Direct answer
Statistical base it is the number and selection of games used to calculate a metric, such as accuracy rate, return, ROI, or performance by league.
It answers a simple question: how many games and from what context did this percentage come from?
Quick summary
| Definition | Number and selection of games used to calculate a metric. |
|---|---|
| Practical formula | Statistical sample = games or entries considered in the selection |
| Where it appears | Performance, Accuracy History, Return History, Active Model, and market filters. |
| Common mistake | Trusting a high percentage without checking whether the sample is small. |
| How to use | Read sample together with accuracy, return, market, league, odds, and game factors. |
Statistical sample is the set of games or entries used to calculate a metric; the smaller the sample, the more careful you should be when interpreting the result.
How it works
An accuracy rate can look great over a few games and far less impressive as the sample grows. Football has variance, late goals, red cards, missed penalties, and scorelines that do not reflect the flow of the game. That is why the sample gives context to the number.
Examples of filters that change the sample:
- all games in a league;
- only games with positive EV;
- only high probability;
- a specific market, such as 1x2 or BTTS;
- home teams playing at home or away teams playing away;
- a specific period of the season.
Statistical sample = total games or entries in the selection Accuracy rate = correct picks / statistical sampleSample is not just volume. It is volume within a clear context.
Example in football
| Scope | Sample | Hit | Return | How to read |
|---|---|---|---|---|
| Brasileirão Série A, 1x2 | 420 games | 55% | +4% | More robust sample, more stable reading. |
| State league, 1x2 | 18 games | 72% | +19% | Explore whether that pattern also appears across a longer period and a larger sample. |
The second selection looks better, but 18 games can inflate the reading. With 420 games, the result has gone through more scenarios and fluctuations.
Statistical sample vs. sample
| Concept | What it means | Example |
|---|---|---|
| Sample | Set of observed data. | Games analyzed from a league. |
| Statistical base | Size and selection of the sample used in the calculation. | 180 EV+ games in the 1x2 market. |
How it appears on Predora.si
On Predora.si, the sample appears in places such as Active Model, Performance by league, Accuracy History, and Return History.
Performance by league
Shows how many games were included in a league's reading.
Upcoming Games with EV+
Helps read the sample of games with positive EV.
Upcoming Games with High Probability
Shows filters above a probability cutoff.
The reading combines Sample + Accuracy + Return + Market + League, not an isolated number.
Quick checklist
- Check the size of the sample. How many games were included?
- Check the filter. Is it all games, EV+, a specific league, or a specific market?
- Compare identical markets. Do not mix 1x2 with goals and BTTS.
- Look at hit rate and return together. High accuracy with poor odds may not pay off.
- Be wary of small samples. It can be an initial signal, not strong proof.
FAQ about statistical sample
Is statistical sample the same thing as the number of bets I placed?
Not necessarily. On Predora.si, sample is the number of games or entries considered in the calculation shown.
Is a larger sample always better?
A larger sample usually makes the comparison more stable. Review it alongside return, odds, and the time period.
Does a small sample invalidate an analysis?
A small sample is useful for forming an initial hypothesis. Compare more observations and periods to see how the pattern develops.
Why does the sample change when I filter a league?
Because the filter changes the selection. The more specific the filter, the smaller the sample tends to become.
Sample gives weight to the number
Sample size helps you distinguish a persistent pattern from a short run. Predora shows the number of games alongside performance so you can judge the depth of the evidence.