plotier

Haven’t posted on X for a while. Will be dropping a post about polymarket bots and the reality behind it

@thejayden Not necessarily true, if you know the correct logic and framework it is possible. But for vibe coders yes this is impossible

“Claude, build me an avg(YES) + avg(NO) < $1 Polymarket bot.” Let me tell you one thing: you have a higher chance of winning the lottery than making this work.

200ms Latency Wall + Jitter • 30% Queue Skip (Fill disadvantage) • 46 Microstructure features per tick

Simulation is 10% Harder. Latency: 200ms Wall. Alpha: Robust. 1 Day on the PMBot dashboard. BTC is absolutely ripping with a 55.6% win rate despite simulated slippage and a 3/10 chance of being "queue jumped" on every fill. Total 12h Profit: ~$148

@jtrevorchapman @Polymarket Same latency wall and penalties as yours for my training. Going to deploy in 5 days time.

Experimenting with the 15m reward harvesting on @polymarket. Honestly, the sub dollar pair cost arb to merge was responsible for a bit more than 90% of the profit. Again, it was an experiment, so ten share batches are small enough where the rewards aren't anything to buy lunch on, but... lets see where it goes from here. All in windsurf and antigravity, opus and flash.

@jtrevorchapman @Polymarket How did your ML/RL model go? I’m in my testing phase and implemented the same training metrics as you and it’s performing well.

@jtrevorchapman was able to decrease creation time and post time on top . Wonder if i could train a model to be like your one and we can compare

@jtrevorchapman Thanks for the feedback, going to try and make my own version for fun. My current version is on the same basis as everyone just optimized to the max. running latency scans from coinbase/binance then having a threshold and farming yes and no for equal shares.

First foray into ML/RL. Polymarket token price prediction.

@jtrevorchapman how does your executor deal with getting fills, obviously since its sim are you assuming all of your orders are getting filled?

46 data points. Realtime. Focused on predicting y/n token price on 15m markets. Gathering the data that chainlink bases their BTC price on. Exp var between .28 and .4. Moved yesterday to RL scalping w 200ms delay, neg spread rewards, etc. will let it cook for a week then deploy w real capital to polymarket.

@hamptonism @grok what is this video

pov: back in 2021 you responded to an instant message on your Bloomberg Terminal from a user you didn’t recognize located in Hangzhou, probing you about the Multifractal Lévy-Stable Process Adversarial Graph Neural Network Covariates Copula-Driven Spatial Autocorrelation Smoothing algorithm that only you and two other quants in New York were supposedly working on for hyper-optimized trans-Pacific drone-freight routes, leveraging stratospheric UAVs with real-time wind shear adjustments and machine-learned port congestion predictors. The message contained a date and the coordinates of a Tsim Sha Tsui dive bar in Singapore, where you exchanged the algos architecture under the table for $30M in Bitcoin private keys split three ways, then went ghost for 4 years. Now it’s 2026, you’re living in Hong Kong in the summer, Seoul in the winter, views of Victoria Harbour, doomscrolling on 𝕏 every night.

“I’m working on a @Polymarket arb bot”

@thejayden Couldn’t agree more, there are strategies that are still untapped where logic beats speed

Onchain data tells you everything you need to know. A few days ago everyone told you to build bots using made-up code screenshots. Now the same vibe-code gurus say you have no chance because you’re competing with Chinese bots. The edge is not always the speed. It’s not always expensive infrastructure. A “perfect” bot makes a few hundred dollars per hour. But if your bot makes tens of dollars per day, for most people that’s another monthly salary, and that’s totally fine. The edge is there. And it’s more visible than ever. I’m not here to put you down or praise your skills. The perfect moment to start is now. Shoot your shot. Regret hurts more.

@jtrevorchapman @Polymarket People don't understand that it will take weeks of development, this is what real results look like

288 15m BTC @polymarket binary directional markets ran through the latest iteration of my script with results I like, real cash, not dry-runs or backtests: • took a position in 174 markets • skipped 114 markets • 11 incorrect predictions BUT nine of the eleven were profitable despite the other side winning due to proper hedging logic. • only TWO losses, and those were partial as they were hedged. • one emergency dump. • global prediction accuracy: 88% • global profit rate: 87% • current session accuracy: 95% • current session profit rate: 90% Currently spend is minimal while iterating. When I’ve done all I can to properly mitigate risk, will turn on the faucet wager wise. The second image shows how this strategy is my favorite of all tested thus far. Profit is cut in half, but once the strategy has been initiated, the market can do whatever it wants. Polymarket bots and scripts are doable. You just have to put in the time. Count on going through hundreds of iterations. Analyze each loss, and how you could have avoided it programmatically. Save ALL data from every market. Btw, $9.67 every market (96/day) is $928/day. Entering only 65% of the markets = $603 I’ll report back on future results (Christmas time = uncommon BTC price action).

@Polymarket Fix your website

BREAKING: Trump administration to begin seizing wages of defaulted student loan borrowers.

@Pharane0 Responded

I picked one person to mentor and things are looking good. I might pick another one in the future but this is the start to something bigger Here are some random aggresive market simulations with my algorithm. check the thread for more screenshots. #polymarket #bot #trader

@Pharane0 All about those gains

A message to the haters🖕