About

A builder and his machines, trading one book by rule.

One systematic operation on crypto perpetual futures. A ruleset scans a fixed universe, takes what it finds, and exits on rule — never on opinion.

Live since 18 Jul 2026 One book One operator No clients

How a trade lives

Six steps. No human touches any of them.
01SignalRuleset scans the universe
02EntryFilled at next open, fixed 1R risk
03ManagedThe stop ratchets by rule only
04ExitStop or trailing rule — no override
05BookedReconciled to exchange fills
06PublishedAppears on the record after 48h
100%of closed live trades reconciled to exchange fills, fees and slippage included
/
48hpublication delay — deliberately useless as a signal feed

The envelope was fixed before the first trade closed

1

Bands fixed 19 July 2026 — the book's first week, before any trade had closed. We simulated the ruleset's own outcome distribution thousands of times and published the resulting percentiles.

2

The live line must earn its place inside. Nothing about the bands has been redrawn, re-fitted or re-scaled since they were fixed.

3

Below the 5th percentile = pre-committed review. It has fired once — August 2026. The review ran; no parameter changed.

A drawdown only carries information when measured against a yardstick fixed while you still had no idea which way it would go. See the live chart →

Every idea meets a locked pass bar

Criteria first, data second, verdict recorded.
01Write the pass criteriaRequired edge, minimum sample, and the control tests it must survive — registered before the idea sees data.
02Run the test onceHalf-split, permutation null, monster retention. Survivors stage on a paper twin running the same signals live.
03Verdict recordedAdopted at a scheduled review or killed and published. Goalposts never move.
16ideas killed and published with their cause of death
·
3survivors in staging — adoption verdict 1 Oct 2026
·
0adopted by exception

What we are, and what we are not

✓ What this is

  • ✓Mechanical — rules compute every decision
  • ✓Outcomes published, wins and losses alike
  • ✓Losses shown at full size, never de-emphasised
  • ✓Dead ideas kept on the record permanently
  • ✓Results judged against pre-registered criteria

✗ What this is not

  • ✗Discretionary — no overrides, no gut calls
  • ✗Investment advice or a recommendation
  • ✗A signal service — the delay makes copying useless
  • ✗A course, group, tier or waitlist. Nothing is for sale
  • ✗A fund. We manage no outside money

Timeline

Reviews are scheduled, never reactive.
Jul 2026Ruleset validated; envelope fixed 19 July — before the first close
18 Jul 2026Live with one book — running continuously since
19 Aug 2026Envelope breach triggered the pre-committed interim review — no parameter changed

Questions we get

Four answers.
Why R-multiples instead of dollars?

One R is the amount lost if a trade's initial stop is hit. Reporting in multiples of that makes the record independent of account size and impossible to flatter by quietly changing leverage.

A +6R trade earned six times the risk it was allocated, whatever the capital behind it. Dollar figures mostly communicate how much money someone has.

Why the 48-hour delay?

A live feed of open positions would turn a record into a product, and that changes what the site is for.

The delay makes the record honest but useless to copy. That keeps the incentive on whether the numbers hold up over time, not on how many people follow them today.

Why publish the failures at all?

Selection is the easiest way to lie with true numbers. Show ten strategies, publish the one that worked, and every individual figure is accurate while the overall impression is false.

The graveyard is the only cheap defence against that — it shows the denominator, not just the winners.

Who is behind it?

One builder and a stack of machines. We stay pseudonymous on purpose.

The argument here should rest on pre-registered criteria and published outcomes, not on anyone's biography or follower count. If the record stops holding up, no amount of credentials should save it.

The record is the argument.

Every closed trade, the envelope, and the graveyard — all on one page.

See the full record →