I've given this subject a lot of thought lately, mainly because I have been writing a lot of reviews - and have come to several (possibly erroneous) conclusions:
* "star ratings" HAVE TO BE relative! One might give a lesser album by a great artist 3 stars, but those are not the same 3 stars one would give a surprisingly okay album by a generally crappy artist
* Musical taste is entirely subjective, but reviewing (for me) has to try be a balance between objective and subjective; just listening to something and saying what you think of it IS also valid of course.
* Objective factors alone (see pie chart below) can never make an otherwise bad album good but subjective factors can.
* 'Classic' albums make a nonsense of all other rules.
Let's examine in more detail, with graphs! (are pie charts graphs?):
Objective factors (see pie chart above) are really only very important when the reviewer doesn't like the music: when you love a song, whether or not the people performing it are technically talented musicians/pitch perfect singers etc is completely irrelevant.
When an album or song (or movie, book etc) is dull or just blatantly abysmal, some comfort can be gained from the knowledge that at least the participants were at least good at the technical aspects of what they were doing, even if they are using those skills for evil.
Although there are many subjective factors that may be relevant; nostalgia for the artist/period, personal associations, all of these amount to either you like it or you don't; simple but not necessarily straightforward.
The positive subjective feeling 'I like it!' can override all else, so that an album which is badly played, unoriginal, poorly recorded and awful even by the artist's own standards can receive a favourable review (though the reviewer will probably want to point out those things)
Meanwhile the negative subjective feeling 'I don't like it' can't help but affect a review, but should hopefully be tempered by technical concerns if (an important point) the reviewer feels like being charitable. S/he may not.
Ideally, a review should be something like 50% objective / 50% subjective (as below) but in practice it rarely happens
The reviewing of reissued classics can be awkward as 'classic' status is completely separate from all other concerns, therefore said classic status can affect ratings just because the album is iconic and everyone knows it. Popularity itself shouldn't play a part in the reviewer's verdict; just because 30,000,000 people are cloth-eared faeces-consumers, it doesn't mean the reviewer should respect their opinion, but s/he should probably acknowledge it, even if incredulously. Sometimes classic status is attained for cultural, rather than (or as well as) musical reasons*, and it should be remembered albums are as much a 'cultural artefact' (in the sense of being a mirror and/or record of their times) as cinema, TV, magazines or any other zeitgeist-capturing phenomenon.
* in their very different ways, Sgt Pepper's Lonely Hearts Club Band,Thriller and The Spice Girls' Spice were all as much 'cultural phenomena' as collections of songs
SO ANYWAY; how does this all work? Some examples:
Recently I offended a Tina Turner fan with an ambivalent review of the 30th anniversary edition of Ms Turner's 1984 opus Private Dancer.
As a breakdown (of 'out of 10's, for simplicity) it would look something like this:
TINA TURNER: PRIVATE DANCER (3OTH ANNIVERSARY EDITION)
* musicianship - 9/10 - hard to fault the adaptability or technical skill of her band
* songwriting - 6/10 - in terms of catchy, verse-chorus-verse efficiency & memorableness these are perfectly good songs, if a bit cheesy & shallow & therefore a waste of Tina Turner
* production - 9/10 - no expense was spared in making the album sound good in its extremely shiny, 80s way
* originality - 0/10 - as an album designed to make TT into a successful 80s artist, it wasn't really even supposed to be original, so hard to fault it in that respect
* by the standards of the artist - 2/10 - in the 60s/70s Tina Turner made some great, emotionally forceful, musically adventurous and just great records. In 1984 she didn't
Overall: 26/50 = 5.2/10
* I don't like it: 1/10 (but not 0, because Tina Turner is a legend and it would be wrong to deny that somehow)
Overall 5.2/10 + 1/10 = 6.2/20 = 3.1/10 = 1.55/5 (round up rather than down, out of respect for Tina) = 2 stars
and in fact I did give the album two stars, though I didn't actually do any of the calculations above; but it's pleasing to find out that the instinctive two stars is justified by fake science.
by way of contrast, a favourite that seems to be an acquired taste at best:
VENUSIAN DEATH CELL: Honey Girl
* musicianship - 1/10 - David Vora's guitar playing is not very good, plus the guitar is out of tune anyway, and his drumming is oddly rhythm-free
* songwriting - 2/10 - the songs on Honey Girl are not really songs, they may be improvised, they don't have actual tunes as such
* production - 0/10 - David pressed 'record'
* originality - 10/10 - Vora doesn't sound like anyone else, his songs are mostly not about things other people sing about
* by the standards of the artist - 9/10 - I like all of Venusian Death Cell's albums, they are mostly kind of interchangeable, but Honey Girl is one of the better ones (chosen here over the equally great Abandonned Race only because of the uncanny similarities between the cover art of Honey Girl and Private Dancer).
Overall: 22/50 = 4.4/10
* I like it: 9/10 (but not 10, because if encouraged too much David Vora might give up and rest on his laurels. Though if he did that I'd like to "curate" a box set of his works)
Overall 4.4/10 + 9/10 = 13.4/20 = 6.7/10 = 3.35/5 (round up rather than down, out of sheer fandom) = 4 stars
And in fact I did give Honey Girl four stars, but I've yet to hear of anyone else who likes it. Which is of course fuel for the reviewer's elitist snobbery; win/win.