I’ve mentioned enough times before how most of my maps have been entirely unplanned. Dust was unplanned. Dust 2 only had the vaguest of plans. Cobble was unplanned. Tides grew out of a single photo I found online. ETC and ETC2 had very little planning indeed.

I’m not a fan of drawing plans on paper. It’s never worked for me - the second I start plonking the design down into the game, I start changing it. Sometimes it’s because the plan is bad. Sometimes the scale is all wrong. Sometimes I forget I even have a plan and just make it up as I go along anyway. None of my maps bear any resemblence to any plan I’ve ever drawn.

On the other hand, some people swear by them. They plan everything down to the tinest detail, scrawling notes and precise measurements down, almost like an architect. Their finished maps are perfect realisations of all the notes and diagrams they’ve drawn. It’s a remarkable achievement. Yet, in doing this, they’ve lost a huge aspect of what level design is and the whole advantage of the physical mapping process:

The ability to change. Adapt. Modify. Correct. Destroy. Trial. Test… and most importantly, undo.

The problem with planning on paper is that all you get is a static 2D representation. 3D if you feel like it. Put the same structure into a map and suddenly you have freedom to explore and consider the intricacies of the creation. It’s the best possible way to work out what to do next - by simply looking at what you’ve done so far.

For me, planning on paper happens at the very basic of levels - to get the premise of the map. To understand the concept and the purpose. It is not to work out where the doors are, where the spawns are or where the enemies are, because I know this is easier and better done with the 3D geometry right in front of me. I can see opportunities that 2D planning wouldn’t have offered me. Meanwhile, 2D lets me see the bigger picture.

I tend to go through three stages. In the first, I try out some random structures and get a general theme going. In the second I work out what I will create with the theme and the basic gameplay concepts. Third, I actually build the thing. Only the second stage involves planning, since it’s using whatever was learnt in the first stage to determine what to do in the third.

Works for me!