Kelly sizing
Given your measured win rate, R:R, and current DD buffer, returns the dollar risk per trade that maximizes log-growth (full Kelly) plus the practitioner-default half-Kelly and the conservative quarter-Kelly. Refuses to size negative-EV strategies.
Risk of ruin
Probability of hitting the DD floor before reaching the profit target, assuming i.i.d. trades with constant risk size and a static DD floor. For trailing-DD firms, expect 10-15 pp lower than this — use the simulator on the firms page for those.
Minimum edge
Backward question: pick a firm + a target pass probability + an R:R, and we'll tell you the minimum win rate you need. Useful before paying a challenge fee — "I want to pass with 50% probability; at 1:1 R:R that means I need at least X% WR. Do I have that?"
Edge significance
Wilson 95% confidence interval on your observed win rate, plus a one-sided test against a null (default 50%). Honest answer to "is my 60% WR over 30 trades real edge, or noise?" The CI is the most honest summary at small sample sizes; the p-value adds the standard significance interpretation at larger N.
DD-distance sizing
Kelly sizing that respects the firm's actual DD geometry (static / trailing / trailing-then-lock) and the buffer health you have right now. As your buffer thins, position size scales down by √(buffer_health) — smooth de-leverage to keep you from running half-Kelly into a near-empty buffer.
Phase-shift sizing
Challenge phase vs funded phase optimal sizing using the convex-payoff insight: in the challenge phase your downside is capped at the challenge fee (sunk cost), so size more aggressively. In the funded phase every dollar of drawdown is real money — back to standard half-Kelly. Returns both sizings + the boost factor that quantifies the asymmetry.
Time-of-day analysis
Group your logged trades by hour or day-of-week and surface the windows where your realized stats diverge. Honest noise handling: buckets with under 5 trades are flagged; best/worst windows only call out buckets with 10+ trades. The most useful tool for finding your own edge from your own data.