Quantato — The Experiment Cockpit

Stage 1 · The board Every hypothesis, one glance Each card carries the one metric that matters at its stage. Click any card to open its cockpit.

Hypothesis2

Registered1

Burn-in2

Running2

Concluded3

Stage 2 · The cockpit Weather NYC temp overs Everything you need to read the state, the evidence, and decide what happens next.

Weather NYC temp overs Running weather paper

Collecting evidence since Jun 1 · anytime-valid test, safe to read at any moment.

−$0.40graduate line · $0+$0.40

evidence band (two one-sided α-bounds — not a 95% CI) graduate at $0 kill below ev_min ($0.05)

Leaning graduate. Lower bound is −$0.004 — it needs to clear $0 to graduate. At the current rate that's roughly 60 more settled contracts (~3 days). No action required; the verdict fires itself.

Settled

180

Net PnL

+$5.58

EV / contract

+$0.031

Wins / Losses

104 / 76

Fill rate

92%

Decision

continue

Loss-harvest status new

● Active 12 losing contracts harvested this cycle · offsets $1.84 against funded PnL

Harvest windowrolling 30d

Wash-sale guardon · 31d cooldown

Lineage new

#41 SOL momentum fade → #57 this experiment → refine into #58 ?

Sharpened from #41 by tightening the entry band. Refine spawns a child hypothesis with a narrower pre-registration, re-validated on fresh data — the parent's evidence never leaks in.

🔒 Pre-registration · locked

entry band8 – 46¢

ev_min (per ct)$0.05

min / max sample20 / 300

alpha0.05

OOS boundary2026-06-01

familyweather

Locked when the experiment was registered. Only title / hypothesis / mechanism stay editable — the bar can't move to fit the result.

Does the follower reprice after a BTC lead move? · 214 qualifying BTC moves observed

ETH follows BTC

KXETH15M · lag window 180s

58%hit-rate · n=142

Above chance. ETH makes a same-direction move 58% of the time after BTC leads — the marker at 50% is coin-flip. This trigger is worth graduating to a live experiment.

SOL follows BTC

KXSOL15M · lag window 180s

—insufficient · n=88

Sample below the 20-move floor for a confident read at this threshold. Keep the monitor running — don't graduate SOL yet.

Reference · the rules What each status lets you do Derived from the server-enforced state machine. Illegal moves are simply never offered.

Status	What you can do	What it does
Draft hypothesis only	Register	Locks the pre-registration so the bar can't move to fit the result.
Registered	Send to burn-inStart (go live)Park	Calibrate a trigger first, or start collecting evidence now.
Burn-in	Start (go live)Park	Graduate the calibrated trigger to a live evidence clock.
Running	RefinePausePark→ passed / failed (auto)	The verdict fires itself from the sequential test. You only pause, park, or spin off a sharper child.
Paused	ResumePark→ passed / failed (auto)	Pick the evidence clock back up, or retire it.
Passed graduated	Fund this edgeRefine	Deliberate money step — re-activates the validated strategy for live trading.
Failed / Parked terminal	Refine	No live actions — but you can spawn a sharpened child from what you learned.

primary the move you'll usually want neutral reversible destructive confirms first, irreversible money real capital new coming this phase auto the system decides

Graduated · the money step Fund a validated edge Only a passed experiment with a strategy can be funded. This is where research becomes real capital.

passed SOL momentum fade graduated · EV/ct +$0.041 · lower bound cleared $0

Funding re-activates the strategy this experiment validated and starts placing live orders inside its pre-registered band. Actual fills still gate on the strategy loop's kill-switch. Leave a field blank to keep the experiment's own sizing.

Max contracts / order Allocated capital ($)