The Headline Factory
38: 71 appearances, with 58.8 expected.
Last 480 draws
p for this question alone: 0.0541
Loading…
Reading the draws and working out what chance predicts.
LotteryOracle · F-44
This archive is fair by construction. Explore it until something looks remarkable, then publish the finding and reveal what the search itself does.
Every machine here is simulated. No real lottery is being accused or predicted.
The Headline Factory
Last 480 draws
p for this question alone: 0.0541
The correction covers every available number, pair and window, including options you did not click. Choosing where to look is part of the search. · headline-search-v2 · seed 20260922
Weekdays belong to a fictional daily calendar. Droughts count consecutive misses at the fixed end of the archive. Both are included in every null search.
For a preselected number or pair, the appearance count is binomial. The experiment searches for unusually high counts across every number, every pair, and three overlapping windows. Each fair replicate repeats that whole search. The adjusted p-value counts replicates whose smallest raw p is at least as extreme. This accounts for choosing a question after seeing the archive, but not for repeatedly changing the seed to hunt for a different archive.
Weekdays belong to a fictional daily calendar. Droughts count consecutive misses at the fixed end of the archive. Both are included in every null search.
pnumber = k/N; ppair = k(k−1)/[N(N−1)]; P(G ≥ g) = (1−k/N)^g; padj = (1 + #{min(pnull) ≤ pselected})/(R + 1)
This archive is fair by construction. Explore it until something looks remarkable, then publish the finding and reveal what the search itself does.
The correction covers every available number, pair and window, including options you did not click. Choosing where to look is part of the search.