Spreadwire

About

Spreadwire finds sharp moves in election markets and assesses them against available headlines. It preserves the original reasoning, evidence and later observations. Prediction-market observations are clearly separated without presuming equivalent contracts or quote types. These AI assessments do not prove causation or future price direction.

Observed processing record, last 30 days

1011retained threshold-crossing moves, before/after the mechanical check
58assessed as “real move”
48assessed as overreaction
78assessed as noise
730failed the 90-second check, with no alert
ElectionRetained movesReal-move assessmentOverreactionNoiseReverted ≤90sOther/pending*
● US 2028411333166
● US Midterms 2026 (final)2941171924413
● Présidentielle 202767634385647078

* minor events explained on demand or awaiting analysis — every row reconciles to the total.

17 markets watched. Observed counts, not predictive-accuracy scores.

What Spreadwire provides

How it works

Detect (statistical thresholds on live prices) → Confirm (90-second re-validation) → Explain (AI verdict grounded in the news, with sources, in ~2 minutes). Full method: methodology.

Press & data

Every verdict has a permalink, machine-readable versions (Markdown / JSON), and a chart image. Spreadwire verdicts may be cited with attribution and a link. Market data remain subject to their respective platforms’ terms of use.

Read the Ossoff case study · Read the observed follow-through record

Who

Built and run independently by @jlcossi — contact: DMs open on X.

Price data: publicly observable prediction-market prices. Spreadwire is independent of these sources.

Version française

Spreadwire analyses publicly observable prediction-market data. These platforms are not authorised in France; Spreadwire does not promote, link to or recommend them. Political-risk analysis, not investment advice.