Imagine you are walking through a gallery of thousands of paintings, some of which impress you and some of which don't. Would you not feel some compulsion to understand why? Now suppose I am walking with you through the same gallery- would you expect me to be impressed by the same paintings and for the same reasons? Either way, how would you account for it?
Appreciation of art is manifestly subjective. Anyone with an analytic bent might be tempted to identify independent dimensions by means of which works of art could be compared. You might consider, for example, originality, the presence of metaphor, the degree to which the depiction is realistic, the range of colours, the composition, the perspective, the apparent skill displayed, the subject matter, whether some emotional resonance is triggered within you, and so on. What weight you attach to each of those is up to you. You might find that some aspects of an artwork are appreciated in the same way and to the same extent by the majority of people who view it. Does that make the evaluation more valid in some way? I think that to assume so would be to confuse validity with popularity, but you are free to take a different view. Ultimately, the only logical way to assess different works of art in a consistent way is to declare what (arbitrary) criteria you are going to adopt, and apply them consistently. You can then say that picture A is better than picture B because you have defined 'better' to mean a particular weighted combination of ratings.