rdba (0xc4d1a863e9cc45d02ba22d3a1ae9ba7822018ce8) Polymarket trader turned $7.3M into $69K in PnL across 292 trades with a 70% win rate — proof that prediction market whales can dominate individual bets while their portfolio bleeds out. rdba is a Polymarket whale operating in the deep end: rank 1571, 207 markets touched, averaging $9.4K per trade with a disciplined 2 trades per day. The stats scream contradiction. Win rate of 69.89%? Elite. Best single trade on New York City Mayoral Election hit $279K profit. But the portfolio value sits at $118K against $7.3M net deposits. ROI reads -93.56%. This is what happens when a Polymarket trader wins most battles and loses the war. The edge here is surgical position sizing with a 14:1 buy-to-sell ratio — rdba loads up on conviction bets rather than scalping noise. Entry price averages 0.78, meaning he hunts underpriced outcomes and holds through volatility. Low risk designation matches the data: max single loss capped at -$100K (worst trade on Reds vs. Phillies (2025-07-04)), which is brutal but controlled relative to the account size. Closed 197 positions while holding 95 open suggests he's not panic-closing winners too early — real discipline on the Polymarket leaderboard. The brutal math: $20M total volume traded, 70% hit rate on individual outcomes, yet somehow the Polymarket strategy turned $7.3M into negative territory. This screams concentration risk or timing hell. Likely scenario: rdba nailed specific prediction markets (elections, sports) but sized into a handful of correlated thesis plays that tanked. The $279K NYC Mayoral win got wiped across other positions. Open 95 positions right now — could be recovery attempt or continued conviction bleeding. Real talk: this Polymarket whale has the win-rate edge and discipline most degens will never touch. But raw PnL tells you prediction market skill doesn't equal portfolio management. He's got $118K left to work with and $95 open bets still grinding. Either he finds the leak or becomes a cautionary tale about how even 70% accuracy fails when you're wrong about correlation.