Entertainment Purposes Only
Not trading advice. Not financial advice. This is a mathematical exercise. No representation is made that any account will or is likely to achieve profits. Trading prediction markets involves risk of loss.
Algorithm Reference Card

KALSHI EUCLIDEAN 7-Step Discretionary Algorithm

Prediction Market Trading Framework · Mathematical Edge Over Intuition

┃ Key Metrics

Round Trip Cost
Spread + Fee ≈ 4¢
Required Edge
> 4¢ to break even
Skip-to-Trade Ratio
~5:1 (80% filtered)
Half-Kelly Ruin Prob.
< 1%
Edge Per Trade
Small but positive
Edge Compounds
Over volume

┃ Order Book Geometry

BID (Buyers) ASK (Sellers) $0.47 Best Bid $0.52 Best Ask MID $0.495 Spread = 5¢ LIMIT ORDER Place at mid-price spread capture adverse move risk Buyers compete upward Sellers compete downward Mid = (Bid + Ask) / 2
Bid Ask Mid / Limit Entry

┃ The 7 Steps

1
Step One

CONTRACT SCREENING

  • 24hr volume > $500
  • Spread ≤ 5¢
  • Expiry: 3–90 days
  • Price: 25–75¢
  • Result: ~15–20 candidates
2
Step Two

FAIR VALUE ESTIMATION

  • Research event causal factors
  • Historical data on similar events
  • Assign probability + confidence (High / Med / Low)
  • Skip if Low confidence
3
Step Three

EDGE CALCULATION

  • EV = (Estimate − Price) − RoundTripCost
  • Only trade if EV > 0
  • Expected skip rate: 4 out of 5
  • Edge magnitude determines position size
4
Step Four

POSITION SIZING (Half-Kelly)

  • Kelly % = Edge / Odds Received
  • Position = ½ Kelly × Capital
  • Low confidence → Quarter-Kelly
  • NEVER size by feel
5
Step Five

ORDER PLACEMENT

  • LIMIT ORDER at mid-price ONLY
  • Never market order
  • Timeout: 15 min (high vol) / 30 min (med vol)
  • Cancel if unfilled, re-screen
6
Step Six

EXIT PROTOCOL (3 conditions)

  • Profit Target: exit at estimated fair value
  • Stop Loss: exit at ½ expected gain adverse move
  • Time Stop: exit after 72 hours regardless
  • Whichever hits first
7
Step Seven

JOURNAL & REVIEW

  • Record every trade (entry, estimate, confidence, size, exit, P&L)
  • Weekly calibration review
  • Adjust estimation process based on actual vs predicted frequencies
  • Compounding skill is the only permanent edge
Geometric Principle
"You do not need to know what will happen. You need to know what things are worth, and what things are worth is visible in the order book to anyone who knows how to read it."