retroactivesource Polymarket trader dropped $197k on prediction markets, hit a 79% win rate across 288 trades, and still got absolutely rekt with a -$66k PnL — the math on how this happens is genuinely weird. retroactivesource sits outside the top 2M Polymarket traders with 79.43% win rate but -38.71% ROI on deposits. That's the core tension: this Polymarket whale is winning more than 4 out of 5 bets and still bleeding money. Whale category, 273 markets touched, mostly low-risk bets at $2.86k average size. The portfolio sits at $82.7k against $295k total deposits — a textbook "right direction, wrong position sizing" case. The edge hack here is dead simple: bet small, bet often, chase noise. retroactivesource fires 5.9 trades per day across esports (LoL mainly), hitting that 79% clip by picking low-variance outcomes where the odds are already tight. Best trade pulled $14.1k on a Bilibili vs JD Gaming BO3 matchup. But the worst trade blew -$21.1k on EDward vs Oh My God — a single loss that wiped out 1.5x the best win. That's the trap: high win rate doesn't survive asymmetric downside. The real Polymarket strategy here is volume-first, edge-last. 288 trades over ~50 days means this is machine-like execution, probably semi-automated or extremely disciplined routing. The 14.45 buy-sell ratio suggests heavy one-sided conviction on certain outcomes. Win rate stays clean, but ROI collapses because the market is pricing these same esports outcomes tighter than retroactivesource's edge can exploit. You're not beating the Polymarket leaderboard by being right 79% of the time if your position sizing punishes you on the 21% that miss. Currently holding 71 open positions across active markets. The -$66k PnL is real and underwater — not a temporary blip. Risk level stays "low" by bet size, but portfolio drawdown says the strategy needs rework. Not everyone survives the drawdown when your win rate masks position sizing disasters.