Social Posts Decode Quant Bots
Recent social-media analyses dissected high-performing quant bots: a Polymarket sports bot reported $7.4M PnL using buy-only engines, volume farming for rebates, and cross‑market hedging, while a separate HFT-style bot claimed roughly $12k/day using Bayesian updates, Kelly sizing, and Stoikov inventory models. Both write-ups present detailed microstructure and sizing approaches that traders are studying publicly. (x.com) (x.com)
Two social-media breakdowns turned opaque trading bots into public case studies, with traders now parsing the math behind their reported profits. (x.com 1) (x.com 2) One post by 0xMovez described a Polymarket sports bot that reportedly made $7.4 million in profit and loss using “buy only” execution, cross-market hedging, and reward capture tied to exchange incentives. A separate post by 0x_Discover described a high-frequency-style bot that allegedly made about $12,000 a day using Bayesian probability updates, Kelly sizing, and an Avellaneda-Stoikov inventory model. (x.com 1) (x.com 2) Prediction-market bots trade contracts that pay $1 if an event happens and $0 if it does not. On Polymarket’s live sports pages, those contracts now cover leagues including the National Basketball Association, National Hockey League, Ultimate Fighting Championship, Major League Baseball, soccer, and tennis. (polymarket.com) (docs.polymarket.com) The exchange also pays some liquidity providers cash rebates in USD Coin when their resting orders get filled, creating a second revenue stream beyond directional bets. Polymarket’s documentation says rebates are paid daily in USD Coin and are based on the share of liquidity a trader provided that was actually taken. (docs.polymarket.com) (help.polymarket.com) That helps explain why a bot might trade both for edge and for incentives. Polymarket’s United States documentation also describes a separate volume incentive program that rewards traders based on their share of trading volume on eligible contracts, though it says that program is “coming soon.” (docs.polymarket.com) (docs.polymarket.us) The technical terms in the posts map to standard quant tools. Bayesian updating means revising a probability estimate as new evidence arrives, and the Kelly criterion is a formula for sizing bets to maximize long-run growth rather than simply maximizing the next trade’s payoff. (ocw.mit.edu) (ideas.repec.org) The Avellaneda-Stoikov model comes from a 2008 Quantitative Finance paper on market making in a limit-order book. In plain terms, it tells a bot to shift its bid and ask prices depending on how much inventory it already holds, so it does not get stuck too long or too short. (ideas.repec.org) (math.nyu.edu) Not every number in the threads can be independently verified from public records in the posts themselves. But the mechanics they describe line up with documented exchange incentives, live sports market structure, and widely used market-making models that have circulated in quantitative trading for years. (x.com) (docs.polymarket.com) (math.nyu.edu) The result is that two X threads from April 2026 are being read less like brag posts and more like open notebooks. They gave the public a rare, line-by-line look at how traders say they are turning prediction-market microstructure into systematic bets. (x.com 1) (x.com 2)
