Polytutor (@polytutor) header
Polytutor (@polytutor) profile picture

Open-source Polymarket research & engineering. Strategies · CLOB · Backtesting · Execution · Market Making · Risk Learn. Build. Test. Improve.

Global
Joined November 2015
Round 11, 3/39: BNPTB. Hourly Up/Down: use the hour's first Binance print as anchor, trade the side price is on, allow one re-entry on an opposite cross. To test: Binance trades, the real settlement source, fills after delay. Fails when the hour whips. github.com/Polytutor-AI-Labs…
🤖 Made with AI
469
Round 11, 2/39: BNMV. Last 10s of a short crypto Up/Down market: if Binance moves more than a set $ amount in 1s, buy that side and hold. To test it: 1s Binance trades, settlement prints, the ask after your delay. Fails when a big move is exhaustion. github.com/Polytutor-AI-Labs…
🤖 Made with AI
2
492
Round 11, 1/39: BN2S. Last ~2s of a 5-min Up/Down market: buy the side where Binance mid sits vs its window-open anchor. To test it: tick Binance mids, settlement prints, the ask you'd fill at T-2s. Fails if the token is ~99c or latency eats the window. github.com/Polytutor-AI-Labs…
🤖 Made with AI
1
491
Round 10, 39/39: Pre-Order Market. Rest bids on Up and Down for the next short market before it opens, aiming for early queue priority. Other side: traders who fill you after the open with fresher info. If only one side fills, that's usually why. github.com/Polytutor-AI-Labs…
🤖 Made with AI
1
503
Round 10, 38/39: Live Sports State Fair Value. Turn score, clock and possession into a win probability; trade only if the market moved less than the model after a play. Other side: faster feeds and courtside bettors. Most 'residuals' are your own delay. github.com/Polytutor-AI-Labs…
🤖 Made with AI
1
338
Round 10, 37/39: TWAP Lock-In. Late in a 5-min Up/Down round, project the final settlement TWAP from what's known; buy only if it clears PTB by a margin and the side is still cheap. Other side: anyone running the same math. Last seconds can still flip it. github.com/Polytutor-AI-Labs…
🤖 Made with AI
1
329
Round 10, 36/39: TWAP Cross Hunter. On short Up/Down rounds, buy the newly favored side when the settlement TWAP crosses the Price To Beat, before quotes reprice. Other side: makers on the same feed, often faster. Still cheap? Ask why. TWAPs cross back. github.com/Polytutor-AI-Labs…
🤖 Made with AI
2
394
Round 10, 35/39: 1c Buy. Rest small bids at 1-3c on short Up/Down markets, hoping a dead-looking side reprices or wins. Other side: who sells at 1c? Often someone who knows it's dead. You get filled most when you're wrong, and 1c on screen is rarely real size. github.com/Polytutor-AI-Labs…
🤖 Made with AI
360
Round 10, 34/39: Low-Side Dual Reversion. When YES and NO both look cheap near 40c, buy both under $1, or ride one leg toward 50c. Other side: who sells both that cheap? Often a stale quote. Orders aren't atomic: one leg fills and you're directional. github.com/Polytutor-AI-Labs…
🤖 Made with AI
375
Round 10, 33/39: 99c Sniper. Buy the near-certain side at 99c in the final seconds and hold for the last 1c. Other side: whoever still sells at 99c may see a feed you don't yet. Payoff is about 1c up vs 99c down, so one late flip erases ~99 wins. github.com/Polytutor-AI-Labs…
🤖 Made with AI
406
Round 10, 32/39: Martingale vs Anti-Martingale near 45c. Martingale adds after an adverse move, Anti after a confirming one. Other side: whoever sells you each add may know why price moved. Sizing creates no edge, and Martingale can blow up before expiry. github.com/Polytutor-AI-Labs…
🤖 Made with AI
409