Not by plan. One question led to one bot. One bot broke, got fixed, broke again. That single bot has been rewritten 61 times.
Here's what the six look like now:
Each one came from an assumption I wanted to test. Most assumptions were wrong.
Built a news analysis bot convinced that news moves prices. Ran 1,056 entries. Average price change after news: -0.012%. Assumption wrong, entire bot built around it.
Trading bot still isn't profitable. Documenting the whole thing anyway — the gap between what I expected and what happened is more interesting than the wins.
Next: the elif bug that silently cut my data in half. No errors. No warnings.
Anyone else running a solo stack like this?