A few days ago I had a shower thought, tying together two current projects: what if I could design an “abbreviated” version of the Architect of Worlds design sequence specifically for the Traveller role-playing game? The idea would be to produce a world generation sequence that uses mechanisms similar to Classic Traveller, and is not significantly more complex or time-consuming, yet avoids the more implausible results of those rules, and is compatible with the full Architect of Worlds design sequence.

So for the past week or so, I’ve been running experiments along those lines, with considerable success. Consider this an interim report on how it’s going, with a high degree of confidence that I’ll have a finished product ready for release within a few more days. When it’s done, I’ll post it as a freebie for Traveller fans and my patrons.

One piece of this has involved a lot of work with Excel spreadsheets.

Microsoft Excel is a very good tool for what’s called “Monte Carlo” simulation. If you’re working with a complex model that has lots of moving parts, and you can’t just compute the distribution of likely outcomes, then the thing to do is run lots of random trials and look at how those turn out. Rather to my surprise, I’ve developed an Excel spreadsheet that does this for Architect of Worlds. It takes a few parameters (the mass, density, and blackbody temperature of a putative world), then it runs through a not-overly-simplified version of Architect of Worlds and spits out 1,000 random planets. Finally, it counts how many of those planets exhibit various combinations of atmosphere and surface liquid coverage. All at the touch of a button.

It’s a rather heady experience. Amazing, the amount of computational power in a simple laptop device.

Working with my Monte Carlo setup has given me all the data I need to produce that abbreviated Architect for Traveller. I’m pretty pleased with the result – it’s a bit more ornate than the original Classic Traveller world design sequence, but most of the complication has been pushed into one big lookup table. I think the meat of the final document is going to be no more than 3-4 pages long.

This side project has also led me to find a few odd cases and possible errors in the current Architect draft, so I’ve made some good progress on what’s going to be the version 0.5 alpha release. My patrons can look for that later this month as well.

Finally, I’m beginning to think I might be able to automate Architect of Worlds with another big Excel workbook. I’ve used crude spreadsheets to do some basic testing of earlier versions of the sequence . . . but I’ve learned a lot about Excel in the past week. This might be useful for any number of other projects.

Progress!