Dynamic Map Elements in RTS Games

read the full article at my blog if you want

One thing I always love to see in an RTS is when its game maps, or at least portions of them, are able to evolve over time in reaction to the actions players take. To a lesser extent, I also enjoy seeing maps which change over time without player input, to force players to adapt to changing situations and threat profiles, but I feel that has less of a solid case in competitive games.

In many RTS games, especially many of the the classic franchises, game maps are largely static. They may become denuded of resources (like trees) over time, which opens up new movement options to target enemy holdings, but in most cases this is not a major tactical consideration across most competitive matches in those games.

There are tactics and strategy games, however, where the map itself is able to be modified by players in a variety of ways: opening up new attack paths or closing them off, changing which areas of the map players care about and want to contest, or removing options from players in specific geographic areas. 

A screenshot from the Steam version of Z which shows what cliffs look like. The blue tank at the top of the image is shooting cliffs to destroy them so that it can pass through. 

The oldest game where I remember seeing what I'm calling a 'dynamic map' is Z, the retro tactical robot game by Bitmap Brothers. In that game, there are cliffs which block traversal by units (see screenshot above). Infantry can take down the cliffs by throwing grenades at them, which of course depletes the limited supply of those valuable pick-ups. Tanks can also fire on cliffs to destroy them. The presence of destructible cliffs might not have had an outsized impact on the game's design, but it certainly did on my imagination as a kid.

Some Additional Examples
Relic's RTS games are more recent examples of games with dynamic map elements. In Dawn of War 2 and Company of Heroes 2, some higher tier units are capable of crushing terrain and cover underfoot or under tread, and some weapons hit hard enough to create craters that can themselves provide cover. Additionally, in these games, when vehicles die they remain behind on the map and can be used as cover or destroyed entirely. Over time in these games, the map devolves into something else, devoid of cover for infantry, covered in craters and the smoking wrecks of tanks and halftracks.

Large units, like the Carnifex, are able to crush map elements by walking through them 

While not seen ranked multiplayer any more, Company of Heroes 2 also launched with weather effects on the game map, where at irregular intervals blizzards would occur, freezing water, creating snow drifts which would slow down infantry, and creating hostile environments where infantry would die without access to a heat source. Some weapons, such as flamethrowers, could melt the snow. This wasn't so much driven by player action, but did force interesting reactions from players in terms of preparing for the blizzards and how the game worked during and even after them. Clearly it wasn't a widely popular system, since it ended up being removed from the multiplayer experience. 

Additionally in Company of Heroes 2, explosives and some weapons are able to destroy ice under troops and tanks, instantly killing them by submerging them in the frigid water underneath. This occasionally has a dramatic effect in winter maps, in addition to changing how units are forced to move while on ice (to avoid holes in the ice). I believe during blizzards, ice could re-freeze as well, creating a cadence of changes as ice was destroyed and re-frozen.

Tanks breaking through the ice in Company of Heroes 2 is one of my favorite examples of dynamic map interactions 

Supreme Commander has a slightly different take on this: In SupCom, when units die, they leave behind a wreck that can be salvaged for resources, seeding the sites of battles with valuable income for whomever reclaims it. This doesn't change the map itself as much, but does change how players interact with the map and which parts of it might be the most valuable to hold or contest.

Company of Heroes games, and now Iron Harvest, have a similar idea in that when units die they drop either the weapon they're crewing, or a weapon they're carrying that then any other infantry squad can appropriate. This allows for both tactical flexibility (since players are able to gain squad types on the fly) and a transfer of resources from one player to the other, since the weapons all cost resources and a player can gain a tool paid for by their opponent at no cost. Actually, C&C 3 does this as well with some vehicles dropping wrecks that can be re-crewed by engineers for quick access to a new unit. 

Company of Heroes 2 goes farther with this than any other game I can think of (including C&C3) with de-crewed tanks having the ability to be captured by other players. Since tanks are so powerful and expensive, scoring one off of an opponent is a major gain. 
I consider these dropped weapons and crew weapons and tanks to be 'dynamic map elements' in exactly the same way that I consider dead units in Supreme Commander to be such: they're free resources that occur in locations on the map that are the result of combat between players, that can advantage either player/team and cause those areas of the map to become temporarily tactically or strategically important, to say nothing of dead vehicles becoming cover for infantry in these games. 

I actually really like Supreme Commander's approach, and have implemented a similar system in my personal game project: SCRAP mod. In SCRAP, when units die, they decay into resources after spending 3 minutes as a burning wreck, giving players time to jockey for control of those resources before actually being able to mine them.

When units like the Aeon Illuminate's Czar die, they create obstacles to unit movement that are also valuable sources of income for either player, creating an area of high value on the map that didn't previously exist 

Even StarCraft 2 and Grey Goo have dipped their toe a bit into maps that can evolve over time with the addition of rocks that block certain map access points; and, in StarCraft 2's case, pillars which can be destroyed to block off access to areas of the map with debris that can itself be destroyed to open up the path again. 

Going into Detail about 'Terrain' 
In most cases, implementations of dynamic map elements are mainly focused on how units traverse the game space. Destructible bridges are a good example of this - in several of the Command and Conquer games, Engineers are capable of re-building bridges by entering a designated structure at either end of the bridge.

Also in most cases, the majority of actions players can take is to remove said dynamic map elements from the game. There's a lot of crushing and burning and demolishing that goes on, and not a lot of growing or shoring up.

Builder units in Earth 2150 can modify map terrain to make bridges, walls, ditches, and more. 

I need to be careful here: I'm specifically using the term "dynamic map elements" instead of "dynamic terrain" because in almost no case is the actual terrain of the map able to be modified. If there's an impassable cliff, it remains a cliff for the duration of a match. By contrast, a 'dynamic map element' is in some ways similar to a player-produced structure: it's something that cannot move, that can be destroyed, but in the case of a dynamic map element it also serves some tactical purpose in the game but can be utilized by either/any player. Alternately, such 'map elements' might be an indirect (to me, it's important that the reaction be indirect!) result of player action: a dead vehicle turning into a wreck, or a heavy attack leaving a crater, that sort of thing.

Why is it important to me that dynamic map elements be created as the result of indirect action? That's a really good question. I think, ultimately, it's because at that point both/all players must react to the changing battlefield as opposed to having the players prepare a changing battlefield to force their opponent to react in order to deal with. I wrote a while ago about categories of action in strategy gaming, and I feel that strategy games have tended to have a sufficient amount of preparatory actions in them (e.g. actions that players build up over time in order to give themselves an eventual leg up in the game) and game maps in particular remain pretty static over the course of any competitive multiplayer match, unless dynamic map elements are present. Adding reactive elements to the map, to me, is a positive thing. Gotta keep players on their toes.

Some games, such as Earth 2150 and, more recently, Zero-K, do allow for a more free-form modification of terrain itself: making mountains, bridging chasms, digging ditches, that sort of thing.

In Zero-K, terrain manipulation can be a pretty big deal 

While really fascinating, I'm not fully convinced that competitive strategy games are always made better by providing players with unlimited freedom to modify the terrain of a map. In Earth 2150, for instance, I rarely if ever made use of the terrain modification tools for any reason.

Of course, Earth 2150 is admittedly an example of 'kitchen sink' design: players design their own units, the game has a 'picture in picture' mode where the player can watch multiple areas of the map simultaneously, there's a system for digging and moving units around in underground tunnels on a separate layer of the map... the list goes on. Zero-K is probably the RTS that I can think of which takes terrain modification most seriously. There are tactical and strategic elements to the game's terrain management system that interest me, even if I haven't played much of the game yet.

I tend to be more in favor (personally) of game systems that are clearly defined and constrained, while still allowing for deep and complex interactions between units. I tend to be a bit wary of more freeform systems, including 'design your own units' type systems in games. I find that it's often easy to find 'optimal' builds in those systems that actually have a smaller number of viable builds/strategies than games with pre-defined systems and also that they're also ultimately too much complexity buying too little depth, with the added risk of that complexity actually damaging emergent gameplay where it's trying to foster it instead.

I would like to be 1000% clear here: I'm not saying that I'm against full terrain manipulation in RTS games. I'm just saying that I'm skeptical of it and have yet to be convinced either way. I may report back once I get more hours into Zero-K as to the results of that experience.

While admittedly more binary, I tend to prefer things like how bridges work in (most) C&C games. They can be destroyed by units (killing anything on the collapsing bridge) and rebuilt by an Engineer going into the bridge's associated control tower. This gives both player a lot of say over when, and how, bridges exist on a game map. Something like bridges or destructible buildings/other terrain objects (like cover in DOW2) to me feels a lot more interactive and straightforward to grok to me than something more nebulous like more general terrain manipulation. 
 
I'm a pretty big proponent of clarity in terms of game systems (you might find this hard to believe if you've played my personal mod project, but I digress). And, to me, concrete objects on the map like buildings, rock to hide troops behind, trenches... these are things that a player can understand and interpret clearly. The lip of a crater will protect your troops from incoming fire; holing up in a building will do the same. Setting the building alight to prevent the enemy from using it or to destroy a bunch of enemies holed up inside, well... that's a pretty clear interaction too. 

 
Wrapping it Up
There's definitely cases that can be made both in favor of, and in opposition to, both what I'm calling "dynamic map elements," "dynamic terrain," and traditional static RTS maps. 
Competitive purists: those that mostly play Age of Empires 2 or StarCraft 2, tend to appreciate the relatively limited rough edges that maps provide that get in the way of the expression of their skill and competitive drive. There's not much in a Brood War map that's going to mess with a player's expectations of how different encounters are going to go, for instance, or the parts of the map that are going to matter over the course of a match.

 
In Command and Conquer Remastered, however, the presence of Tiberium on the map can have some impact (this is more true in Tiberian Sun, where Tiberium regenerates faster and can be more lethal). In the older games, players don't have much option to control where Tiberium is, but as the maps clear out of Tiberium, they become much safer for infantry to traverse, which can have some impact on the progress of an individual match.

So, to me, the biggest con for dynamic map elements is that they might not fit into the game based on the overall design goals. Though again: even StarCraft 2 has managed to find limited expressions of this sort of thing in the form of destructible debris.

Ultimately, I think it's a great thing to have dynamic elements on the game map, whether it's as simple as rocks that block paths, which can be destroyed to open them up, or whether it's more complex like defensive tools that can serve either player, and be destroyed by abilities (like grenades). When the map responds to player action, it creates emergent depth in the game by giving the players new areas of the map to fight over (whether it's for resources, or for actual weapons of war to steal from their opponent), or creating new ways to react on certain areas of the map by removing or adding defensive options, or new approaches to dealing with their opponents. 

On average, I feel like it would be good for maps in RTS games to feel more like a living place and less like a static game board. They should evolve based on player actions, with crashing helicopters digging up turf, buildings burning down, walls crumbling as shots or vehicles pass through them. I ultimately like this for mechanical/systems reasons, but I imagine it would also be cool visually.

What do you think about the topic of dynamic map elements? Should maps be more static, like in StarCraft, or should they evolve more like in Relic's RTS? Is more freeform terrain manipulation like Zero-K or Earth 2150 better, or going too far?

Thanks for reading.

read the full article at my blog if you want