Reading the bracketology forums, and making my own choices (brackets + my own top 100 in progress), made me think about the subject of this thread.
I would like you to answer that question
What, in your opinion, makes a good song ? I sorted a certain number of criteria in alphabetical order.

sort you own list from the more to the less relevant, and of course add comments if you feel like it
add other criteria if you feel they are absent from this list

- Arrangements and orchestration
-History/Exposure of the song (e.g. : has it been overplayed ? )
- Importance of the song in music history
- Importance of the song in your personal history
- Innovation
- Instrumental performance
- Lyrics
- Melody (apart from orchestration)
- Musical genre (some people are allergic to certain styles or cherish other)
- Singer’s personality
- Vocal performance(not only technically speaking)


I know perfectly well that when you love something, you don’t really know (and can tell) why, but the brackets are the results of choices, and they are interesting because they reveal patterns in every one’s tastes.
So I don’t know if it’s the stupidest idea in the forum, but try to play that game (it’s nothing but a game). After all, we are here because we love music and…games.
Of course what’s interesting is not the game itself, but what it will raise (conversations, comments, arguments), so be free to give your comments and explain your choices

I’ll start myself to answer

1.Emotional power (the song has to be contagious, it has to give me the chills)
2.Melody/Music (the Beatles)
3Vocal performance (I love to sing and to hear strong voices, that’s why I love r&b, and distinctive vocalists like Robert Wyatt, Neil Young or Tom Waits)
3.Lyrics (Springsteen, Waits, Reed, Bowie…)
4.Importance of the song in your personal history (I wrote a book about that)
5.Instrumental performance (specially acoustic guitar and piano)
6.Importance of the song in music history
7.Innovation
8.Arrangements and orchestration (a song may be well orchestrated, but without true soul and feelings it’s a hollow shell)
9.Musical genre (I love MUSIC before all, even if I prefer roots music)
10.Singer’s personality
11. History/Exposure of the song (I don’t give a damn if it has been overexposed or used in ads )

I think you got a nice order there, except for maybe the 4th spot which I would put a little lower. (Then again, I have no real songs that have an importance in my history, so what do I know.. :))

what I mean when I say "importance in your history" is the memories it will bring when you hear it
But thinking about it I don't think it is such a good criteria because it applies only to songs you already know, and not to songs you discover.
So feel free to count that one out

The order looks good to me also. But I'd completely remove #10 - singer's personality; it's not even a factor.

For me, this aspect has absolutely no bearing on my enjoyment of music, whether it be a song, an album, a particular artist, etc. For a few reasons....

At the end of the day, a good tune is a good tune. I don't care if the lead singer kills puppies in his spare time. Ok, that might be an extreme example, but if he's a great songwriter, I could really care less what his personality is like, or how good his character is.

I'm an Oasis fan, and have been from the beginning. I know that they may not be completely relevant today, but at one time they were on top, and they were led by two of the most antagonistic, self-centred and egotistical brothers in rock history: Liam and Noel Gallagher. But Noel sure knew how to write a pop song (even if he was ripping off his idols) and Liam could sing the hell out of it. When I'd read about them getting in pubs fights or getting kicked off airplanes (and banned from certain airlines), of course I'd shake my head, but it didn't prevent me from enjoying their music or buying their albums. The quality was there, and a few of their songs can arguably be considered classics.

Another thing to consider: as the music-buying public, we only know what the media reports to us. One of our favorite artists or singers could be an abusive father or husband in his private life, but we'd never know it.

Their no rules for what makes a good song. It's more complex.
But for me, as a Beach Boys fan, I can't imagine a good song without a beautiful melody and harmony.
Sometimes personal history can also play a role. I like some songs that I would probably find crappy if they were not associated to a period of my life, people I love.
In general I hate the sound of the 80s music, but some songs of the 80s still have some echo in me (some Inxs,Spandau Ballet,A-ha,Tears for Fears,Simple Minds,etc...) because I was born in the late 70s and so the 80s were the years of my childhood and as everyone knows the childhood is important and influencial in a person's life.

I've heard many explanations of what makes a good song. Even more for what makes a hit song. Here are a couple ways of looking at it:

Jerry Wexler (I think it was him, maybe not) said a song hits you in either the head, the heart, the feet (danceability), the hands (athleticism) or the groin. A good song hits you in two of those. A classic does three or more.

I often write reviews of tracks and I always follow this format: vocals, song-writing, style, production. That's really what matters. Everything else is trivial by comparison.

A major rep told me the top two things he (and most other professionals) look for in any artist are vocal ability and image. So that means his idea of a good song is one with strong vocals and image. Song-writing ability (if they write their own songs) comes third.

Really it comes down to the definition of a good song. What exactly is meant by good? Profitable? Prevalent? Functional? Complicated? Skilled? Acclaimed? Most likely to win bracketology? There are very different answers for each criteria.

in this particular exercise, i meant a good song is a song that moves you
the one you would give maximal rating
a song you wouldn't change if you were the artist or the producer

Genre is the most important factor.

I’m gonna rebel a bit and argue that you can’t actually establish a set of criteria that works for every song. For example, look at these four songs:

Miles Davis, “So What”
The Dead Kennedys, “Holiday in Cambodia”
Britney Spears, “Toxic”
Dusty Springfield, “You Don’t Have to Say You Love Me”

Now, I think these are all good songs—I wouldn’t skip past them if they came up on my iPod—but can anybody claim with a straight face that there’s a set of criteria that works equally well for all four of them?

I mean, imagine that this was a bracket and you tried to apply one criterion only:

-Musicianship? Miles wins hands down, the Kennedys and Britney are far behind.
-Vocals? Dusty wins.
-Danceability (the American Bandstand standard)? Britney.
-Songwriting? Depends completely on what kind of song—“Holiday in Cambodia” is the best-structured pop song despite being loud, fast, and discordant. “So What” has an absolutely perfect structure, but it’s partly improvised. If production is important, then it’s Britney…if you want a great ballad, Dusty.
-Emotional impact is really problematic; I find all of these songs emotionally engaging, but the emotions are completely different. Same thing with style—all of these artists have a lot of it, but in wildly different ways.

And so forth. And this is actually a pretty simple example; I could have made the exercise even more complicated by including, say, a rap song with a lot of sampling.

I guess what I’m saying is that you have to take any song on its own merits (if any). Or, to put it another way, “genre” HAS to come first, because without it, you don’t know which other criteria are important.

This is NOT to say that some genres are somehow better than others; I don’t think that idea makes any sense—although in practice, everyone likes some genres better than others. But the goal of the music critic, or critical listener, ought to be to try to understand as many different genres as possible (and what criteria should and should not be applied to them). Otherwise we’re just fighting for turf.

well those songs all have in common that they are unique, creative, have strong image and skilled songwriting.

Schleuse,

What you said about genre was not wrong : I have underestimated it a bit. What I meant by giving it this place was that I like music in general;i know a lot of music lovers who are definitly against one genre or another.
But let's take this week's brackets : If I had correctly applied my criterias, I think Girls & Boys (a song that I find danceable, funny, well structured and produced) would have won against Whole Lotta Love I think that awful breaks really ruins the whole song which is an awkward attempt at over-electrifying Chicago blues; but Plant isn't Muddy, and the lyrics sound a bit ridiculous coming from an Englishman.
Anyway, I placed it above Blur because of that sound, that rock'n roll sound that I really prefer to disco and electro dance.
But in the other hand, I put soft cell in 2nd position the other week, ahead Mott the Hoople. That was personal history and the obvious link with Motown.
So I guess genre (tradition vs modernity ?) might play a bigger role than I thought.

Another telling characteristic of a great song is durability. I am most impressed by songs that grab my attention the first time I hear them, but then still sound fresh years later after many listens. Example: California Dreamin' We've all heard it many times, probably, but it still easily won its bracket. Obviously it holds up for a lot of us.

Other songs might sound good to me at first, but then get old fast and I never want to hear them again.

Miles and Britney, together at last. I'm starting to feel better about having Bob Dylan and Starship both on my iPod.

Each to their own, right?

What about Kylie Minogue and Leadbelly (on mine ?)
Oh those funny shuffles..

Oh, that's a good one nicolas. Yes, shuffle is an amazing (and funny) thing.

Makes me wonder how many of us who have iPods use the shuffle feature. It's the most used function on mine, that's for sure.