The thing to know here is the bayes factor. That’s the true positive rate divided by the false positive rate. In this context, it’s the percent of successful articles that have a property (like using the number 10) divided by the number of unsuccessful articles that have that property. This removes any advantage a property gets from being more common.