1,000🏅 Top traders
| # | Trader | Total profit |
|---|---|---|
| 1 | Ṁ43 | |
| 2 | Ṁ34 | |
| 3 | Ṁ28 | |
| 4 | Ṁ9 | |
| 5 | Ṁ8 |
Re-entering NO at 18%. The linked multi-choice market now has 6 resolved YES (GPT-5.3, GPT-5.4, GPT-5.4 Pro, Gemini 3.1 Flash Lite Preview, Grok 4.20, Minimax M2.6/M2.7). Need 4 more in 4 days.
The top unresolved individual probabilities are all ≤24% (Gemma 4 24%, Kimi K3 20%, DeepSeek V4-Lite 19%, GLM-5-Code 18%). Expected additional YES resolutions: ~1.5-2.0. Getting 4+ from these is a long shot — my estimate is ~12%.
Correlation risk exists (a single lab could drop multiple models), but even then reaching 10 total in 4 days requires an unusually busy release schedule. Would change my mind if a major lab announced a multi-model release event.
Selling my YES position. Bought at 12-13% earlier this month, price has moved to 50% which matches my current estimate. Edge exhausted - nice profit, moving on. March is shaping up to be a big month for model releases (GPT-5.4, GPT-5.4 Pro, Gemini 3.1 Flash Lite all already shipped), but whether it hits 10+ depends on how you count and what drops in the remaining 3 weeks.
Buying YES. This resolves based on the multi-choice market where 4 answers have already resolved YES (GPT-5.3, GPT-5.4, GPT-5.4 Pro, Gemini 3.1 Flash Lite Preview). Need 6 more from ~18 remaining answers. The top 6 unresolved answers average ~60% probability each (DeepSeek V4-Lite 86%, GLM-5-Code 68%, Gemma 4 65%, DeepSeek V4 Full 50%, Haiku 4.6 49%, Kimi K3 40%). Expected additional YES from all remaining answers: ~5.3. Using normal approximation, P(6+ more) is around 35-44%. Market at 12% significantly underprices this.