Markets

One endpoint, verbatim

Many trials list two or more primary outcomes. TRAILBLAZER-ALZ 2 lists two. A Placebo market names exactly one, copied word for word from the registry, with its alpha (0.05 unless the registry states another) and the direction that counts as a win. YES needs that measure's first prespecified analysis to report a p-value under alpha, with superiority, in that direction.

The frozen spec

The market is the JSON below. Its keccak256 is the specHash the keeper passes to createMarket, so anyone can check that the question on the page is the one people staked on. The market id on chain is the keccak256 of the NCT id.

{
  "v": 1,
  "id": "NCT06125951",
  "nct": "NCT06125951",
  "question": "Will XanaMIA meet its primary endpoint at alpha 0.05?",
  "measure": "Effects of 10 mg Xanamem on integrated cognitive and functional abilities",
  "alpha": 0.05,
  "direction": "lower",
  "deadline": "2027-12-31",
  "ladder": [
    "registry",
    "sponsor",
    "fda",
    "stopped",
    "void"
  ],
  "expected": 2,
  "domains": null
}

id 0x07a584a4e1d9d7866ed0b8d4328a0c4ea31ba7e8caaba318758db0cc6a8e5c16
specHash 0x3db9c49b1b0b61d36b2f06e1b085fbc212c3af6b216930a9919829aef27c3a3a, over 339 bytes, keys sorted, no whitespace

Registry drift

Sponsors edit their registry records. The market keeps the record it was born from; the site shows any later change as a diff beside the market, and the change never alters the question.

The launch set

Guided dates, such as XanaMIA's topline in November 2026, are printed as the sponsor's guidance with its source, never as a registry fact.