The Risk Takers Podcast Why Some Stats Predict Better Than Others & Kalshi "Death Markets" | Ep 145
50 snips
Mar 4, 2026 A nerdy deep dive on which sports stats actually predict outcomes and how to separate signal from noise. A LoL example shows why raw counts can mislead and when to use proxies and standardized inputs. A detailed look at a controversial Kalshi Iran leadership market, settlement rules, and the problems with last-traded-price resolutions. Brief industry news, trading mechanics, and ethical concerns around death and war markets.
AI Snips
Chapters
Transcript
Episode notes
Start Market Making By Posting Simple Limit Orders
- Anyone can begin market-making on Kalshi by posting limit orders and monitoring flow; full designated market-maker agreements require uptime and quoting obligations.
- SP noted many profitable traders start on the API or with selective manual quotes.
Build A Bottom Up Model Then Validate Against Market Lines
- Build your own bottom-up model first, then use market lines to validate or find missing signals.
- SP prefers starting with sport data, then comparing to books/exchanges to identify systematic gaps.
Model Closing Lines As A Faster Route To Edge
- Alternatively, modeling the close/closing line can be an easier path to a pragmatic edge if you lack deep sport-specific data.
- John Shilling recommends predicting closes to set opening prices or exploit market inefficiencies quickly.