Know further.
Implausible measures human prediction. Under sealed conditions. Against chance.
Every day you act on instinct.
On people. On markets. On what happens next.
Some of those calls are right.
Nobody has ever shown you how many.
Chance is the line.
Guess blindly and half your calls still land. Every score we give you is measured against that line. Nothing else counts.
We don't ask you to believe in intuition.
We don't dismiss it either. We built an instrument to measure it.
Take a closer look
The instrument.
Five parts. Each one exists to stop you, us, or luck from flattering the result.
- Blinded protocolThe server assigns conditions, including silent controls.
- State inductionA short game quiets deliberate thinking first.
- Sealed targetCommitted with a cryptographic hash before you choose.
- Prediction ledgerEvery call sealed, timestamped and chained to the last.
- MetamodelLearns which of your conditions move your hit rate.
One session
A few minutes.
Four movements.
01 Engage
A fast game pulls your attention out of your head.
02 Absorb
The rhythm slows. Deliberate thinking goes quiet.
03 Descend
The target is sealed. You still can't see it.
04 Signal
Stillness. One call, on first instinct.
Your turn
A or B?
This target was sealed when you opened the page.
sha256 sealing
First instinct. Go.
One call means nothing. Hundreds show whether your hit rate sits above 50%, and when it moves.
Two doors. One instrument.
Your instinct is a skill. Or it isn't. Now you can know which.
For people
Find your hit rate.
For anyone who trusts their gut and would like proof either way. Short sessions on your phone, scored honestly over weeks.
- Three launch gamesFlow Pong, Grid Oracle, Flash
- Three kinds of trialRandom targets, sealed images, real events
- Your MetamodelWhen you're sharpest, session counts shown
- The CohortEarned by sessions, never bought
For organisations
Put the instrument to work.
Score the instinct your decisions already run on. Query the trained Metamodel on your own questions. Or study the phenomenon with us.
- The AuditScore your people's instinct
- Metamodel accessA calibrated view on your questions
- Research accessFor universities and labs
Integrity
Built to be doubted.
Extraordinary questions need ordinary rigour. If our results can't survive a sceptic, they aren't results.
- Sealed before you choose.Targets are committed by hash, so nothing can change after the call.
- Scored against chance.Every number names its baseline. No baseline, no number.
- On the ledger.Every call joins a chained, tamper-evident record. Your data is encrypted at rest and in transit.
- Session counts on everything.A hit rate from twelve sessions says so. Patterns stay exploratory until the data earns more.
- Null results published.If the data says chance, you'll read it from us first.
- Separated at source.Your experiment data sits under a random ID, apart from your name and email. Population patterns need at least thirty sessions per group.
The long question