The numbers of ミニロト that deviate most from the uniform expectation (sample: 20): 31 (19.41%), 12 (12.13%), 29 (12.13%), 26 (8.86%), 1 (6.55%).Data includes draw #1394 of 07.07.2026.
A high χ² only means a number deviated from uniformity more than others in the past — it describes history, not a forecast: in a fair lottery, drum bias is usually small, and it does not affect the probability of the next draw. But as a selection system the method works: if you want to pick numbers by statistics rather than at random, clicking a number in the table adds it to the combination generator.
χ² (Pearson's goodness-of-fit test) compares each ball's actual frequency with the one expected from a perfectly uniform drum. Nearby, related views of the same bias: Z-Score →Frequency →
χ² = Σ (Oᵢ − Eᵢ)² / Eᵢ
where for each ball:
- Oᵢ — the ball's actual draw frequency;
- Eᵢ — the frequency expected under uniformity = (k / m) × N, where k is numbers per draw, m balls in the drum, N draws;
- in the table each ball is shown as its share of the drum's total χ² (%).
Where to next
Number frequency
The actual frequencies of ミニロト numbers — the very Oᵢ that χ² is built from.
OpenZ-Score
The same ミニロト bias, but in standard deviations — how far a number stands out in σ.
OpenRuns test
Whether ミニロト has non-random streaks and droughts — another randomness test of the draw.
OpenBenford's law
First digits of ミニロト draw sums vs the Benford baseline — another check of shape.
OpenFrequently asked questions about ミニロト
What does a high χ² for a number mean?
That its draw frequency in ミニロト deviated from the uniform expectation more than others over the period studied — higher or lower. This describes the past, not a prediction: χ² does not affect the next draw.
Should I play ミニロト numbers with a high χ²?
A high χ² gives no future advantage — the draw is random and all combinations are equally likely. But the method works as a selection system: numbers with a notable deviation can seed a combination in the generator — that is picking by a system, not predicting a draw.
Does a Pearson bias prove ミニロト is unfair?
Usually not. Over a finite history a small χ² appears for any fair drum — that is noise. Real bias would show as a persistently high total χ² on a large sample. To double-check, use the runs test and Benford's law.
Is this the Pearson correlation coefficient?
No. This is Pearson's goodness-of-fit test (χ²), which checks the uniformity of ball frequencies. The Pearson correlation coefficient (r) measures the relationship between two variables — a different method, unrelated to this page.
How many draws are needed for the calculation?
The larger the archive, the more reliable the conclusion: on a short window χ² jumps around and noise is easily mistaken for bias. Premium opens the full archive to compute across all draws.