This post is supported by Patreon.

The MOBA genre of games grew into popularity with the WarCraft 3 mod Defense of the Ancients or Dota Allstars, and took off like wildfire with the advent of Riot Games’ League of Legends. Now, roughly 7 years after the launch of League, MOBAs are a saturated genre, with their own sub-genres, spinoffs, and so on. More or less coincident with this, or perhaps slightly before, was the beginning of the real-time strategy genre’s slow slide into its present state of relative obscurity.

While I would like to shed gallons of digital ink on why this might be, I won’t for two reasons. First, for my current purposes it is enough to note that MOBAs are a highly popular competitive phenomenon, and real-time strategy games (with very few exceptions) are not. Secondly, I have not done extensive research into the topic and feel that I have little light to shed on it.

For a long time, I was preoccupied with the question of how MOBAs seemingly supplanted RTS. I am no longer interested in, specifically, that question or line of thinking. What’s become quite interesting to me is the modern RTS that seek to merge the multi-unit management, base expansion, and/or harvesting of real-time strategy games with some combination of heroes, creep camps, itemization, map design, or in-match leveling found in MOBA games.

This is, I am convinced, the incorrect approach.

I believe that RTS have a lot to learn from other genres, MOBAs amongst them. I believe in the beauty of real-time strategy game systems, and that games such as Offworld Trading Company are clear indications that the genre can experiment, test its bounds, and grow in exciting ways.

I also believe that surface-level MOBA mechanics are a red herring. There are things to be learned from MOBAs, but they’re deeper down than their implementation. I realize this may take some explaining.

Let’s get started.

Who’s Even Doing This?

This article was prompted in part by the short life and abrupt demise of a game called Guardians of Atlas. When I first saw Guardians of Atlas (after months of anticipation over a game at least theoretically designed in part by Day[9]) my first and enduring thought was “League of StarCraft.” The game’s aesthetic evoked a combination of MOBA (and here, reference League of Legends and DOTA2) and… Protoss.

Mechanically, the game was a true and somewhat bewildering mishmash of MOBA and RTS systems. The players chose a hero and a loadout of units from one of 4 factions. They loaded into maps in teams, and killed creep camps of various types to accrue resources. Resources were spent on upgrades – a seemingly endless series of upgrades that took the players’ focus off of the action and felt entirely too fiddly for their own good.

While I did not get a chance to play much of the game, my impression was that units served little purpose unless kept in reasonable proximity to a given player’s hero. There were few points of contention on the map, and little for units to do aside from kill enemy units or creep camps. This, in spite of the choice offered by unit loadouts, really served to make the game feel sadly shallow.

Besides GoA, there are several existing or cancelled games out there that take a greater or lesser nod from both RTS and MOBAs. A partial list includes:

• Supernova. This game uses mostly MOBA mechanics, but allows players to upgrade and choose the composition of lane units. I think this is an interesting twist on the MOBA model, though I have not had the opportunity to actually play
• Battleline: Steel Warfare. This game feels like an ARTS (action realtime strategy) version of World of Tanks. Players take objectives and push a “battle line” across the map while controlling 1-4 MOBA-like tank “heroes”
• Dawngate. Dawngate (now cancelled) was a MOBA that tried to streamline its item shop, and included RTS-like resource harvesters that could be harassed or captured by either team.
• DropZone. This is another hybrid MOBA/RTS, though mechanically it sticks much closer to the MOBA. Players can control 3 heroes at a time, and can customize them with gear and pilots which can have a dramatic impact on their role in game.
• Servo. Servo felt very much like an RTS, with base building and economic expansion, but the titular Servos were very MOBA-like hero units. The game felt like an arcadey sort of Warcraft 3. Many of the game’s maps evoked the feel of MOBAs, with respawning creep camps and powerful turrets as obstacles.
• AirMech. While many might consider AirMech to be very much a MOBA-type game, I disagree strongly. The AirMech itself is less a ‘hero’ than a choke point for how much a player can do: the AirMech is busy with non-combat economic and logistics actions, even if it is a powerful hero unit with a specialized combat role. I consider this the most successful example of hybridization
• There are more out there. Demigod was another game I consider successful at hybridizing MOBA and RTS mechanics.

Trying to Copy the Symptoms

Now that we have some concrete examples of the sorts of games out there trying to reconcile or merge real-time strategy games and multiplayer online battle arenas, let’s talk about their areas of focus.

Heroes

The hero is the cornerstone of the MOBA. In essence though: heroes are divided, loosely, into roles and support their team in its goal in a unique way. Fundamentally, real-time strategy armies don’t function the same as heroes.

Even with units having unique roles within an army, it is inescapable that units can occupy unique points in space and time within a match. Armies can be subdivided to contest, harass, decoy, zone/control or capture various points on the map. Individual units can be killed, multiple units can act independently from one another.

In MOBAs, virtually all heroes are a single thing that impacts a single discrete area of a map at a single point in time. There are some few heroes that can teleport across a map, or cast map-wide abilities, or in some way contest, observe, or otherwise impact multiple areas of the map.

To an extent it is possible to allow RTS players to specialize into a role within a match, but the roles imposed upon MOBA heroes don’t fit once unit counts get above about 10 or so. Especially once production structures and resource-based economy comes into play, analogies to MOBA conventions just start to break down.

Bottom line: geographic dispersal of HP, damage, abilities, and innate characteristics is fundamentally incompatible with MOBA-style hero roles. RTS dynamics of expanding and producing structures is fundamentally incompatible with MOBA-style interaction dynamics.

Porting heroes into an RTS can cause more challenges than that, of course. It’s all too easy for heroes to be so powerful that units don’t matter, or heroes so weak that they aren’t particularly important. To me, as an RTS player, the latter is perfectly acceptable. I don’t consider this argument compelling though – it’s a balance issue, not a design issue. I think that fundamentally the problem isn’t one of number crunching, but of the fundamental incompatibility of what MOBA heroes do and how they relate to one another versus how RTS armies interact.

Now, that’s not to say that heroes don’t work in an RTS. That’s an entirely different argument and one that notable minds in RTS have discussed for years. I’m not trying to argue this one way or another. What I am saying, emphatically, is that heroes don’t solve any problems in RTS, and that MOBA-style hero roles are incompatible with the realities of strategy games.

Items + Customization

Much ado has been made of itemization in MOBAs. Old-school LoL- and DOTA- style itemization is seen as a significant hurdle to learning. They offer false choices, are difficult to read visually, and end up requiring painful systems that do things like suggest builds for new players to try to guide them in the right direction.

Itemization, in essence, is a spin on RTS progression systems: research, mostly. Items and leveling are the primary forms of progression in MOBAs. Here’s the problem, and I consider it to be endemic to MOBAs in general: Virtually all progress made in MOBAs is irreversible. Experience learned (in most games) is experience kept. Money spent on in-match items can virtually never be lost. Money earned, in most cases, can never be lost (DOTA is a major exception in many of these cases, and god do I love it for that).

A major difference here is that in RTS, virtually every advantage gained can be lost. Structures can be destroyed (and rebuilt – something again that most MOBAs eschew) and their conferred advantages erased. Units can be killed, et cetera et cetera. Not so in MOBAs. MOBAs are all about team momentum, all about the rate of gains. There are, in many cases, much more limited opportunities for a reversal of fortune.

That being said, I do think in general that RTS can have too many points of failure. Too many advantages conferred that can be lost or squandered. I think there’s a middle ground here. But importing flawed itemization/scaling models into strategy frameworks aren’t solving any problems.

Map Design

In terms of map designs, let’s face it. MOBAs, aside from Heroes of the Storm, tend to not be very inspired. Most of them, the vast majority, are spinoffs and slight modifications of the original DOTA map. But, as before, the central contention here is not about the specifics, the fine details of jungle pathing.

MOBA maps are divided and constrained. At best, they feel like entering an RTS halfway, in a stalemate and fighting slowly and by degrees to victory without the aid of superweapons. RTS maps are never, or virtually never, like this. They’re open fields of possibility, waiting to be contested and won. There may be resource nodes, or objectives, or neutral structures that block movement and line of sight, garrisonable buildings, or whatever. But almost never will they specifically and intentionally limit access to the majority of the map area.

RTS can accommodate maps like this, but here’s the point. Though their control schemes are often similar, the actions players are asked to take and the fundamental logic driving those actions are profoundly different. This is the central contention of my issue with hybrid games: RTS and MOBAs are, at their core, in pursuit of vastly different ends. This is why the surface mechanics of either are so difficult to shoehorn into the other, or to meld peacefully into one.

I am absolutely positive that it is possible to make a good game with elements of both MOBA and RTS. I don’t really want to argue down that road, though it would be so easy to veer off in that direction. All I want to say is that, from the perspective of someone passionate about RTS and dedicated to expanding the genre in terms of both innovation and playerbase, MOBA mechanical systems cause at least as many problems as they solve. Real answers, substantive lessons, do abound in the MOBA. But the surface mechanisms aren’t where they lie.

The Real Lessons

Objective-Driven, Role-Based Gameplay

In MOBAs, while the other team is the primary obstacle in the way to victory, they’re not, really, the objective of the game. They can certainly feel that way, and I’ll freely admit to reaching just a smidge here, but the real objective is the other team’s Ancient/Base/what have you. Teams that focus on map objectives tend to do better in MOBAs, all else being equal. Towers, inhibitors, barracks, etc – these are all objectives that players are really aiming for. Now, since they tend to all be in a single base area, and since the other team typically has to be dead for your side to make significant headway, this might seem a bit semantic. But bear with me.

A major stumbling block for RTS players psychologically, tends to be that the main and entire focus is on destroying your enemy. And an RTS loss can be a death of a thousand cuts: lost workers must be replaced, a cost of resources and time and harvest rate/income. Destroyed structures make for a less effective army build/rebuild cycle. Inefficiently traded battles mean further losses down the line to infrastructure, compounded with the enemy’s lowered repair and replacement rates. This snowballs into a mental slugfest where losing just makes you feel stupid. The other player was smarter than you, at least this time.

Adding objectives other than “kill the other player” opens up a lot of avenues in a game. Suddenly, the other person’s stuff isn’t quite as important. It’s an obstacle that must be overcome in pursuit of your higher goal. Objectives make room for different skillsets (like heroes in MOBAs that are good at taking down towers, vs heroes that are good at farming the jungle, vs heroes that become wrecking balls in the later game) to be successful. Then, if a player prefers turtling up and holding territory, they can contribute in a meaningful, obvious way to a victory.

Giving players different things they’re allowed to be good at, allowing different play styles to be successful, is a recipe to bring players into a game and keep them. RTS tend to force players to be good at a wide range of skills simultaneously. One place where there’s definite room for improvement is keeping players engaged and challenged while still allowing them to fall into a comfortable niche.

Ways to consider doing this that don’t fall into the camp of “hero” might include things like Company of Heroes 2’s Commander model, where players can pick specialized units and abilities to bring into combat. Alternately, some type of unit loadout or deck, as in the Wargame titles or the upcoming ARTS Tooth and Tail.

The bottom lines here are that MOBAs demonstrate the value of having both player and map be high value targets: this produces a more cadenced pacing naturally, with focus seesawing between map objectives and in fighting the enemy. It allows for different types of success to be measured, and opens up the possibility for different types of unit with specializations not seen in mainline RTS. It also opens up opportunities for players to specialize within a team, something at which MOBAs excel and RTS do not.

To conclude, though. While I think specialization into a particular role is an admirable goal, “specializing into a role” in a game where you have 10, 20 or more moving pieces is really a matter of degrees more than anything. An individual hero or unit’s only moving the dial a little bit towards the general role a player might have. In RTS-style games, a player will always be more of a generalist than in games where they only have 1 character to manage.

A perfect example of this (that might not be widely accessible, sorry) are the differences between End of Nations and Command and Conquer 4. C&C4 tried to cram RTS units and abilities into 3 roles: offense, defense, and support. Support players had all of the air units and support powers, defense players had infantry, turrets and artillery, and offense players had mainline combat units. And it felt wrong. As a player and game designer, I didn’t like feeling crippled, especially as an offense player. The other roles felt more interesting and fleshed out, with units that could interact differently with terrain, defensive options, etc.

End of Nations took a much more refined approach to specialization. Each class in EON had 2 or 3 things it did well, and players could build out a unique loadout pretty much however they wanted. All classes got a mix of support powers, turrets, and various class of unit. There were options all across the board, even if one class’ turrets might be innately stronger or another class’ support powers might feel broadly a bit more potent, they all had the same basic categories of tool.

In games where you control, build, harvest and produce with multiple moving pieces, you cannot be as pigeon-holed within a single role. And any game with that many moving pieces needs to allow the player to take advantage of that fact – this was one of the hard lessons of Guardians of Atlas. Armies as an extension of a hero are much less interesting than heroes as an extension of an army.

… Hopefully, I’ve made the case well enough that “role specialization” and “role specialization as seen in MOBA” games are actually pretty different things, and worthy of the distinction.

Teams!

Another psychological stumbling block for prospective RTS players is ladder anxiety. As mentioned above, it can be anywhere from humbling to downright humiliating to lose a match in StarCraft or Company of Heroes. This is greatly ameliorated by the presence of other people, with whom the player can share the glories of victory and the ignominies of defeat.

It’s simply not as bad to lose if you can blame it on your friends/fellow players. And it’s still basically just as good to win if you do it with a group. Now, most RTS support team play, but virtually none of them design or balance around it specifically. Team dynamics really impact unit, map, economical, balance and pacing design choices in various and sundry ways that can’t fully be accounted for in games designed intentionally to support 1v1 as its default format.

The inherent problem here is, of course, that virtually no RTS is designed from the ground up as a team game. Having some degree of role specialization is a way to encourage team play, but very few games have tackled the idea of team play where each member of each team controls 20+ moving parts at once. There are a number of essential challenges to solve, but making each player more autonomous while still being able to contribute to the greater whole in a unique way (per my example above) feels like a good first step. And, again, not to harp too much, but map objectives separate from player infrastructure/economy/base can help here too. More things for players to contest means more things for units to do, and more types of way for players to contribute and succeed.

Personalization

This one is pretty big. People like having ‘their’ own content. Their army, their hero, looking the way they want. They want to collect rare and pretty toys to show off and use when they see fit. RTS have nothing, really, for players who also happen to be avid collectors. Since the halcyon days of the RTS, unit design, faction design, etc have trended towards the forgettable. Even narrative design tends to fall by the wayside a bit in RTS, which lends itself to forgettable faction and unit designs. Personalization is about letting the player define their narrative, their role in the game’s fantasy. Giving the player power over their presentation and interaction with the game. RTS needs to step up its game here (and copying most MOBA’s lack of cohesive storyline isn’t a way to do that).

MOBAs?

In the end, RTS and MOBAs are different animals trying to do different things. There’s value to be gained from them, but it shouldn’t surprise you that the game I outlined with my lessons doesn’t… really… sound like a MOBA. If anything, it kind of sounds like World in Conflict, or maybe reminds you of End of Nations? Or Wargame?

MOBAs speak to some general truths that many gamers prefer. Strength in numbers, the pride of ownership, the feeling that we have self-selected our tools and made them special to us. These are things that RTS do not communicate as well, though they can. The MOBA is currently a largely fixed system, with many immutable elements and assumptions tangled within. RTS are better served at discerning the foundational ideas of MOBAs and adapting them than they are by aping surface traits like heroes or map towers.