George research / ongoing series
Keep the failures.
They protect the edge.
Every study starts with a reason the setup might work, then tries to kill it with unseen dates, controls, conservative fills, and real option prices. A refusal is a result.
The evidence tape
Research notes
12 frozen studies. New work appends; old misses stay visible.
Repointing our flow read at the raw tape would have made it worse
One vivid morning suggested the raw aggressor tape was the truer source. Across fifty-three independent disagreements the proposed replacement is the weaker of the two, so the swap is refused.
REFUSED · the proposed replacement is the weaker signDealer gamma regime does not separate our highest-conviction reads
A widely-held idea says directional signals should be trusted less when dealer positioning is in its dampening state. Across a month of live reads the dampened group performed slightly better, not worse.
NEGATIVE EVIDENCE · the regime split carries no signal hereA rate-of-change filter for large prints selected the losing half
Volume thresholds on a cumulative total are really a statement about the time of day, so a rate-of-change test looks like the obvious upgrade. Run silently for a month alongside the live rule, it picked the worse prints.
REFUSED · stays in shadowBuying the mirror contract after a washout is a coin flip
After a same-day contract prints a violent low and recovers, buy the opposite side at the same strike. Sixty-five paper trades later it is break-even, and it does not survive dropping its best day.
REFUSED · paper clamp stays onThe cheap far-out rung is a lottery ticket, and the ticket got cheaper
Attaching a low-priced far-from-the-money contract to our highest-conviction calls. Forward evidence confirms it: a decent peak on the way, and nothing at all at the bell.
REFUSED as a hold · stays a labelled display lineNone of the four opening shapes earned a playbook
Classify the first minutes of the session into one of four shapes, then run the shape’s playbook. Fifty-eight forward instances later, no shape clears a bar — and the best one is the shape nobody proposed.
REFUSED · no shape promotedThe early warning is as good as the full signal — and that is not a promotion
A quiet heads-up fired when a large position is only part-way unwound. It grades level with the loud signal it precedes, which makes it real and still does not justify making it loud.
NEGATIVE EVIDENCE for promotion · stays quietThe guardian ratchet: stop giving winners back
A one-way trailing stop for same-day options. It sleeps until a trade is up 15%, then never lets the result go backwards.
ADOPTED · RUNNINGThe independent SPX 0DTE edge search
Eight causal ideas, real SPXW prices, conservative fills—and no candidate earned access to confirmation.
NO EDGE FOUNDRepeated later-session prints were episodes, not independent signals
Thousands of rows collapsed to a handful of daily leaders, and the best-looking cell still lost to controls.
REFUSEDQuote-attributed aggressor side did not rescue the opening print
Even the strict SOLD cohort was tiny, unstable, and negative for the proposed call fade.
REFUSEDThe opening put-print fade looked real—until 2024
A promising recent-window result failed its independent dates and oldest regime.
REFUSED
House rules
What “real” means here
- Only information available at the decision time.
- Exact same-day option contracts and observed bars.
- Adverse fills, stop-first grading, unresolved missing paths.
- No-feature and opposite-side controls beside the setup.