Forecasting on POLYGYM N = 375
0.178Brier ScoreInfoDelphi
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| InfoDelphiCategory=Ours, Deliberation=Iterative, Rounds=R=22026.07 | 0.178 | 77.9 | |
| Zero-shotCategory=Single-Agent2026.07 | 0.201 | 66.7 | |
| Standard DebateCategory=Multi-Agent2026.07 | 0.202 | 73.9 | |
| InfoDelphiCategory=Ours, Deliberation=None (no deliberation), Rounds=R=12026.07 | 0.206 | 74.1 | |
| Chain-of-ThoughtCategory=Single-Agent2026.07 | 0.216 | 69.9 | |
| MoACategory=Multi-Agent2026.07 | 0.22 | 70.9 | |
| Self-ConsistencyCategory=Single-Agent2026.07 | 0.221 | 69.6 | |
| Halawi et al.Category=Single-Agent2026.07 | 0.238 | 65.3 | |
| SuperforecasterCategory=Single-Agent2026.07 | 0.242 | 68.8 | |
| Crowd EnsembleCategory=Multi-Agent2026.07 | 0.245 | 65.9 | |
| AIA ForecasterCategory=Multi-Agent2026.07 | 0.247 | 67.7 | |
| Seq. BayesianCategory=Single-Agent2026.07 | 0.256 | 62.1 |