Data Sources and Corrections: How Odds2Win Records Odds and Results
How Odds2Win handles prediction data
Odds2Win keeps the event record, published selection, publication price and eventual result as separate parts of the prediction history. This page explains where those records originate, how prices are retained, how results are graded and how factual errors are corrected.
The purpose is to preserve a reviewable historical record. A settled prediction should be assessed against the market, selection and price recorded with the original publication rather than a later market price or a revised interpretation created after the result is known.
Where event information comes from
Event records used on Odds2Win prediction, data and results pages originate from published Odds2Win prediction records. Each record can contain the sport, participants, competition and scheduled start time captured for that prediction.
These stored fields are the reference used when a published prediction is later grouped, filtered or matched to a result. Odds2Win does not attribute an event record to a third-party data provider unless that provider was actually recorded with the source data.
- Sport: identifies the event category used for classification and filtering.
- Participants: the teams, players or competitors attached to the original prediction record.
- Competition: the league, tournament, cup or event label stored with the prediction when available.
- Scheduled time: the event date and start time associated with the published record when available.
Missing information is not silently reconstructed from an unrelated fixture or assigned to a provider that was not recorded. If a field cannot be supported by the stored prediction record, it should remain unconfirmed until it can be checked.
Published prediction records can be reviewed at https://odds2win.bet/predictions/.
Where the odds come from
The price used for a prediction is the decimal odds stored with that published prediction. When the bookmaker or odds source was recorded at publication, that source can be displayed alongside the price.
Older or incomplete records may contain a stored publication price without a named source. Odds2Win does not retrospectively assign a bookmaker, sportsbook or odds provider to those records when the original source was not retained.
Decimal odds can be converted to raw implied probability using 1 ÷ decimal odds × 100. The result describes the probability implied by that price before adjustment for the wider market margin. It is not an independently verified probability that the selection will win.
Market prices can change after publication. A later price therefore does not automatically indicate that the stored publication price was incorrect and does not replace it merely because the market subsequently moved.
How the original publication price is preserved
When a prediction has valid odds stored with the published record, those odds become its publication-price reference for historical results tracking.
If the line moves after publication, the original stored price remains the historical reference. This prevents a later, more favourable or less favourable number from being substituted simply because it became available after the prediction was published.
If no reliable publication price is stored, Odds2Win does not create a historical price later solely to complete ROI, net-unit or other price-dependent calculations. Such a prediction can remain part of the historical record without being counted as a priced graded pick.
| Record | How it is treated | Archive effect |
|---|---|---|
| Valid publication odds | Retained as the prediction's historical reference price. | Can be used in price-dependent performance calculations after grading. |
| Later market movement | Does not replace the stored publication price merely because the line changed. | The original stored price remains the reference. |
| No reliable stored price | No historical price is invented retrospectively. | The record can be excluded from price-dependent metrics. |
How a prediction result is determined
A prediction record is graded against the outcome required by the market and selection originally published. The final event score alone may not be sufficient when the selected market has additional conditions or a defined settlement period.
The review therefore starts with the original event, market and selection before the applicable result is assigned. Different markets can require different grading logic, including 1X2, Draw No Bet, totals, handicaps and other market types.
Identify the original event. Confirm that the result belongs to the participants and event represented by the published record.
Read the original market and selection. The record is graded against what was actually published rather than a different market that happens to point in the same general direction.
Confirm the applicable outcome. Match the event result to the period, line and conditions relevant to the published selection.
Assign the archive verdict. Once the outcome is sufficiently clear, the prediction record can be classified as won, lost, push/void or another applicable archive status.
Odds2Win grading describes the status of a published prediction record for the site's historical tracking. It does not determine, replace or override the settlement rules applied by an individual bookmaker to a customer's wager.
An unresolved record is not converted into a win or loss simply to complete performance statistics. Where the available information is insufficient for reliable grading, the record can remain pending review until its status is resolved.
The public results archive is available at https://odds2win.bet/results/.
How postponements and cancellations are handled
A postponed event is not treated as a completed loss. A cancelled event is also kept separate from an event that has been completed normally. Maintaining these states separately prevents unresolved fixtures from being presented as settled prediction results.
| Status | Archive treatment | Performance effect |
|---|---|---|
| Completed | Compare the confirmed outcome with the original market and selection. | Grade once the relevant outcome is sufficiently clear. |
| Postponed | Keep the prediction unresolved while the event remains pending. | Do not force a win or loss. |
| Cancelled / void | Record the non-standard event status when confirmed. | Do not count the record as a normal win or loss when it is treated as void. |
| Conflicting or incomplete data | Hold the prediction for review rather than infer an outcome. | Keep it outside final grading until the conflict is resolved. |
When the event status, market scope or applicable outcome remains uncertain, keeping the record unresolved is preferable to assigning an unsupported win or loss.
How to report an error
Readers can report a factual issue involving an event, market, stored price, result, verdict or other information published on Odds2Win. Reports are easier to review when they identify the exact record and disputed field.
- the exact Odds2Win page URL;
- the event or prediction concerned;
- the specific field or statement that appears incorrect;
- the information you believe is correct;
- supporting evidence or a reliable source when available.
General information about Odds2Win and its correction process is available at https://odds2win.bet/about/.
Correction requests submitted to Odds2Win concern information published on Odds2Win. Issues involving a betting account, payment, wager acceptance or bookmaker settlement must be handled directly with the relevant operator.
How corrections are reviewed
A reported factual issue is checked against the original prediction record and the information relevant to the disputed field. The purpose of a correction is to restore an accurate historical record, not to improve a prediction after its outcome has become known.
Locate the original record. Identify the relevant page, event, market, selection and stored publication price.
Review the disputed field. Compare the published value with the available evidence relevant to that specific issue.
Correct only the affected information. A factual correction should not be used to rewrite unrelated prediction reasoning or replace the historical publication price because the result is already known.
Update dependent archive figures when required. If an amended price or verdict changes a priced graded record, performance calculations that depend on that record should use the corrected data.
Record material public corrections. Changes that materially affect event identity, price, market, result, verdict or derived archive figures can be added to the correction journal below.
Correction journal
This journal records material factual corrections made after the launch of this public log. Minor spelling, typography, formatting or layout changes that do not alter the meaning of a prediction record do not need a journal entry.
No entries have been added to this public correction journal since its launch.
What a material correction entry records
| Date | Affected page | Correction | Archive effect |
|---|---|---|---|
| — | No entry since launch | — | — |