Prediction markets pay on contract language, not on real-world outcomes. Being directionally right is not enough.
Iran stops fighting
→ You assume YES
"Signed and announced by both parties"
→ Verbal deal = NO
The contract wording is the bet. Reality is just input into whether wording triggers YES or NO.
Find the resolution source
Look for 'resolves according to [X]' — that source is authoritative. Your news source is not.
Find the resolution trigger
Is it an announcement? A signed document? A published report? A single agency decision?
Find the resolution deadline
Exact date and time zone. Events near the deadline edge are the highest wording-trap risk.
Ask: can the event happen in a way the wording doesn't capture?
This is the directionally-right-contractually-wrong test. If yes, size down or avoid.
| Dimension | Kalshi | Polymarket |
|---|---|---|
| Wording style | Formal / legal | Varies (community + editorial) |
| Resolution source | Designated source (per market) | UMA oracle / designated source |
| Dispute path | Kalshi support / CFTC | UMA dispute / committee |
| N/A resolution | Yes (rare — ambiguous events) | Yes (rare — ambiguous events) |
| Where to find criteria | Market rules tab | Resolution source field |
Prediction markets are not betting on what you think will happen. They're betting on whether the contract wording triggers YES.
Reading the contract takes 30 seconds. Not reading it can cost you everything you were right about.
