TiralumeFrequency counts appearances
A number's frequency is the proportion of draws in the sample in which it appears. The same principle applies to stars.
Fictional example
If a number appears in 12 draws out of 100, its observed frequency is 12 %.
This percentage describes only the selected sample. It does not mean that the number has a 12% chance of appearing in the next draw.
A shorter window may show more noticeable variation; a longer window summarises a different period. Keep the same window when comparing two figures.
A gap measures a past absence
The gap is the number of draws since a value last appeared. A gap of zero means that it appears in the latest draw in the sample.
Fictional example
A gap of six draws means that six draws have taken place since the value last appeared.
This absence does not create a debt that chance must repay. A number that has not appeared for a while is not “due”. A run of appearances does not demonstrate that a number will keep appearing either. These measures remain observations, with no demonstrated predictive value.
Read pairs and associations in context
A pair counts how often two values have appeared together. Triples and associations between numbers and stars follow the same principle of co-occurrence. An association can seem remarkable after just a few appearances; it can also result from chance.
The observed/expected ratio shown in the app compares appearances with a uniform model. It is neither a significance test nor proof of a predictive relationship. Always check the sample size alongside these figures.
Compare compatible periods
The number of available stars has changed during the game's history. To avoid directly mixing different sets of available stars, Tiralume's analyses use draws eligible under the twelve-star rules. Older draws can still be viewed in the history.
Open the app, choose a window and read the explanations shown for each measure.