| income_carry | 50 | |
| 29 | ||
| cycle_recovery | 25 | |
| rerating | 22 | |
| share_gain | 20 | |
| capacity_expansion | 17 | |
| turnaround | 7 | |
| quality_at_price | 4 | |
| commodity_cycle | 1 |
| foundry-logic | 26 | |
| 24 | ||
| semicap-equipment | 23 | |
| memory | 16 | |
| hyperscaler | 13 | |
| staples-dividend | 13 | |
| energy-dividend | 11 | |
| passives | 8 | |
| healthcare-dividend | 7 | |
| dc-infra | 6 | |
| utility | 5 | |
| power-thermal | 5 | |
| telco-dividend | 4 | |
| tech-dividend | 3 | |
| reit | 2 | |
| trading-house | 2 | |
| power-semis | 2 | |
| networking | 1 | |
| neocloud | 1 | |
| div-materials | 1 | |
| div-industrials | 1 | |
| div-comm-services | 1 |
| probability_weighted | 119 | |
| base_only_no_probabilities | 32 | cannot be scored — predates probability weighting |
| 24 |
| analyst-aisemi | 115 | |
| analyst-income | 59 | |
| selftest-writer-boundary | 1 |
| conviction 5 vs 3 — do high-conviction calls actually do better? | insufficient data (n=0, need ~30) |
| model x coverage — does the producer model change the answer? | insufficient data (n=0, need ~30) |
| adversary value-add — do broken verdicts predict worse outcomes? | insufficient data (n=0, need ~30) |
| Ben's overrides — is he right to overrule the system? | insufficient data (n=0, need ~10) |
| dimension disagreements as predictors | insufficient data (n=0, need ~30) |
| Kelly switch criterion | insufficient data (n=15, need ~15) |