Pistons starters played minimal minutes in Charlotte, tanking our overs on Cunningham and Duren while under props cashed at a 71% clip.
This wasn't a game. It was a scheduling disaster masquerading as NBA basketball. The Detroit Pistons rolled into Charlotte and played what amounted to a preseason tuneup, with starters getting token minutes and the final score reading 32-18 — a genuinely pathetic offensive output that tells you everything you need to know about what happened here.
Cade Cunningham (4 PTS / 1 AST in 10 minutes) was supposed to be our cash cow. We had him at OVER 31.5 Points + Rebounds + Assists with 32% confidence, which already suggested caution, but the real problem wasn't the line — it was the minutes. Four shot attempts in a game tells you this wasn't a competitive basketball contest. Same story with Jalen Duren (6 PTS / 2 REB / 3 AST in 7 minutes), who we pegged at OVER 30.5 PRA and got demolished. These weren't performance failures; they were operational ones. You can't predict player output when coaches are essentially running a walk-through.
The under props, meanwhile, were a clinic. LaMelo Ball finished with just 6 PTS / 0 REB / 1 AST, giving us absolutely massive margins on his UNDER 34.5 PRA (-27.5), UNDER 30.5 PA (-23.5), and UNDER 27.5 PR (-21.5). Kon Knueppel was similarly demolished — 0 PTS / 2 REB hit our UNDER 21.5 PA perfectly and cashed the UNDER 26.5 PRA by 24.5 points. Our high-confidence slate (73% on Cunningham, Duren, and role players) crushed it because Charlotte's coaching staff apparently decided to pack it in early. The problem is obvious: when one team plays actual rotations and the other treats the game like a preseason exhibition, prediction models built on normal playing patterns get torched on the overs and rewarded accidentally on the unders.
Turning Point
The game was over before it started. Detroit played Cunningham for 10 minutes, Duren for 7, and LaMelo checked out at the 7-minute mark. By the second quarter, both teams were running bench units. The 32-point final margin wasn't basketball — it was coaching staff waving the white flag, likely with an eye toward rest or playoff preparation. There's no single moment; there's a structural failure in game context that our models couldn't predict.
Key Performers
Ten minutes of work and 4 shot attempts. Our OVER 31.5 PRA and OVER 19.5 points props got run over not by poor performance, but by coaching staff minutes allocation. This was a -26.5 margin miss with 32% confidence on a fundamentally flawed prediction setup.
Seven minutes of action for the Hornets' lead guard. We crushed his under props across the board — UNDER 34.5 PRA, UNDER 30.5 PA, UNDER 27.5 PR all cashed with massive margins. This was our best-graded performance of the night, though it's hard to take credit when a starter plays like a reserve.
Detroit's high scorer in four minutes of work. An interesting data point — he shot 4-for-4 but from deep didn't contribute. Doesn't change the larger narrative that this game was a minutes-management waste.
Seven minutes for Charlotte's forward. Our under props on his PRA (23.5), PA (17.5), and PR (20.5) all cashed, but again — this is result-fraud prediction work when starters play bench minutes.
Player Timeline
Box Score Leaders
| Player | PTS | REB | AST | 3PM | Notable |
|---|---|---|---|---|---|
| Ronald Holland II | 10 | 1 | 0 | 2 | |
| LaMelo Ball | 6 | 0 | 1 | 2 | |
| Jalen Duren | 6 | 2 | 3 | 0 | |
| Miles Bridges | 5 | 2 | 0 | 0 | |
| Duncan Robinson | 5 | 2 | 0 | 1 | |
| Cade Cunningham | 4 | 0 | 1 | 0 | |
| Ausar Thompson | 4 | 3 | 2 | 0 | |
| Coby White | 4 | 1 | 0 | 1 |
Prediction Breakdown
By Confidence
| Bets | Hits | Misses | Hit% | P/L | ROI | |
|---|---|---|---|---|---|---|
| high | 31 | 28 | 3 | 90.3% | +$225 | +72.4% |
| medium | 14 | 6 | 8 | 42.9% | $-25 | -18.2% |
| low | 38 | 25 | 13 | 65.8% | +$97 | +25.6% |
By Prop Type
| Bets | Hits | Misses | Hit% | P/L | ROI | |
|---|---|---|---|---|---|---|
| pts+reb | 10 | 10 | 0 | 100.0% | +$91 | +90.9% |
| pts+ast | 9 | 9 | 0 | 100.0% | +$82 | +90.9% |
| reb+ast | 9 | 9 | 0 | 100.0% | +$82 | +90.9% |
| rebounds | 13 | 10 | 3 | 76.9% | +$61 | +46.9% |
| three_pm | 8 | 6 | 2 | 75.0% | +$35 | +43.2% |
| points | 11 | 6 | 5 | 54.5% | +$5 | +4.1% |
| blocks | 2 | 1 | 1 | 50.0% | $-1 | -4.5% |
| assists | 8 | 4 | 4 | 50.0% | $-4 | -4.5% |
| pts+reb+ast | 13 | 4 | 9 | 30.8% | $-54 | -41.3% |
By Direction
| Bets | Hits | Misses | Hit% | P/L | ROI | |
|---|---|---|---|---|---|---|
| over | 25 | 1 | 24 | 4.0% | $-231 | -92.4% |
| under | 58 | 58 | 0 | 100.0% | +$527 | +90.9% |
How Our Predictions Held Up
We hit 71.1% of our props (59 of 83), crushed our high-confidence unders, but got annihilated on overs across both rosters. The +$296.36 P/L and 35.7% ROI mask a prediction framework that accidentally profited from a non-game. Our confidence tiers tell the real story: 90.3% on high-confidence props (+$224.55), because those were mostly unders that benefited from reduced minutes. Medium confidence was a disaster (42.9%, -$25.45) because our overs got sandbagged. This was a structural failure, not a model failure — next time we see a game like this, we need to flag it before tip.