Iterative Development for Economy Calculator

This iterative build was to test the economy calculator at different time intervals. I created an array to store information on 5 basic protoss units – Nexus, Pylon, Gateway, Probe and Zealot. I then wrote some quick calculators to accept a time, in seconds, and return the amount of minerals available at that time. The display is fairly minimal, just showing economy, but various units and builds were constructed at certain times.

The calculation is based on finding out how many probes exist at this time, how long they have existed, and multiplying that by 5/6. 5/6 is the number of minerals gathered per second per harvester (before saturation, at least). Once I have found the total minerals gathered, I get a list of all units that have begun construction at that time and subtract it from the total and return it.

The formula for calculating wealth seems okay, but the initial array contained far more data than I ended up using and the formulas were a bit too hard coded. Things like MULE and Chrono Boost are not being accounted for, and supply count is ignored as well.

I like the idea of a function being passed a time value and returning a snapshot of the build status at that time, and the array storing the build information also containing when each item was started. This system would fairly easily allow new commands to be inserted mid way through the build.

I definitely need to make the calclations more dynamic, though. I’m also still unsure of how to handle the starting Nexus and 5 Probes. This system puts the build time at -100 and -17, respectively, and always adds 700 minerals to the mineral count.

All in all, this iteration did show some interesting ideas, particularly with where to store data on all of the units. I ended up sliding it in as an array inside the build class, but that seems impractical. I should probably set a race as a class, units as a class, buildings as a class and special abilities as a class. My next experimentation will involve testing those methods for viability.

  1. No comments yet.
(will not be published)