A particular nucleus may last a shorter or longer time than its half-life, but in a large sample, almost exactly half of the nuclei will have decayed after a time equal to one half-life.
Similarly, if we have a very large number of radioactive atoms of one type (say, uranium), there is a specific time period, called its half-life, during which the chances are fifty-fifty that decay will occur for any of the nuclei. Radioactive Decay: This graph shows (in pink) the amount of a radioactive sample that remains after several half-lives have passed.
Still, astronomers can use the numbers of craters on different parts of the same world to provide important clues about how regions on that world evolved.
On a given planet or moon, the more heavily cratered terrain will generally be older (that is, more time will have elapsed there since something swept the region clean).
(credit: modification of work by NASA/GSFC/Arizona State University) Bear in mind that crater counts can tell us only the time since the surface experienced a major change that could modify or erase preexisting craters.
Estimating ages from crater counts is a little like walking along a sidewalk in a snowstorm after the snow has been falling steadily for a day or more.
One way to estimate the age of a surface is by counting the number of impact craters.