Chapter 13 Key Takeaways: Finding and Quantifying Your Edge
Core Concepts
1. Edge Is the Foundation of Profitable Trading
Edge is the systematic difference between your probability estimate and the market price: $E = q - p$. Without edge, every trade is a coin flip minus fees. Most traders do not have consistent edge, and the first step toward profitability is honestly assessing whether you do.
2. Expected Value Is Simply Your Probability Minus the Price
The EV of buying YES at price $p$ when the true probability is $q$ is just $q - p$. The EV of buying NO is $p - q$. Every trade decision reduces to this calculation. If EV is positive, the trade is worth considering. If negative, walk away.
3. The Kelly Criterion Tells You How Much to Bet
For buying YES at price $c$ with estimated probability $p$: $$f^* = \frac{p - c}{1 - c}$$ Kelly maximizes long-run geometric growth. It automatically sizes larger bets for bigger edges and smaller bets for smaller edges. It never recommends betting when there is no edge.
4. Use Fractional Kelly in Practice
Full Kelly is too aggressive for real-world trading due to estimation uncertainty, correlated bets, and psychological tolerance. Half Kelly achieves 75% of full Kelly's growth rate with dramatically reduced drawdowns. Quarter Kelly is appropriate for beginners or uncertain estimates. When in doubt, bet less -- over-betting is far worse than under-betting.
5. Know Where Your Edge Comes From
Edge decomposes into calibration, information, model, and timing components. Knowing which sources drive your profits helps you invest in your strengths, avoid your weaknesses, and detect when your edge is eroding.
6. Probability Estimation Is the Hardest Part
Everything depends on the quality of your probability estimate $q$. Base rate analysis, reference class forecasting, quantitative models, and estimate combination are practical tools for producing well-calibrated estimates. Always attach a confidence interval to your estimate.
7. Edges Decay Over Time
As markets mature and competition increases, edges shrink. Information edges vanish when information becomes public. Analytical edges erode as competitors adopt similar methods. Behavioral edges are most durable but still diminish. Successful traders maintain a pipeline of strategies at different lifecycle stages.
8. Track Everything
A systematic edge tracking system -- logging trades, computing realized edge, decomposing P&L, measuring calibration -- is essential. Human memory is biased; only systematic tracking reveals your true performance. If your tracking shows you do not have edge, stop trading for profit.
Formulas to Remember
| Concept | Formula |
|---|---|
| Edge | $E = q - p$ |
| EV of buying YES | $\text{EV} = q - p$ |
| EV of buying NO | $\text{EV} = p - q$ |
| EV as % return (YES) | $\text{EV\%} = (q - p) / p$ |
| Kelly fraction (YES) | $f^* = (p - c) / (1 - c)$ |
| Kelly fraction (NO) | $f^* = (c - p) / c$ |
| Fractional Kelly | $f = \alpha \cdot f^*$ |
| Brier score | $\text{BS} = \frac{1}{N}\sum(q_i - o_i)^2$ |
| Ruin probability (full Kelly) | $P(\text{reach fraction } x) = x$ |
| Ruin probability (half Kelly) | $P(\text{reach fraction } x) \approx x^2$ |
Rules of Thumb
- If you cannot articulate your edge in one sentence, you probably do not have one.
- A 3-5 percentage point edge is meaningful in prediction markets. Edges below 2 percentage points are often consumed by fees and uncertainty.
- Never risk more than 5-10% of your bankroll on a single trade, regardless of what Kelly says.
- Your total exposure should not exceed 40-50% of your bankroll to maintain reserves for new opportunities and drawdown recovery.
- You need at least 50-100 resolved trades to assess whether your edge is real and statistically significant.
- Over-betting at 2x Kelly produces zero growth. Over-betting beyond 2x Kelly produces negative growth. When uncertain, always err on the side of smaller bets.
- The market is right more often than you think. Treat large apparent edges with suspicion -- they may indicate that you are missing information rather than that the market is wrong.
Common Mistakes to Avoid
- Confusing entertainment with investment. If you are trading for fun, set a fixed entertainment budget and do not expect positive returns.
- Anchoring to your initial estimate. Update your probabilities when new information arrives. Do not defend a position just because you already own it.
- Ignoring fees. A 5% edge minus 3% in fees is only a 2% edge. Always compute net-of-fee EV.
- Trading too many markets. Spreading your analysis across dozens of markets dilutes your edge. Focus on markets where you have genuine comparative advantage.
- Ignoring correlations. Five bets on related political events are not five independent bets. Treat correlated positions as a single, larger bet for sizing purposes.
- Survivorship bias in self-assessment. You remember your brilliant calls and forget your embarrassing misses. Let the data speak.
- Failing to retire losing strategies. If 100 trades show no edge, the strategy does not work. Stop, reassess, and try something different.