Recently I posted somes older game art I did for a client on pixeljoint, here's the image for Pix as well

I got a few questions about how I pick colors, something that I've been asked a lot, generally, but more important to me lately because I have something useful to answer with.
Never was much for colours, honestly. On paper I am a black and white, ink or wash tone artist. I literally learned color through pixel art, and the idiosyncratic aspect of me was very interested in 'symbolic colors' from early on. I wasn't so much interested in realistic representation with color, instead I would use, say, a green tint as some sort of visual language (green for me for example is usually sickness, evil, depression, stuff like that - very common interpretations).
So, influenced by certain demoscene artists at the time and from the pixelation kind of way of doing things (which was at that time being in a constant feedback state, it wasn't set in stone) I would spend a lot of time trying to
1. get palettes to be small, usually sub 16 colors, and
2. get a lot of hue representation in there, what we ended up erroneously calling 'hue-shifting'.
Good sense on which hues to choose to move towards and through which grey steps I never felt comfortable that I could explain or 'teach' in any capacity. This isn't what this thread is about. When people, 5, 8 years ago would ask me how I choose colours, I didn't have anything interesting to say, I don't think.
What I can tell you now is that my choice at that time of introducing a lot of grey slots between more saturated colors created a very certain 'look' that I wanted to move away in the last few years. And one of the tools I now use to do this (and I used it in this piece as well), is that I go in photoshop and use
color balance somewhere in the middle of creating a palette.
I'm sure you can do it in some other program, but the concept is the same. RGB sliders for shadows, mid and highlights, in a 24bit color space, regardless if you're going to end up with 16 colors or what.
Why is this useful?
Building on my ideas about symbolic colors, an overarching tint to a picture can be very useful, especially a balance between two tints. Above the attempt was between the yellow/orange in the earth and the purple and cyan in the shadows and sky. For me that's a picture of summer. But it didn't start out with this colorspace. At one time the hand-picked shades were more in the middle of what's going on here:

I did a lot of color correction, then back to manually changing every palette entry, then back to color correction, then back to manual editing.
What is the benefit of this?
It removes a lot of the muddy gray identity that a lot of my past art was left with. Nowadays I don't have this problem as much anyway because I do not use unlimited bit depth for my palettes so everything ends up quite saturated anyhow, but I know most pixel artists at this time are choosing their colors from a very large bit depth, so some sense of overarching tint would be useful for their process.
The theory is very simple - when you're adding +2 to red in the middle part of the values, if it finds a 20,20,20 color (a gray) it's going to make it 22,20,20, so from a gray it is now a gray-with-a-bit-more-red. If you take it really easy with this process, you can get an environmental effect, something like a soft bloom to work even for your pixel art.
As an activity, you could post pixel art (esp background art and tiles) that you think could benefit from this, before and after color correction and manual calibration.
Is this making sense, do you get what I'm going on about?