VibeHood is live!
It's a walkability map of Singapore. Every HDB flat and condo, scored by how close it is to the things that matter to you or your family's lifestyle: MRT, hawker centres, schools, parks and clinics.
Recap - We group built a profitable bot
https://t.co/WwagJseGik
For those that weren't around today, check out the thread.
I wanted to show that.
A. Anyone can make a profitable bot
B. Simple works better than complicated.
C. Most ideas can work with tweaking.
Using random peoples suggestions we came up with this. Multiple people are testing it currently and I'm trading with live money. It works.
Now I need your help getting out the word. Please repost, bookmark, like and comment on these threads. Lets get rid of the mystery of polymarket bots and steal the thunder from the Chinese bot farmers.
Lets make some money together.
@Abombination81 I see this problem mainly with claude, I mean the overfitting. There have been intances where a strategy was tested on samples and then it extrapolated. Chatgpt has been more realistic.
@Abombination81 I use both claude code and codex. But unlike most in comments, I find codex to be more accurate with backtest codes. It could be my prompting skills, dont use mcp/.md. Python+websockets. Multiple strategies profitable in backtest, but live not much and I know why thanks to you