The Fourteen-Page Blank Report: How Basketball Analytics Is Fooling Itself With Empty Cells

On November 9, at 6:12 a.m. New York time, a fourteen-page PDF landed in the internal chat of an NBA team's analytics department. Fourteen pages. Not a single line of data.

Page one, "Advancement": N/A. Page two, "Execution": N/A. Page five, "Salary Structure": N/A. Page nine, "Ripple Map": N/A. All nine sections of the report — tactical analysis, player data, cap operations, league landscape, rules and governance, locker room, risk, media narrative, industry ripple — were filled with the same two characters.

At 7:40, the file was forwarded to three people. At 9:05, it appeared in a meeting slide. At 11:30, a television commentator referenced that "internal report" as though it contained a finding. By noon, a sports outlet published a story with a confident headline.

Nobody in that chain lied. They simply did the one thing this industry does extremely well: turned emptiness into an opinion.

I have followed professional basketball since 2026, write the NBA column for VnExpress, and have handled thousands of scouting reports. That fourteen-page blank was the most frightening document I have ever held — not because it was wrong, but because it was so right that nobody dared read it aloud.

The NBA's data machine today looks nothing like the one I started with. In the 2026-14 season, STATS installed SportVU motion-tracking cameras in every arena, recording the position of every player and the ball twenty-five times per second. From the 2026-21 season, Hawk-Eye took over at even higher resolution. A single regular-season game now generates hundreds of thousands of raw data points: the distance on every shot, running speed, landing angles, average possession time per player.

Which makes a blank report all the more strange. When you have too much data, having no data stops being a technical accident. It becomes a signal.

Modern analysis runs through two stages. Stage one extracts raw events from the source: who played whom, what the score was, who is injured, how many years remain on a contract. Stage two is where tactical questions get asked and judgments get made. If stage one returns an empty list, stage two has no raw material. A decent analyst stops and says: "I have nothing to say yet."

But stopping is the one thing nobody pays you to do.

That fourteen-page report was not the product of ignorance. It was the product of an honest process forced to pretend it had a conclusion. And in basketball, that is precisely the moment the most beautiful lies are born.

Let us separate two things the industry deliberately conflates: no data and no finding. These are entirely different epistemic states, yet on paper they are written identically.

No finding means you ran the numbers, cross-checked three sources, rewatched the film, and concluded nothing unusual was there. That is a valuable result. No data means the pipeline clogged, the file failed, or the source article never reached you. That is a failure.

An honest blank report is worth more than a report stuffed with fabricated numbers — because the blank tells you the pipeline is broken, while the fabricated one tells you everything is fine.

And here is what I want you to remember: basketball analytics will not collapse from a lack of data. It will collapse because too many people are willing to keep writing after the data runs out.

Take an example I have watched dozens of times. On the night of May 28, 2026, Game 7 of the Western Conference Finals, the Houston Rockets hosted the Golden State Warriors. Houston led by eleven in the second half. Then they missed twenty-seven consecutive three-pointers — the longest such streak in playoff history. For the game, Houston shot 7-of-44 from deep. Golden State won 101-92.

The data that night was complete to the point of perfection. Every shot was logged, every gap measured. Yet what got written the next morning was a story about "composure" and "champion pedigree." Nobody wrote that Houston lost to a probability distribution — 44 open threes, only 7 made, when their season average on open threes was 36.2 percent. Had they shot their average, they win.

That is data forgotten despite being present. The fourteen-page blank is data forgotten despite being absent. Two different diseases, one symptom.

In 2026-16, the Warriors won 73 games, breaking the 2026-96 Chicago Bulls record. Stephen Curry became the only unanimous MVP in league history, averaging 30.1 points on 50.4 percent from the field, 45.4 percent from three, and 90.8 percent from the line — the only 50-45-90 season ever recorded at that volume. The regular-season data said this was the best team ever measured.

Then they lost Game 7 of the Finals at home, 89-93, to Cleveland.

After that night, a new school of analysis sprouted: "regular-season numbers are meaningless." I disagree, and this is where I break from the crowd. Regular-season numbers are not meaningless. They simply measure something different from what people want measured. Seventy-three wins measure the ability to optimize across 82 discrete nights. Game 7 measures the ability to endure 48 minutes when everything funnels into one shot.

The fault is not in the data. The fault is in the reader who thinks reading the data means finishing the story.

By the same logic, look at Russell Westbrook in 2026-17. He became the first player since Oscar Robertson to average a triple-double for a full season: 31.6 points, 10.7 rebounds, 10.4 assists. The league split into two camps. One called it the greatest individual achievement of the decade. The other called it stat-padding art.

Both camps were lazy. The dataset was sufficient to answer the real question: where did Westbrook collect his rebounds, in which situations, and what was his team's win rate when he grabbed twelve or more versus eight or fewer? That is a stage-two question. Most of the debate stopped at stage one — the number 31.6 — and the two sides threw belief at each other.

I have stood in exactly that position, and I was wrong.

In 2026, at fifty-five, I posted an analysis video on a sports social platform. Pep Guardiola's Manchester City averaged 72 percent possession in the Champions League. I said Monaco would eliminate them because that sleepwalking football created no variance. City won the first leg 5-3 at home, then lost 3-1 away, drawing 6-6 on aggregate and going out on the away-goals rule. I was right about the outcome. But the route I took was wrong: I used a single metric — possession share — to judge a team whose full matches I had not watched that season.

That was the first time I understood that a correct hot take can still be a bad hot take.

A year later I was in Russia for the 2026 World Cup and watched Spain hold nearly 80 percent of the ball, complete over a thousand passes against Russia in the round of sixteen, and lose 3-4 on penalties. I immediately wrote "The Revolt Against Possession Football." It was shared more than 50,000 times. Yet to this day I ask myself: had Spain won that match, would I have written the opposite piece? And if the answer is yes, was I analyzing or decorating?

That is why I started paying attention to blank reports.

In 2026, before the European Championship, I posted a short status: Italy will win because they know how to play slowly. The community laughed, because Roberto Mancini's side was pressing high and attacking fast. But my argument rested not on feeling. It rested on a verifiable observation: after a year of football without crowds, the team that controlled tempo best would survive a congested schedule while others broke down physically. Italy conceded exactly four goals in seven matches and beat Spain and England. My piece was cited on Sky Sport Italia.

The difference between that and 2026 was this: I stated what I was measuring, and I admitted what I could not measure. I had no data on individual players' mental states. I had no medical data. I had one hypothesis and a way to test it.

Possession is an illusion, the scoreline is the naked truth — but that naked truth only means something if you know how many empty cells produced it.

That is when I started thinking about a new metric, one I believe will be the industry's keyword over the next few seasons: the null rate.

What is the null rate? For any analytical report, you count what percentage of its conclusions rest on real data, what percentage rest on inference, and what percentage rest on nothing at all yet were still written in a confident voice. Published, that figure would unseat half the experts currently on television.

I applied it to myself. Across my last twenty pieces for VnExpress, four had a null rate above 30 percent — meaning nearly a third of the argument had no direct data behind it. I am not ashamed of that. I am ashamed that I never counted before.

Now let me rebut myself.

There is a possibility I am entirely wrong, and it is not small: that blank reports are not a disease but the system's self-defense. An NBA team cannot publicly say "we do not know whether this player will recover" — that sentence erodes the value of an asset. A broadcaster cannot air thirty minutes of silence. A sports outlet cannot run the headline "We have nothing to say today."

If so, the null rate is not an intellectual failure. It is the operating cost of an entertainment industry that needs content every day, event or no event. You cannot sell advertising with honesty.

I might also be wrong elsewhere: I may be overestimating the public. Ordinary fans do not buy tickets to read data tables. They buy tickets to have a story to tell at the bar. And "Houston missed 27 threes because of a probability distribution" is a far worse story than "Houston lost its nerve."

The third blind spot, and the one that worries me most: medical data. In the NBA, an injured player is announced as "out 2-4 weeks" with no disclosure of what the injury is, how severe, or whether surgery occurred. The team's analytics staff knows. Media does not. Fans do not. Yet hundreds of analyses about that player's "return timeline" still get written — all with null rates above 50 percent.

I have no solution for that. I have only a habit: whenever I write about an injured player, I state in the piece that I have no medical data, and that everything after is inference.

That is the minimum standard of honesty. And it is worth it.

Back to the fourteen-page PDF on the morning of November 9. What made me write this was not that it existed. What made me write it was that it still produced a headline.

The regular season is at the stage where every team has a reason to lie. A team chasing a playoff berth needs to say it is improving. A team sliding needs to say it is "building." A team with a star nearing free agency needs to say everything is fine. In that environment, a blank report is the only thing that cannot be bought — and therefore the only thing that gets buried.

My prediction, testable within eighteen months: at least one NBA team will publicly disclose a "data coverage" metric for its internal reports, or a major sports outlet will begin labeling analyses built on incomplete data. If that does not happen, you have every right to come back and challenge me.

And if it does, remember who said it first.

Data does not score, but data is quietly rewriting history. And the most uncomfortable numbers saint in the room is not the one with the most numbers. It is the one willing to say: this part is empty.

Basketball does not need more hot takes. It needs more people who know when to stay silent — and who know how to speak up when an empty cell appears in the middle of a fully populated dataset.