How do *you* design a map?

Get inspired by real life locations, video games, or otherwise.
Find an area you think would be enjoyable to fight in and build around it.

Sketch on paper, draw an angle if something looks bland.
Make a new height level if gameplay feels too dull.

Make your chokepoints have different engagement ranges (short/medium/long), St. Marc in CSGO is a great example, so would Badwater Basin be in TF2.

I have a lot of non-standard examples myself, though this worked for me recently on ctf 4tokyo.

"You don't have to change much to change everything."

The first thing I ask myself is, "what's my hook?" That is to say, what is the defining characteristic of my map. Badlands, for example, has deep trenches and tall platforms, with an emphasis on extreme height differences.

Next, I take to a sketchbook and plot out the main floor plan, where the majority of play takes place. Then the 2nd level, bottom level, etc. Once I'm happy with that, then I start drawing aesthetic details. How the Intel room would look from the perspective of the hallway leading up to it, a Control Point and it's surroundings, etc.

Then I take to Hammer, and whatever finds its way in there to fill in the gaps is usually what I end up going with.

... and other times I just completely wing it in Hammer with absolutley no plan whatsoever.

Generally I like to have a basic idea and overall goal in mind before starting. This could be something as simple as an image of an area I have in my head or a full on sketch. After that I like to feel a map out, see where it wants to go and iterate on shapes and ideas till I get something playable. Once a project as reached that stage, it's all up to testing and more iteration.

Something I learned a while back was to not get attached to anything. Everything can be changed, everything can be replaced, everything is adjustable regardless of how much you like it or how much work you've put into it. It's good to make big sweeping changes in an attempt to actually fix a layout problem then small ones that don't actually address whatever issue there may be.

When it comes down to it though, I work on a project till I run out of ideas, then move on to the next. I may come back, I may not, and while this may lead to many unfinished projects, each one of those is a learning experience to apply to whatever you may come up with next.

Make spawnroom, throw in random blocks to get to a CP. Publish A1. Delete every brush and remake for A2.