How random are the KENO results? The Shannon entropy of the number distribution over the last 20 draws is 6.18 of 6.32 bits (97.8%). The frequency distribution is close to uniform — the results are consistent with a random draw. Data includes draw #3056 of 27.07.2026.
The maximum log₂N is reached when every number appears equally often; any shortfall from the maximum means a frequency skew. To see exactly which numbers appear more and less often, open the frequency table: Number frequency →
Shannon entropy is a measure of uncertainty of a random variable from information theory (Claude Shannon, 1948). Applied to a lottery, it compresses the entire frequency table into a single number: the closer the entropy is to its maximum, the more uniformly the numbers are drawn and the fewer pronounced skews the draw history contains.
H = −Σ pi·log₂ pi
where:
- H — entropy in bits;
- pᵢ — the share of hits of number i among all balls drawn in the sample (frequency ÷ sum of frequencies);
- H_max = log₂ N — the maximum for a perfectly uniform distribution of N numbers; the page compares entropy to this maximum as a percentage.
Where to next
Number frequency
The actual KENO number frequencies — the very shares pᵢ the entropy is computed from.
OpenRuns test
KENO streaks and droughts: are there non-random runs in the sequence of hits.
OpenPearson test (χ²)
The same question about KENO frequency skew — via a formal goodness-of-fit test.
OpenShannon index
Each KENO number’s contribution to entropy over two periods — which balls create the diversity.
OpenShannon entropy questions
What is Shannon entropy in a lottery?
It is a measure of how uniform the frequency distribution is. We take how many times each KENO number has been drawn, convert the frequencies into shares and compute H = −Σ pᵢ·log₂ pᵢ. The maximum log₂N is reached when all numbers appear equally often; the lower the value, the more the draw history is skewed towards particular numbers.
Are the KENO results random?
Entropy answers this question from one angle — frequency uniformity: a value near the maximum is consistent with a random draw. A small shortfall on a short sample is expected and is not a sign of manipulation. For the full picture, randomness is checked with several tests: streaks and droughts (runs test) and the Pearson χ² test.
Can entropy be used to pick numbers?
No — it is a diagnostic of the whole KENO distribution, not a tipster: entropy has no “recommended numbers”, and its value does not change the odds of any combination. If you want to pick numbers based on statistics rather than at random, start with the frequency table and the tools built on it.
How does Shannon entropy differ from the Shannon index?
Entropy is a single number for the whole distribution: how uniformly all the balls are drawn together. The Shannon index breaks the same quantity down per ball: each number’s contribution to entropy over two periods. Entropy answers “is the draw random overall”, the index — “which numbers make it diverse”.
What does sliding entropy show?
Entropy recomputed in a window of the N most recent draws that moves across the KENO history. It shows how uniformity changed over time: dips in the chart are periods when frequencies in the window were skewed. Smaller windows fluctuate more — that is a property of the sample, not a signal.