Datasets:

lowskillcoding
/

pdata-prediction-markets

+---
+license: cc-by-4.0
+pretty_name: "pdata.world prediction markets (Polymarket, Kalshi and 6 more venues)"
+language:
+- en
+tags:
+- prediction-markets
+- polymarket
+- kalshi
+- manifold
+- forecasting
+- finance
+- tabular
+size_categories:
+- 100K<n<1M
+configs:
+- config_name: resolved
+  data_files: data/resolved.parquet
+  default: true
+- config_name: markets-open
+  data_files: data/markets-open.parquet
+- config_name: events-open
+  data_files: data/events-open.parquet
+---
+# pdata.world prediction markets
+Prices, volumes and results for prediction markets on eight venues — Polymarket, Kalshi,
+Manifold, Myriad, Limitless, Predict, Opinion and Gemini — in one schema. The data is
+collected by [pdata.world](https://pdata.world), which polls each venue every 10 to 60
+minutes, and is published under CC BY 4.0.
+This snapshot was taken on 2026-10-07. It has three parts:
+| Config | Rows | What one row is |
+|---|---|---|
+| `resolved` (default) | 450,216 events, with their top 1,752,957 markets (at most 10 each) | An event pdata stopped tracking after it closed, with each top market's reported result |
+| `markets-open` | 417,132 | A market open for trading: price, 24h move and volume |
+| `events-open` | 38,356 | An open event, which groups the markets of one question |
+```python
+from datasets import load_dataset
+resolved = load_dataset("lowskillcoding/pdata-prediction-markets", "resolved", split="train")
+markets = load_dataset("lowskillcoding/pdata-prediction-markets", "markets-open", split="train")
+# or with pandas
+import pandas as pd
+df = pd.read_parquet("hf://datasets/lowskillcoding/pdata-prediction-markets/data/markets-open.parquet")
+```
+## Rows per venue
+- **resolved:** kalshi 194,286, gemini 118,368, polymarket 115,407, predict 15,216, limitless 6,438, opinion 344, manifold 148, myriad 9
+- **markets-open:** polymarket 268,382, kalshi 112,611, gemini 11,900, manifold 10,848, predict 10,667, limitless 1,435, opinion 1,255, myriad 34
+- **events-open:** polymarket 18,596, kalshi 13,751, predict 2,211, manifold 1,672, gemini 1,273, limitless 547, opinion 272, myriad 34
+## Read this before using the numbers
+- `probability` and prices are on a 0–1 scale.
+- **Volume is not comparable across venues.** `volume` and `volume_24hr` are in each venue's
+  own unit: shares for Polymarket and contracts for Kalshi (each pays $1 at resolution),
+  play-money mana for Manifold, and each other venue's figure as it reports it. pdata's
+  report [How eight venues count volume](https://pdata.world/research/prediction-market-volume)
+  explains each one.
+- **In `resolved`, a market's `probability` is the last price pdata saw**, which is usually
+  after the market settled. It records the result; it is not a forecast. For prices before
+  close, use the API's history endpoint while a market is live (history is deleted 10 days
+  after close) — see [How accurate were prediction-market prices a day before the
+  result?](https://pdata.world/research/prediction-market-accuracy) for one way to do it.
+- **`result` is null where the venue never reports one.** Of the 1,752,957 archived markets,
+  1,006,292
+  carry a result.
+- **A gap in `resolved`:** events that closed between 2026-06-02 and 2026-08-11 are missing,
+  because they were pruned before a backup was taken. Events closed in the 10 days before
+  the snapshot are in neither file yet.
+- pdata groups Kalshi's sibling events for one game under a single group event
+  (`is_group`), whose volume is the sum of its children's.
+## Fields
+### resolved
+Every event pdata has stopped tracking live, with its final result: the record pdata keeps when it prunes an event, 10 days after it closes. Events closed more recently are in neither file yet. Kept since 2026-08-31; events pruned before then were rebuilt on 2026-10-07 from backups taken 2026-05-09 to 2026-08-29. Events that closed between 2026-06-02 and 2026-08-11 are missing: they were pruned before the next backup was taken.
+| Field | Meaning |
+|---|---|
+| `source` | Venue: polymarket, kalshi, manifold, myriad, limitless, predict, opinion or gemini. |
+| `id` | The venue's event id, unique within source. |
+| `title` | Event title when pdata pruned it. |
+| `categories` | pdata topic categories (e.g. Politics, Sports), from rules and AI tagging. |
+| `closed_time` | When the event closed (UTC): the venue's close time, else its end date, else when pdata last saw it. |
+| `markets` | The event's top markets by lifetime volume, at most 10: id, question, result (as the venue reported it; null where the venue never says), final probability (0-1) and lifetime volume (venue unit). |
+| `url` | The event or market's own page on the venue's website; null when it cannot be built. |
+| `pdata_url` | pdata's page for the event, https://pdata.world/events/{source}/{id}. |
+| `pruned_at` | When pdata wrote this record (UTC): when it pruned the event, or 2026-10-07 for an event pruned before 2026-08-31 and rebuilt from backups. |
+### markets-open
+Every market open for trading on the 8 venues pdata tracks, one per line: price, 24h move and volume. A market is one yes/no question, or one outcome of a multi-outcome event.
+| Field | Meaning |
+|---|---|
+| `source` | Venue: polymarket, kalshi, manifold, myriad, limitless, predict, opinion or gemini. |
+| `id` | The venue's market id, unique within source. |
+| `event_id` | The parent event's id (same source); joins events-open on (source, id). |
+| `event_title` | The parent event's title. Read question with it: an outcome of a multi-outcome event can be a bare label. |
+| `question` | The market's question as the venue titles it. |
+| `label` | Short outcome label where the venue gives one (e.g. a candidate's name). |
+| `categories` | pdata topic categories (e.g. Politics, Sports), from rules and AI tagging. |
+| `outcomes` | Outcome labels in venue order, usually ["Yes", "No"]. |
+| `outcome_prices` | Price of each outcome, 0-1, in outcomes order. |
+| `probability` | Implied probability of the first outcome (usually YES), 0-1. |
+| `prob_24h_change` | Change in probability over the trailing 24h, on the same 0-1 scale (0.05 = up 5 points). Null when there is no snapshot in that window. |
+| `volume` | Lifetime trading volume. Venue-reported, in the venue's own unit and not converted across venues: shares for Polymarket and contracts for Kalshi (each pays $1 at resolution), play-money mana for Manifold, and each other venue's figure as it reports it. |
+| `volume_24hr` | Trading volume over the last 24 hours. Venue-reported, in the venue's own unit and not converted across venues: shares for Polymarket and contracts for Kalshi (each pays $1 at resolution), play-money mana for Manifold, and each other venue's figure as it reports it. |
+| `liquidity` | Liquidity as the venue reports it, in the venue's own unit. |
+| `open_interest` | Open interest as the venue reports it, in the venue's own unit. |
+| `start_date` | When the market opened, as the venue reports it (UTC). |
+| `end_date` | Scheduled close or expiry, as the venue reports it (UTC). |
+| `status` | The venue's own status string, verbatim; null for Polymarket. |
+| `countries` | Countries the market concerns (full English names, at most 3), from pdata's AI tagging. [] means none applies; null means not tagged yet. |
+| `updated_at` | When pdata last wrote this row (UTC): every fetch from the venue, and again when pdata recomputes a derived field. |
+| `url` | The event or market's own page on the venue's website; null when it cannot be built. |
+| `pdata_url` | pdata's page for the event, https://pdata.world/events/{source}/{id}. |
+### events-open
+Every event open for trading on the 8 venues pdata tracks, one per line. An event groups the markets of one question; markets-open holds its markets.
+| Field | Meaning |
+|---|---|
+| `source` | Venue: polymarket, kalshi, manifold, myriad, limitless, predict, opinion or gemini. |
+| `id` | The venue's event id, unique within source (pdata builds the id of a group event). |
+| `title` | Event title as the venue publishes it. |
+| `categories` | pdata topic categories (e.g. Politics, Sports), from rules and AI tagging. |
+| `start_date` | When the event opened, as the venue reports it (UTC). |
+| `end_date` | Scheduled close or expiry, as the venue reports it (UTC). |
+| `volume` | Lifetime trading volume. Venue-reported, in the venue's own unit and not converted across venues: shares for Polymarket and contracts for Kalshi (each pays $1 at resolution), play-money mana for Manifold, and each other venue's figure as it reports it. |
+| `volume_24hr` | Trading volume over the last 24 hours. Venue-reported, in the venue's own unit and not converted across venues: shares for Polymarket and contracts for Kalshi (each pays $1 at resolution), play-money mana for Manifold, and each other venue's figure as it reports it. |
+| `open_interest` | Open interest as the venue reports it, in the venue's own unit. |
+| `mutually_exclusive` | True when exactly one of the event's markets can resolve YES; false when they are independent questions; null when the venue does not say. |
+| `market_count` | How many markets the event has (for a group event, across its child events). |
+| `is_group` | True for an event pdata built to group sibling Kalshi events of one matchup. |
+| `parent_event_id` | The pdata group event this event belongs to, or null. |
+| `synthetic` | True for an event pdata created to wrap a market the venue lists without one. |
+| `updated_at` | When pdata last wrote this row (UTC): every fetch from the venue, and again when pdata recomputes a derived field. |
+| `url` | The event or market's own page on the venue's website; null when it cannot be built. |
+| `pdata_url` | pdata's page for the event, https://pdata.world/events/{source}/{id}. |
+Timestamps are UTC. The manifest each file was built from is in [`manifests/`](manifests/).
+## Updates and versions
+A new snapshot is added monthly; each is tagged with its date (this one is `2026-10-07`),
+so `load_dataset("lowskillcoding/pdata-prediction-markets", "resolved", revision="2026-10-07")` always returns the same
+rows. For daily data, the same three files are rebuilt every night at
+<https://api.pdata.world/bulk/> as gzipped JSON lines.
+## Other ways to get the data
+- REST API, no key: <https://api.pdata.world/docs>
+- MCP server for AI assistants: `https://api.pdata.world/mcp`
+- Read-only Postgres with price history:
+  `postgresql://pdata_readonly:pdata_public_2026@pg.pdata.world:6433/pdata_new`
+  (the password is published on purpose)
+- Python client and examples: <https://github.com/noskillcoding/pdata-client>
+- Research built on this data: <https://pdata.world/research>
+## Licence and citation
+CC BY 4.0. Credit "Data: pdata.world" with a link to <https://pdata.world>. The underlying
+market data remains subject to each venue's own terms. pdata.world is independent and not
+affiliated with any venue.
+```bibtex
+@misc{pdata_world_prediction_markets_2026_10_07,
+  title        = {pdata.world prediction markets: prices, volumes and results from eight venues},
+  author       = {{pdata.world}},
+  year         = {2026},
+  note         = {Snapshot 2026-10-07},
+  howpublished = {\url{https://huggingface.co/datasets/lowskillcoding/pdata-prediction-markets}}
+}
+```

build.py ADDED Viewed

	@@ -0,0 +1,78 @@

+"""Convert pdata.world's nightly bulk files (gzipped JSON lines) to Parquet
+for the Hugging Face dataset. Usage: build.py BULK_DIR OUT_DIR
+Line 1 of each bulk file is a manifest; it is written to OUT_DIR/manifests/
+and the rows to OUT_DIR/data/<kind>.parquet with an explicit schema, so a
+column never changes type between monthly snapshots.
+"""
+import gzip
+import json
+import sys
+from datetime import datetime
+from pathlib import Path
+import pyarrow as pa
+import pyarrow.parquet as pq
+TS = pa.timestamp("us", tz="UTC")
+STR_LIST = pa.list_(pa.string())
+SCHEMAS = {
+    "markets-open": pa.schema([
+        ("source", pa.string()), ("id", pa.string()), ("event_id", pa.string()),
+        ("event_title", pa.string()), ("question", pa.string()), ("label", pa.string()),
+        ("categories", STR_LIST), ("outcomes", STR_LIST), ("outcome_prices", pa.list_(pa.float64())),
+        ("probability", pa.float64()), ("prob_24h_change", pa.float64()),
+        ("volume", pa.float64()), ("volume_24hr", pa.float64()),
+        ("liquidity", pa.float64()), ("open_interest", pa.float64()),
+        ("start_date", TS), ("end_date", TS), ("status", pa.string()),
+        ("countries", STR_LIST), ("updated_at", TS), ("url", pa.string()), ("pdata_url", pa.string()),
+    ]),
+    "events-open": pa.schema([
+        ("source", pa.string()), ("id", pa.string()), ("title", pa.string()),
+        ("categories", STR_LIST), ("start_date", TS), ("end_date", TS),
+        ("volume", pa.float64()), ("volume_24hr", pa.float64()), ("open_interest", pa.float64()),
+        ("mutually_exclusive", pa.bool_()), ("market_count", pa.int64()), ("is_group", pa.bool_()),
+        ("parent_event_id", pa.string()), ("synthetic", pa.bool_()),
+        ("updated_at", TS), ("url", pa.string()), ("pdata_url", pa.string()),
+    ]),
+    "resolved": pa.schema([
+        ("source", pa.string()), ("id", pa.string()), ("title", pa.string()),
+        ("categories", STR_LIST), ("closed_time", TS),
+        ("markets", pa.list_(pa.struct([
+            ("id", pa.string()), ("question", pa.string()), ("result", pa.string()),
+            ("probability", pa.float64()), ("volume", pa.float64()),
+        ]))),
+        ("url", pa.string()), ("pdata_url", pa.string()), ("pruned_at", TS),
+    ]),
+}
+def convert(bulk: Path, out: Path, kind: str) -> tuple[int, dict]:
+    schema = SCHEMAS[kind]
+    ts_cols = [f.name for f in schema if f.type == TS]
+    with gzip.open(bulk / f"{kind}.jsonl.gz", "rt") as f:
+        manifest = json.loads(next(f))["_manifest"]
+        rows = []
+        for line in f:
+            r = json.loads(line)
+            for c in ts_cols:
+                if r.get(c) is not None:
+                    r[c] = datetime.fromisoformat(r[c])
+            rows.append(r)
+    extra = set().union(*(r.keys() for r in rows)) - set(schema.names)
+    if extra:
+        raise SystemExit(f"{kind}: fields not in the schema: {sorted(extra)}")
+    table = pa.Table.from_pylist(rows, schema=schema)
+    pq.write_table(table, out / "data" / f"{kind}.parquet", compression="zstd")
+    (out / "manifests" / f"{kind}.json").write_text(json.dumps(manifest, indent=2, ensure_ascii=False) + "\n")
+    return table.num_rows, manifest
+if __name__ == "__main__":
+    bulk, out = Path(sys.argv[1]), Path(sys.argv[2])
+    for kind in SCHEMAS:
+        n, m = convert(bulk, out, kind)
+        size = (out / "data" / f"{kind}.parquet").stat().st_size
+        print(f"{kind}: {n:,} rows, {size/1e6:.1f} MB, as_of {m['as_of']}")

data/events-open.parquet ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:c45a06740626ce59edab83c3e6445a98eeaee5b530161cac1185604c57970149
+size 2274937

data/markets-open.parquet ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:e4767a1d5bf157a5ec22f48141fd2d291e6b6673ca763d9f50cd3d4a53e726be
+size 14508185

data/resolved.parquet ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:55c186949651828200806adc37ed2e53b2eeb543038d6a82b3a37e145ae8180e
+size 31722264

make_card.py ADDED Viewed

	@@ -0,0 +1,166 @@

+"""Write README.md (the dataset card) from the manifests and the Parquet files."""
+import json
+from collections import Counter
+from pathlib import Path
+import pyarrow.parquet as pq
+REPO = "lowskillcoding/pdata-prediction-markets"
+KINDS = ("resolved", "markets-open", "events-open")
+here = Path(__file__).parent
+man = {k: json.loads((here / "manifests" / f"{k}.json").read_text()) for k in KINDS}
+tables = {k: pq.read_table(here / "data" / f"{k}.parquet") for k in KINDS}
+as_of = man["resolved"]["as_of"][:10]
+snapshot = as_of
+def fields(kind: str) -> str:
+    rows = "\n".join(f"| `{k}` | {v} |" for k, v in man[kind]["fields"].items())
+    return "| Field | Meaning |\n|---|---|\n" + rows
+def by_venue(kind: str) -> str:
+    c = Counter(tables[kind].column("source").to_pylist())
+    return ", ".join(f"{s} {n:,}" for s, n in c.most_common())
+n = {k: tables[k].num_rows for k in KINDS}
+legs = sum(len(x) for x in tables["resolved"].column("markets").to_pylist())
+card = f"""---
+license: cc-by-4.0
+pretty_name: "pdata.world prediction markets (Polymarket, Kalshi and 6 more venues)"
+language:
+- en
+tags:
+- prediction-markets
+- polymarket
+- kalshi
+- manifold
+- forecasting
+- finance
+- tabular
+size_categories:
+- 100K<n<1M
+configs:
+- config_name: resolved
+  data_files: data/resolved.parquet
+  default: true
+- config_name: markets-open
+  data_files: data/markets-open.parquet
+- config_name: events-open
+  data_files: data/events-open.parquet
+---
+# pdata.world prediction markets
+Prices, volumes and results for prediction markets on eight venues — Polymarket, Kalshi,
+Manifold, Myriad, Limitless, Predict, Opinion and Gemini — in one schema. The data is
+collected by [pdata.world](https://pdata.world), which polls each venue every 10 to 60
+minutes, and is published under CC BY 4.0.
+This snapshot was taken on {as_of}. It has three parts:
+| Config | Rows | What one row is |
+|---|---|---|
+| `resolved` (default) | {n['resolved']:,} events, with their top {legs:,} markets (at most 10 each) | An event pdata stopped tracking after it closed, with each top market's reported result |
+| `markets-open` | {n['markets-open']:,} | A market open for trading: price, 24h move and volume |
+| `events-open` | {n['events-open']:,} | An open event, which groups the markets of one question |
+```python
+from datasets import load_dataset
+resolved = load_dataset("{REPO}", "resolved", split="train")
+markets = load_dataset("{REPO}", "markets-open", split="train")
+# or with pandas
+import pandas as pd
+df = pd.read_parquet("hf://datasets/{REPO}/data/markets-open.parquet")
+```
+## Rows per venue
+- **resolved:** {by_venue('resolved')}
+- **markets-open:** {by_venue('markets-open')}
+- **events-open:** {by_venue('events-open')}
+## Read this before using the numbers
+- `probability` and prices are on a 0–1 scale.
+- **Volume is not comparable across venues.** `volume` and `volume_24hr` are in each venue's
+  own unit: shares for Polymarket and contracts for Kalshi (each pays $1 at resolution),
+  play-money mana for Manifold, and each other venue's figure as it reports it. pdata's
+  report [How eight venues count volume](https://pdata.world/research/prediction-market-volume)
+  explains each one.
+- **In `resolved`, a market's `probability` is the last price pdata saw**, which is usually
+  after the market settled. It records the result; it is not a forecast. For prices before
+  close, use the API's history endpoint while a market is live (history is deleted 10 days
+  after close) — see [How accurate were prediction-market prices a day before the
+  result?](https://pdata.world/research/prediction-market-accuracy) for one way to do it.
+- **`result` is null where the venue never reports one.** Of the {legs:,} archived markets,
+  {sum(1 for x in tables['resolved'].column('markets').to_pylist() for m in x if m['result'] is not None):,}
+  carry a result.
+- **A gap in `resolved`:** events that closed between 2026-06-02 and 2026-08-11 are missing,
+  because they were pruned before a backup was taken. Events closed in the 10 days before
+  the snapshot are in neither file yet.
+- pdata groups Kalshi's sibling events for one game under a single group event
+  (`is_group`), whose volume is the sum of its children's.
+## Fields
+### resolved
+{man['resolved']['description']}
+{fields('resolved')}
+### markets-open
+{man['markets-open']['description']}
+{fields('markets-open')}
+### events-open
+{man['events-open']['description']}
+{fields('events-open')}
+Timestamps are UTC. The manifest each file was built from is in [`manifests/`](manifests/).
+## Updates and versions
+A new snapshot is added monthly; each is tagged with its date (this one is `{snapshot}`),
+so `load_dataset("{REPO}", "resolved", revision="{snapshot}")` always returns the same
+rows. For daily data, the same three files are rebuilt every night at
+<https://api.pdata.world/bulk/> as gzipped JSON lines.
+## Other ways to get the data
+- REST API, no key: <https://api.pdata.world/docs>
+- MCP server for AI assistants: `https://api.pdata.world/mcp`
+- Read-only Postgres with price history:
+  `postgresql://pdata_readonly:pdata_public_2026@pg.pdata.world:6433/pdata_new`
+  (the password is published on purpose)
+- Python client and examples: <https://github.com/noskillcoding/pdata-client>
+- Research built on this data: <https://pdata.world/research>
+## Licence and citation
+CC BY 4.0. Credit "Data: pdata.world" with a link to <https://pdata.world>. The underlying
+market data remains subject to each venue's own terms. pdata.world is independent and not
+affiliated with any venue.
+```bibtex
+@misc{{pdata_world_prediction_markets_{snapshot.replace('-', '_')},
+  title        = {{pdata.world prediction markets: prices, volumes and results from eight venues}},
+  author       = {{{{pdata.world}}}},
+  year         = {{{snapshot[:4]}}},
+  note         = {{Snapshot {snapshot}}},
+  howpublished = {{\\url{{https://huggingface.co/datasets/{REPO}}}}}
+}}
+```
+"""
+(here / "README.md").write_text(card)
+print("README.md", len(card), "chars")

manifests/events-open.json ADDED Viewed

	@@ -0,0 +1,31 @@

+{
+  "publisher": "pdata.world",
+  "url": "https://pdata.world/data",
+  "attribution": "Data aggregated by pdata.world (https://pdata.world) from 8 prediction-market venues; CC BY 4.0; source data © each venue.",
+  "license": "CC BY 4.0",
+  "license_url": "https://creativecommons.org/licenses/by/4.0/",
+  "file": "events-open",
+  "description": "Every event open for trading on the 8 venues pdata tracks, one per line. An event groups the markets of one question; markets-open holds its markets.",
+  "as_of": "2026-10-07T07:58:31.137721Z",
+  "as_of_note": "When this export started. Rows are read over a few minutes; each row's own timestamp says when pdata last wrote it.",
+  "format": "JSON lines: this manifest, then one record per line.",
+  "fields": {
+    "source": "Venue: polymarket, kalshi, manifold, myriad, limitless, predict, opinion or gemini.",
+    "id": "The venue's event id, unique within source (pdata builds the id of a group event).",
+    "title": "Event title as the venue publishes it.",
+    "categories": "pdata topic categories (e.g. Politics, Sports), from rules and AI tagging.",
+    "start_date": "When the event opened, as the venue reports it (UTC).",
+    "end_date": "Scheduled close or expiry, as the venue reports it (UTC).",
+    "volume": "Lifetime trading volume. Venue-reported, in the venue's own unit and not converted across venues: shares for Polymarket and contracts for Kalshi (each pays $1 at resolution), play-money mana for Manifold, and each other venue's figure as it reports it.",
+    "volume_24hr": "Trading volume over the last 24 hours. Venue-reported, in the venue's own unit and not converted across venues: shares for Polymarket and contracts for Kalshi (each pays $1 at resolution), play-money mana for Manifold, and each other venue's figure as it reports it.",
+    "open_interest": "Open interest as the venue reports it, in the venue's own unit.",
+    "mutually_exclusive": "True when exactly one of the event's markets can resolve YES; false when they are independent questions; null when the venue does not say.",
+    "market_count": "How many markets the event has (for a group event, across its child events).",
+    "is_group": "True for an event pdata built to group sibling Kalshi events of one matchup.",
+    "parent_event_id": "The pdata group event this event belongs to, or null.",
+    "synthetic": "True for an event pdata created to wrap a market the venue lists without one.",
+    "updated_at": "When pdata last wrote this row (UTC): every fetch from the venue, and again when pdata recomputes a derived field.",
+    "url": "The event or market's own page on the venue's website; null when it cannot be built.",
+    "pdata_url": "pdata's page for the event, https://pdata.world/events/{source}/{id}."
+  }
+}

manifests/markets-open.json ADDED Viewed

	@@ -0,0 +1,36 @@

+{
+  "publisher": "pdata.world",
+  "url": "https://pdata.world/data",
+  "attribution": "Data aggregated by pdata.world (https://pdata.world) from 8 prediction-market venues; CC BY 4.0; source data © each venue.",
+  "license": "CC BY 4.0",
+  "license_url": "https://creativecommons.org/licenses/by/4.0/",
+  "file": "markets-open",
+  "description": "Every market open for trading on the 8 venues pdata tracks, one per line: price, 24h move and volume. A market is one yes/no question, or one outcome of a multi-outcome event.",
+  "as_of": "2026-10-07T07:58:31.137721Z",
+  "as_of_note": "When this export started. Rows are read over a few minutes; each row's own timestamp says when pdata last wrote it.",
+  "format": "JSON lines: this manifest, then one record per line.",
+  "fields": {
+    "source": "Venue: polymarket, kalshi, manifold, myriad, limitless, predict, opinion or gemini.",
+    "id": "The venue's market id, unique within source.",
+    "event_id": "The parent event's id (same source); joins events-open on (source, id).",
+    "event_title": "The parent event's title. Read question with it: an outcome of a multi-outcome event can be a bare label.",
+    "question": "The market's question as the venue titles it.",
+    "label": "Short outcome label where the venue gives one (e.g. a candidate's name).",
+    "categories": "pdata topic categories (e.g. Politics, Sports), from rules and AI tagging.",
+    "outcomes": "Outcome labels in venue order, usually [\"Yes\", \"No\"].",
+    "outcome_prices": "Price of each outcome, 0-1, in outcomes order.",
+    "probability": "Implied probability of the first outcome (usually YES), 0-1.",
+    "prob_24h_change": "Change in probability over the trailing 24h, on the same 0-1 scale (0.05 = up 5 points). Null when there is no snapshot in that window.",
+    "volume": "Lifetime trading volume. Venue-reported, in the venue's own unit and not converted across venues: shares for Polymarket and contracts for Kalshi (each pays $1 at resolution), play-money mana for Manifold, and each other venue's figure as it reports it.",
+    "volume_24hr": "Trading volume over the last 24 hours. Venue-reported, in the venue's own unit and not converted across venues: shares for Polymarket and contracts for Kalshi (each pays $1 at resolution), play-money mana for Manifold, and each other venue's figure as it reports it.",
+    "liquidity": "Liquidity as the venue reports it, in the venue's own unit.",
+    "open_interest": "Open interest as the venue reports it, in the venue's own unit.",
+    "start_date": "When the market opened, as the venue reports it (UTC).",
+    "end_date": "Scheduled close or expiry, as the venue reports it (UTC).",
+    "status": "The venue's own status string, verbatim; null for Polymarket.",
+    "countries": "Countries the market concerns (full English names, at most 3), from pdata's AI tagging. [] means none applies; null means not tagged yet.",
+    "updated_at": "When pdata last wrote this row (UTC): every fetch from the venue, and again when pdata recomputes a derived field.",
+    "url": "The event or market's own page on the venue's website; null when it cannot be built.",
+    "pdata_url": "pdata's page for the event, https://pdata.world/events/{source}/{id}."
+  }
+}

manifests/resolved.json ADDED Viewed

	@@ -0,0 +1,23 @@

+{
+  "publisher": "pdata.world",
+  "url": "https://pdata.world/data",
+  "attribution": "Data aggregated by pdata.world (https://pdata.world) from 8 prediction-market venues; CC BY 4.0; source data © each venue.",
+  "license": "CC BY 4.0",
+  "license_url": "https://creativecommons.org/licenses/by/4.0/",
+  "file": "resolved",
+  "description": "Every event pdata has stopped tracking live, with its final result: the record pdata keeps when it prunes an event, 10 days after it closes. Events closed more recently are in neither file yet. Kept since 2026-08-31; events pruned before then were rebuilt on 2026-10-07 from backups taken 2026-05-09 to 2026-08-29. Events that closed between 2026-06-02 and 2026-08-11 are missing: they were pruned before the next backup was taken.",
+  "as_of": "2026-10-07T07:58:31.137721Z",
+  "as_of_note": "When this export started. Rows are read over a few minutes; each row's own timestamp says when pdata last wrote it.",
+  "format": "JSON lines: this manifest, then one record per line.",
+  "fields": {
+    "source": "Venue: polymarket, kalshi, manifold, myriad, limitless, predict, opinion or gemini.",
+    "id": "The venue's event id, unique within source.",
+    "title": "Event title when pdata pruned it.",
+    "categories": "pdata topic categories (e.g. Politics, Sports), from rules and AI tagging.",
+    "closed_time": "When the event closed (UTC): the venue's close time, else its end date, else when pdata last saw it.",
+    "markets": "The event's top markets by lifetime volume, at most 10: id, question, result (as the venue reported it; null where the venue never says), final probability (0-1) and lifetime volume (venue unit).",
+    "url": "The event or market's own page on the venue's website; null when it cannot be built.",
+    "pdata_url": "pdata's page for the event, https://pdata.world/events/{source}/{id}.",
+    "pruned_at": "When pdata wrote this record (UTC): when it pruned the event, or 2026-10-07 for an event pruned before 2026-08-31 and rebuilt from backups."
+  }
+}