There is also the multiple-endpoints principle to think about. The likelihood of this particular set of galaxies forming a ring is very low. The chance of some set of galaxies among all the billions in the sky doing this is much higher.
Then we notice and cherry-pick only the one interesting data point, we never notice all the mundane ones.
It's always difficult to tell if a popular-science article is really describing something unusual or if it's using selective perception to create the illusion of one. (I have no idea in this case.)
> The chance of some set of galaxies among all the billions in the sky doing this is much higher.
Of course in relative terms it's much higher, but it doesn't matter—what matters is the absolute value. 10^-100 is much larger than 10^-10000, but if something with the probability of 10^-100 happens, it's still "astonishing."
The probability of a particular planet has a shape of pyramid is so low. And yes, the probability of finding any planet in the universe that has a shape of pyramid is much higher, but still very low. If one was found, scientists would freak out.
It's always difficult to tell if a popular-science article is really describing something unusual or if it's using selective perception to create the illusion of one. (I have no idea in this case.)