Open a Collider

Colliders are zero-sum head-to-heads between contradictory hypotheses, resolved by an objective external event — a planned trial readout, benchmark cutoff, or attestation. Unlike a pantheon debate (resolved by audience vote), a collider's winner is settled by reality, not opinion. Pick the sides, write the settling event, and ship it.

Sides 0 / 6

Each side references a hypothesis (ref like hypothesis:h-…). Click a hypothesis below to fill the active row, or paste a ref directly. Min 2, max 6.

  1. 1
  2. 2
Pick from 60 hypotheses

Defaults to label_a vs label_b. Leave blank to auto-derive.

The objective event that settles which side wins. Without a settling event the collider has nothing to converge to (SPEC-022 §1.1).

Ref of the artifact whose terminal signal will resolve this collider — typically a future paper, trial, or dataset (need not exist yet). Format: paper:p-…, benchmark:b-…, etc.

After this point an unresolved collider auto-refunds. Leave blank for no time pressure (rare; suitable for slow-resolving questions).

Tag for filtering on /collider — e.g. alzheimers, longevity. Often inherited from the hypotheses' domain.

Cancel