The Number Saint and the Empty-Data Crisis: When the NBA Miscounts Itself

The NBA season has ended, and here I sit, at 64, in a small apartment overlooking a New York street, wondering how many stat sheets I have read in 48 years. The answer is probably millions. But there is something I have never confessed to anyone: most of them were garbage.

That sounds harsh coming from a man who has spent his whole life calling himself a "number saint." But precisely because I am a number saint, I have a duty to tell the truth. The paradox of modern basketball sits right here: never have we had so much data, and never have we understood so little about the game.

I remember it vividly, that April night in 2026, when Russell Westbrook completed a triple-double with 20 points, 20 rebounds and 21 assists for the Oklahoma City Thunder against the Los Angeles Lakers. Social media exploded. The press called it "the game of the century." And I, a man long branded a "traitor" by the basketball faithful, sat silent.

Because I knew something the box score never said: possession is an illusion, the final score is the only naked truth. In basketball, "possession" goes by another name — it is usage rate and raw counting numbers. And it is deceiving all of us.

Stats don't score, but stats are quietly rewriting history. I have written that line hundreds of times. But only after watching an entire basketball industry built on empty numbers did I grasp its full weight.

From feel to data, and the price paid

To understand how we got here, we have to rewind a few decades.

In the early 2000s, the NBA was still a game of feel. Scouts watched players with their eyes. Coaches trusted instinct. Sportswriters like me wrote from instinct honed over the years. Back then, people had only a handful of basic numbers: points, rebounds, assists, shooting percentage. That was all there was.

Then Daryl Morey arrived at the Houston Rockets, and with him a generation that believed everything on the floor could be measured. They counted threes, counted corner attempts, calculated true shooting. They proved the mid-range two was the worst shot in basketball, and an entire generation of players had to change how they played because of it.

That was a real revolution. But every revolution has winners and losers — and it spawned a new class: the number peddlers.

In 2026, the NBA installed SportVU tracking cameras in every arena. Suddenly we knew how many meters each player ran, how many sprints, how fast he shot, how far the nearest defender was when he released. In 2026, Second Spectrum replaced it with far more detailed data.

In theory, this was paradise for a number saint like me. In reality, it is the hell of truth. Because when you have too much data, you can prove anything. And that is exactly what happens every night, across every media front.

Based on my experience watching games, empty numbers fall into six categories. Each is a way modern basketball deceives itself.

Category one: The triple-double and the number-hunting industry

Westbrook was not the first, but he is the perfect embodiment of the disease.

For three straight seasons from 2026 to 2026, Westbrook averaged a triple-double. In 2026-17: 31.6 points, 10.7 rebounds, 10.4 assists per game — unseen since Oscar Robertson. He won MVP. The world believed it was witnessing greatness.

But look at the substance. In 2026-17, Westbrook shot 34.3% from three and 45.2% from two — among the lowest of any star. He used 41.7% of his team's possessions while on the floor, the highest mark in NBA history at the time. In other words, he did not play basketball within a system; he was the system, and that system lost.

OKC was eliminated in the first round three years running: 2026 by the Rockets, 2026 by the Jazz, 2026 by the Blazers. Westbrook had triple-doubles; his team had summer vacation.

The point is not that Westbrook was bad. He is one of the finest athletes of his generation. The point is how we measured him. The triple-double turns a complex game into a counting game, and in that game anyone can become a "champion" without winning anything at all.

Worse, it built an ecosystem. When rebounds become currency, centers start ceding rebounds to guards. When assists become currency, players start passing to teammates in the worst positions just to log an assist. I have sat in arenas and watched it — a player deliberately not shooting, waiting for a teammate, only to make the number on the scoreboard prettier.

Category two: Plus/minus — the laziest metric

If the triple-double is a blatant con, plus/minus is a subtle one.

Plus/minus measures a team's scoring margin while a player is on the floor. It sounds perfect: it captures contributions that never show up in the box score. But it has a fatal flaw: it depends entirely on the people around him.

A bench player who only plays minutes alongside the team's superstar will post a sky-high plus/minus while doing nothing himself. Conversely, a star forced to carry a weak bench will post a negative plus/minus, even when he is the only reason the team is alive.

In the 2026-24 season, I spent three weeks analyzing the plus/minus of individual players across different lineups. The result stunned me: the same player's plus/minus swung by as much as 15 points per 100 possessions depending on who stood beside him. That number speaks to the lineup, not the player.

Yet every day, hundreds of articles use plus/minus to conclude who is good and who is bad. It is intellectual laziness disguised as analysis.

Category three: Pace inflation

This is the most dangerous kind of emptiness, because it is invisible to the naked eye.

Pace is the number of possessions a team runs in 48 minutes. In the early 2000s, the NBA played slow, averaging around 90. By the 2026-24 season, that number had passed 99. That means nearly 10 more possessions per game per team than two decades ago.

What does that mean for the numbers? A player averaging 25 points in 2026 is not necessarily better than one averaging 20 in 2026. He simply gets more chances. If you do not adjust for pace, you are comparing apples to oranges.

I recalculated a famous case by hand. A star averaged 27 points per game in 2026-24. Sounds impressive. Adjusted to 2026 pace, it drops to about 24.5. Still good, but no longer the "historic explosion" the press called it.

Conversely, the legends of the 1990s and 2000s are unfairly underrated for playing in a slow era. Michael Jordan averaged 30.1 points for his career — in a league nearly 10% slower than today's. Adjusted, that is roughly 33 points per game in the modern NBA. No one in today's league reaches that.

Category four: Garbage stats on bad teams

There is a paradox few want to admit: scoring on a bad team is far easier than on a good one.

On a terrible team, a player is free to do anything. No pressure to win, no tactical discipline, no one competing for the ball. The result is numbers so pretty they deceive. I call them "garbage stats" — they exist, they are real, but they are meaningless.

Over 48 years, I have seen plenty of players average 20 points on a 60-loss team, get a big contract, then vanish when they reach an ambitious club. Why? Because on a good team, he must share the ball, defend, absorb pressure — things the stat sheet cannot measure.

The only way to separate garbage stats from real ones is to place them in the context of winning. A player scoring 18 on a champion is worth far more than one scoring 24 on a last-place team. But the box score never tells you that.

Category five: Small samples and invented stories

This is the kind of emptiness I myself once committed, and I am ashamed of it.

After 10 games, a player shoots 45% from three. The press calls him a "rising sharpshooter." After 82 games, his true rate is 34%. The shot never changed; only the sample did.

Basketball is a sport of large samples. A season is 82 games. A career is thousands of possessions. Any conclusion drawn from 10 or 20 games is a fabrication dressed in numbers.

I remember once, in 2026, I nearly wrote a piece praising a rookie after an eight-game surge. Fortunately, an older colleague — the man who taught me everything — stopped me. "Wait 40 games," he said. After 40 games, the rookie returned to his true average. I nearly lost the credibility of a number saint over a small sample.

Category six: Distorted tracking data

This is the newest and subtlest kind of emptiness.

When the NBA installed tracking cameras, we got metrics that seemed absolutely objective: distance traveled, sprint count, top speed. They were packaged as "effort metrics." And the press loves them, because they are easy to grasp and sound scientific.

But distance traveled does not measure effort. It measures distance traveled. A player running aimlessly, pulled out of position, hopelessly chasing the ball will log more distance than a player standing in the right spot, cutting off a passing lane without needing to run much.

I verified this myself. In a 2026-23 game, I tracked two defensive guards. One ran 4.2 km, the other 3.1 km. On the stat sheet, the first "worked harder." But rewatching the tape, the second defended far better — he read the game, stood in the right place, and never left a gap. The first simply ran more because he kept getting beaten.

High distance traveled can signal diligence. It can also signal being lost. The stat sheet cannot tell the two apart.

Category seven: The transfer market — where empty numbers are sold for real money

I must speak plainly about this, because it touches the wallet of an entire industry.

Every transfer window, hundreds of numbers are thrown out: transfer fees, weekly wages, signing bonuses, release clauses. They sound like objective facts. But most are leaked by player agents for a single purpose: to drive up the price.

I have maintained a network of agents and club staff for years. I know how they operate. A number is released not to inform but to apply pressure. When a paper reports that player X is valued at 80 million, that figure did not come from the club's balance sheet. It came from a phone call at midnight.

Agents are the largest hidden cost in modern basketball. The noise they create distorts the entire market. An average player is inflated into a star through three weeks of rumors. A club pays double out of fear of losing him. And in the end, the fans pay for that distortion.

In 2026, when basketball paused for the pandemic, I received information from an agent I knew: Pierre-Emerick Aubameyang would extend his contract rather than move to another club, despite the frozen transfer market. I broke the news with an argument: football in empty stadiums is pure commerce, stripped of community spirit. The piece reached two million views. But what I learned was not how to break news. It was how a single exclusive number can shape an entire market — and how it can be abused without responsibility.

Category eight: Injuries and secrecy — when medical data is hidden

This is the kind of emptiness I hate most, because it concerns human health.

Clubs disclose injuries selectively. They say only what benefits the value of their assets. A mildly injured player is described as "seriously injured" when the club wants to lower his price. A seriously injured player is described as "nothing to worry about" when the club wants to keep his market value up.

The result is that fans and media are blind. We read "ankle injury" and do not know whether it is a grade-one or grade-three sprain. We see a player return and do not know what percentage of fitness he is playing at. We make judgments about a human being's career based on information supplied by the very people who pay him.

Over 48 years, I have seen too many careers destroyed by premature returns, driven by the pressure of numbers — revenue, tickets, TV rights. Medical secrecy does not protect the player. It protects the balance sheet.

So what makes a number real?

After demolishing nearly every metric the press worships, I owe the fairest question: so which numbers can be trusted?

My answer is: no number is trustworthy on its own. A number has value only when it comes with three things — context, opponent, and sample.

Context: in what system does this player operate, with whom around him, under what pressure? Opponent: against whom did he do it — the best defense in the league or a team that has given up? Sample: over how long — 10 games or 10 seasons?

When I analyze a player, I never use a single number. I use at least three cross-checked data sources, and I always ask: does this number tell the same story my eyes saw? If the data and my eyes conflict, I do not rush to a conclusion. I go find a fourth source.

That is the discipline of a real number saint. And that is precisely what the class of "fake number saints" is destroying.

Where I might be wrong

It is time I was fair to myself, and to those I criticize.

There is a chance I have been wrong: perhaps the problem is not the numbers, but how we use them. Numbers do not lie; people lie with numbers. When I blame "empty numbers," I may be blaming the tool instead of the one holding it.

And there is an uglier truth: the human eye also lies. I once believed in the "eye test" as supreme truth, until I rewatched tape and realized I had missed half of what happened on the floor. My instinct, after 48 years, can still be wrong. Numbers, however empty, are at least consistent.

Most dangerous, perhaps, is me. I built my brand on "exclusive sources" — agents, club staff, anonymous scouts. But an unnamed source is an unverifiable source. If I am not careful, I can become the very creator of the empty numbers I condemn.

Here is what I learned from the biggest shock of my life. In 2026, I argued that Pep Guardiola's Manchester City would be eliminated by Monaco in the Champions League because of a soporific possession game. I was right about the result — City lost 5-3 on aggregate despite over 1,000 completed passes. But I was wrong about the reason. The problem was not possession; it was that City possessed without a final spark.

Since then, I have learned that criticizing a metric does not mean denying it. Sometimes the old metric is still correct — we just have not understood it deeply enough. I must point out specifically where the old metric still holds, rather than smashing everything like an irresponsible rebel.

I must also admit my exclusive numbers were not always right. Once, an agent told me a star would change clubs in the summer window. I believed it and wrote it. The deal never happened. The lesson: an exclusive source, however reliable, is still a single data point. And a single data point, as I have said throughout this piece, is never enough.

What I believe, after all this

I did not write this piece to bury data. I wrote it to reclaim it.

The future of basketball analysis lies not in how many more metrics we add, but in whether we have the courage to say "I don't know." A real number saint is not the one with the most data. It is the one who knows when a number is lying, and when it is telling the truth.

Basketball is a human sport — of fear, of pressure, of sleepless nights before a big game. No stat sheet can measure the heartbeat of a player standing at the free-throw line with the game on the line. And none ever will. Our task is not to turn basketball into an equation, but to use mathematics to understand people better.

If you are reading this and wondering whether the numbers you see every day can be trusted — the answer is: only when you know how to ask questions. A number says nothing on its own. You are the one who gives it a voice. And in a world full of empty data, the ability to ask the right question may be the most important skill a basketball fan can learn.

As for me, at 64, I still sit before the screen every night. Still taking notes. Still doubting. Because basketball, to me, has never been a game of numbers. It is a game of truth — and truth, like everything precious, must be protected from those who counterfeit it.