What a half-life actually measures
Imagine a sample containing a very large number of atoms of one radioactive isotope. Some of those atoms will decay, changing into other atoms or forms of the same element. The isotope’s half-life is the time it takes for the number of undecayed atoms to fall to half its starting value, in the statistical sense: it describes the expected behavior of a group.
Suppose a sample starts with 1,000,000 undecayed atoms and the isotope has a half-life of one hour. After one hour, about 500,000 are expected to remain. After another hour, about 250,000 are expected to remain—not zero. Each interval halves the number still present, so a half-life is not the time required for a sample to finish decaying. For an actual sample, the counts may differ slightly from those expectations.
An atom has no half-life countdown
A particular atom does not carry a clock that rings at the end of one half-life. Its decay is unpredictable: it might happen soon, much later, or not within any period we choose to watch. The half-life tells us the chance that an atom will still be undecayed after a given time, not the moment at which that atom will change.
For the usual model of radioactive decay, an undecayed atom has the same chance of decaying during the next short interval whether it has existed for a moment or for many half-lives. Surviving an hour does not make it “overdue.” This can feel surprising because many familiar objects wear out as they age. Radioactive decay, however, is not usually a gradual wearing-down process inside a single atom.
Why large groups are more predictable
Think of repeatedly tossing a fair coin. You cannot know whether one toss will land heads, but with many tosses you can make a useful prediction about the proportion of heads. Radioactive atoms are not coins, and their probability of decay over a particular interval depends on the isotope. Still, the comparison captures the difference between an unpredictable individual event and a reliable pattern across many opportunities.
With only a handful of atoms, chance has a visible effect. If a group contains ten undecayed atoms and one half-life passes, exactly five need not remain. With an enormous group, the fraction remaining will usually be much closer to one-half. The same principle applies when measuring radiation from a sample: counts collected over equal stretches of time can vary even when the sample’s underlying rate of decay is changing smoothly.
What makes the fraction shrink
The key is that the chance of decay applies to atoms that have not yet decayed. At the start, every radioactive atom in the sample is available to decay. Later, fewer undecayed atoms remain, so fewer decays are expected during the next equal interval. That is why the expected number falls by a constant fraction over each half-life rather than by a constant number of atoms.
This pattern is called exponential decay. It also explains why a sample can keep producing decays long after several half-lives have passed. After three half-lives, the expected fraction of the original radioactive atoms remaining is one-eighth. After four, it is one-sixteenth. Those fractions become smaller, but they do not suddenly reach zero. In a real, finite sample, the last radioactive atom will eventually decay; the half-life alone cannot tell us when.
Using half-life without overreading it
Half-life is useful whenever we need to connect a measurable group-level change with elapsed time. For example, scientists can compare the amount of a radioactive isotope in a sample with other information about how that isotope entered or left it. Such reasoning depends on assumptions about the sample’s history, not just on knowing a half-life.
The central distinction is simple: half-life gives a dependable description of the expected pattern for many atoms, while each decay remains an uncertain event. If half a large sample is expected to remain after one half-life, that does not mean we can identify which atoms will survive. It means that randomness at the atomic scale can still produce a clear, useful pattern when we look at a group.
