The Sauce · Money & Markets
Prediction-Market Research Edge (Small Stakes, Real Math)
Use AI to research underpriced event-market questions and place small, disciplined positions where you have an information edge — treated as a skill, not a slot machine.
Submitted by @@evhunter · 2026-06-02
💸 Highly variable / can lose money/mo
Editorial estimate — no community reports yet. Be the first.
$100–$500 bankroll (risk capital only)
Startup cost
Advanced
Difficulty
Days–weeks (highly variable)
Time to $1
The stack
Some tools in this stack may be affiliate links — Flux may earn a commission. Workflow rankings are editorial and never paid.
The playbook
- 1
Confine yourself to 1-2 domains where you genuinely know more than the average bettor (a sport, a regulatory area, a niche industry) — edge comes from specialization, not breadth.
- 2
For each candidate market, research with Perplexity to form YOUR probability, then have Claude argue the opposite side hard to find what you're missing.
- 3
Only bet when your estimated probability differs meaningfully from the market price (a real expected-value gap), and size each position small (1-3% of bankroll).
- 4
Log every bet — thesis, your probability, market price, outcome — and review weekly to see if your calibration is actually beating the market.
- 5
Compound slowly from a bankroll you can fully afford to lose; cut domains where your logged calibration is no better than the crowd.