You guys are all over-simplifying.

I don't think you can even really compare the "hardness" of these things.

For one thing, we can pretty easily measure someone's skill at playing an instrument by comparing accuracy to the sheet music or a reference performance.

You can't do anything like that with writing music. The popularity of a song does not tell you anything about the skill required to write it. In fact, it tells you more about the skill of the performer, the production quality, instruments used, pre-existing popularity of the performer, familiarity of sound*, and marketing.

Songs used in movies, TV, and radio become popular. Ones that aren't don't. Very few exceptions. This isn't just true for music, it's true for everything. If you never hear it or see it you're never going to like it no matter how much you would like it.

The fact is, there is waaaaaay more music than can ever be marketed or heard on a large scale.

It doesn't take me long to find extremely unpopular music that I like. Even though I think the artist is great, I know there's very little chance they will ever become popular. Not because they suck but because there's only room for so much popular music. You can't have 10,000 songs all popular at once, it just doesn't make sense.

We can't all be astronauts or pro-athletes, either, even if we are all perfect for the job(s).

As with anything, the best way to measure something is to ask the experts. Have a large group of music writers judge the skill of each other, and you can probably get a pretty good measurement.

But I don't see anyone doing that.

If you use popularity, at best you are judging skill at writing popular music, not general skill. At worst you're just measuring the factors I listed earlier and pretending you're measuring skill.


*It's been shown scientifically that people like music because they recognize it. Your brain is literally trained over time to become accustomed to the music you listen to, and recognizing the same patterns leads to enjoyment. This is why if any of you (probably) listen to very foreign music you won't like it. It won't even sound like music to you because your brain is physically not adapted in a way to recognize it musically. This is especially true of foreign music that uses a different scale with more or less notes.

Disclaimer: you are (probably) not a robot and your brain's musical recognition configuration is not the only thing that determines whether you will like a song. You can think too, so critical evaluation and emotions all play a role. You may not like a song because it sounds too unoriginal, for example. But generally you will like a song because it is both familiar and interesting in a new way.

Marketing isn't random. If a song is marketed it means someone with money thinks it is good enough to become popular. Also, any reasonably familiar-sounding song that gets airtime will become popular. I don't think you can dispute that.



That's a pretty good way to measure the likability of a song fairly.



My experts are not evaluating how "good" the music is, only the skill required to write it. People who have experience at a task are obviously the best judges of how hard it is to do.



Not all music is made for the general population. Some artists avoid doing that on purpose. If a song is popular and experts say it took no skill to write then it is both. No one is wrong.

I don't know. I don't know what songs you are referring to.



I don't think so either.



If it's semantics then just replace skill with creativity in what I said before and I think it still works.

I think it's a combination of both though.

The problem with music in particular is there just aren't that many combinations of possible things to write. There's only a few notes/chords to work with. You can pretty much brute force it until it sounds good if you have a lot of time. And if what you think sounds good is the same stuff that gets popular then you can write popular music. (Then you need it performed, produced, and marketed -- the expensive parts).

As with any experiment, feel free to modify it to lessen the effects of bias. For example, weed out the data where the experts didn't like the music personally. Even better, require that they first choose songs they like, and then order those by apparent skill. Unless all the experts can't separate what they like from what they find most skillful, you should get useful results.

Quality as it pertains to what? A shoe is a very low quality mitten. Music doesn't have a concrete use case so you have to define one before you can determine quality. You can select popularity but then let's just call it that and stop saying quality.



I don't think we have to use a bad method at all. I think we should just clearly define the terms.

I'm sure some would, but all of them? I can tell you out of the music I like which songs are very low-skill. And they're certainly not the ones I like least. There's also plenty of music I hate but I can tell you required skill to write. I can't be the only one.

Even if most of them fuck it up, only a small amount need to get it right to show up statistically.

I don't think there is a an ethereal, non-specific good for anything.

With shoe quality, it's safe to assume we're talking about durability, comfort, protection etc.

With music quality, the fairest measurement I can think of is something like: how highly regarded is it among people who like it? How much skill and creativity went into it's production?

Even with shoes, you could have ambiguities. For example, what if I want a shoe to run a race in once? Suddenly durability is out the window and comfort and speed are all that matters.

I think music is the same thing. Its quality depends on what you want to use it for. If you want to use it to entertain, inspire, comfort or motivate a particular person or people then it's ability to do that is its "quality".

Quality is always context-sensitive.

I mean "experts" in the sense of "expert witness" not in the sense of "best in field."

If you run a study and invite music writers as volunteers then you will get the experts I'm talking about. Just people with experience in writing music at all. You could probably weed out some volunteers with some simple tests but I'm not even sure that's necessary. I don't have any reason to believe a statistically significant number of volunteers would be lying. If that were the case, no studies would ever work.

I think your example of mathematicians is funny because I could probably trick the average person into thinking I'm a math expert without really knowing any math, but I couldn't do the same with music writing.

I know.

But since then we also started talking about skill, creativity, and the difficulty or rareness of playing versus composing music.

What if an unintended audience really likes it?

(Warning! Not serious post!)