What's the Chance of Getting a Random Peak at Any Number?

Researchers looked for the most common occurrence and found something around 28 million years. They ran a bunch of statistical tests to see like, how likely is it to see something at 28 million years? And they found there's very unlikely to see this kind of peak. The problem with this method of statistical analysis is that they're only calculating the probability of seeing a peak from random events at28 million years. Whereas these random events also could have generated peaks at 30 million years and 15 million years.