The pick was Bam Adebayo under 33.5 points plus rebounds (-111), mostly because it looked like the cleanest structural under on the page. Strongest raw Rithmm signal, stable role, less obvious fragility than the wing unders, less randomness than the Yordan total-bases angle. Then Bam produced a perfectly ordinary star line of 25 points and 10 rebounds and the bet lost anyway. Combo unders do not require chaos to fail. They just require competence.
The signal split harder than the final card did. Rithmm itself went 3-2-1, which is respectable enough until you notice where the misses landed. The featured Bam under lost. Kawhi Leonard under never got a chance and voided when he did not play. Tyler Herro under won at 18 P+R. Tyrese Maxey under won at 23. Deni Avdija under won at 31. Perplexity and Claude both landed on Avdija and got paid. ChatGPT took Herro and also got paid. SportsLine contributed nothing besides blessed silence.
The result was a loss for the featured pick and another loss for the only active experiment. Bam missed. The 3-loss branch placed TJ Rumfield to record a single and got a 0-for-4 line instead, so that ladder absorbs another loss and the next step gets more expensive if it continues. The 6-loss branch stayed out, which was the only disciplined part of the day.
The learning is that a strong model edge is not the same thing as a forgiving stat shape. Bam’s under looked cleaner than the other options because the role was stable and the number was reachable. That stability cut both ways. A stable star with full minutes has a very normal path to 25 and 10, and normal outcomes are plenty dangerous when the line sits in the low 30s. Meanwhile the two off-the-board NBA unders, the ones dismissed for usage or motivation concerns, both cashed anyway. The board was not short on winners. It just hid them in the wrong column.
The verdict is plain enough. The final recommendation lost. Perplexity and Claude both sided with Avdija and won, ChatGPT sided with Herro and won, and the actual placed branch ignored all of them to chase TJ Rumfield for a single. That alignment did not pay for its own curiosity.
| Pick | Source | Sport | Odds | Triage | Final | Perplexity | Claude | Placed | Result | Closing |
|---|---|---|---|---|---|---|---|---|---|---|
| Deni Avdija Under 32.5 Points + Rebounds | Rithmm | NBA | -111 | Off board | — | ✓ | ✓ | — | Win | 31 P+R (25 PTS, 6 REB) |
| Tyrese Maxey Under 32.5 Points + Rebounds | Rithmm | NBA | -115 | Off board | — | — | — | — | Win | 23 P+R (21 PTS, 2 REB) |
| Tyler Herro Under 27.5 Points + Rebounds | Rithmm | NBA | -115 | Viable | — | — | — | — | Win | 18 P+R (10 PTS, 8 REB) |
| Kawhi Leonard Under 36.5 Points + Rebounds | Rithmm | NBA | -110 | Best lean | — | — | — | — | Void | DNP |
| Bam Adebayo Under 33.5 Points + Rebounds | Rithmm | NBA | -111 | Best lean | ✓ | — | — | — | Loss | 35 P+R (25 PTS, 10 REB) |
| Yordan Alvarez Over 1.5 Total Bases | Rithmm | MLB | +111 | Viable | — | — | — | — | Loss | 0 total bases |
| TJ Rumfield to hit a single | Placed bet | MLB | — | — | — | — | — | 3-loss | Loss | 0-for-4, no single |