Rob (robyrt) wrote in dreamermaxi,

Number-Crunching an RTS

(cross-posted to my lj)

For the last few days I've been working on a master list of the key variables in the RTS Total Annihilation. (My group plays the Uberhack balance mod, so I had to make the chart myself from the stats at BSR's R&D.) What makes TA conducive to analysis is its open-ended structure: potentially infinite self-contained economy, no rock-paper-scissors balancing or research trees, and a genuine physics model. The five variables I chose as the real keys to effectiveness were range, damage, speed, armor, and cost.

Right now, my formula goes like this:
1. Adjust values to equivalence for the Raider, the "average" medium tank unit.
2. Weight range 3x as heavy and damage as 1.5x as heavy as speed and armor.
3. Divide by total to give effectiveness in terms of number of Raiders.
4. In a separate column, calculate cost (metal + 1/60 energy, the exchange rate available at the beginning of the game) and divide effectiveness by ln(cost), to reflect the increasing ease of obtaining resources as the game progresses.

Since my website is down, instead of giving you the chart, I'll simply say that the game's super-units (rapid-fire artillery, giant robots, blue lasers) score leagues above the rest of the crowd, as expected, followed closely by the heavy-hitters we all know and love from experience. Scout units and light defenses rank at the bottom, as expected. The one update I need is a way to factor in the difficulty of using siege cannons, which right now rank in the top 10 of mobile units despite a hefty 75% damage cut.

Once the site is back up, I'll make another post with a fixed chart and my SmartDraw diagrams of the four phases of an RTS game (early, mid, late, end).
  • Post a new comment


    default userpic