Wayward Strategy

My first SC2 Arcade project: “Remnants”


A couple months ago, I wrote up a thought experiment about a strategy game in which no resource was plentiful, a so-called “scarce-resource RTS.” Since then, I’ve been working on a mod for StarCraft 2 intended to prototype and experiment with this idea. Mostly for my own edification, I’m going to write up a brief and broad overview of the project, outlining the theme, salient features, inspirations, and my overall goals for the project. I won’t be sharing this widely on social media, since again this is mostly for me. Thanks to anyone who takes the time to read this and/or shares their feedback.

I’m going to approach this in what is a logical order for me: starting with the game’s setting and inspirations, moving on to the themes I’m hoping to capture mechanically, then on to the game’s salient systems, factions and faction design, and the conclusion which will contain a rough timeline for implementation. As much as I fancy myself an ersatz game designer, I fancy myself one who starts with a thematic or story concept and builds mechanics around that, instead of starting with a mechanical system and building a story around those parts. I am given to understand that these are two common approaches to games design, but have yet to grok the second approach in a meaningful way.

The above image is from The Meatly. Check out his site! It’s awesome. Should the original author express frustration in my copying of his work in this location, I will happily take it down. Hopefully the fact that I’ve sourced it will be sufficient.

the working title of this mod is Remnants, by the way.

Inspiration and Setting

So, let’s talk setting + inspirations. Being a ‘scarce resource’ RTS, it seemed fitting that the setting be post-apocalyptic. 2 of my original inspirations were in fact Cormac McCarthy’s novel The Road, and by Tim Burton’s film 9. The Road is notable in a number of ways, especially for being set in a post-apocalypse scenario in which all readily available resources (in its case, mostly food) have long since been removed from obvious places like stores and private homes, a huge contrast from settings like the Walking Dead in which it is relatively easy to find sustenance and weapons, only procuring them is difficult thanks to hordes of zombie Walkers. In The Road, we’re far enough into the apocalypse to see stores picked bare and homes looted years long past, and merely finding sustenance on a day to day basis is nigh-impossible. Now, this doesn’t translate well into RTS mechanics, but its pressures are something I very much am interested in exploring.

In the film 9, we witness the struggle of a small group of social, intelligent machines mostly intent on survival against the monstrous threat of a single artificial intelligence intent on harvesting their unique power source as a resource for its continued production. All weapons and creatures encountered are cobbled together out of the leftovers of a previous civilization that is little understood by the current inheritors of the world.

Also, the 2008 WALL-E contained some scenes in the early minutes of WALL-E repairing himself with pieces of his broken-down garbage compacting comrades, and clear indications of his programming mutating over the years, giving him something closely approaching sentience.

Let’s collect all of these things above to produce an elevator pitch of sorts for the setting: a small group of social machines, cobbled together from the leavings of a higher civilization, struggling against resource scarcity in the context of a post-apocalyptic world.

In the context of this being a StarCraft 2 Arcade title, I’m allowing that these ‘higher civilizations’ are in fact the Terrans and the Protoss, since for the time being I have to use SC2 models for terrain, units, doodads, structures etc. Ideally, I have other plans, but this is sufficient for the time being. So, we have a setting in which a number of machines have gained sentience on a planet left barren by the constant warring of Protoss and Terran forces, trying desperately to survive.


I went into this above, but let me state again: one main theme of this game will be resource pressure. I specifically want to create a strategy game in which a constant flow of income is not to be expected. Resource gathering will be in fits and starts. My ideal situation is that players would locate and mine out a resource cache in short order, then move on to the next while making hard decisions on how to spend the resources they just obtained.

Further, the overall resources on a map will be limited in a way that you don’t see in StarCraft 2 proper. In StarCraft matches, it’s theoretically possible to mine out an entire map’s worth of resources, but it seldom happens. In Remnants, my intent is that in an average game, there should be no extant on-map resources about 10 or 15 minutes into a match (this to be adjusted based on testing).

To prevent the exhausting of on-map resources from setting too hard a limit on match length, units will ‘drop’ 50-75% (to be adjusted based on testing) of their cost worth of resources when they die, allowing for new resource fields to be seeded on battlefields. I’m waiting to see a bit on this design decision, as it seems like it will create a slippery slope that favors the victor of a given engagement. I have some mechanics I think might ameliorate this somewhat, but this still remains a point of design contention.

Another theme I hope to carry through the design is hard choices with nontrivial outcomes. I will go into this, as with the previous theme, in more detail in the following section, but at the core I want the player to be faced with mutually exclusive choices, both economically and tactically, that have nontrivial outcomes. As a for-instance, in the Dawn of War games in particular, often choosing an upgrade for a unit will be mutually exclusive between 3 alternate options, and once that upgrade is researched for a squad, the 2 other options are no longer available. You also see something similar in the upgrade system in Grey Goo, where players are only able to pick 1 of 3 options per upgrade module, though the option can be later removed and re-picked.

So, elevator pitch: players will be pressured to make hard choices with limited resources and mutually exclusive upgrade paths.


I went into the resource system for this game in my thought experiment, but let me summarize and expound briefly here. the game will include 2 resources: scrap, and power. The game’s 2 factions will treat the power resource differently, but will utilize scrap as the main component in unit production and upgrades.

Scrap is a resource that is found in fields, much like Trees/Wood in the Age of Empires or WarCraft series. Also like these games, scrap can be either harvested or destroyed, though mostly it will be higher tier units that are able to destroy scrap with any reasonable speed. Unlike StarCraft 2, scrap harvesting operations will be undertaken with tier 1 combat units (much like Undead faction ghouls in WarCraft 3) and the drop-off point will be the player’s non-replaceable, mobile Constructor (the other faction will have a somewhat different drop-off point).

The ‘story’ here will be the Constructor and its retinue moving from scrap patch to scrap patch, harvesting, producing and upgrading. Most construction and upgrading will also be done via the Constructor, though it can produce structures which will expand upon its core functions.

For both factions, all buildings require Power. Power is more of a system than a discrete resource in its own right. The Power system has 3 components: Generators, conduits and Power Nodes. Nodes work like Protoss Pylons, projecting a limited field that allows for construction. However, the player must capture Generators to activate corresponding Nodes. The Generator and Node are connected by a Conduit, which can be destroyed (though each faction will have units that can rebuild destroyed Conduits).

Thus, both Power and Scrap can be harassed (by destroying Scrap that could otherwise be used by an opponent, or by capturing enemy Generators or destroying enemy Conduits).

Furthermore, many units will have Power reserves they can use for various abilities. However, unlike in StarCraft 2, units will not generate their own Power slowly over time, and must return to Power Nodes to refill their power reserves. This will serve as a sort of ammunition for many units, especially more expensive high tier units, and should serve as a method to limit unit effectiveness without outright destroying them.

In addition to “destroy enemy Constructor” (or its analog) each map will have an alternate win condition. Each map will have 2 Substations and one Mainframe, representing the knowledge of a lost civilization (in this case, the Protoss). The faction in control of this knowledge will win the match. To control the Mainframe, the player must capture and hold both Substations. Once the player captures both Substations, they must hold the Mainframe for 6 minutes (non-contiguous) to achieve victory. That is, if they hold the Mainframe for 3 minutes, then capture it again, they must only hold the Mainframe for 3 additional minutes. This is made more difficult by the Mainframe only being obtainable by a player holding both substations simultaneously.

One of my main goals is to encourage small-group combat in many areas of the map simultaneously. In having Scrap fields, Power Generators, Conduits and Nodes, and having mobile resource gathering and unit production operations, I hope to create maps with numerous dispersed points of contention that players must focus on simultaneously.

Factions + Faction Design

I am planning 2 factions, the Constructors and the Dendrites. The Constructors are a ‘race’ if the term may be applied, of civilian construction robots, essentially mobile 3d printers, that have become self-aware enough to have some self-preservation instinct. The main focus of the Constructor is to upgrade itself and to create an army of lesser robots to fend off anything that might do it harm. Over the course of the match, a constructor will go from a mostly harmless mobile drop-off point to a considerable fighter in its own right, with the player able to choose from a variety of weapons, defensive options and modules that increase its versatility and potency.

Constructors will have a number of structures they can produce to expand their core capacity. As a single-thread harvester/unit producer/builder, it will behoove the player to build factories and repair centers to offload and supplement these core functions.

The Dendrites are, essentially, self-aware power generation systems and will function somewhat differently from the Constructors. The core of the Dendrites is the Axon, a structure that requires power to function as a unit production structure, but serves as the drop-off point for harvesting operations anywhere on the map. The Axon can ‘unpack’ into a type of Dendrite, so any surviving Dendrite-class unit can become an Axon, making the system somewhat more survivable than an early-game Constructor. Unlike the Constructor, however, Dendrites are pretty weak and very specialized: for research, unit upgrades, and structure production, the player must ‘spend’ his fragile and expensive Dendrites much the way the Zerg have to spend Drones.

I haven’t finalized the unit lists for either faction, but anticipate having 4 ‘core’ units per faction each with 3 possible upgrades, of which a player will be able to choose 2 over the course of a match, meaning there are 16 overall units per faction (not counting the Constructor) of which 10 should be available at any given time. Likewise, the Constructor and Axon/Dendrite should have a total of 15 upgrades, with enough mutually exclusive choices that they would only be able to implement 10 per match total. A unit would be ‘spent’ to turn into its upgrade variant (as Zerglngs into Banelings).


That’s a high level look at what I’m (attempting to) build. I think the idea has some merits, and certainly some issues that will become more apparent as I continue work on the mod. I welcome any feedback you may have, and thanks much if you actually made it this far.

The game’s first pass will include a functioning, basic map and the Constructors faction, or most of it. The second pass will include map refinements, and the full Constructors faction. the 3rd pass will include the Dendrites. Subsequent to this, I would like to add:

  1. The ability to pick from 4 different maps + choose faction
  2. leaderboards
  3. somewhat competent AI
  4. a complete game-within-a-game with tutorial, single player campaign and functional multiplayer
  5. custom models for at least Dendrite class units and the Constructor
  6. Eventually, a version of the game not based on SC2 whatsoever + drop any SC2 references in lore.

Who knows if it’ll get that far. It may only ever be 1 map with Constructors vs Dendrites. But a man can dream, can’t he?