Comment by astrophysician

In my experience I’ve never found an instance where you would use Brier scores over cross entropy/Bernoulli/Binomial log likelihoods. Does anybody know a concrete example when you would prefer Brier??