The framework comes down to five criteria: recent form, the context of the match, head-to-head records, statistical data specific to the sport, then reading the odds as the last filter. The first four belong to a qualitative, sporting analysis; the fifth is a quantitative reading of the market. Keeping them separate is what makes comparing them useful. Each criterion also has its moment: the first four can be worked on days ahead, the fifth belongs to the final half-hour.
The framework at a glance
Five criteria, four qualitative and one quantitative. The first four describe the match; the fifth describes what the market thinks of it.
This page describes the sequence: what to look at when, between the fixture being announced and kick-off. The methodological framework — which steps make up a prediction, and in which order — is covered by how to make a sports prediction.
| Criterion | What it brings | Its limit |
|---|---|---|
| Recent form | The current state, beyond the table | Depends on the level of the opposition faced |
| Context | What is at stake, fixture list, travel, motivation | Hard to quantify |
| Head-to-head records | Recurring patterns of play | Obsolete if the squads have changed |
| Statistical data | Indicators comparable from one match to another | Describe the past, not the match to come |
| Market odds | The aggregated consensus of the operators | Never says why the price is what it is |
What to look at when, before kick-off
The five criteria are not consulted at the same moment. Some are stable days ahead, another only exists an hour before the match.
| Moment | What is available | Why then |
|---|---|---|
| Three to two days out | Recent form, standings, head-to-head records | That data is already stable: there is no reason to wait |
| The day before | Stakes, fixture congestion, likely rotation | The context of the match takes shape |
| About an hour before, depending on the competition | Official line-ups, confirmed absences | This is when the missing information becomes public |
| The final half-hour | Odds and the direction of their movement | The price now incorporates everything above |
A criterion consulted too early is not wrong, it is incomplete: an analysis frozen two days out ignores, by construction, an absence announced on match day.
Criterion 1 — recent form
Recent form is read over five or six matches, weighted by the level of the opposition. Three wins against teams at the bottom of the table and three wins against teams at the top do not describe the same state.
Two traps come up systematically. The first is overweighting the last match, usually the freshest in the memory. The second is reading a run of results without looking at the manner: a team that wins three times in the closing minutes does not necessarily have the same real level as a team that wins three times comfortably.
Criterion 2 — the context of the match
Context often explains what form does not predict. It cannot be encoded in a statistic, which makes it all the more useful: it is the part of the analysis where an attentive human reading still adds something.
- What is at stake — a team already qualified or already doomed does not approach the match like a team still fighting.
- Fixture list — a match slotted between two important dates is rarely prepared for in the same way.
- Travel and recovery — distance covered, days of rest, any time difference.
- The point in the season — the start of the season with squads still finding their rhythm, the end of the season with accumulated fatigue.
Criterion 3 — head-to-head records
A head-to-head history is only worth something if it is recent and substantial enough. Two matches four years ago, with different squads and different coaches, no longer describe anything.
When the history can be used, what it brings is less the result than the pattern: a style of play that systematically neutralises the other, a recurring tendency in the volume of goals or points scored.
Criterion 4 — statistical data
The relevant indicators depend on the sport and the market you are targeting. Expected goals in football, the surface and the percentage of points won on serve in tennis, the pace of possessions and offensive efficiency in basketball.
All this data shares the same limit: it describes what has happened. To be projected forward, it assumes that conditions remain comparable — which is exactly what criterion 2 qualifies.
Responsible gambling — an analysis framework does not remove the risk. Sports betting carries a risk of financial loss and presents, according to the Autorité nationale des jeux, the highest individual risk of problem gambling among regulated activities. Analysing a match does not make any bet risk-free. Set a budget, do not try to recover your losses and use the limit-setting or self-exclusion tools available. Learn more about responsible gambling.
Criterion 5 — reading the odds, last
The odd comes in at the end of the process, as a filter and not as a starting point. Consulted first, it steers the reading of the four previous criteria: you then look for the data that explains the displayed price, instead of forming an independent opinion.
Consulted last, it allows a real comparison. Three situations:
- Agreement — the analysis and the market arrive at the same order of likelihood. The opinion is more solid, it is not more certain.
- Moderate disagreement — information is probably missing on one side or the other.
- Strong disagreement — the market usually sees something the analysis is missing. Prices in online betting markets absorb available information quickly, including information that is not yet public.
The detail of converting an odd into a probability is covered on implied probability, and the interpretation of variations on odds movements.
What this framework allows — and what it does not
Three levels never to be confused.
- What it lets you do — organise a reading of a match, make explicit the information you have kept, and spot the places where the analysis and the market diverge.
- What it lets you estimate — an order of likelihood between the outcomes of a market, and the relative strength of a line of reasoning.
- What it never lets you conclude — that an outcome is going to happen. A complete, well-filled framework remains an opinion about an uncertain event.
Bringing an odd into an analysis, without copying it →
Back to the full sports betting predictions guide →
Dig into the market
Odds movements are only part of the story. Here are the next topics to read.
How to make a sports prediction
A five-step method, from collecting data to comparing it with the market.
See the methodOdds and predictions
How to use an odd as analysis data, without mistaking it for a certainty.
Understand the linkOdds movements before a prediction
What an odds movement signals, and what it never says.
Read a movementSports betting predictions: analyse before betting
The full cluster guide: data, odds, reliability, biases and the role of AI.
Read the guideFrequently asked questions
How many criteria should you cross-check?
Enough to cover form, context and the data specific to the sport — so four to five families. Beyond that, the risk is no longer missing information but keeping only the information that confirms an intuition.
Should you ignore your intuition?
No, but you should treat it as a hypothesis to be checked. An intuition that no data supports remains an intuition.
Are statistics enough on their own?
No. They describe what has happened, not the context of the match to come: what is at stake, the fixture list, a last-minute absence. None of that can be read in a series of figures.
Why compare with the odds at the end rather than at the start?
Because an odd consulted too early steers the analysis. Formed beforehand, the opinion can be compared with the market; formed afterwards, it often does no more than paraphrase it.
What should you do if the criteria contradict each other?
The contradiction is information. It usually indicates that the market you are targeting is more uncertain than it looks, which is in itself a conclusion of the analysis.
Does OddScore provide tips?
No. OddScore compares the odds of several bookmakers, removes the margin and tracks how they evolve. The platform publishes no betting selections and provides no staking advice.
Sources & methodology
This page draws on the French regulatory framework for sports betting published by the Autorité nationale des jeux, on the economic research devoted to price formation in online betting markets, and on the odds-analysis methodology developed by OddScore.
- Separate the qualitative analysis (form, context) from the quantitative reading of the market.
- Weight each piece of data by its relevance to the market you are targeting, not by its availability.
- Consult the odds last, to compare them with an opinion already formed.
- Treat a contradiction between criteria as a result of the analysis, not as a failure.