The likes of Mozart, Bach and Beethoven tend to dominate the 'top composers' charts, because they wrote so much -and so much of what they wrote is recorded over and over. They therefore account for a large number of recordings in my collection and when every recording has been played at least once, they're going to be 'top composers' in terms of hours-played or number-of-plays, almost without trying! So, perhaps a more interesting way of looking at 'who are the top composers' is to assess what proportion of their recordings in my collection have been played. If Benedetto Marcello is in my collection with a single recording that lasts 600 seconds, but I've played that recording 4 times and thus clocked up 2,400 play-minutes of his work, that would mean 400% of Marcello's collection has been played. Mozart may have 1,000,000 seconds of recordings in the collection, but if I've played only 900,000 seconds of that, Mozart would have a mere 90% play-rate.
It's no more a 'valid' way of looking at top composers than anything else, but it does produce quite different results from the other ways of thinking about what counts as 'top', as you can see below. This data is refreshed once per day, at 11pm.
There is, of course, the corollary view: the composers for whom the least recordings have been played as a proportion of all their available recordings. They would look like this:
In an ideal world, every composer should be at least at the 100% mark: you'd want to play every recording you own, regardless of composer! I'm currently not far off that, but there is still (as of May 2025) a way to go, with Wagner significantly under-played, simply because his stuff is so long it takes a special sort of mindset to listen to it!