When the Spreadsheet Comes Back Empty: The Ethical Line of a Basketball Analyst

Late one November night, I sat in front of two screens in a small apartment in Queens. One screen replayed a game between Zadar and a mid-table Italian side, slowed to 0.25 speed. The other was my spreadsheet: fourteen variables tracking ball movement, deflection positions, and the efficiency of every pick-and-roll type. I ran the aggregation macro and waited. The spreadsheet returned exactly one line — blank. The error was not in the data entry: the tracking system on that independent streaming platform had failed in the third minute, and for the remaining thirty-seven, every number I needed simply did not exist. I sat there, fingers hovering over the keyboard, and recognized a familiar temptation knocking: fill the gap with something that sounds plausible.

A low-tier game on a small screen, and I see a whole universe in motion — but this time the universe was silent. For someone who analyzes tactics for a living, silence is the most uncomfortable state. Not because I have nothing to say, but because my profession pays me to always have something to say.

Modern basketball analysis runs on an assumption almost nobody questions: data always exists, you just need the right tools to extract it. Since 2026, when SportVU was installed in every NBA arena, each step a player takes became a coordinate; by 2026, Second Spectrum took over tracking and turned every possession into thousands of data points per second. In Europe, EuroLeague and the Spanish league adopted optical tracking later, but by 2026 most top competitions had gone digital.

In the American market where I work, data is an industry. Tracking companies sell access to teams, broadcasters, and statistical platforms alike. Every season a new metric is born, along with a demand to explain it to the public. But most of the content labeled data-driven that readers consume comes with no methodology. People cite a number the way they cite a proverb, without asking where it came from, how many games it covers, or who computed it.

Yet most of the world's basketball does not live in top competitions. There are hundreds of second- and third-tier leagues, regional competitions, games streamed on a single independent platform with one camera in the stands. That is where I started, and it is where data constantly disappears. This asymmetry creates a paradox: the less data there is, the easier it becomes for an analyst to fill the gaps with guesswork, because no one has enough information to disprove the opposite.

Low-tier games have no optical tracking. No SportVU, no Second Spectrum. They have one camera, one commentator, and sometimes a frame blurred by arena lighting. For me, that was the best laboratory. Because when numbers are not ready-made, you are forced to learn to read the game with your eyes first, and only then look for a way to measure it. That order matters.

Based on my experience following games, I spent nearly two years — from 2026 to 2026 — collecting video of four hundred games from EuroLeague, the VTB United League, and the Spanish league to build a defensive database. The original goal was ambitious: fourteen variables per possession. The reality: more than a third of the games had to be discarded for missing camera angles, missing shot clocks, or cameras simply placed wrong. Four hundred sounds impressive, but the number actually usable was closer to two hundred sixty. I learned the first lesson of the trade: volume is not reliability.

I am not calling for anyone to abandon data. Quite the opposite. But there is a fundamental difference between two kinds of analysis that look identical on the surface.

The first begins with a question: what data do I have, and what does it allow me to conclude. The second begins with a conclusion: I believe this, now let me find numbers to prove it. Both end in an article with charts. Only one of them is real analysis.

In the Zadar game I mentioned, if the tracking system had worked, I would have measured the seven-beat ball-movement cycle the home side used to attack the weak corner of a 2-3 zone. I saw it with my own eyes. I rewound it twelve times. But with an empty spreadsheet, I had two honest choices: say that I saw this by eye but had not measured it, or stay silent. The third option — writing a fluent analysis with estimated numbers dressed as precise ones — is the option I had to actively refuse.

The biggest blind spot in the analysis industry is not a shortage of data, but the pressure to fill missing data with a story that sounds reasonable. That pressure does not come from laziness. It comes from an incentive structure: algorithms favor content with clear conclusions, readers want answers, and an article that says it does not know rarely spreads.

Every tactical system is born from a detail everyone saw but no one noticed. The problem is that when that detail is not recorded, people start to imagine it.

Defense is the final language; only those patient enough to listen to four hundred games in a row can translate it. I learned that while rewatching games of teams no one remembers. They have no stars, no highlights, but they have a system. And that system only appears when you are willing to sit long enough.

I once spent three weeks analyzing how France used an inverted ball-screen with Rudy Gobert in the Tokyo 2026 Olympic final. The mechanism is subtle: they did not use it to create scoring space, but to force the American defense to choose between two equally bad situations — step up or drop back. But what stands out more is what I could not prove: I counted seventeen specific possessions, yet no tracking system recorded the reaction time of the opposing center. I had to time it by hand, frame by frame, and admit an error margin of at least a tenth of a second. A tenth of a second — in elite basketball, that is the distance between a perfect defensive possession and an open shot.

The blind spot is not on the diagram; it lies between two movements that no one measures.

Numbers are not fireworks. They are witnesses. And an honest witness must be allowed to say they do not remember clearly, or that they were not there.

The irony is that the sports-analysis industry is moving the opposite way. The boom in predictive models and composite metrics creates the feeling that everything can be quantified. But precisely because models grow more complex, readers are less able to verify the inputs. A beautiful chart with clean axes can hide a dataset of only three games. No one checks. No one asks about sample size.

I do not watch a game as a spectator; I read it as a text of deliberate mistakes. And in that text, white space is part of the content too. A good analyst is not the one who fills every blank, but the one who knows which blanks are allowed to remain.

In 2026, when Brittney Griner was released after two hundred ninety-four days detained in Russia, I was interning at a sports data analytics firm in New York. The whole office talked about geopolitics and the future of foreign players. I could not stop thinking about how all our models suddenly became meaningless in the face of a humanitarian crisis. I wrote a piece about the limits of pure analysis, and leadership called it off-topic. But that was the lesson: some gaps data can never fill, and admitting that is a professional skill, not a failure.

So when my spreadsheet returned a blank line that November night, I did what I believed was right: I logged the date, logged that the tracking system failed in the third minute, logged what my eyes saw but my hands had not measured. Then I shut the machine down.

Basketball is a sport of gaps — gaps between two movements, between two decisions. And perhaps the analyst's real task is not to fill every gap with a number, but to show readers which gaps are worth looking into themselves. The question for this season is not which team has the most data, but who among us is brave enough to say they do not yet know.