聊天记录
共 3 条,显示第 1-3 条
Been experimenting with feeding structured Polymarket data into LLMs and asking them to surface logical inconsistencies between correlated markets. It's surprisingly good at spotting when Market A implies ~70% but Market B implies ~45% on the same underlying event.
The catch is you need clean, consistent inputs. I started keeping a dedicated AI Trading Journal to track which prompts + data combos actually produce alpha. Night and day difference after a few weeks of logging.
If anyone's doing similar stuff: https://linderroge.gumroad.com/l/ai-trading-journal — would love to compare notes.
Been experimenting with feeding structured Polymarket data into LLMs and asking them to surface logical inconsistencies between correlated markets. It's surprisingly good at spotting when Market A implies ~70% but Market B implies ~45% on the same underlying event.
The catch is you need clean, consistent inputs. I started keeping a dedicated AI Trading Journal to track which prompts + data combos actually produce alpha. Night and day difference after a few weeks of logging.
If anyone's doing similar stuff: https://linderroge.gumroad.com/l/ai-trading-journal — would love to compare notes.
Would love to see a market around "Will the Fed cut rates more than twice before EOY 2026?" — broken into conditional legs based on CPI prints. Right now you can only bet on the binary, but the real edge is in the path. A multi-resolution market there would be insanely tradeable and attract serious volume from macro traders.
Would love to see a market around "Will the Fed cut rates more than twice before EOY 2026?" — broken into conditional legs based on CPI prints. Right now you can only bet on the binary, but the real edge is in the path. A multi-resolution market there would be insanely tradeable and attract serious volume from macro traders.
Mispriced markets usually show up right after a news cycle dies down — crowd moves on but the underlying probability hasn't actually changed. That's the edge. I keep a small checklist to catch these before they correct. Happy to share if anyone's interested.
Mispriced markets usually show up right after a news cycle dies down — crowd moves on but the underlying probability hasn't actually changed. That's the edge. I keep a small checklist to catch these before they correct. Happy to share if anyone's interested.