pic related is my bot running the exact same strategy across 20+ perps. this taught me something. notice how there is no consistency in profit or losses despite it using the exact same strategy, some perp types are up some are down. in fact it looks completely random despite being the same strategy just on different perps. wtf is up with that? you would think a strategy that works... just works!
im sorry to inform you OP but ur losing moneys...
>>62687487>you would think a strategy that works... just works!yes, but this is the important part: your strategy doesn't work, so it ... just doesn't work! spooky how that happens.
>>62687487>6 losing trades>3 winning tradesUmmmmmmm... Last I checked, "working strategies" have a higher than 33% winrate.
>>62688711retard
>>62687487Based on this, if you simply inverted your strategies you would be deep in profit
>>62687487do you have to pay for the bot?
>>62689673sort of. i vibe coded it with AI. so i paid $20/month for the AI.if you are interested in doing the same anyone can do it at this point without any technical skill. just download ChatGPT App to your PC and pay $20/month to use the best models (GPT 5.5 or newer depending) and ask it to code a trading bot for you. All you really gotta know is where you want it to trade (i chose solana perps on jupiter and Gains Network or something like that) and ask it to come up with different strategies for you.AI came up with the 20+ strategies and runs them all simultaneously. make sure to ask it to make the bot have Paper Trading at first before real trading that way you can keep winning strats and drop losing ones. any time AI asks you to do some technical shit just tell it to do it itself.it will take 20+ passes for the AI to get it right, so it does what it can in 1 pass then you ask 19 more times for it to improve it. only then will your bot be truly ready. and sometimes you should ask the AI what it thinks would improve it or make the system more sane, it will tell you, and you can cherry pick from them if you really want to
>>62690868further advice would be to tell AI to make sure the bot supports plugging in multiple trading venues, so you can trade multiple venues simultaneously (Jupiter, Gains Network, and whatever other perps providers there are out there) then in addition to that tell the AI you want to be able to run many different strategies simultaneously, and each strategy has its own bankroll (i gave each strategy $100), and that the AI knows you want the bot to include paper trading as well as real trading but start paper trading.periodically ask the ai to delete (or disable) losing strats and keep winning strats, also occasionally ask AI to come up with entirely new strats, so you are preferably always running 20+ strats at once. hopefully this will create some sort of self improvement feedback loop, but i am yet to hit that point myself if that point even exists. i just started.i'd call it a shotgun approach
>>62688885and this guy brings up an interesting point i'd like to try, if a strat is losing, ask the AI to invert the strategy and add it as a new strategy and see what happens
>>62687487>you would think a strategy that works... just works!yes I suppose if you were an idiot you might think that. you're trading different underlyings which are not perfectly correlated.
>>62687487its almost like.. you dont have any edge
>>62688844Indeed. What makes a strategy "good" or "successful"? That it wins more than loses? That it makes money instead of loses it?Look at the code. It's beautiful. The logic. It's elegant. And if you want, just inverse if you want opposite day. Simply divine.If it just works, it just works. It couldn't be more simple to make money
>>62690909that's not why he called him a retard, retard. he called him a retard because win rate is irrelevant, you can lose 99/100 and still be incredibly profitable if you're wins are big or losses small enough. it's more about skewness of returns than 'win rate'eg open a $50 wide iron condor on SPY any given day and you'll have a very high win rate, but you'll pay it all back and more when it loses
>>62688675He's saying that if strategy didn't work, it would consistently not work. Instead, it seems to both work and not work. His mistake is assuming that working means green and not-working means red when in reality, working means consistent, period, and not-working means random. If it were consistently red, you just short the strategy instead. It's the consistency that is valuable
