Research Note №1 · A career autopsy · rev. v2

The Rise and Fall of Ja Morant

A statistical account, 2017–2026.

Francesco Freedman · The Gravity Report · the GRAVITY player-value model · published July 10, 2026 · revised July 11, 2026 (v2)

Abstract. We measure every season of Ja Morant's career — college, rise, peak, and collapse — with one rating system applied the same way to every player in every season. Three findings organize the paper. The rise was real: from an average-ish rookie to the 97th percentile of the league in 2021–22. The collapse was about missing, not choosing: measured across shot zones, 88% of his 2025–26 efficiency crash came from worse conversion and only 12% from where he shot — while his playmaking hit a career high. Much of the "decline" is too small a sample to judge: the three-point collapse covers just 85 attempts — too few to distinguish genuine decline from ordinary variance — and his free-throw shooting, the purest touch test, actually improved, significantly. The fall of Ja Morant is, by the numbers, mostly an availability story: 79 games played in three seasons. His rank of #140 of 644 prices that risk, and the number is robust to the model's own judgment calls — every variation tested lands him between #112 and #147. It probably underprices the talent, and the paper shows exactly where the difference lives.

Principal findings

Missing, not choosing: of his effective-field-goal collapse, 88% was shot-making and only 0.7 points shot-selection (an exact zone-level Shapley split — with its scope stated honestly: zone data cannot see shot difficulty within zones).
The sample-size problem: the three-point collapse is 20-of-85; its 95% interval [.158, .336] contains his career norm. The one split that moved beyond doubt was his free throw — upward, to .897.
The absence ledger: 79 of a possible 246 games across three seasons (32.1%) — the 2023–24 season alone: a 25-game suspension, a nine-game return, then 48 games lost to the shoulder.
The stress test: seven model variations (overrides off, equal season weights, different anchors, reliability stack deleted) land him #112–#147. #140 is not an artifact of any single judgment call.

Page 4 of the paper: Morant's seven-season career arc in impact and percentile terms, and the corrected absence ledger. — **The arc, and the ledger (p. 4).** Seven seasons in one currency — the 2021–22 peak at the 97th percentile, and the absence ledger that explains #140 better than any shooting chart.

Page 9 of the paper: the availability-times-impact matrix pricing every 2026-27 future, and the conclusion. — **The price of every future (p. 9).** The availability × impact matrix: health and form are near-perfect substitutes at the bottom of the table, and top-50 requires both.

In plain terms: #140 is a defensible price for the delivery risk. It is probably the wrong forecast of the talent. Both statements are the same mathematics, read in two directions.

📄 Read the full paper (PDF, 12 pp) 📥 Read it in the inbox LaTeX source 🏀 The 644

Cite this note

Freedman, F. (2026). "The Rise and Fall of Ja Morant: A Statistical
Account, 2017-2026." The Gravity Report, GRAVITY Research Note No. 1, v2.
https://thegravityreport.com/notes/morant/

@techreport{gravity2026morant,
  author      = {Freedman, Francesco},
  title       = {The Rise and Fall of Ja Morant: A Statistical Account, 2017--2026},
  institution = {The Gravity Report},
  type        = {GRAVITY Research Note},
  number      = {1},
  year        = {2026},
  month       = {7},
  note        = {v2, revised July 11, 2026},
  url         = {https://thegravityreport.com/notes/morant/}
}

Revision note (v2, July 11, 2026), following external review: the absence ledger was corrected (the January 2024 shoulder injury cost 48 games, not 73 — the other 25 were the suspension, listed on its own row; denominator 246, not 249); the Wilson-interval, free-throw, and Shapley-split language was tightened to what the evidence supports; the bundled forward scenarios were replaced by an availability × impact matrix; and a model-sensitivity appendix was added. The errors were ours. Nothing in the correction changes the ranking itself.