MQ-000631
Celta v Santander · La Liga · Sat 19 Sept · 16:30 UTC
Every prediction is sealed with a hash before kick-off, in a table the database refuses to update or delete. Results are graded from the official score; a correction never overwrites silently — the previous state is written to an append-only correction log first, and the record shows it. The sealing time is WinPIQ's own server clock: no outside party has timestamped it.
Version 1
- Fair price
- 2.13
- Market price
- 1.78
- Probability gapWinPIQ's probability minus the market's probability, in percentage points — only where a market price exists. · It's the plain-language size of the disagreement between WinPIQ and the market. · 56% versus 49% is a gap of 7 percentage points. Full definition →
- 6 percentage points
- Generated
- Tue 15 Sept · 18:08 UTC
- Published
- Tue 15 Sept · 18:08 UTC
- Locked
- Sat 19 Sept · 14:32 UTC
- Model artifacts
- 0168d70ef9af
Publication recordTue 15 Sept · 18:08 UTC83b2f…5931
{
"confidenceScore": 75,
"dataQualityScore": 49,
"dataSnapshotId": "cmu2zp0jh00w6r6ru2hyiv8af",
"edge": -0.059758,
"fairOdds": 2.13,
"featureVersion": "f0.3",
"fixtureId": "cmtxtl84t0062r6dt7a5lfp5c",
"generatedAt": "2026-09-15T18:08:50.298Z",
"market": "1X2",
"marketOdds": 1.78,
"modelProbability": 0.468787,
"modelVersion": "ens_v0.3",
"predictionId": "MQ-000631",
"previousPredictionId": null,
"publishedAt": "2026-09-15T18:08:50.298Z",
"selection": "HOME"
}Lock recordSat 19 Sept · 14:32 UTC2b0cd…db88
{
"confidenceScore": 75,
"dataQualityScore": 49,
"dataSnapshotId": "cmu2zp0jh00w6r6ru2hyiv8af",
"edge": -0.059758,
"fairOdds": 2.13,
"featureVersion": "f0.3",
"fixtureId": "cmtxtl84t0062r6dt7a5lfp5c",
"generatedAt": "2026-09-15T18:08:50.298Z",
"market": "1X2",
"marketOdds": 1.78,
"modelProbability": 0.468787,
"modelVersion": "ens_v0.3",
"predictionId": "MQ-000631",
"previousPredictionId": null,
"publishedAt": "2026-09-19T14:32:05.948Z",
"selection": "HOME"
}Where this prediction came from
The inputs and versions behind this row, as they stood when it was published. Rows we do not hold are marked rather than hidden.
- Probability vector
- HOME 46.9% · DRAW 27.6% · AWAY 25.5%
- Model version
- ens_v0.3
- Feature version
- f0.3
- Calibration version
- iso_v2
- Calibration scope
- league
- Feature snapshot taken
- 2026-09-15 18:08Z
- Information cutoff
- 2026-09-15 18:08Z
- Data coverage
- ADVANCED
- Lead time
- H72_PLUS
- Analysis standard
- FULL
- Price at lock
- 1.78
- Sources consulted
- not yet recorded by this product
- Signal policy version
- not yet recorded by this product
- Segment health at publication
- not yet recorded by this product
The hash shows this payload is unchanged since it was sealed: recompute it and it matches, or it does not.
It does not show WHEN it was sealed. That timestamp is our own server's, in a table the database refuses to update or delete. No outside party has attested it.
Version 2 · previous MQ-000631
- Fair price
- 2.27
- Market price
- 1.79
- Probability gapWinPIQ's probability minus the market's probability, in percentage points — only where a market price exists. · It's the plain-language size of the disagreement between WinPIQ and the market. · 56% versus 49% is a gap of 7 percentage points. Full definition →
- 8.5 percentage points
- Generated
- Thu 17 Sept · 12:40 UTC
- Published
- Thu 17 Sept · 12:40 UTC
- Locked
- Sat 19 Sept · 14:32 UTC
- Model artifacts
- 0168d70ef9af
Publication recordThu 17 Sept · 12:40 UTCd1791…7017
{
"confidenceScore": 76,
"dataQualityScore": 49,
"dataSnapshotId": "cmu5iufbj009ar6fx48wtczkl",
"edge": -0.084784,
"fairOdds": 2.27,
"featureVersion": "f0.3",
"fixtureId": "cmtxtl84t0062r6dt7a5lfp5c",
"generatedAt": "2026-09-17T12:40:09.042Z",
"market": "1X2",
"marketOdds": 1.79,
"modelProbability": 0.44001,
"modelVersion": "ens_v0.3",
"predictionId": "MQ-000631.v2",
"previousPredictionId": "MQ-000631",
"publishedAt": "2026-09-17T12:40:09.042Z",
"selection": "HOME"
}Lock recordSat 19 Sept · 14:32 UTC419b4…f189
{
"confidenceScore": 76,
"dataQualityScore": 49,
"dataSnapshotId": "cmu5iufbj009ar6fx48wtczkl",
"edge": -0.084784,
"fairOdds": 2.27,
"featureVersion": "f0.3",
"fixtureId": "cmtxtl84t0062r6dt7a5lfp5c",
"generatedAt": "2026-09-17T12:40:09.042Z",
"market": "1X2",
"marketOdds": 1.79,
"modelProbability": 0.44001,
"modelVersion": "ens_v0.3",
"predictionId": "MQ-000631.v2",
"previousPredictionId": "MQ-000631",
"publishedAt": "2026-09-19T14:32:05.948Z",
"selection": "HOME"
}Where this prediction came from
The inputs and versions behind this row, as they stood when it was published. Rows we do not hold are marked rather than hidden.
- Probability vector
- HOME 44.0% · DRAW 29.1% · AWAY 26.9%
- Model version
- ens_v0.3
- Feature version
- f0.3
- Calibration version
- iso_v2
- Calibration scope
- league
- Feature snapshot taken
- 2026-09-17 12:40Z
- Information cutoff
- 2026-09-17 12:40Z
- Data coverage
- ADVANCED
- Lead time
- H48_72
- Analysis standard
- FULL
- Price at lock
- 1.79
- Sources consulted
- not yet recorded by this product
- Signal policy version
- not yet recorded by this product
- Segment health at publication
- not yet recorded by this product
The hash shows this payload is unchanged since it was sealed: recompute it and it matches, or it does not.
It does not show WHEN it was sealed. That timestamp is our own server's, in a table the database refuses to update or delete. No outside party has attested it.
Version 3 · previous MQ-000631.v2
- Fair price
- 2.36
- Market price
- 1.79
- Probability gapWinPIQ's probability minus the market's probability, in percentage points — only where a market price exists. · It's the plain-language size of the disagreement between WinPIQ and the market. · 56% versus 49% is a gap of 7 percentage points. Full definition →
- 10.8 percentage points
- Generated
- Sat 19 Sept · 09:10 UTC
- Published
- Sat 19 Sept · 09:10 UTC
- Locked
- Sat 19 Sept · 14:32 UTC
- Model artifacts
- 0168d70ef9af
Publication recordSat 19 Sept · 09:10 UTC5fa84…6dde
{
"confidenceScore": 76,
"dataQualityScore": 49,
"dataSnapshotId": "cmu867bna008yr60nai97bos0",
"edge": -0.107878,
"fairOdds": 2.36,
"featureVersion": "f0.3",
"fixtureId": "cmtxtl84t0062r6dt7a5lfp5c",
"generatedAt": "2026-09-19T09:10:32.294Z",
"market": "1X2",
"marketOdds": 1.79,
"modelProbability": 0.423896,
"modelVersion": "ens_v0.3",
"predictionId": "MQ-000631.v3",
"previousPredictionId": "MQ-000631.v2",
"publishedAt": "2026-09-19T09:10:32.294Z",
"selection": "HOME"
}Lock recordSat 19 Sept · 14:32 UTC46fbb…11d3
{
"confidenceScore": 76,
"dataQualityScore": 49,
"dataSnapshotId": "cmu867bna008yr60nai97bos0",
"edge": -0.107878,
"fairOdds": 2.36,
"featureVersion": "f0.3",
"fixtureId": "cmtxtl84t0062r6dt7a5lfp5c",
"generatedAt": "2026-09-19T09:10:32.294Z",
"market": "1X2",
"marketOdds": 1.79,
"modelProbability": 0.423896,
"modelVersion": "ens_v0.3",
"predictionId": "MQ-000631.v3",
"previousPredictionId": "MQ-000631.v2",
"publishedAt": "2026-09-19T14:32:05.948Z",
"selection": "HOME"
}Where this prediction came from
The inputs and versions behind this row, as they stood when it was published. Rows we do not hold are marked rather than hidden.
- Probability vector
- HOME 42.4% · DRAW 28.1% · AWAY 29.5%
- Model version
- ens_v0.3
- Feature version
- f0.3
- Calibration version
- iso_v2
- Calibration scope
- league
- Feature snapshot taken
- 2026-09-19 09:10Z
- Information cutoff
- 2026-09-19 09:10Z
- Data coverage
- ADVANCED
- Lead time
- H6_12
- Analysis standard
- FULL
- Price at lock
- 1.79
- Sources consulted
- not yet recorded by this product
- Signal policy version
- not yet recorded by this product
- Segment health at publication
- not yet recorded by this product
The hash shows this payload is unchanged since it was sealed: recompute it and it matches, or it does not.
It does not show WHEN it was sealed. That timestamp is our own server's, in a table the database refuses to update or delete. No outside party has attested it.
Version 4 · previous MQ-000631.v3
- Fair price
- 2.62
- Market price
- 1.79
- Probability gapWinPIQ's probability minus the market's probability, in percentage points — only where a market price exists. · It's the plain-language size of the disagreement between WinPIQ and the market. · 56% versus 49% is a gap of 7 percentage points. Full definition →
- 15 percentage points
- Generated
- Sat 19 Sept · 12:11 UTC
- Published
- Sat 19 Sept · 12:11 UTC
- Locked
- Sat 19 Sept · 14:32 UTC
- Model artifacts
- 0168d70ef9af
Publication recordSat 19 Sept · 12:11 UTCe0873…5a7e
{
"confidenceScore": 57,
"dataQualityScore": 49,
"dataSnapshotId": "cmu8coe1c002yr6gwvtfs3fhm",
"edge": -0.14976,
"fairOdds": 2.62,
"featureVersion": "f0.3",
"fixtureId": "cmtxtl84t0062r6dt7a5lfp5c",
"generatedAt": "2026-09-19T12:11:26.597Z",
"market": "1X2",
"marketOdds": 1.79,
"modelProbability": 0.382014,
"modelVersion": "ens_v0.3",
"predictionId": "MQ-000631.v4",
"previousPredictionId": "MQ-000631.v3",
"publishedAt": "2026-09-19T12:11:26.597Z",
"selection": "HOME"
}Lock recordSat 19 Sept · 14:32 UTCa269b…bae8
{
"confidenceScore": 57,
"dataQualityScore": 49,
"dataSnapshotId": "cmu8coe1c002yr6gwvtfs3fhm",
"edge": -0.14976,
"fairOdds": 2.62,
"featureVersion": "f0.3",
"fixtureId": "cmtxtl84t0062r6dt7a5lfp5c",
"generatedAt": "2026-09-19T12:11:26.597Z",
"market": "1X2",
"marketOdds": 1.79,
"modelProbability": 0.382014,
"modelVersion": "ens_v0.3",
"predictionId": "MQ-000631.v4",
"previousPredictionId": "MQ-000631.v3",
"publishedAt": "2026-09-19T14:32:05.948Z",
"selection": "HOME"
}Where this prediction came from
The inputs and versions behind this row, as they stood when it was published. Rows we do not hold are marked rather than hidden.
- Probability vector
- HOME 38.2% · DRAW 28.1% · AWAY 33.7%
- Model version
- ens_v0.3
- Feature version
- f0.3
- Calibration version
- iso_v2
- Calibration scope
- league
- Feature snapshot taken
- 2026-09-19 12:11Z
- Information cutoff
- 2026-09-19 12:11Z
- Data coverage
- ADVANCED
- Lead time
- H1_6
- Analysis standard
- FULL
- Price at lock
- 1.79
- Sources consulted
- not yet recorded by this product
- Signal policy version
- not yet recorded by this product
- Segment health at publication
- not yet recorded by this product
The hash shows this payload is unchanged since it was sealed: recompute it and it matches, or it does not.
It does not show WHEN it was sealed. That timestamp is our own server's, in a table the database refuses to update or delete. No outside party has attested it.
WinPIQ is not affiliated with any gambling operator. All information is analysis and probability estimates only — not advice, and not a promise.