Sixth Ball

Number frequency statistics

Read this first. Every draw is random and independent. These tables only describe what has already been drawn. A number that appeared more often in the past is not more likely next time, and one that appeared less often is not "owed". No table, chart or pattern on this page can tell you what the next numbers will be.

We count how often each white ball and Mega Ball has come up, period by period, and then test whether the differences between numbers are bigger than chance would produce.

Era boundaries inferred from the data, not from official rules. Only the current number range has been checked on an official page; earlier ranges have not, so for consistency each period below is the stretch of draws over which the largest white ball and largest Mega Ball seen stayed the same. Boundaries are placed where the history is much more likely under two different number ranges than one (likelihood ratio of at least e6, about 400 to 1) and can be off by a few draws. See Methodology.

Loading draw history from NY Open Data…

How to read the chi-square result

If every number is equally likely, each one should come up about the same number of times: the expected count is the number of balls drawn divided by how many numbers there are. Real counts always wander around that. The chi-square statistic adds up the squared gaps between observed and expected counts (each divided by expected), and the p-value says how often a fair draw would give gaps at least that large.

Why this does not help you choose numbers is covered in why number patterns don't work; the odds for one line are on the odds page.

Questions

Does a number that came up often have a better chance next time?

No. Each draw is independent of the ones before it. A frequency table describes the past only; the chance of every number in the next draw is the same.

Why split the history into periods?

The number ranges have changed over the years. Counting a number that could not be drawn in some years against one that could would make the table meaningless, so we count each period separately.

What does the p-value mean here?

It is the probability of seeing differences between numbers at least this large if every number really were equally likely. Values that are not tiny (say above 0.01) mean the counts look just like fair random draws.