prediction_markets.markets

The prediction_markets.markets table is the cross-venue market dimension. Grain: one row per binary market — Polymarket condition_id, the combo condition ID for Polymarket Combos (multi-leg parlays, is_parlay = TRUE), or Kalshi ticker. Every row carries a venue column and uses venue-neutral column names so you can query both venues with a single statement. Market names and attributes live here; prediction_markets.trades and prediction_markets.ohlcv_hourly carry facts only and join on (venue, market_id). Every market is presented from its Yes side. On Polymarket, Yes is the token in position 0 (outcome_index = 0), the side the venue’s settlement resolves against, so yes_outcome_id, resolved_outcome and settlement_value_usd always agree. Kalshi covers every single market the venue has listed, traded or not, plus every parlay with at least one fill.

For markets without a natural Yes (team vs team, candidate vs candidate), Yes is simply whichever side the venue listed first. Check yes_outcome_name before reading the Yes side as an affirmative answer.

Table Schema

Column	Type	Description
`venue`	`VARCHAR`	Venue identifier (`polymarket` or `kalshi`)
`market_id`	`VARCHAR`	Market identifier, unique within a venue: Polymarket `condition_id` in 0x-hex, the combo condition ID for Polymarket Combos, Kalshi ticker. Joins `prediction_markets.trades` and `prediction_markets.ohlcv_hourly` on `(venue, market_id)`
`event_id`	`VARCHAR`	Parent event grouping related markets. Always populated on Kalshi; on Polymarket only for markets that belong to a multi-outcome (neg-risk) event. NULL for Combos
`series_id`	`VARCHAR`	Recurring series the market was created from. Kalshi only; NULL on Polymarket, which has no series concept
`yes_outcome_id`	`VARCHAR`	Identifier of the Yes side: the token at `outcome_index = 0` on Polymarket, the market itself on Kalshi, where both sides trade on one instrument
`no_outcome_id`	`VARCHAR`	Identifier of the No side: the token at `outcome_index = 1` on Polymarket. NULL on Kalshi
`yes_outcome_name`	`VARCHAR`	Label of the Yes side as the venue shows it, usually `Yes`, `Up` or `Over`. For markets without a natural Yes it is whichever side the venue listed first
`no_outcome_name`	`VARCHAR`	Label of the No side as the venue shows it
`question`	`VARCHAR`	The market question. For Combos the generated combo name, NULL until every leg is labeled
`event_name`	`VARCHAR`	Title of the parent event. NULL for standalone Polymarket markets and Combos
`category`	`VARCHAR`	Unified topic category: `sports`, `crypto`, `politics`, `finance`, `technology`, `culture`, `weather`, `world`, `health`, `other`. Parlays take the category of their legs, or `mixed` when the legs span several categories
`category_native`	`VARCHAR`	The venue’s own classification: Polymarket’s comma-separated tags, Kalshi’s single coarse category. NULL for Polymarket Combos
`link`	`VARCHAR`	URL of the market on the venue’s website. NULL for parlays on both venues, which have no public page
`status`	`VARCHAR`	Lifecycle state: `active` while open for trading (including Kalshi markets not yet opened), `closed` when trading has stopped or the venue delisted the market without a result, `settled` once resolved
`is_resolved`	`BOOLEAN`	TRUE once the market has a final result. A resolved market is always `settled`; a Kalshi market can be `settled` without a published result, in which case this is FALSE
`is_parlay`	`BOOLEAN`	TRUE for multi-leg parlays: Kalshi parlays and Polymarket Combos. Filter `is_parlay = FALSE` for like-for-like single-market comparisons
`leg_count`	`INTEGER`	Number of legs in a parlay. NULL for single markets
`legs`	`VARCHAR`	The parlay’s legs as a JSON array. On Kalshi each leg names its event, market and required side; on Polymarket each leg is a `question: outcome` label. NULL for single markets
`is_mutually_exclusive_event`	`BOOLEAN`	TRUE when at most one market in the parent event can resolve Yes, such as a multi-candidate election (Kalshi mutually exclusive events, Polymarket neg-risk events). FALSE when unknown
`frequency`	`VARCHAR`	How often the market’s series recurs: `fifteen_min`, `hourly`, `daily`, `weekly`, `monthly`, `quarterly`, `annual`, `one_off` or `custom`. Kalshi only; NULL on Polymarket. Useful for excluding high-frequency series from cross-venue aggregates
`market_start_time`	`TIMESTAMP`	When trading opened. For Combos, when the combo was registered on-chain
`market_end_time`	`TIMESTAMP`	Scheduled close of trading. On Kalshi this reflects later changes to the close date. On Polymarket it is the venue’s published end date, which is often inaccurate; use `settlement_ts` and `status` for the actual lifecycle. NULL for Combos
`settlement_ts`	`TIMESTAMP`	When the market settled. NULL until resolved
`resolved_outcome`	`VARCHAR`	How the market resolved: `yes`, `no`, `void` for a refund, or `scalar` for a fractional payout (Kalshi scalar markets, Polymarket Combos with a voided leg). NULL while unresolved
`settlement_value_usd`	`DOUBLE`	Payout per Yes contract at settlement: `1.0` when Yes won, `0.0` when No won, `0.5` for a void, a fraction for scalar settlements. NULL while unresolved
`strike_type`	`VARCHAR`	Structure of a Kalshi threshold or range market (`greater`, `less`, `between`, and so on). NULL on Polymarket
`floor_strike`	`DOUBLE`	Lower strike of a Kalshi threshold or range market. NULL on Polymarket
`cap_strike`	`DOUBLE`	Upper strike of a Kalshi range market. NULL on Polymarket
`rules`	`VARCHAR`	Resolution rules as published by the venue. NULL for Polymarket Combos, whose outcome follows from their legs
`source_updated_at`	`TIMESTAMP`	When the market record last changed at the venue; for Polymarket Combos, when Dune last recomputed the combo from its legs. Safe to use as a sync watermark
`_updated_at`	`TIMESTAMP`	When this row was last written by the pipeline. Unchanged rows keep their stamp

Table sample

Example queries

-- Active single markets by category across both venues
SELECT
  venue,
  category,
  COUNT(*) AS active_markets
FROM prediction_markets.markets
WHERE status = 'active'
  AND is_parlay = FALSE
GROUP BY 1, 2
ORDER BY 3 DESC

-- Markets settled in the last 7 days with their Yes-side payout
SELECT
  venue,
  market_id,
  question,
  yes_outcome_name,
  resolved_outcome,
  settlement_value_usd,
  settlement_ts
FROM prediction_markets.markets
WHERE is_resolved
  AND settlement_ts >= NOW() - INTERVAL '7' DAY
ORDER BY settlement_ts DESC
LIMIT 50

Curated Data

Solana

Hyperliquid

Non-EVM Networks

Protocols

EVM Networks

Partner Data

Dune Index

prediction_markets.markets

Table Schema

Table sample

Example queries

​Table Schema

​Table sample

​Example queries

Table Schema

Table sample

Example queries