Auditing Detroit's 23-0 Run: The Scoreboard Doesn't Add Up, and the Real Question Is Bigger Than Tatum's 6-of-23

Detroit beat Boston 121-113. The report says Detroit closed on a 23-0 run over the final eight-odd minutes. I took the three scores printed in that same report. Eight minutes left — 89-99. Ausar Thompson's fadeaway — 106-100. Final — 121-113. Put those three numbers on one sheet of paper and the 23-0 disappears. Between 89 and 106 Detroit scored 17 and conceded one; that is a 17-1, not a 23-0. A 23-0 only fits if the score at the fadeaway was 112-99, not 106-100.

I kept the log for eleven weeks and nobody asked to see it. Here it is the other way round: the log is printed in the report and it does not reconcile with itself. I audited the obituary before I agreed to write it — I did the same with the UBA in 2026. So I read this game as a scorecard, not as a headline.

The shape of the game itself, in this early-season fixture, is reasonably clear. Boston pushed Detroit to 50-72 across the first thirty minutes — a 22-point gap. Boston conceded only 18 in the second quarter, meaning Detroit scored 18 in that stretch and still trailed; that is the signal ordinary reporting files away as a "slow start." By the end of the third Boston had restored an 89-99 lead, with Paul George already on 28 of his 38. Then came the last eight minutes.

What the box score records: Jayson Tatum 6-of-23, 1-of-10 in the second half, 11 assists. Cade Cunningham 22, Jalen Duren 22. Off the bench, Duncan Robinson 21 (14 in the fourth) and Jalen Green 19. Ausar Thompson 11 points, 11 rebounds, 5 assists.

One piece of paperwork first. The player the report calls "Oscar Thompson" carries a profile — Detroit, 11/11/5, rookie-scale timeline — that matches Ausar Thompson, the fifth pick of 2026, almost exactly. A single name error does not change the story of a game, but it does tell you what level of proofreading this report received. And when you are about to price a future number, the level of proofreading matters.

The most diagnostic figure in this game is not the 32-point swing but the 22-point deficit. Detroit was roughly 22 down across the first thirty minutes. That is closer to its true level. The 32-point swing over the last fifteen is something else — a tail event produced by an opponent's collapse and a bench shooting night arriving together. A swing that size happens in a single game; it does not repeat across a season.

Tatum's line is the most readable thing on the whole card. Twenty-three field-goal attempts were maintained, 11 assists came, and the shot went 1-of-10 after halftime. This is the picture of a slump in which the process survives and the result breaks — historically the less worrying version of a slump. A player who cannot get the ball out of his hands does not post 11 assists. And a run like 23-0 requires eight to ten consecutive empty possessions; a 6-of-23 night from the number-one option is the most plausible engine for that.

What was not given should also be on the record. There are twelve numbers on this scorecard and three of them are doing all the work. Tatum's point total appears nowhere — only 6-of-23 and 11 assists. Printing 6-of-23 while dropping the points is not neutral editing; it enlarges the inefficiency picture and conceals whether he still finished around 15. George's 38 comes with no FGA, no 3PA, no FTA. Thirty-eight on thirty shots and thirty-eight on twenty have opposite implications, and the report does not let me tell the difference. No OffRtg, no DefRtg, no pace, no lineup plus-minus. No efficiency claim in this report can be verified. The editor said nobody reads possession logs. Here the log itself is printed incomplete.

Detroit's 40 bench points — Robinson 21, Green 19 — are the largest and least projectable component of the result. League bench output normally sits between 25 and 32; 40 is an outlier. The shooters who make 40 one night make 12 the next. Detroit's closing group reads as one big and four out — Cunningham, Robinson, Green, Thompson, Duren. Plenty of shooting, thin rim protection.

Thompson's 11/11/5 is the most transferable number on the card. Rebounds and assists are stickier than shooting; they survive efficiency regression in both directions. Alongside it sits something the report never states: Thompson's 11 boards plus Duren's work at the five imply Detroit won the glass, and that is the mechanical prerequisite for a 23-0 — defensive rebound, then transition. Team rebound totals are simply not given.

Here is the uncomfortable part. The run the report turns into a story is, on my reading, a variance event rather than a system, and its playoff transferability is low. Two concrete reasons. First, a 23-0 depends on the opponent's shot variance, and playoff defence compresses that variance. Second, bench-shooting comebacks are the first thing a targeted defence takes away — rotations shorten, close-outs tighten, and across a seven-game series the opposing staff starts hunting Robinson and Green at the other end.

And one thing I cannot reconcile: the score sequence does not produce a 23-0, yet the entire story stands on that 23-0. It could be a detail error, or a definition-of-run error — but when a report's central claim does not survive its own numbers, my job is to soften the claim, not repeat it.

The second point sits outside the box score. The Boston roster picture this report implies — Tatum, Brown and a wing on 30%+ of the cap — does not fit the CBA framework I know. A team past the second apron cannot aggregate outgoing salaries and cannot take back more than it sends. Absorbing a $45-50M salary therefore requires sending out roughly the same amount in a single, non-aggregated structure — which in practice forces a core piece out. No such teardown is recorded anywhere in the report. The claim may be wrong, or it may post-date my information. Either way the conclusion holds: no model can be built on this roster claim.

Detroit's real concern is not this game but the extension cliff. Cunningham's extension is already on the books; Duren's and Thompson's rookie-scale extensions are in the queue. That is where the savings of a cheap core run out — and it will decide more about where Detroit stands in three years than any 23-0 run.

The next game answers the question: do Detroit's first thirty minutes still look like the 22-point version, and do Robinson and Green repeat 40 or fall back to 15? George's 38 has no value until his shot attempts are counted. And Tatum — if 23 attempts and 11 assists hold, there is nothing to worry about; if the attempts start falling away, then something beyond 6-of-23 has happened.