That’s good insight. One of the artists that I’ve been into recently (who I just posted about in the other music thread) is Baths, and he throws a lot of unpredictable wrenches into his songs. After a while you start to realize there are specific motifs that get used over and over to create that unpredictability though. I think it’s really about developing certain things that work, outside the predictable set of musical tropes. Then, when the toolbox is big enough, you can throw in these well-refined motifs and not have it seem like you’re doing exactly that. Related comic I saw on reddit a couple minutes ago: Also I’m gonna check out that BoC album. I’ve listened to them a bit but haven’t found a good place to really dive in.
Whilst from one perspective what veen said isn't wrong , it's not definitive. There's plenty of incredible and interesting music that eschews the quality of 'catchiness' and the predicitable-but-not-too-predictable approach. That's not to mention forms such as electroacoustic/acousmatic music which are unbound from traditional notions of pitch, rhythm and instrumentation, which are all things we would would consider essential foundations to basically the entirety of western popular music forms. Instead they explore texture, sound gestures, and the transformation/processing of sound sources. Here's 4 different examples of unpredictable and un-catchy music from 4 different genres: Oneohtrix Point Never aka Daniel Lopatin, aka the pioneer of Vaporwave - Americans (experimental ambient) The Physics House Band - Abraxical Solapse (prog rock) Dennis Smalley - Wind Chimes (acoustmatic music) Doom Salad - Relax Drank (jazz fusion) As you can hopefully hear, all interesting pieces of music with great composition, musical moments, and ideas. Yet I don't think you would rush to call them 'catchy' or 'predictable'. The point is this: don't let anyone tell you what music should be. The only person who decides that is you. That's not to say that you should shut your ears to criticism or feedback. Of course you shouldn't. But personally I become wary when people start making definitive statements like "music is interesting if..." (not hating on you veen!). Aim to listen as much music as possible and seek to understand the works and ideas of those who you admire and/or are better than you. After some time (a lot of time) you will start to form your own conception of how catchy, predictable or <insert any other musical quality here> your music needs to be.
I think you're right too - it depends greatly on what you want to do with your music. My advice came directly from the book The Song Machine which is about the history of pop music. I figured that a good, reachable first goal would be to make a song that would be good enough to share on Facebook or something like that. Then, once you got the basics down, write songs that are unpredictable and explore new avenues.