first twelve tour-level matches on grass. A 67% win rate reads well. But across twelve matches the confidence interval around that figure is wide enough to cover both a genuine grass specialist and a player who has never learned to move low. On grass the ball skids and stays low, so a good serve can mask every weakness in lateral movement. Twelve matches cannot separate the two. My own experience tracking matches across three consecutive grass seasons suggests roughly two hundred tour-level matches are needed before a sample means anything, and very few players ever accumulate two hundred on grass. The same problem surfaces in the rankings. Because points roll off a 52-week window, a ranking position is a photograph, not a curve. A top-ten player can be standing on a very thin points structure: two Grand Slam semi-finals, 780 points each, together accounting for most of the total. If both events fall in the same stretch of the following season, the defence pressure arrives at once. In the press room people call that a form crisis. In the stats room people call it a calendar. There is a larger blind spot in how points defence is told: it is usually described as a law of physics, when what actually decides results is the body. A player grinding through three consecutive weeks on three different surfaces switches from clay to grass inside twelve days, and across that stretch ankles, knees and Achilles tendons absorb very different load profiles. Injuries in the surface-transition window are the hardest to forecast, and the least represented in analytical tables. Meanwhile the governing bodies are erasing the interpretive gaps around the match itself. The 25-second serve clock shortens the waiting. Off-court coaching, trialled widely across ATP and WTA events from the 2026 season and later extended to the Grand Slams, turns tactical exchange into an official part of the match. Electronic line calling removes disputes over the chalk. Each change fills a gap while creating a new class of data the industry has not yet learned to read: how often a player glances at the box, when a coach signals, the rhythm of the pauses. So why do the empty cells keep getting filled? Because in this industry a filled cell is always rewarded and an empty one is always punished. Headlines with data travel further than headlines admitting the data is not there yet. When the incentive sits on the side of filling, the cost lands on the reader, slowly, over years. Readers are never told they are consuming a very weak prediction, so they store it as confirmed fact, and are surprised when it dissolves. The mechanism repeats at every level of the sport. When Carlos Alcaraz and Jannik Sinner split almost all the major titles across a few seasons, the media instantly had a label ready for an entire era. On the WTA side, Iga Świątek and Aryna Sabalenka are routinely slotted into a manufactured rivalry, while each keeps a very different schedule and a very different preferred surface. With Novak Djokovic, the longest career in the sport’s history has been misread through short 52-week windows. The label arrives first, the data second, and by the time the data lands nobody remembers the old label. I once sat in a production room in Moscow and heard a male colleague laugh when I asked a footballer whether he felt sad after winning. He said women like to turn everything into poetry. The piano in Moscow taught me that winning is not the only thing worth recording. Years later, making a documentary series about empty stadiums during the pandemic, I understood that what he called poetry was a different analytical move: paying attention to what was left blank. The man who says “women don’t understand football” has verified nothing about the woman in front of him; he is only describing his fear that an empty cell might survive unfilled. They told me I don’t understand football, but I understand what it does not say. In recent years, much of what tennis does not say has been the most important part: an undiagnosed injury, a withdrawal with no press conference, a sample too small that still gets built into a story. An empty court turns out to have its own sound of memory. On a printed sheet, an empty cell has its own language too: it says the writer once had a chance to stop, and chose not to. When an analytical tier returns a null payload, that null is a product, not a system error. The question is whether this industry has the nerve to print a sheet with five blank lines, and to let readers decide that sometimes the most honest thing a stat sheet can do is stay silent.