The Apron Era: Why Klay Thompson Left Golden State and the Spreadsheet Wrote It Two Years Early

On July 6, 2026, Klay Thompson — born in 2026, a four-time NBA champion with the Golden State Warriors — moved to the Dallas Mavericks via a sign-and-trade worth 50 million USD over three years. Most fans saw an emotional moment: a legend leaving after thirteen seasons. My spreadsheet had already logged this line in July 2026, when the Warriors extended Jordan Poole on a four-year, 128 million USD deal and Andrew Wiggins on a four-year, 109 million USD deal in the same week. At that point, their payroll had crossed a threshold no team can walk back. In the 2026-24 season, Golden State spent roughly 209 million USD in salary plus more than 170 million USD in progressive tax. Nearly 380 million USD for a season that ended in the play-in. Thompson left because the payroll ran out of room. The real story of the 2026 summer is not in Dallas; it sits in the words "second apron" inside the new collective bargaining agreement.

The 182.8 million USD line

From the 2026-24 season, the NBA and the players' association (NBPA) introduced two hard spending thresholds: the first apron at roughly 172 million USD and the second apron at roughly 182.8 million USD. Crossing the first apron costs a team its full mid-level exception and the right to take back more salary than it sends out in a trade. Crossing the second apron triggers harsher penalties: no aggregating multiple player salaries in one deal, no sending cash, no using trade exceptions carried over from prior years. Most severe of all, if a team stays above the second apron in two seasons within four years, its first-round pick is pushed to position 30.

The meaning is clear: the league wants to stop the superteam model. A wealthy team used to bundle three or four contracts for a star, pay the progressive tax, and keep spending. Now every step is locked. This structure turns the payroll from a tool into a court order: it decides which team may keep whom.

Based on my experience watching Warriors games across their dynasty, I noticed something the box score never shows. This team won through roster depth, and roster depth is the first thing the apron erodes. Between 2026 and 2026, they could pay mid-level money to five or six quality players. By 2026-24, they had only four reliable names, with the rest on minimum contracts.

Breaking down the numbers

Klay Thompson averaged 17.9 points per game in 2026-24 and shot 38.7 percent from three — the lowest mark of his career to that point. My spreadsheet did not stop there. Watching Warriors games that season, the more telling change was his role: he ran less, received the ball in tougher spots, and his rate of uncontested threes fell sharply.

Focusing only on individual decline misses the main point. Dallas signed Thompson because it needed an experienced shooter next to Luka Doncic and Kyrie Irving — two players who draw defensive attention. In that system, Thompson does not need to create his own shot; he only needs to stand in the corner and wait. It is simple addition: a player who once shot above 40 percent from three when freed from the ball, placed next to two bright satellites.

Compare the two salary structures. Dallas carried roughly 165 million USD in payroll in 2026-24, below the first apron, meaning it retained its full mid-level exception and the ability to execute a sign-and-trade. Golden State was the opposite: it could only complete a deal by sending out more salary than it took back, nearly impossible for a free agent like Thompson. What the media called "Dallas stealing Thompson" was really this: Dallas had room on the spreadsheet that Golden State no longer had.

The NBA's progressive tax is a tiered mechanism. Each 5 million USD above the threshold is taxed at a higher rate, and a team paying the tax in three of four seasons faces an added penalty — the repeater tax. Golden State fell into that group. Above the second apron, it faced two walls at once: cost and operations. In a season when broadcast revenue was fixed by the collective agreement, spending nearly double the league's average payroll stopped being strategy and became a gamble.

Why did the Warriors not keep him? Paying Thompson 15 to 17 million USD a year would have forced a choice: keep Thompson and lose the ability to sign a defensive forward, or let him go and preserve flexibility. In the apron era, flexibility is worth more than a 34-year-old shooter. Dallas accepted a risk of its own: Thompson turns 35 during the life of the deal, and his injury history — two major surgeries in 2026 and 2026 — attaches a probability to any three-year contract. But a mid-tier team can absorb physical risk, which sports science can manage, rather than structural risk, which it cannot.

I trust numbers more than people — because people know how to lie, while numbers only know how to be wrong. The Warriors' 2026-24 numbers were clear: with Thompson on the floor, the team's net rating was lower than when he sat — the first time in his career. That is evidence the system around him changed, not evidence he was useless. Golden State was no longer the team of moving passes; it was a team rebuilding around a playmaking forward.

The spreadsheet does not lie — only those too lazy to read it fool themselves. Dallas read the same data and saw a different value.

The contrarian angle

The story the media tells is tidy: loyalty is dead. Thompson was the last of the legendary core to leave, and fans assume the reason was money. That reading is wrong. The cause was not money but structure. The new CBA was built to make over-spending teams pay with their own future competitiveness.

Look at the nature of the penalties. Banning salary aggregation means a team above the apron cannot bundle three players for a star as before. Pushing the draft pick to the end of the round means a wealthy team must trust long-term investment through the hardest path. The league does not ban success; it taxes it.

There is another blind spot few mention. The media blames Thompson for his play-in performance, or the Warriors' front office for letting him go. Both sides are right in their own way, and this structure forces the rightness of one to be the wrongness of the other. That is the mark of a system designed perfectly so that both cannot win at once.

In the locker room, players like Draymond Green talk about basketball, about fire, about commitment. I respect that. But the spreadsheet does not talk about fire. It talks about row three, column seven: 182.8 million USD. Above that number, no commitment is large enough to protect a player from the very system that made him.

A prediction with an expiration date

I am publicly betting: within 18 months — that is, before January 2026 — at least one other high-spending team will be forced into a sign-and-trade to escape the second apron, just as the Warriors did with Thompson. The clearest candidates: the Boston Celtics and the Phoenix Suns. Both have crossed the first threshold, and both carry large contracts nearing expiry for players born between 2026 and 2026. For the Celtics, the question is whether they can keep their 2026 championship roster if they hit the second apron in 2026-26. For the Suns, the question already has an answer.

The next domino

Numbers do not cut across the story — they tell a different one, and they are rarely wrong. The Klay Thompson deal is the first chapter of a new era, where big teams are forced to choose between present and future through ruthless figures. Fans may cry, players may rage, but the league set the rules and those rules are working exactly as designed. Mid-tier teams will benefit: they do not need to spend the most, only to spend the smartest.

The only question I keep asking myself: when will a team be brave enough to say out loud that in the apron era, keeping a legend may be the most selfish decision of all? The spreadsheet wrote the answer long ago. We simply have not read it.