Detroit Sports Reporter

• Tigers10-8
• Lionsnone yet
• Pistonsnone yet
• Red Wingsnone yet

August 9, 2026 · Detroit Pistons · Analysis

The Pistons just made the biggest two-year climb in the NBA since 1993. I went looking for the crash that is supposed to follow it

Three summers ago the Pistons lost 68 games. This spring they won 60, finished first in the Eastern Conference, and got run out of their own building 125-94 in a Game 7 they hosted. Then the offseason: Jalen Duren's restricted free agency stalled, Detroit's search for a second scorer alongside Cade Cunningham ended with no deal, and four rotation players left in Tobias Harris, Isaiah Stewart, Caris LeVert and Marcus Sasser.

So the worry going around is specific, and it is not stupid. Teams that rise that fast are supposed to fall. The roster that won 60 is not quite the roster coming back. The East got harder.

The first half of that is the part people assert instead of checking, so I pulled every NBA season since 1991 and went looking for teams who climbed the way Detroit climbed.

There is no team who climbed the way Detroit climbed

Detroit went from 14 wins to 44 to 60. Out of 995 other three-season spans in the sample, the number that gained as much as Detroit's 46 is zero. The closest is Philadelphia in 2018, at 42.

Two things about that table that change how you should read it.

Four seasons in this window were not 82 games: the 1999 lockout, 2012, and the two covid years. Everything here is converted to a wins-per-82 pace, marked with a dot. That conversion is necessary in one direction and misleading in the other. It stops a 50-game season from looking like a collapse, but it also manufactures climbs nobody watched. San Antonio's 1999 row reads as a 61-win team and they won 37 games, in a 50-game season, and their "-8" the next year was really 37 wins becoming 53. Phoenix in 2021 won 51, not 58. Read the dotted rows as rates, not as seasons that happened.

The other thing is the three-way tie at +32. All three are in the table because cutting the list at a round number would have dropped two teams who got better the next year and kept the one who got worse, which would have made the right column look like something it is not.

What those teams did next

Take the fifteen teams with a climb of 30 wins or more that have a following season on record. Their median change the next year was minus 1.0 wins. Now take every season in the sample at a pace of 58 wins or better: median change minus 6.0, with 69 of 83 declining.

On those two numbers the big leapers look better than ordinary very good teams. That comparison is worthless, and the table above shows why. The leapers' peaks run from 43 wins to 66, with a median of 53. A 53-win team regresses less than a 62-win team for a reason that has nothing to do with how either of them got there: there is less above it to fall from. Comparing a group of mostly-53-win teams against a group of all-60-plus-win teams measures the difference in their peaks, not the difference in their histories.

So match them. For each of the fifteen leapers, take every non-leaping team-season whose win pace was within three wins of that leaper's, and ask what those teams did the following year. That is the number the leaper has to beat.

Median difference: plus 2.7 wins. Nine of fifteen beat their matched group.

Nine of fifteen is what a coin produces; the odds of nine or more heads in fifteen flips are about three in ten. Resample those fifteen differences twenty thousand times and the middle 95 percent of the median runs from minus 2.0 to plus 7.2 wins, an interval with zero sitting inside it. Drop every span that touches a shortened season and ten cases remain, median plus 2.8, seven of ten positive, which is the same non-answer.

The two largest positives are Miami in 2010, who added LeBron James and Chris Bosh, and Oklahoma City in 2024, a very young team that kept improving. Those are not evidence that leaping helps. They are evidence that what happens to the roster in July matters, which everyone already knew.

So the crash is not a thing the record shows. The careful version of that sentence matters, because with fifteen cases spread from minus 11 to plus 14, this design cannot see an effect smaller than about five wins in either direction. It rules out the big scary version. It does not prove the climb is meaningless.

Two more holes worth naming. The matched pool is non-leaping teams at the same win level, which by construction means teams that arrived there flat or falling, and a team that dropped to 52 from 58 is not the same animal as one that rose to 52 from 45. And the fifteen differences are not independent, because leapers with similar peaks share most of their control seasons, so that bootstrap interval is a little narrower than it should be. Cutting the other way: selecting teams on a huge positive swing partly selects on a lucky peak season, which should drag the leapers' next-year numbers down. That bias runs against the leapers and they still came out even.

What does carry information

One thing, and it is unglamorous. Detroit is a 60-win team now, and that is itself the prediction.

Of the 86 team-seasons in the sample that finished within three wins of a 60-win pace, the median lost 5.2 wins the following year. 62 of the 86 declined at all. 24 of the 86, better than one in four, fell below 50 wins.

That is the base rate and it applies to Detroit exactly as it applied to everyone else. It is also milder than the mood suggests. Five wins off 60 is 55 wins. The scenario Pistons fans are bracing for is not the median outcome, it is the one-in-four outcome, and one in four is a real risk rather than a prophecy.

The call

Detroit finishes next season between 52 and 58 wins. That is the number that gets graded, and it is deliberately narrow enough to be wrong: the base rate alone would let me say "fewer than 60" and be right 62 times out of 86, which is not a call, it is a hedge.

The limit on all of this is roster continuity, and I did not measure it. Every number above treats a team as a win total, so it cannot tell a 60-win team bringing everyone back from a 60-win team that lost four rotation players and has an All-NBA center whose contract is unsettled. Duren is still listed on Detroit's roster today. If that changes, none of these base rates knew about it.

That is the thing worth watching this fall. It was never the 46.

Standings from ESPN's public API, every season from 1991 through 2026, pulled and analysed by scripts/nba_leaps.py; the chart comes from the same run via scripts/nba_leap_chart.py. Franchise spans are joined across relocations, so Seattle to Oklahoma City counts as one team. Offseason reporting via PistonPowered and Sports Illustrated; Game 7 via ESPN. Not betting advice.