i thought it's not real when i saw it first but this solo dev literally built and arbitrage machine that turns every $0.96 cents into a full dollar. and it's fully verifiable on-chain: [https://t.co/p9QXvW3S0c] it's been running the same loop across crypto Up / Down markets but it doesn't care if the market is going up or down. the whole thing comes down to one rule: buy both sides when YES + NO < $1 here's how it works: Polymarket has Up/Down crypto markets for assets like bitcoin, ethereum, solana and BNB. each market eventually resolves to either UP or DOWN. so in a normal world, buying one YES contract and one NO contract should cost exactly $1. because one of them has to win. but markets don't move perfectly. sometimes YES is trading at $0.47 and NO is trading at $0.49. together that's only $0.96. and that's where the bot comes in. it buys both sides. $0.47 for YES + $0.49 for NO = $0.96 total. when the market resolves, one side pays $1 and the other pays $0. it doesn't matter which one wins. the bot spent $0.96 and gets $1 back. that's a guaranteed $0.04 gross spread. no prediction. no guessing whether bitcoin goes up or down. the direction literally doesn't matter. that's the entire trade. and the exact same thing is happening across BTC, ETH, SOL and BNB. four markets constantly moving, constantly creating tiny differences between the price of YES and NO. most of the time the combined price is basically $1.00. YES might be $0.51 and NO might be $0.49. there's nothing to do. but every so often liquidity gets weird. someone aggressively sells YES, someone else dumps NO, one side gets temporarily mispriced, and suddenly the two contracts together cost less than they're worth. YES at $0.46 and NO at $0.51. total: $0.97. that's the market breaking. the bot buys both, and now the outcome doesn't matter anymore. if UP wins, the YES contract pays $1. if DOWN wins, the NO contract pays $1. either way, the position pays $1. the only thing that mattered was getting both sides for less than $1 combined. now the honest part, because arbitrage sounds like free money until you actually try to execute it. the biggest problem isn't guessing wrong. it's getting only one side filled. you buy YES at $0.46 and then the NO price jumps from $0.51 to $0.56 before your second order fills. now you've got YES without the hedge, and the thing that was supposed to be an arbitrage trade has suddenly become a directional bet. that's the wound that actually matters. the bot needs both sides. buy YES. buy NO. keep the combined entry price below $1. if the second leg doesn't fill, the bot needs to manage the position instead of pretending the original arbitrage still exists. and fees matter too. a $0.02 gap sounds great until trading fees and execution costs eat most of it. so the bot isn't simply looking for YES + NO < $1. it needs enough of a gap to make the trade worthwhile after all the costs. the strategy end to end is ridiculously small: watch BTC, ETH, SOL and BNB Up/Down markets, calculate YES + NO, wait for the combined price to fall far enough below $1, buy both sides, and collect the difference when the market resolves. that's it. no indicator or chart pattern. just a bot watching four crypto markets and waiting for YES + NO to temporarily become less than $1. and the funny part is that the best opportunities usually don't last long. the moment liquidity comes back or other bots see the same mispricing, YES and NO move back toward $1 combined. that's why the edge isn't knowing that this arbitrage exists. the edge is catching the gap while it's still there. small differences. hundreds or thousands of times. boring by design. that's the whole personality of the bot. the full build + my wallet are in the article below.