What Are We Fighting For? Winning and Victory Conditions in RTS

Welcome back, commander!

It's been quite a while since I was able to find (make? steal?) time to write an article in my own blog. Turns out, being the lead game designer on an RTS game can occasionally be somewhat time consuming? And while I'm ridiculously thankful for my excellent contributors, it feels really good to spill some digital ink of my own as well! I'm glad to be back, and thank you for giving this post a look.

While I am the lead designer on Tempest Rising and I'm sure anything I write will be interpreted in that context, I'm also very interested in the greater RTS and strategy game landscape That's always been the point of this blog, after all! With Dune Spice Wars releasing a major update lately, not to mention the contentious Company of Heroes 3, and even Age of Empires 4 out there as well, this has been a pretty good time to think about and discuss the mechanics of how games are won.

My Discord group recently had a spirited discussion about gameplay objectives in RTS games, and I realized that I haven't written anything concrete on the topic the way I have about Dynamic Map Elements, Research, Fog of War, or Hard Counters, to name a few. Time to fix that!

Table of contents:

In this article, I am going to cover the pros, cons and purpose of various RTS methods for determining who wins and how players are asked to go about securing a victory. In short, I'm trying to find a good vantage point on the topic both as a player and as a designer. This is less of a "I'm going to fix Victory Conditions in RTS" sort of thing, but rather "let's explore the space and see if we can learn something valuable. along the way"

I see game design, as with most forms of design and indeed most acts of creation, as a series of compromises and tradeoffs, and systems like victory conditions are tools in the designer's toolkit. Knowing what tool to use and how to use your tools is far better than saying that a screwdriver is better than a hammer.

My hope is, even if you're absolutely convinced that the only possible good way to handle RTS is annihilation style victory conditions (e.g. 'breaking all their toys') or whether you love Company of Heroes style capture points, or one of the other victory systems we're going to look at, that there will be something of value you can gain from or add to this discussion (feel free to comment at the bottom of this post or join us on Discord!).

Victory conditions - a starting point

In a strategy game, players typically begin with very few pieces (that is: units and/or buildings and/or resources) and increase the number of things they have and can do in the game over time: gaining some kind of store of value (money, Gold, Wood, Food, Vespene) and ways to expend that store of value (factories, barracks, archery ranges, research laboratories) in order to acquire various ways to change the game state (units, support powers, etc).

Please note that this doesn't constitute me attempting to list all of the categories of thing in a strategy game; I'm just trying to make a point about how games are won or lost.

So, if you're out of units, you mostly can't do much to your opponent. If you're out of factories, you can't spend resources to make more units. If you're out of money, you can't use your factories or make more factories, or units, or whatever.

What I'm saying is, if you're deficient in any one of the above areas, you might be able to get to the point where you can't really continue to participate in the game meaningfully, and the baseline cutoff for "losing" is usually something like: You have no ability to affect the game any more. No units, no money, no factories, et cetera.

Regardless, the "loss" condition seems to be the more important part here, and the "win" condition is more "I'm the last one left who hasn't reached the loss condition." We can consider this kind of the default or starting win condition, since it arises naturally out of the systems in an RTS. In games with multiple win conditions, this is typically called something like "annihilation" though I like to think of it more in terms of "breaking all the other players' toys." The last person who still has some unbroken stuff, they won.

What's the issue with the default win condition?

Anyone who's played Command and Conquer before knows the pain of missions which require you to hunt down every last enemy soldier on the map. Being forced to scour every corner in search of that one guy preventing you from closing out the level and getting on with things can be a really frustrating experience.

This begs the obvious idea that there might be some reasonable subset of "everything" that a player can lose where we can assume they've lost the game. For instance, losing all of your structures is a common loss condition: no matter how large your army size, "no buildings" means you're out of the running, and it's all over for you.

While it's an easy jump to make from "you're in the game until all of your stuff is gone" to "you win by destroying these certain things for all other players" this still leaves us in a little bit of a pickle regarding what it looks like to remove enough stuff from your opponents to constitute a win. And we'll go into that a bit in our next section, since games like Total Annihilation and Age of Empires 4 tackle the edges of that space.

But there's another aspect of this, one more core to the very idea of "breaking all of your opponent's toys" that I want to address.

"The way that you win in StarCraft is to instill your opponent with despair and they go 'I'm fucked' and then they leave the game" - Sean "Day[9]" Plott

As RTS commentator Day[9] has quite astutely pointed out, the effective outcome of the idea of breaking all of your opponent's stuff is that you put the game into what the other player reads as an unwinnable state and get them to forfeit.

Due to the way RTS economies and scaling work (we discussed this above) you have this chain of income > ways to spend money > pieces you use to win the game in virtually every RTS that is highly susceptible to disruption. I talk about this sort of thing in my articles on time-based disadvantages and equilibrium in RTS.

The main way you actually win most of the time is not by breaking all of the other players' toys, but by breaking enough of them so they feel like they can't catch up. Since the more harvesters you have, and factories you have, and more units you have, the easier it is to continue to accrue advantages over a player who has to try to catch up, and there is often no real meaningful way to attempt to do so after losses have accumulated above a certain threshold. It's one reason I'm very interested in the early game phase of RTS and how important that is to get right.

I have to admit I have something of a philosophical issue with this sort of gameplay. To not put too fine a point on it, it just feels really bad to have your ability to even play the game you're trying to... play stripped away from you over time. I've likened it before to a fighting game where you lose access to moves, and move more slowly, as your character takes damage. I think there are actually fighting games that do this, but that's a bit beside the point I'm trying to make. I hope this example illustrates the point at least enough to get it across.

All of that to say:

OK, I think we've just about driven that one into the dirt, I'd say. Let's go on and look at some other ways that games have handled winning and losing and see if we can get any pros and cons of these systems, as well as determining if they might be "better" than the default.

The Total Annihilation/Supreme Commander model

Above, we were talking about the idea of what subset of "all of the opponent's stuff" that can be destroyed before a loss feels unfair. We actually have several examples of this sort of thing happening in RTS and I think it's interesting to take a look at how these games play out and consider their pros and cons.

Carrying this conversation to the next step, I want to start with the Supreme Commander/Total Annihilation formula: each player has unit vaguely analogous to the King from Chess: a single piece where the player simply loses if that piece is lost to them. In most TAlike RTS, the commander or ACU or what have you is very strong in the start of the game, and unlike a hero unit actually becomes more and more vulnerable and weak over time; in TAlikes this is due to later game units scaling up to the point where the starting power of the ACU becomes weak in comparison. While I think it's a great formula overall, there are some definite downsides to this win condition (or more properly loss condition) that really bear mentioning.

While my gut reaction is that this sort of Gordian Knot solution to a game is a good thing... Desperate, come-from-behind moments of tension are great gameplay and keep the game from feeling too one-sided to the player who's behind. But that sense of unfairness when a player who's otherwise doing well just... loses out of nowhere, that is a real pickle in and of itself.

Dipping again into the realm of game feel, it feels really bad when a player perceives they're doing well in the game, only to get an abrupt loss out of nowhere because their "king" unit gets sniped. There are some ways to deal with this, such as the ACU getting an 'escape' function like in Supreme Commander 2, where there is an upgrade which allows the head of the ACU to eject and then rebuild into a full commander later on, or if the player is given more than 1 king unit. For instance, if each player had 3 critical units to protect rather than just the 1.

Earlier, I promised to talk about Dawn of War 3, and now is the time to do that. See, Dawn of War 3 did just what I'm suggesting here. They give you multiple things to protect: 2 shield generators out on the map, 2 Turrets closer to your base, and then your Core right up next to your main command center building.

Here's the thing though. First off, it was a bad fit for the Dawn of War formula. Secondly, it feels a lot like how winning works in the MOBA genre, which turned off a lot of core RTS players. Thirdly, it was just weird and cumbersome to have to figure out all the little rules about what order things can be killed in: once at least 1 shield generator is taken out, then any turret can be killed. And the Core is vulnerable once at least one turret is killed (is what I remember? May the shield generators protect their respective turret?)

Regardless: it's complicated, it made a bunch of RTS players feel like MOBAs were co-opting their genre, and it just wasn't what people wanted out of Dawn of War. Lastly and perhaps most importantly, many of its maps were set up such that map control (which in Relic style games is synonymous with resource dominance) also allowed you to protect your objective structures pretty well. So this thing that was supposed to decouple winning from a player's ability play the game ironically just reinforced the dominance of the winning player most of the time.

So while this sort of "kill multiple things that don't directly have to do with the player's ability to play the game" win condition might be (to me at least) a step in the right direction... there are better and worse ways to go about doing this. And Dawn of War 3 is definitely a cautionary tale in this regard.

The actual tabletop Warhammer 40K game has a win condition that would be a decent idea for RTS to try: critical locations spread over the map, and whomever controls the most of these when a pre-defined game timer runs out (the default is 6 turns in tabletop, I believe) is the winner.

Innovative TA-like Zero-K (which is free on Steam if you want to try it out) has a Commander unit whose death does not lose the game. Rather, its Commander generates a significant amount of resources (up to 30% of early game income) such that losing one early is effectively economically fatal, though the impact of its loss is reduced as player economy increases over the course of a match.

Which brings us to Age of Empires 4, where we can talk both about the concept of Landmarks, as well as introduce the idea of multiple win conditions. This, by the way, is turning into one of those pieces where I feel like each individual thing I talk about could itself be spun into an entire unique article. Ah, if I but had the time...

Age of Empires 4: the allure of multiple win conditions

Regarding Age of Empires 4, there are a couple things I want to look at. First and foremost, its "main" condition: Landmarks. In Age of Empires 4, players construct buildings called "Landmarks" to advance to their next Age rather than paying for a game-level upgrade as is done in Age of Empires 2. And while I generally think that's a cool thing in the first place, it also comes with another interesting facet:

The player loses when all of their Landmarks, as well as their starting Town Hall, are destroyed. This win condition (or loss condition) has the advantage of shortening already-lost games as well as provides remarkable focus for each player in terms of what target(s) to go after or protect.

It also has the remarkable effect of providing more targets and more buffer from a loss as the game progresses, something I find missing in the more traditional Supreme Commander model, where the longer the game goes on the easier it is to find and kill that solitary ACU with nukes, Experimental units, or what have you. So overall I find this a compelling middle ground between "kill the King" and "break all of the other team's stuff" in terms of generating a win or loss condition in a game.

This also seems like a good time to bring up the idea of having multiple win conditions in a game, and the challenges of doing it well.

Age of Empires 4 provides the player several explicit ways to attempt to win the game. First, players have the standard "break all the toys" win condition, with the caveat that they can shortcut it through killing Landmarks. Then of course, there's the Age of Empires classic Wonder, which the player must build and then defend for 6 minutes before getting a win.

And lastly, they have Sacred Sites spread around the map. Sacred Sites are vaguely like Victory Points from Company of Heroes. In Age of Empires 4, priest units can (sometimes with an upgrade required) capture Sacred Sites. Each Sacred Site generates Gold, but if a player holds all 3 of them then they'll get a victory timer similar to having a Wonder.

Given the sometimes protracted later game stages of Age of Empires matches, I really love the forcing function that Sacred Sites can have to cause the game to get a move on, already. Rather than being a solid way to close the game out, these Sites can be an impetus to force players to contest each other's holdings and nudge things towards conflict and conclusion in a more concrete way than being worried your opponent might outstrip your economy.

One thing that Age of Empires gets right that we see fail in other games is making (at least some of) the win conditions be viable enough to bend the gameplay due to the possibility of them happening. As I said above, one player trying to go for Sacred Sites can make another feel the need to interrupt the attempted win condition, which can kick the game back on track and keep engagement going and the unit mill churning. A player going for a Wonder will likewise push another to act, and even if a vanishingly small percentage of games are decided by these conditions, their very presence can encourage positive gameplay outcomes.

Sadly, it's not quite that simple.

Dune Spice Wars and the challenge of multiple win conditions

It would be nice if we could just say "well, throw multiple ways to win at the player(s) and let them sort everything out!" It's even better if the different ways to win are somewhat unrelated, so that a player who's losing by sheer military might could go for another way to win, some kind of research or political or subterfuge victory instead.

I should note that a lot of RTS actually do something like this: most RTS that do something other than "annihilation" for winning the game also allow this condition to occur. For instance, both Company of Heroes and Iron Harvest use both VP tickers (we'll discuss these next!) as well as allowing the destruction of the opponent's HQ to result in a win. But, Dune Spice Wars is a great example of the challenges inherent in this sort of attempt to throw options at players.

(Here is a Dune Spice Wars review from IGN: If I do one here on the blog, I'll replace this).

In Spice Wars, there are technically several ways to win: First, take out enemy HQ buildings. Then, use Solaris (money, one of the main resources) to buy ownership in CHOAM, basically converting money into a win. Then, players who specialize in spying can attempt to assassinate the enemy by basically attempting to purchase and run a series of support powers against the enemy commander. Players can be elected to the Dune Governorship in a periodic lottery system where players compete over buffs and debuffs similar to the Food Network show Cutthroat Kitchen

the Choam governorship sees players competing to see who can dump the most money into this meta-layer interface to hit the shares threshold first

I might have actually missed one or two win conditions, but the main way to win is by gaining a victory resource called Hegemony: you get it for controlling territories on the map, and there are a number of other things you can do to gain this resource, but here's the thing. Hegemony is far and above the easiest way to win, to the point where even if players are pursuing one or more of the other ones, it might not even force the other players to attempt to react just due to how the game is put togethers.

Part of the problem is that several of the win conditions are so separate from the rest of the game that there's not much players can do to even attempt to stop an alternate win beyond maybe pushing harder for their win condition: someone going for CHOAM and another player going for Hegemony might be able to largely ignore each other and just see who can race to get their thing across the finish line first.

Anyways, I might be getting a bit in the weeds. The main point I'm trying to drive at here is that the game doesn't do its multiple win conditions justice because they're not particularly well integrated with each other, and take a very long time or are hard to interact with beyond the player who's ahead being challenged in some other sphere. Ironically, fewer and quicker win conditions would probably result in more dynamic, interesting gameplay where players feel more agency.

Now, one interesting thing about Dune Spice Wars, as well as about Sacred Sites and the win conditions in Relic's games are that you can actually talk about them in terms of things the player wants to do, rather than something the player wants to avoid in fear of incurring a loss. They actually have some kind of goal outside of breaking all of someone else's stuff.

The opponent as obstacle

One thing I like to say about the "default" win condition is that it paints your opponent as the objective of the game. Your whole goal, often your only goal as a player, is to wreck their stuff and take them out of the game. In formulas like those seen in Company of Heroes, Iron Harvest, or Deserts of Kharak, the players main objective is to secure things on the map, and their opponent exists as effectively a thing in the way of completing that primary objective.

Relic hit on a kind of beauty with the win condition they iterated on with Company of Heroes and Dawn of War 2.

I think one of the most important things to note about the Relic model of VP tickets is that they design their entire gameplay formula around the idea of this sort of win condition. They fill the map with Points of Interest and objectives, and give the player relatively few units with which to contest them. They scale the economy automatically with population, to attempt to keep players on a roughly even footing given roughly equal skill levels, they make the base very hard to break in the early to midgame (interestingly, Age of Empires does this as well to an extent!)

But, more so than the other formulas we've discussed, the Company of Heroes formula actually has a fairly complete game without the other player directly involved, to the point where often pushing the player off of areas of the map results in the dominant player just continuing to contest the map since that is the real main, sure way to win. Base assaults do happen, particularly in matches where one player is consistently ahead or in desperation, but it's far more common to have matches end based on the VP ticker. Now, we still might have that Day[9] style "well I can't come from behind and overcome this VP advantage so screw it I quit!" moments, but the way the economy (doesn't really) scale can allow matches to seesaw back and forth quite a lot.

I like to think about game design in terms of how often the best and worst outcomes occur. Sure, there are really one-sided blowouts in COH2, but the game allows 2 relatively evenly matched players to slug it out for a long time, and gives a player on the back foot actual opportunities to claw their way back into the running.

Now, it's not all rainbows and butterflies in Opponent as Obstacle land, of course. Or else, we'd see these dominating the RTS win condition landscape, which is obviously not happening.

This article is going on pretty damn long so I'm going to try to be brief about it.

First off, these sort of win conditions have to be taught a lot more than the default win condition. In Deserts of Kharak, players can win by gathering "artifacts" from the map and returning them do drop zones in various areas of the map. Why artifacts? Why drop zones? What are drop zones and why are they out in arbitrary places on the map? The game describes all of this, of course, but the point is that it requires a whole extra layer of explanation and teaching before the player understands how or why any of this is important. It's less intuitive, less easy to grok, and that can be a stumbling block for a game both psycholocially and mechanically.

Now, the other issues are:

While attractive, and I think there's a lot of promise in this sort of thing, there are some big challenges to implementing them gracefully into an RTS. And the best way seems to be to carefully plan the game systems to support the win condition, as well as ensuring that the various victory conditions are balanced well enough against annihilation/default.

In conclusion...

I write to get my thoughts the heck out of my head and to attempt to elicit conversation in the RTS space on what I see as important topics. While I do have my preferences (and I try to state those up front while providing context for why I think those preferences bear out in practice) this article is less an attempt at persuasion than it is an attempt to explore the topic of win conditions along with you.

I currently consider "breaking all your opponent's toys" or the default win condition as effectively the worst solution to RTS win conditions... except all the other win conditions which have been tried. It's the most popular because it's simple, straight forward, understandable, and doesn't require any rickety overhead solution to be bolted onto the rest of the game. It's just the natural outcome of the gameplay rules. As bad as it can feel, at least it makes intuitive sense and doesn't usually surprise you with an out-of-nowhere loss as can happen when an ACU gets sniped.

I do, however, remain suspicious of any win condition which is more easily identified as a "loss condition" applied to all other players or teams in a game, as well.

Personally, I am interested in seeing where future RTS can take the ideas like those of Age of Empires 4 and honestly the rest of that series: critical units or structures that the player accumulates over time bolstering up as a defense against a loss condition, coupled with a late-game economic win condition (like Wonders) with an on-map backup win condition which becomes viable in the later mid-game stage (like Sacred Sites) that can serve as a forcing function for stalemates. I think it's a great triumvirate where there are a lot of permutations that can be incorporated into existing game formulas.

That being said, I don't really feel done with this. In fact, I have several thousands of words on the topic that I had to cut from this article in an effort to constrain its size. So, don't be surprised if I follow this up at some point with a Part 2.

See you then. It's good to be back.

Hell, it's about time.