HOW IT ARGUES
Three moves. No guesses.
01 — YOU PITCHenglish, or a chart
"You showed me winners — that's what eyes do. I'll test every look-alike since 1993, losers included."
02 — IT INTERVIEWSambiguity earns a question
"Here's what I heard — every dial is adjustable."
QUESTION 1 OF 1 — I DON'T GUESS
Two exits could apply at 21 days — take whichever hits first, or profit target only?
03 — IT ATTACKS ITSELFthen the verdict
Backtest — real bid/ask, never midreal bid/ask, never mid
○Test on data it never sawlast 30% kept hidden
○Test each time periodrolling ~2-month windows
○Reshuffle the trades 1,000×how much was luck?
○Nudge the settings ±20%does it survive small changes?
▶The honest verdictgrounded in the numbers above
live numbers stream while it runs — never a loading bar
MOST BACKTESTERS TELL YOU WHAT YOU WANT TO HEAR
Refusal is a feature.
Twelve closed trades in a year isn’t evidence — it’s anecdotes. Skeptic ships the numbers, withholds the blessing, and shows exactly what unlocks a verdict.
VERDICT — THE HONEST READVERDICT WITHHELD
12 closed trades is not a sample, yet the strategy claims a perfect win rate and zero drawdown.
The spec is valid. The engine ran it. But only 12 trades closed inside the 1-year window — below the 15-trade evidence bar. Numbers shown, blessing withheld.
TWO HONEST WAYS TO A VERDICT
re-run on a longer window — unlocks at ≥ 15 trades (has 12)edit the spec — make the entry fire more often
SPY .15Δ short put · 12 trades in a 1-year window · ran Jul 14 ’26
THE RANGE — RECENT VERDICTS
refused, damned, or blessed — whatever the evidence supports.
rendered from real stored runs — every number computed, none decorative
Research tool, not financial advice.
RECEIPTSevery fallback disclosed · gaps stay gaps, never interpolated
The data earns the verdicts.
—
SESSIONS ON TAP
0DTE
MINUTE-LEVEL FILLS
bid/ask
NEVER MID
—
CHAINS SINCE
nightly lake · every fallback disclosed · gaps stay gaps, never interpolated
COMING NEXTWHERE THIS GOES — COMING SOONautonomous session — replayed
Coming next: itIt takes the trade itself.
Enters, manages, exits — and narrates every call as it makes it. Still refuses what the evidence can't support.
SPY · 5m · demo feedfills from the stored run
REASONING — LIVE
Mar 4 · scanning the tape — hunting rich premium at 30Δ
Mar 4 · SOLD 1× SPY 512P @ 2.31 (bid 2.29 / ask 2.35)
why · target 50% · time exit 21 DTE · theta does the work
Mar 18 · BOUGHT BACK @ 1.12 — target hit · +$115 · held 14d
Apr 8 · next entry looks juicy — but the spread is 31% of mid. standing aside.
edge spent. every call logged + replayable — no human touched this trade.
demo replay of a stored run · research tool, not financial advice
PRICING
Yup, this is the whole price list.
1free backtest — no account, no cardfree backtest, right nowno account, no card
5free with an account — still no cardfree with an accountyes account, still no card
COMING SOON
$10 → 50one-time — no subscription, no tiersone-time, fifty moreno subscription, no tiers
“Not enough evidence” refunds the credit — you only spend on a graded verdict.
same if a run fails on our end — the credit comes back on its own