Like-for-like criteria
| Criterion | Pre-match | Live |
|---|---|---|
| Information set | Data available before play | Pre-match data plus observed match state |
| Update pace | Usually slower | Can change after each event or clock update |
| Suspension | Around material pre-event changes | Common around goals, penalties and other key events |
| Latency sensitivity | Lower | Higher: event, feed, model and display delays matter |
| Price record | Opening, current, closing | Timestamp and match state are essential |
| Availability | Depends on limits and market | Can disappear or re-open at a different price |
A peer-reviewed Bayesian in-play football model demonstrates one transparent way to update team-strength estimates with in-match information. It does not describe every operator's production system.
Same label does not guarantee same state
A pre-match Home price and a 30th-minute Home price refer to different information sets. Even two live prices captured at the same match minute may differ if one feed has processed a card or goal attempt that the other has not. Store match clock, score, red cards and capture time with the odds.
Suspension and acceptance
An in-play selection may be suspended while an event is checked. A click before suspension is not enough to prove acceptance; the receipt and operator rules control. Betfair's exchange rules provide one current example of in-play market operation and settlement.
Illustrative update
At kick-off, odds of 2.00 have a raw reciprocal of 1 / 2.00 x 100 = 50.00%. At half-time with the score 1-0, an illustrative live price of 1.40 has a reciprocal of 1 / 1.40 x 100 = 71.43%. The 21.43-point change reflects a new price under a new match state; it is not a measure of model accuracy.
Make the comparison at equivalent decision points
Pre-match and in-play prices are conditioned on different information sets. Once play begins, score, time remaining, cards, substitutions and suspensions can alter both probability and market availability. Comparing a pre-match 2.20 with an in-play 1.80 is therefore not a simple claim that one source became less generous.
For a useful record, capture the pre-match price at a defined cutoff and the in-play price with match minute, score, disciplinary state, market status and available stake. If the market was suspended during a goal or review, record the gap rather than assigning the first reopened quote to the earlier moment.
Like-for-like fields
| Field | Pre-match snapshot | In-play snapshot |
|---|---|---|
| Market | Exact label and period | Same label and period |
| Time | Stated cutoff | Match clock and timestamp |
| State | Scheduled event information | Score and material match state |
| Price | Executable for stated stake | Executable for stated stake |
| Rules | Current market rules | Current in-play rules |
This structure supports analysis of how a price responded to observed events. It does not show that waiting is generally better, because the later opportunity may disappear, shorten or be unavailable at the intended size.
Related resources
Use in-play betting for the concise definition or odds movement for timestamped price comparison.
Continue learning
- Next guide: Short Odds vs Long Odds
- Related guide: Starting Price vs Fixed Odds
Assumptions and limitations
The example is illustrative and ignores market margin. Live systems, delays, bet acceptance and settlement differ by operator and jurisdiction. No universal claim about lower margin, better value or faster reaction is made.
GambleAware
GamCare
Gambling Therapy
GAMSTOP