When the Data Is Empty: Basketball Analysis and the Temptation of Baseless Conclusions

On the night of November twelfth, I sat in front of a screen with a spreadsheet open. The left column held player names. The right column held efficiency metrics. Between them lay a long, flat stretch of white, without a single number. The data-collection program I had installed returned an empty result, on the very night I had to go live with a forty-minute podcast about the game that had just ended. I could have made things up. That is the easiest thing in the world. I could have said this player had an astonishing three-point efficiency, that team's defense played with unprecedented intensity, the coach made a brilliant tactical adjustment in the fourth quarter. Not one of the thousands of people listening live could open a book and verify anything in that moment. I could have spoken smoothly, confidently, fluently, and the broadcast would have been perfect. But I did not. I opened the podcast with the exact sentence I had been thinking for the three hours before: Tonight I have no data, so tonight I will say less than usual. That was the moment I understood that my profession, the one I have pursued for nine years, holds a temptation greater than any other. The temptation to speak when you have nothing in your hands. Data does not lie, but the people who read it do. And on the day I saw an empty spreadsheet, I realized the worst reader of data might be me, if I let the fear of silence defeat my discipline. Before going further, I need to rebuild the context. Over the past decade, the business of basketball analysis has changed entirely in nature. In the old days, a post-game commentary was written with feeling, with memory, with the glow of a beautiful play. The writer remembered that team A scored more, remembered that player B hit a three at the decisive moment, and that was enough to write a piece of praise. No one asked what the real efficiency was, no one asked how many quality chances the team created, no one asked how many points the defense conceded per hundred possessions. Then the data era arrived. Professional teams began hiring analysts, data engineers, people who could turn every possession into a line of numbers. Fans grew used to terms like true shooting percentage, pace, defensive rating per hundred possessions. Sports media, once in need of a writer who could tell a story, now needed people who could both tell a story and read numbers. I belong to the first generation of that wave. I studied economics, I learned to read tables, and I brought that skill into basketball. But here is the flip side no one talks about. When the whole industry shifts to chasing data, the pressure to produce content multiplies. Every game, big or small, must have analysis. Every night, whether anything happened or not, must have an article. Every player, even one who plays seven minutes, must have a story. And when the pressure for content exceeds the volume of real data, people begin doing something no one taught them: they begin to fill the blanks. That is when I saw the strange parallel between a failed data pipeline and a failed writer. Both return the same result. A set full of conclusions that sound very convincing, but beneath which there is not a single data point to hold them up. I have seen this many times. I have seen thousand-word analyses of a player the author had never watched for five full games. I have seen conclusions about a team's collapse built on three games at the start of a season, that period when any serious analyst knows three-point rates are unstable and every conclusion sits in the noise. I have seen praise for a rookie after he scored twenty points, without anyone noticing he needed twenty-five shots to do it. That is not analysis. That is decorated blank space. And this is what I want to go deepest into, because it is the core of the matter. In basketball, as in any probability-driven sport, incomplete information is not a rare situation. It is the default state. One game only gives you forty-eight minutes. A ten-game stretch is still not enough to tell a good player from a lucky one. Sports scientists have shown it takes hundreds of attempts before a player's three-point rate begins to approach its true value. Yet the media concludes after fifteen attempts. Which means the professional, technically speaking, constantly works in a state of missing data. The question is not how to always have enough data, because that is impossible. The real question is: when data is not enough, what do you choose? Do you choose silence or do you choose to speak? I choose the following principle, and I apply it to every piece I write: I do not pass judgment on a player without at least five games to verify it. Five games is not a sacred number. It is the minimum threshold so that I do not deceive myself. Sometimes I need twenty games, sometimes a full season is still not enough. But five games is the line below which I am obliged to tell the reader that I do not yet have grounds. This is harder than people think. Hard because the entire system incentivizes you to do the opposite. Audiences want answers, not questions. Broadcasters want headlines, not caution. Algorithms want decisive content, not hesitation. A piece saying there is not enough data to conclude will get far less engagement than a piece asserting something forcefully, even if that something is wrong. This is a paradox anyone in the content business in the twenty-first century must live with. I think about the failure of the giants. Giants do not collapse because they are weak, but because they forget they were once small. And I realized that forgetting does not only happen to teams. It happens to those who write about them. When a writer becomes famous, when the follower count rises, when each piece is read by hundreds of thousands, that writer begins to forget the feeling of the early days. The feeling of having to open a book to check, to rewatch the tape three times, to hesitate before asserting. Fame creates a kind of false confidence, and that confidence is the most dangerous blank space of all. I personally paid for this lesson. In twenty twenty-one, at the Olympics held in Tokyo after a one-year postponement, Japan's men's national basketball team entered the tournament with enormous expectations. It was the first time in history the country had two players competing in the American professional league at once. I wrote an analysis staking my reputation on the team reaching the quarterfinals. I relied on the glow, the fame, the confidence of someone who had followed this basketball nation since she was a sixteen-year-old girl. The result everyone knows. The team lost all three group games, one by twenty points. And when I sat down to reread my notes, I found the reason. I had ignored the defensive data. I was so focused on the offensive glow that I did not look at the key number: the team's defensive rating at the time was very low, reflecting a system that could not withstand world-class opponents. That number was there, right in my dataset. I did not read it, because the story I was telling was more appealing than the number. I wrote a long public self-reckoning, admitting the mistake and re-analyzing the opponents' defensive systems. But more important than that reckoning was a new framework I built afterward. From then on, I never make predictions based on a player's fame. I assess every team through three pillars: offense, defense, and physical condition. Those three pillars, combined, give me a truer picture than any glow. And I drew a lesson I think is the most important for anyone in this profession: fame is only yesterday's story. Today's data is the truth. But wait. I must push back against myself here, because if I only say that, I fall into another trap, the trap of the one who believes data is everything. Data is not everything. Data is a tool, and like any tool, it has limits. There are things in basketball that numbers cannot measure. Leadership in the locker room. The ability to stay calm under pressure in the final minute. The bond of a group over a long season. Those things exist, they affect results, and they are in no spreadsheet. So my principle is not blind data-first. It is: when data is enough, let data lead. When data is not enough, let data stay silent. And when data is not enough, do not fill the blank with stories that sound good. This is where I want to address how people bend numbers to excuse failure. I have seen it too many times. A team loses, and someone finds a number to say they actually played well. They scored more in the paint, they had more assists, they controlled the ball more. Those numbers may be real, but they are meaningless detached from context. Controlling the ball a lot with harmless sideways passes says only one thing: the team held the ball a lot and did not know what to do with it. This is why I always say possession percentage is the most deceptive metric in sports. It is a beautiful number. It makes a loss look like an injustice. It lets people say we deserved to win. But basketball, like any sport, does not reward deserving. It rewards those who put the ball in the basket more. And this is the most counterintuitive part. I argue the greatest enemy of truth in sports is not wrong numbers. The greatest enemy is right numbers placed in the wrong context. Because a wrong number is easy to catch. But a right number, carefully presented to lead the reader to a wrong conclusion, is far more dangerous. It has the power of truth, but serves a lie. I have seen this in the analysis of Japan's youth basketball league, the market I care most about. This is where the country's mainstream media barely looks, as if gazing at a white field of snow and assuming nothing lies beneath. But I found gold in Japan's youth league, where everyone else saw only snow. In twenty seventeen, when I was just sixteen, I happened to watch a youth-league game and was drawn to a player of unusual height for his age. I began building a spreadsheet tracking his scoring efficiency and defensive impact across fifteen games. When he moved to the American college league, I held a detailed data trove no Japanese sports outlet had. But what I learned was not that I had predicted a talent. What I learned was that the market was severely lacking in quantitative analysis of young players, and when data is missing, people fill it with prejudice. People said a player could not succeed because he came from a small school. People said he was only tall but not skilled. Those conclusions were based on no data at all, they were based on feeling. And feeling, in these cases, is often wrong. Japan taught me that treasure is always there, you just need the patience to dig. But it also taught me the opposite: you can dig forever and find nothing, and in that case, the honest thing to do is put down the shovel and say you have found nothing. The temptation of a contrarian is always to deliver a shocking conclusion. I know this, because I am that kind of person. I like counterintuitive angles. I like pointing out that the majority is wrong. I like the feeling of standing in a cheering crowd and saying you are looking at the wrong place. But that pleasure, unchecked, turns me into a flip-flopper. Someone who argues just to argue, not because the data forces him to. So I set a rule for myself: every shocking conclusion must have a clear evidence threshold behind it. If I want to say a team is collapsing despite winning, I must point to the number that shows it. If I want to say a star is declining despite scoring a lot, I must show where those points come from. Without an evidence threshold, my counterintuition is just a performance. And here is where I return to that empty spreadsheet that night. Because of all the lessons from nine years in this trade, the hardest is not learning to read numbers. The hardest is learning to say I do not know. People praise decisiveness. People reward confidence. Across the entire modern content ecosystem, from social media to video platforms, what rises to the top is what is certain, forceful, unhesitating. A strong opinion draws more engagement than a hesitation. A decisive conclusion spreads faster than a question. And this creates a perverse system of incentives: it rewards those who say the most, not those who are most right. In such a system, the honest professional is disadvantaged. They say less. They admit limits. They let a blank space be shown publicly, instead of filling it with a lie that sounds good. And in the short term, they rank behind the loud. But I believe in the long term the opposite is true. Because the only thing a content creator accumulates over time, the only thing that cannot be stolen, is credibility. And credibility is not built by saying things that sound convincing. It is built by saying things that are true, and by admitting what you do not yet know. I have watched very famous writers collapse after a single false piece. Not because they were wrong, but because people discovered they had fabricated. In my industry, being wrong can be forgiven. Being false cannot. A person can misjudge a game, a player, a trend, and still keep readers' respect if he is honest. But once people know he is willing to fill blanks with numbers that are not real, every piece he writes thereafter becomes worthless. That is why I choose to keep a certain hesitation in me. I choose to keep the blank spaces. I choose, whenever I lack enough data, to tell the reader I lack enough data, rather than planting in their ear a conclusion I cannot defend. The story of the empty spreadsheet is not a story about a technical error. It is a story about a moral choice. Because every time we fill a blank with a baseless statement, we do not only deceive the reader. We deceive ourselves. We build a habit, and that habit, over time, erodes the very thing we need most to do this work: the ability to see the truth. The failure of a giant is a gift to the observer. But that gift has value only if the observer is honest enough to open it carefully, rather than tearing it apart and scattering the scraps of paper everywhere. So what comes next? This is where I want to pose a few questions to myself and to those in my trade. As analytical tools grow stronger, as artificial intelligence can write smooth-sounding analysis from a thin dataset, the line between real analysis and fake analysis will blur more than ever. Readers will find it increasingly hard to tell a data-based piece from one generated by carefully decorated blank space. In such a world, the value of the genuine professional will not lie in the ability to write brilliant sentences. It will lie in the ability to say something no one wants to hear: I do not know. It will lie in the discipline of holding the blank spaces. And it will lie in the honesty brave enough to look at an empty spreadsheet and say that before I tell you about it, I must open it one more time. When the whole world stops, I choose to start from zero. That zero may be the zero of views, of engagement, of praise. But it is also the zero of self-deception. And between those two zeros, I choose the second. Every time. For the rest of my career.