Saturday July 24, 2021
I've recently had some thoughts about the competitive 1v1 ladder scene in RTS games. Specifically, I recently spent some time trying to get into Supreme Commander via the fan community Forged Alliance Forever. From that experience, I want to compare the starting conditions for a game in Supreme Commander to some other well known games.
I find the start of games in Supreme Commander to be different from the start of other games. And I'm not sure that it is a positive difference. So if you wanted to do a truly optimized build order in Sup Com, there are a number of map based factors you might consider. Here are some that I've recognized:
To take advantage of these factors optimally, you need to rearrange or change your build order. If there's a Hydrocarbon deposit nearby, you'll want to send engineers to it early to take advantage. You'll have way more power when it's done, but less in the meantime because you're probably not building the same number of normal power generators as you normally would. This changes the sequences of things you build during the first few minutes of the game. Similarly, the arrangement and distance of early Mass Extractors is anything but uniform, which leads to a wide variance in when you start to get mass income.
But that's not the end of mass variance in the early game. There's reclaim, which can allow you to do things that you could never normally do, because it nets a very large quantity of resources while it's still around to harvest. Taking advantage of reclaim means tasking engineers to go out and harvest it instead of doing other things. Or maybe you build engineers just for that job, in which case more engineers are getting built at an earlier time. My point is that all of these factors can justify substantial changes to your build order.
Supreme Commander maps have a wide variance in all of these factors. Let's look at another game for a point of comparison. So the largest esports RTS is Starcraft 2. A starting base in that game will likely have a mineral line, probably with 7 deposits. It has 2 gas deposits. These things have very little variance in their position and distance. It will probably be on an elevated platform with a single, defensible, wall-able entrance. Your starting position will probably have a natural expansion directly next to it. All the resources of the map are going to be in "bases" with the same configuration as your starting base.
Basically, no matter what map you start on in Starcraft 2, the map is unlikely to substantially change the first 3 to 4 minutes of your build order. But that's not true in Supreme Commander. Weirdly, Sup Com isn't the only game with this many effectual axis of variance in starting conditions. Age of Empires 2 also has many analogous properties. The placement of almost all of the resources on the map is partially randomized in AOE2. When you start, you don't know exactly where the berry bushes will be, where you'll find your starting sheep, where your initial gold deposit is, or where the first clumps of forest will be. You get the picture. And yet there is a big difference between AOE2 and Sup Com in this category.
It's that Age of Empires 2 randomizes it's maps. Resource locations are always somewhat randomized. Even when you play the same map twice, the resource locations will slightly change. So even though this type of variance exists, it's something you can't plan for beyond using a build order with enough slack to deal with it. I will note that even though exact resource locations in AOE2 are randomized, their distances have seemed remarkably consistent to me. Even so, the randomization does mean minute optimization a build order won't translate to the next game.
Supreme Commander does not randomize it's maps at all, so it does not have this property preventing planned minute optimization. As a result, you can reliably optimize build orders to every single map. You can configure the timings of your decisions on a per map basis making use of every little property of that particular map. And unlike Starcraft 2, there's a lot to take advantage of in that category.
I'm not going to pretend that most people really do take advantage of the ability to optimize build orders that closely in the game, especially outside the top of the ladder. But all this does make the game daunting to learn. I can't just "learn a build order" and play it on ladder, unless I'm willing to ignore the idiosyncrasies of the particular map. And that's probably the right call, especially while I'm still learning as a noob. But it sure is disappointing and unsatisfying to realize that. It's a trade off, one that I'm sure many of the game's long time competitive fans really enjoy.
In conclusion, if you were designing an RTS, you would have some choices. Randomizing these factors would create a higher skill requirement in reading the situation and improvising reactions to these differences. Keeping these factors the same per map between matches creates a higher skill requirement in planning and optimization between games. Standardizing these factors across all maps smooths out the learning curve and makes lessons learned more likely to apply to other situations.