#Testing & data
7 public posts on this topic.
Record
I put Friday's net call flow through one more filter — separating direction from expiration cleanup
Net call flow fell all day, yet the S&P 500 finished higher. A year of data suggests Friday's net call number carries expiration-driven selling that has little to do with direction.
Record
I Started Writing the Grading Sheet Before the Hypothesis
Writing down a call isn't enough. By the evening I read it in whatever way flatters me. Fixing the grading rules before the open made the verdict — right direction, not enough distance — write itself.
Record
I used yesterday's data to predict this morning's gap — the correlation was 0.000
41.6% of a trading day's movement happens while I cannot see a screen. And that stretch does not remember yesterday.
Record
I re-ran a 15-year backtest and 85.6% became 53.8%
I built a rule, traded it, and withdrew its justification the same week. What broke was not the market — it was the *open* column I had used.
Record
I pull VIX1D from Cboe now
Accurate inputs make accurate calculations. A record of a 38% gap from a free source.
Record
The day I questioned a 100% win rate
One ticker filter, 329 alerts re-scored, and a validation system that got better for it.
Record
Meet Billie
Keeping only the signals that pass validation. The AI trading partner I built.
🔒 2 trading-journal entries on this topic are in the subscriber-only trading log →