The ratio is going to depend enormously on what needs doing.

If the shop is doing a little bit of application development but handling hundereds of production servers with big databases and serious availability and disaster recovery requirements, it needs more DBAs than developers.

On the other hand, if it's doing a fair bit of development (which includes schema design) and only has a couple of dozen small to medium production servers to support it ought not to need any DBAs at all, the developers should have automated just about all the things that a production DBA would do.

I guess the ratio should vary from all DBAs and no developers to no DBAs and all developers, depending on what's being done, so the idea that there is some single correct or best ratio is not one I tend to believe in.

Anyway, I have a theory that good database developers can easily pass for DBAs, while good development DBAs can easily pass for developers. And I think the sort of overspecialisatin that produces developers (except very young ones who haven't yet got around to it) who haven't learnt to do database stuff or DBAs (again except very young and inexperienced ones) who can administer and organis recoverability and high availability haven't learnt to develop a schema or high performance sets of queries is a bad thing. So the concept of a ratio of DBAs to developers is a bit meaninglss if you are talking about what I think of as competent not overspecialised people.