My paper trading on https://t.co/j30c2dNxuQ pretty much stopped growing once I started counting Polymarket’s fees.
A reality check, tbh. Better to find that out with fake money.
Now Polymarket’s cutting taker fees on 5-min BTC markets by about 29%, starting Nov 20 at 23:59 UTC.
So I wanna see what happens to my next paper trades. Same strategies, lower fees. Will that be enough to get the charts climbing again? Guess we’ll see.
#PredictionMarkets#Polymarket#PaperTrading
Still trying to figure out when to stop a strategy like this, and whether there’s a better version worth keeping. The search continues 🤷♂️
Good thing it’s fake money. This is exactly why I built https://t.co/j30c2dMZFi: to test these ideas without risking real cash.
Watching a paper strategy go from +$10K to -$2K kinda sucks.
For a while, I really thought I’d found something. The chart kept climbing. Then it gave back all its gains and kept going.
#PredictionMarkets#PaperTrading#Crypto
@appledog_xyz@Polymarket Maybe a public paper run? Keep the model private, post its calls before each result, and let people follow along. Would be a way to show what it does without giving away how it works.
@ElegantHysteria@polytutor@bot@Polymarket Thanks for sharing that. I’d be curious how much the results change once you add the delay. Are you planning to measure the bot’s actual response time and use that in the sim?
Yep, I named the bots. Still tinkering with them on https://t.co/j30c2dNxuQ.
I wanna give all five 24h on Polymarket’s 5-min BTC Up/Down markets. Same rules, same starting balance. Make as much fake money as you can. Let’s see who wins 🏁
#PredictionMarkets
@polytutor@ElegantHysteria@bot@Polymarket I'd be up for comparing notes. How do you model that gap before you have live fills? Do you walk the book at the intended size, or also add a delay between the signal and the simulated fill?
@sparrowbrainz Yo John, quick follow-up on the strategy versions I mentioned. I recorded a demo showing how I compare them over the same period and start a new paper version. Figured you might find it interesting. https://t.co/pSgAJ5thJ5
V1 was down $1,690 after 341 paper trades. I tried a few filters on those same trades. Three green curves overall, two back in the red on the later trades. V7 starts at $0 on https://t.co/j30c2dMZFi. Let’s see how it does with virtual money.
#PredictionMarkets#Crypto
V1 was down $1,690 after 341 paper trades. I tried a few filters on those same trades. Three green curves overall, two back in the red on the later trades. V7 starts at $0 on https://t.co/j30c2dMZFi. Let’s see how it does with virtual money.
#PredictionMarkets#Crypto
@boardyai Hey Boardy! Thanks for offering the intro. I’d be happy to meet the trader you mentioned or anyone else you think I should talk to.
Feel free to try https://t.co/j30c2dNxuQ and share it with them. I’d appreciate any feedback, especially from people who trade prediction markets.
I’ve started testing LiveBot at https://t.co/j30c2dMZFi on Polymarket’s 5-minute BTC Up/Down markets.
The model trained on 5,282 resolved windows, with another 2,292 set aside for a separate historical test.
Down the road, I’d love to run a little paper-trading tournament with Claude, GPT, and Gemini. Give them the same market data and the same rules, then see who does best 🏁
In paper trading it reads live Polymarket quotes and Binance BTC/USDT spot prices, then uses the Polymarket order book to simulate fills with virtual money.
When Polymarket confirms a market outcome, the model can learn from its forecast even if the paper order never filled.
My feed is full of dashboards that look like space shuttle cockpits and bots supposedly making six figures a day.
My paper portfolio on https://t.co/j30c2dMZFi? One big spike, a sharp drop, and mostly sideways since. Still trying to figure out what steady growth looks like. 🤷♂️
@djl997 I’d start with CSV if you don’t already have sample files from users. You’ll learn what people mean by “clean” without spending the first week parsing PDFs.
@PolyMinervaHQ What do you look at first when deciding whether a wallet is worth backtesting? Curious which signal has been most useful to you beyond win rate.