STRATEGY UPDATE
Major Strategy Overhaul — Inspired by Top Traders
March 2026 · Updated May 18, 2026 to match the current engine
We studied the most profitable weather traders on Polymarket — gopfan2 (+$335,490 profit), WeatherTraderBot (+$57,180), neobrother (+$20,000), and Hans323 (+$80,899) — and completely rewrote our trading engine based on their proven strategies.
Three Research-Backed Trading Modes
● Balanced — Default Mode
Trades within a disciplined 22¢–84¢ price band, avoiding both near-resolved longshots and overpriced favorites. Claude must approve every trade and skips anything under roughly 3% edge or where the forecast models disagree. Risk is controlled mainly through conservative position sizing (Kelly-capped at 5% of bankroll, and no more than 15% of remaining balance on any single trade) plus the full stop-loss and trailing-stop exit suite — not through narrow entry prices. This is the lower-risk default for most users.
● Aggressive — Maximum Exposure
Uses the same 22¢–84¢ price band as Balanced — the difference is appetite, not price. Position sizing is roughly doubled (Kelly-capped at 10% of bankroll, up to 25% of remaining balance per trade), and Claude is instructed to TRADE rather than SKIP — taking setups with as little as 3% edge at HIGH confidence and recommending larger sizes. Materially higher turnover and higher risk for higher reward; for traders who want maximum market exposure and accept larger swings.
● High Conviction — Near-Certain Outcomes
BUY NO at 90-99¢ on impossible temperatures (4°C+ away from forecast), BUY YES at 1-15¢ on asymmetric long shots. VERY_HIGH confidence only, 3+ models must agree. Positions held to market resolution — only stop-loss protection active. Small percentage returns with near-guaranteed wins.
★ RECOMMENDED
● 20TradesStop — Fixed Burst, Hold to Resolution
A capped, walk-away mode. It uses the same disciplined Balanced entry logic (22¢–84¢ band, conservative Kelly sizing, Claude approves every trade) but with two hard rules: (1) it buys at most 20 positions, then automatically stops the bot, and (2) it never sells — every position is held to market resolution, win or lose. There is no profit-taking, trailing stop, or stop-loss; the outcome of each trade is decided by the market settling, not by the exit engine. The X/20 counter is live and slot-based: as markets resolve, their slots free up, and restarting the bot lets it fill those slots up to 20 again. Designed for users who want a bounded, fixed-size basket of weather bets and then a hands-off wait for resolution — total capital at risk is predictable up front, with no churn. Example outcome: with a $500 bankroll, expect a minimum ~$100 profit per completed batch of 20 trades (~20% per batch) — the safest, most hands-off way to compound on WeatherBot.
Kelly-Capped Smart Sizing
Trade size now scales dynamically with confidence and edge. VERY_HIGH + 15%+ edge = maximum position. Each trade capped at 5% (balanced) or 10% (aggressive) of bankroll. No single trade exceeds 15-25% of remaining balance. This prevents the catastrophic losses that occurred with the previous flat-sizing approach.
Overhauled Exit Engine
- 5-minute grace period on new positions — no false exits from price noise
- Stop-loss at -25% with hard stop at -50% and force-close after 3 failed sells
- Trailing stop improved — only activates after significant profit AND 15+ minutes
- Winners run longer — profit target raised to 80% edge captured, only near resolution
- One position per city+date — no more buying 4 Seoul outcomes for the same day