LeBron to the Sixers — and why we didn't crown them

The Sixers added LeBron James to a core that already had Joel Embiid, Tyrese Maxey, and — as of two weeks ago — Jaylen Brown. On paper it's a four-star roster, one of the most top-heavy in the league, and the kind of move that makes people pencil in a parade.

Our job isn't to pencil in a parade. It's to put an honest number on it and defend it.

Our NBA model reads last season's point differential, so it can't see a summer trade until we tell it. Roster moves enter as a documented, dated adjustment — a win-total delta with a written reason, logged where anyone can check it. Brown was worth +12 wins of judgment on July 11. LeBron we added today.

The result: Philadelphia goes from 9.7% to win it all to 16.7% — a jump from fringe contender to the model's clear East favorite and #2 title pick in the whole league, behind only the defending favorite.

Here's the part we're proudest of, because it's where forecasting goes wrong. Our first pass credited LeBron with the same kind of bump a star normally gets, and the model spat out 25% to win the title and a 65% shot at the #1 seed — a coronation. We backed it down, on purpose.

Two reasons. First, the math: title odds are convex. Near the top of the league, championship equity piles up fast, so a modest win bump doesn't add to the odds — it multiplies them. A number that looks reasonable as 'wins added' can explode into a fake favorite. Second, the basketball: LeBron is 41. He's a fourth star on an already-loaded roster, which is diminishing returns, and the whole season still swings on Embiid's health. Four high-usage stars is a fit question, not a cheat code.

So we kept LeBron's marginal bump small — +3 wins, not +6 — and the title number landed at 17% instead of 25%. A clear favorite, not a lock. If that's still too high, the season will say so.

The Sixers are now the team to beat in the East and a genuine title threat — but the model is deliberately refusing to hand them the trophy in July. Every number here is timestamped to our Ledger and will be graded against what actually happens. If we've got them too high (or too low), you'll be able to see it, in public, no backdating.

Think we're too cautious — or not cautious enough — on an aging superteam? Go look at the board and tell us where we're wrong.