Direct answer
Predictive factors are data and statistical signals that help a model estimate the probability of a scenario in football.
Match factors help explain what supports the model’s view and where current evidence points in a different direction.
Quick summary
| Definition | Statistical signals used to interpret a prediction. |
|---|---|
| Examples | Home/away, recent form, PPG, goal difference, xG, shots, clean sheets, corners, and sample. |
| What they are for | Compare the model's reading with the context of the analyzed market. |
| What they are not | Evidence to compare with the model’s probability estimate. |
| On Predora.si | They appear in the Team Comparison and in AI + Statistics. |
How predictive factors work
A predictive factor is a signal that helps answer a question about the match. This question changes depending on the market.
There is no magic factor that works equally for everything. A team can have good signals for match result and weak ones for goals, or good attacking volume without being reliable to win.
- In 1X2, the question is who has more strength to win or hold a draw.
- In Over 2.5 goals, the factors look at creation, pace, and goal tendency.
- In BTTS, the focus is understanding whether both teams have a real path to scoring.
- In corners, the signals are usually linked to volume, pressure, and corners generated.
Is there a formula for predictive factors?
There is no single formula because each factor measures a different part of the game. The practical logic is combining model, context, sample, and price.
Contextualized reading = model probability + market factors + sample quality + odds priceExample in football
Imagine a Serie A match in the Over 2.5 goals market. The prediction becomes more comfortable if both teams score and concede frequently, shoot a lot, play at an open pace, and the sample is not too small.
If the model gives a high probability but the teams create few shots and tend toward low-scoring games, investigate what drives the difference.
How it appears on Predora.si
At Predora.si, the factors provide transparency to the match reading. They allow you to open the analysis and understand why a market seems interesting or why it may be riskier.
- The quick comparison shows factors relevant to the selected market.
- AI + Statistics opens a view with Prediction, Season, Home/Away, Last matches, H2H, League, and Sample.
- The analysis distinguishes aligned, mixed, and conflicting factors.
How to interpret aligned or conflicting signals
Factors show convergences, divergences, and uncertainties regarding the estimate.
| Scenario | Reading |
|---|---|
| Aligned factors | Most of the important signals point in the same direction as the prediction. |
| Mixed factors | Some signals support the model and others differ. Explore which match factors explain the gap. |
| Conflicting factors | The central factors conflict with the prediction and the scenario becomes trickier. |
Common mistakes
Looking at one factor in isolation
A strong home team at home can be a good sign, but it alone does not decide the match.
Using the right factor in the wrong market
Offensive volume can help with goals, but it does not automatically decide the 1X2 market.
Practical checklist
- Choose the market you want to analyze.
- See the probability and EV of the Active Model.
- Open the Comparison between Teams or AI + Statistics.
- Check the sample and main factors.
- Classify the scenario as aligned, mixed, or conflicting.
- Compare the available odds with the estimate for the same market.
FAQ about Predictive factors
How do predictive factors help evaluate a bet?
Predictive factors help explain the match evidence behind a probability estimate. Compare where they support the model and where they differ.
Does every factor have the same weight?
No. A factor can be central in one market and complementary in another.
How should I interpret factors that contradict the forecast?
A difference between the factors and model gives you a focused research question. Explore which assumptions explain the gap.
Are predictive factors the same thing as statistics?
Not always. A statistic only becomes a useful factor when it helps answer the question of the market being analyzed.
How does Predora.si help with this reading?
The app organizes factors by market to contextualize the model's reading without treating the AI as a black box.
Factors are context, not orders
Predictive factors gain context when they appear alongside probability, EV, sample, and odds, even when they contradict the estimate.