Back in August, I built an engine that I could plug in different strategies in and it would run them based on the priority I gave. It handled entry, exit, trade management, and logging. My plan was to start with SPY scalping and I've calculated over 20 indicators on 1-min bars going back to January 2020. The trouble is that I've never been able to find any correlation between indicators and price to get a positive pnl. I toyed with it until I got frustrated so now I'm lost and haven't touch it in weeks. Since then, I've been trying to watch memlab's youtube videos (his trading accelerator videos below) and Al Brooks price action videos, but haven't made it far into either. https://www.youtube.com/@memlabs-research

@Teun S oh man, everything. i probably can't see the forest for the trees. below are all the signals i have defined for each 1-min bar. yeah, software engineering. i was a java programmer for 10 years but have been in management or program management for the last 15. i hadn't touched code since 2013 but put the hat back on and picked up python pretty quickly. ema10, ema20, ema50, atr, bollinger, candle body ratio, vwap, vwap deviation, rci, macd, volume surge, time of day, and the slope of any of them that would make sense. my thought was to calculate them all for every bar for any regime and i could discard what wasn't needed (assuming perf wasn't that big of an issue while i backtested). i did regression to see if any of them correlated to any direction price change more than 50% of the time on 1, 5, 10, 15, and 30 minute timeframes. completely inconclusive. my regression work is probably fatally flawed but i got burned out on it. maybe you're right about needing to try a different vehicle.

I voted for Thursday 6pm so it doesn't encroach on the weekend, but the next 2 Thursdays are holidays so those probably won't work.