couchdicks/ PAPER LAB 03
CONTACTING RECORDERS

FORWARD-SHADOW / THREE CHAINS / ZERO REAL CAPITAL

Paper Trader
Observatory

A public flight recorder for our Robinhood, Base, and Avalanche trading experiments. Controls and challengers see the same evidence. Nothing promotes itself.

MODEPAPERNOT FINANCIAL ADVICE
LOCAL MARKS ONLYNO REMOTE SIGNAL APINO SYNTHETIC FRAMESHUMAN PROMOTION ONLY
01

LIVE FLEET

Three recorders, one standard of proof.

Waiting for first public read…

Reading paper ledgers…
02

THE TOURNAMENT

Control versus one isolated mutation.

Forward-only from

The challengers are entering the arena…
03

THE LOCK

A pretty backtest cannot open this door.

  1. 60candidate closes
  2. 20×3closes in every purged fold
  3. 2/3positive folds
  4. ≥1.15profit factor
  5. ≤50%top-winner concentration
  6. 7 daysminimum forward soak

Passing means review ready, not deployed. A human still inspects the ledger, data coverage, drawdown, and failure modes.

04

METHOD / READ THIS PART

What this page is—and isn’t.

Observed. DEX market marks are sampled by the paper traders. A persistent watchlist keeps real assets visible long enough to form causal 2m and 5m frames. Missing observations stay missing.

Paired. The deployed control and challenger replay the same forward evidence with the same sizing, friction, and exit machinery. Each challenger changes one rule.

Constrained. These are simulated fills, not audited investment returns. Rug-risk gates are currently disabled and called out in the chain desks. Nothing here is a recommendation.