From Impossible Creatures to Homeworld to Dawn of War, Relic Entertainment is known for pushing the boundaries and redefining the borders of the real-time strategy genre. If they didn’t actually invent “full 6 degrees of freedom” in RTS, the phenomenal Homeworld certainly took the concept and showed its potential to a wide audience. And if they didn’t invent squad-based unit groups, terrain modifiers to unit armor, and hero units with cinematic interactions, Dawn of War certainly brought those things together into a mix that was so compelling it’s still being actively played by a dedicated community.

As Relic has released games, they’ve had their fair share of flak for some of what they’ve made. Dawn of War II received criticism for its complete lack of base-building mechanics, smaller squad size relative to the first installment of the game, the number of maps the game launched with, and the lack of factions compared with the (complete, with all expansions) first game. Company of Heroes 2 likewise received broad criticisms related to the differences between it and the first game in the series, especially centered around the Commander system and changes to its unit design philosophy (the lack of bike-style units comes to mind) – the weather system in the game was also quite controversial, with many players preferring maps that did not make use of the game’s ‘blizzard’ or snow storm mechanic. Likewise, they are widely considered to be heavy-handed on balance changes, often missing the mark or overtweaking when attempting to fix or tweak issues.

Relic’s latest title, Dawn of War III, is the center of a controversy seemingly greater than either of the two I mentioned above. And, from my perspective, a fair bit of that controversy is at least a bit facetious. Here’s what I want to do: first, I’m going to give context into how I weight mechanics and design in RTS. Then, I’m going to provide some thoughts about the various controversies surrounding Dawn of War III. Then, I’m going to provide my own thoughts about the game’s strengths and issues, in the context of what I look for in RTS. I’m hoping through this process to provide a fair and open look into the quality and relative merits of this contentious title.

Just out the gate: I don’t expect this article to persuade anyone or change their minds. I don’t think I’m going to convince anyone of the merits of a game they already dislike. But perhaps it might serve to explain what I see in it, where I think the design might be improved, and generally what I see as the good, and bad, design decisions and implementation that went into the game.

Let’s get to it.

The Context

In the space of reviewing and commentating on games, the main points of contention are not specifically regarding the presence or absence of particular game features or systems. More important, most important, is how the commentators weight and rate the importance of a game’s systems. How they choose comparisons to illuminate their points to readers. What the commentator or player is looking for from a game, and how do the game’s features line up or differ from those desires? And, as a reader, how do the commentator’s/reviewer’s desires line up with the player’s own?

When I approach an RTS, I primarily look for a couple of major things that determine my interest in it. First and foremost, I’m looking for faction variety: the closer each faction is mirrored to the others, the more gameplay comes down to brute resource efficiency and the perfect execution of tightly controlled build orders. To me, that’s less interesting than leveraging what’s unique about your own army to monopolize on your opponent’s unique vulnerabilities, while trying to mitigate or overcome your own. Armies of widely different philosophy tend to lead to more diverse gameplay, as each match has the potential to go in a larger number of different directions.

I look for the opportunities for counter-play that a game permits: it’s just not fun to have things -happen- to you. My philosophy is that games should permit a high variety and large number of gradations in success/failure states for any given action, with the most extreme success states being high-risk and high-reward. As an ancillary to this, I look for abilities or situations within a game that are flexible and can build organically one on another to deliver a wide variety of outcomes: a decent example of this is in Command and Conquer 3, where the Brotherhood of Nod can use a flying unit to boost the range of the Nod laser artillery. Another example is the Spellbreaker from WarCraft 3, who has the ability to move debuffs from allies to enemies, and buffs from enemies to allies. That ability in particular has tremendous implications and broad (if passive, in this case) application.

Increasingly, I look for games that allow some sort of playstyle customization options: choosing your hero character(s) in WarCraft 3 or Dawn of War II; picking Commanders and Bulletins in Company of Heroes 2, picking your Commander character in Halo Wars 2, choosing cards for a deck in Duelyst or a Division (and associated loadout) in Steel Division – there’s something undeniably attractive about choosing in a very personal way how you approach the challenges of a game: it’s one of the selling points of PGs and even, increasingly shooters. While it doesn’t fit in every RTS (StarCraft 2 has a good approach for this by limiting this type of choice to its non-competitive co-op game mode) it’s something that has appeal to me if done well.

Lastly, I look for what I’ve come to call Equilibrium or homeostais in real-time strategy games. That is, systems which attempt to organically slow or offset the slippery slope to keep games competitive for longer and allow some sort of mechanism to prevent games from being effectively decided by a single engagement across a more broad swath of cases. This comes from my perception that losing via an objective (my opponent held points A, B, and C longer than I did, or my opponent gathered 1000 greebles faster than me) feels less harsh than losing because you made a handful of mistakes early in a match and therefore were destined to lose. This, in particular, is incredibly hard to get right but can feel so perfect when implemented elegantly.

I am personally happy to wait while features such as observer mode are implemented in RTS, provided I’m sufficiently charmed by a game’s mechanics, and graphics are a secondary matter to me; My PC is middling-to-low end, and I often have to crank down the settings on graphically or computationally complex games so they’ll run well on my machine. Similarly, story and single player tend to play second fiddle for me, as my primary interest is in competitive multiplayer.

Talking Through Dawn of War III’s Design in Light of its Common Criticisms

One contention I have seen is that Dawn of War III doesn’t have enough factions. The more outrageous complainants will compare the game to the final products of both Dawn of War I and II, where each game with all its expansions had six or more factions, each with unique faction and resource mechanics, hero characters, et cetera. Leaving this obviously bad-faith argument aside – it’s completely unreasonable to compare a newly launched game to a game with 2 or more expansion and years of development behind it – that would be like complaining a new MOBA only had 20 or 30 heroes compared to the 100+ that DOTA2 and League of Legends currently boast – It’s still a little hard for me to swallow this as compelling.

Though Dawn of War III launched with 3 factions instead of Relic’s typical 4 for Dawn of War games, the amount of accessory content launched alongside these factions is, in my view, pretty darn solid. Each faction launched with 9 different Elite units (analogous in some ways to heroes in the previous Dawn of War games, or to to WarCraft 3’s hero units) which have a meta-level tech tree that unlocks cosmetic upgrades and expanded access to certain army doctrines. They have a personal doctrine which serves to alter the play profile of certain unit types, and can unlock an additional one which the player can choose outside of combat. All that to say, Elite Units are a pretty beefy feature set, both in terms of their potential impact on play options, and in terms of faction content. I feel that this has gone overlooked in common criticisms of Dawn of War III.

Obviously, more content is better when a game launches than less. But I feel that dinging Relic for launching with 3 factions instead of 4 is kind of petty, especially given the large number of options represented by all of those Elite units (and Doctrines). My hope – somewhat farfetched at this point, given that the game has proven to be vocally disliked and its active player base is pretty small – is that they’d add more factions over time. Heck, I’d buy Chaos, Necrons, or Imperial Guard factions if they were released in line with the model Relic adopted for Company of Heroes 2.

The issue of how many factions Dawn of War III launched with is one where context comes into play. If your context requires Dawn of War III to mirror its predecessors in order to be considered a good game, this is likely to bother you. For me, at launch, 3 factions of widely different gameplay approach, bundled alongside plenty of experientially different customization options, felt like a solid approach in general. Clearly it was intended to be the foundation of more factions to come.

This, of course, touches on one of the sore points of the Dawn of War III controversy – whether the game serves as a worthy successor to the storied franchise, and whether it’s a deserving Warhammer 40K game. That will be something I continue to touch on throughout the article, and hope to address satisfactorily by the end. At the core though, for me, good RTS design will tend to win out over adherence to a franchise, even one I love.

Elite Units, and Readability in Design

Elite Units themselves, of course, are points of contention. They exist and are accessed via a separate resource from the game’s standard system, respawn like MOBA heroes (sort of) and the ones that become available in the late game have an outsized impact on how the match is played. They aren’t quite like heroes in Dawn of War I or II, lacking sync kills and itemization in particular.

One of the hallmarks of Dawn of War III’s design that differs from previous Relic titles is readability of mechanical design. First and foremost, Relic was shooting for a game where the results of any action the player takes are quite clear and understandable. In previous titles (Dawn of War II and Company of Heroes 2 stand out as prime examples of this) it was sometimes difficult to precisely predict the effects of an action. Some examples of this are: projectiles missing/impacting terrain and having widely inconsistent results. Think of the Zoanthrope’s artillery shot, or an ISU-152’s shell scattering. Or, especially in Company of Heroes 2, the cover system is occasionally a bit unclear. what happens to a squad inside a building when you throw a Molotov Cocktail into it? does the whole thing die? Which models get killed? Or if an attacker is approaching at an oblique angle to a fence, how can each player tell whether the defending squad is going to take reduced damage from the cover system?

Sure, these are edge cases. And you can fall on either side as to which approach you prefer – I’m not trying to imply that Dawn of War III’s cover system and stealth brush are superior to the cover system in Company of Heroes 2. But it is certainly simpler to understand the results of interactions that take place in Dawn of War III, and that, I think, is what Relic was shooting for with this title. From Elite sync kills to cover changes, to Elite units being defined in terms of MOBA-style hero role description, clarity and readability seem to be the hallmarks of every aspect of the design of this game.

So, getting back to those Elite units… I think they’re a solid system. Elites are far more customizable than Company of Heroes 2’s Commanders, for one thing. They provide interesting temporal choices and have some of the most dynamic interactions I can remember in RTS. Seriously, here. Almost every Elite unit’s abilities have multiple applications: from Wazmakka’s traktor beam that can protect and recall allies or slow and re-position enemies, to Macha’s ability variations regarding whether she’s thrown her spear, to Warp Spiders being able to allow armies to bypass terrain (and do bonus damage) to Angelos’ anti-projectile shield, most Elite units have abilities that change the calculus of combat depending on how and when they’re used, and synergize very well together, and with the armies they support and augment.

Speaking of augmenting armies, as an RTS player I really think the passive effects each Elite has on the armies are a really nice touch. It’s not enough that choosing an Elite loadout provides players with a set of interesting tools to use in combat, each Elite can also improve or alter the use case of the player’s regular units as well. This can allow mirror matches to play to unique strengths and weaknesses, even as both players have access to the same mechanics and unit types. You might not agree with me, but I’m a really big fan.

I’m somewhat less satisfied with the overall Elite resource itself, and the common linearity of how Elites come onto the battlefield. Elite points coming from contestable points on the map feels too binary to me, and reduces the effectiveness of one of the best things about how Elites are implemented in Dawn of War III – they exist apart from the resource system in the game, and therefore can serve as force equalizers if one player has fallen slightly behind. Tying Elite resource generation directly to territory control curbs this, as the player with more territory will have overall more resources as well as a higher chance to bring out early late-game Super Elites. It’s a compound problem. A better or more interesting system might have been to allow players to sacrifice an existing resource or energy node by building an Elite Point generator on top of it, or a high-cost structure which passively generates Elite points over time. That would allow players on the back foot the gamble of having reduced income in order to bring Elites out more quickly, or to risk a superweapon strike. Or, the choice of converting selected points on the map into either resource nodes or Elite nodes. I feel like the system, overall, is not fully successful in achieving its task and drags the whole Elite system down with it.

Super Elites are a category all their own. While I appreciate that they can serve, in the late game, as a ‘fun’ choice that is flashy, punchy, and potent, I feel that the power of these units can easily overshadow and overbalance the entire rest of the mechanics of the game. While Experimentals in Supreme Commander (especially the weaker, cheaper ones) felt like a natural evolution of the game’s systems, Super Elites tend to feel like a wrench in the workings of the rest of the game. They’re almost impossible to overpower without a commensurate unit on the opposing side, which I see as the primary issue with them. They force the uninteresting choice of meeting their standalone force (standalone since they can be produced without resources) in kind. Choosing a Super Elite should be a more risky value proposition, in my opinion.

You can defeat Super Elites without other Super Elites, but in practice the game becomes much more balanced if each player/team has equal access to these monstrosities. Regular units and elites are too fragile in comparison, it can be very difficult to cost-effectively combat Super Elites without one of your own. My assessment is that this is likely a balance issue however, and could be remedied with stat changes.

Phases, Map Design, and MOBA Mechanics

Here we get into the most substantial criticisms of the game. The scale of the game, the phases that transition matches from intimate combat to huge blobs of units, the often highly lane-oriented map design, and most importantly, the game’s win condition have been widely criticized by players and commentators of all stripes. Much like a Genestealer eating an Imperial Guardsman, let’s take this one bite at a time.

Dawn of War III is a testbed for a relatively uncommon feature in RTS games: it has a match-level mechanic called Phases that has a number of gameplay effects that change how the game operates across the course of a match. As a match wears on, there will be times when players’ income increases: these milestones are the game’s escalating combat phases. Later phases make it easier to afford a larger army, at the cost of units providing less of their cost as a refund when they are killed.

One side effect of this is that players tend to have a relatively small stable of unit types available in the early game, at it is only really cost-effective to build higher tier units once Energy income starts picking up into the mid game. This is kind of in keeping with Relic’s standard mechanical systems though: Company of Heroes 2 has a similar setup where higher tier units are added into the mix more gradually, and Dawn of War II has fewer gradations of unit additions that happen over a longer period of time.

One good thing to come from this (at least, in my eyes) is that the early game is incredibly brutal. A handful of squads balancing on a knife’s edge, leading to fewer, more tentative combat situations until the Elites start coming out. Early combat feels somewhat reminiscent of Dawn of War II to me, though of course since cover is a completely different beast the specifics of how combat goes down is different. Then, once income increases, the game becomes more lively, with armies crashing against each other multiple times as players vie for advantage. The mid-game is easily my favorite part of a Dawn of War III match, since resources are still scarce enough to matter, and the Elite/Army balance works out in the most interesting ways. Honestly, few games feel like they have as big of a focus shift as they progress as Dawn of War III, which I see as a positive (though as I mentioned above with Super Elites, the endgame can be much more fraught due to their disruptive influence).

Phases, however, don’t really meet the goal they’re trying to achieve. Phases are intended to represent kind of a ‘soft reset’ of the game, allowing the player or team who’s behind a chance to catch up. It seems more often to simply prolong the slippery slope as the player/team who has more territories will still be able to build the larger and more survivable army in the end. Honestly, Listening Posts are partially to blame for this: a player who’s able to take and reinforce a resource node in this manner is able to put a severe roadblock in their opponents’ ability to take that territory back, especially in the already fraught early game and into the midgame.

A brutal and quick ending is frustrating, but not as frustrating as a long and drawn out foregone conclusion. Dawn of War III’s current design tends to lead more towards those foregone conclusions, especially in team games.

Map Layout in Dawn of War III was highly lacking in 2v2 and 3v3 maps when the game launched. They were very lane driven, with long narrow corridors whose primary options presented were whether to advance or retreat. It was boring, restrictive, and so, was highly subpar. The maps released since launch have shown that Relic understands this: they’re much improved, allowing more approaches to points of interest, which allows for more interesting counter play, escape routes, and generally more fulfilling combat situations. If more maps are released, I’d expect the improvements to continue with future maps.

One of the major consistent complaints about Dawn of War III is its win condition. With its turrets and central core structure, it’s undeniably related to the win condition of MOBA games. But, let’s unpack that. What are players being asked to do in DOW3 multiplayer matches?

They’re asked to go into enemy territory and kill buildings. As in MOBAs, what you can kill is staged as the player moves towards their enemy’s base structure: first the generator, than the turret, then the Core. I don’t consider “it’s like MOBAs” to be an adequate complaint or critique of this feature, so I’ll attempt to address it more thoroughly.

As with phases, the tri-stage win condition serves more to delay a defeat than to allow the losing team to turn a match around. This is because, while the Turret and Core are relatively easy to defend and come with their own defensive abilities, the aggressor team has no additional challenges inherent in attacking these structures. A player can focus on taking and holding the map, building up a sufficient resource advantage to be able to continually trade armies in a more and more cost efficient manner, ultimately culminating in a series of assaults that their beleaguered enemy just doesn’t have the income to turn around.

Super Elites are intended to help with this, a resource-agnostic army killer. And to an extent this can work, especially if one player gets really lucky or isn’t too far behind. In practice, however, since both players/teams have equal access to Super Elites this typically results in the player with the largest army/best economy coming out on top with the Super Elite mostly serving to slow down the inevitable defeat rather than allowing the player on the back foot to make a comeback. To be 100% clear: Super Elites, as with the Phases system and turret destruction win condition, all can allow for the player/team on the back foot to make a comeback or win when behind. Unfortunately, these systems currently aren’t well balanced against the game’s economy which means at least as often as they allow for equilibrium and rubber banding, they just shallow out the slippery slope curve.

Back to the win condition though. During my interview with then-design-lead Phil Boulle, I asked about the change in the win condition from territory control to this ‘MOBA-ish’ modified base destruction system. From that piece:

He also mentioned that they’d taken some lessons from Company of Heroes 2 “The other big change was moving away from Victory Points in match play.” This one pricked my ears up, as I am in general a pretty big advocate for Victory Point-based systems in RTS design. “This was an interesting mode” he said “but… it wasn’t about blowing shit up. And Dawn of War is about blowing shit up. Victory Points created this premium of building up around the point to create an unbeatable wall, and we wanted to… encourage an aggressive situation that had an explosion at the end.”

Now, in Company of Heroes 2 and Dawn of War II, there’s no real analog for the Listening Post. In those games, most points are pretty freely capturable throughout the match – the exception being Fuel or Munitions caches, or the Power Generator. But these tend to be relatively minor speed bumps in the big picture, an investment that pays off in the long term while potentially slowing down a capture just enough to allow a player to respond. This accelerates into the mid-game as bigger guns come out and the nodes can be brought to neutral in a more timely manner.

And what does this have to do with the win condition? Well. In Dawn of War III, as in most of Relic’s games, the win condition in nominally outside the purview of the game’s economic system: you just have to kill the 3 buildings, in order, and you win. The difference is that with Dawn of War II and Company of Heroes 2, the win systems are out on the map while still being physically separated from the game’s economic system. In Dawn of War III, on most maps the generator, turret and Power Core are behind most of the map from your opponent, meaning that matches tend to become kind of a fight over the middle in a slow crawl across the map towards your enemy’s map objectives (this is not the case for some maps – many of the 1v1 maps in particular largely get around this and lead to much more interesting gameplay overall).

So in theory, there’s nothing particularly wrong with being asked to kill player-owned structures as a win condition in an RTS. In practice, the locations of those buildings tends to structure the game strongly in favor of a player who’s already in the lead in terms of map control, and map linearity can really discourage players/teams who are on the back foot from attempting to ‘snipe’ the enemy’s objectives. Additionally, late-game army sizes mean that a player or team who’s ahead may have enough units to deal with or stall attempted incursions into their base areas while their main force and Elite units traverse the map to pincer the enemy in a disadvantageous position.

I think that making it harder to hold territory while defending one’s objective structures would be the ideal design solution here: maps that allow a player to hole up defending their resource sectors along with their objective structures leads to ugly slippery slope scenarios and really hurts the game. Also, one of the Eldar’s faction characteristics is the ability to teleport around the map, ostensibly to allow for hit-and-run strikes. The longevity of Listening Posts seems to counteract this, allowing Eldar forces to flee from fights without the commensurate advantages of being able to easily take or hold territory.

The Good, the Bad, and the Ugly – Additional Perspectives on Dawn of War III’s Design

I don’t feel like I’ve seen much love for Dawn of War III’s faction design, but it’s one of my favorite things about the game. The Orks, Space Marines and Eldar have widely and wildly different approaches to the game, leaving lots of space for players to approach play in their preferred way and have experientially different approaches to the game.

Eldar seem to me to fare the worst in terms of faction mechanics: their shields certainly have an impact in combat, but are a passive benefit that can pale in comparison to the punchy benefits derived from Space Marine and Eldar faction mechanics. Drop Pods and the Standard are reactionary abilities designed specifically to punish the enemy, and exist to tip combat in the favor of Space Marine armies. This leaves encounters with Space Marines able to be dictated by SM players, as they have all of the pacing controls they’d want on-hand out of the gate. It feels to me like the designers of these faction mechanics didn’t properly weight the existence of high-impact reactionary tools when putting these mechanics into the game. My preference would be to either decrease the frequency with which these abilities can be used, decrease their potency, or increase the Space Marines’ reliance on perfect execution of those abilities in order to carry the day. Right now, they kind of feel like gravy, giving Space Marines access to tools and power that no other faction can bring to bear.

Reactionary abilities, especially consistent ones that are hard to defend against, have an outsize impact on RTS combat.

The Orks have a very interesting mechanic: the ability to loot and utilize scrap to upgrade their units means the player gets to prioritize their approach to the game, and encourages unit preservation while tying it to a faction that’s not really designed for unit preservation. It’s a fun, wild system that manages to feel strategic and bring a real weight to the decision. It does get fiddly in the late game, where large armies are annoying to loot up to sufficient power, but it feels like a really excellent balance of strategy, luck, and skill that fits pretty well into the context of Ork lore.

The Eldar’s teleportation ability, where they can relocate structures around the map and link them with webway portals to move their forces across the map instantly… It feels fun and potent and interesting. But, it’s often not nearly as useful as one would hope it could be. Eldar currently seem fated to give up ground and constantly flee from the potency of their opponent’s forces, constantly on guard lest a drop pod, jump infantry squad, or the like, break their lines and start slaughtering them. Fleet Foot, granted by a structure they can place on the map, can be useful, but it relies strongly on the player’s reaction time to make sufficient use of. Where the other factions get ways to leverage advantages, the Eldar’s innate advantages rely heavily on the constant skill and attention of the player.

I do think that most of this could be handled by balance (stat) tweaks, but as someone to whom the Eldar really appeal, it is kind of frustrating at the moment. But it’s not a foundational flaw, and could be corrected if Relic still has manpower committed to the game (I have no clue whether or not they do).

Retreat, Retreat!

The deviance from classic Dawn of War, and indeed Relic game design methodology that’s remained a thorn in my side has been the removal of a Retreat mechanic. Without the ability to easily pull units out of combat, the early game has become much, much more brutal than it is in virtually any other Relic RTS. Any lapse in concentration in the early game can be fatal, especially once early Elites hit the field. A squad of Stormboys can completely wreck early game forces to an almost unrecoverable degree. It can be fun and interesting and challenging, but also stressful and disheartening and feel intensely unpleasant for a player to feel unable to come back into a game because of a single encounter in the early game. This has improved somewhat with a recent patch that slightly reduced weapon lethality across the board. Part of the problem is the lack of a viable retreat mechanic, and part lies in the fact that Elite units can make such a huge impact when they come on the field. A 2-point Elite can represent an enormous power differential in the early game when only one player has an Elite on the field.

As with virtually every change implemented in Dawn of War III, Retreat was removed as a way to increase predictability and clarity of game mechanics and outcomes. Retreat had some funky (that’s a technical term) edge cases in the previous Dawn of War and Company of Heroes games, and was occasionally exploitable by lining up kill squads along retreat paths. But it was so integral to unit survivability that I really, really question its removal. There are so many cases in Dawn of War III where I desperately wish to be able to evac my army from a bad position, where a retreat button could minimize the damage I’d take. Ah, well.

Heresy! The Conclusion

Well, this has been a long-winded rant if ever I’ve written one. Five Thousand rambling words about the relative merits and issues with an RTS that isn’t widely played and was widely ill-received by players. What can I say? I prioritize my writing really well. Laser focused on the issues.

So where do I stand after all that palaver? To what conclusion is this mass of text leading? If there’s a takeaway I want to provide, it’s this: Dawn of War III is not a terrible game in its own right. It has a lot of really interesting systems. Its faction design, though I’ve talked through my issues with it, is ambitious and interesting. Its personalization systems (Elite Units and Doctrines) are similarly solid (though I do have quibbles, as I pointed out above). The game has solid bones, even if there are problems with some of the meat laid on those bones (map design: laney maps and poor treatment of the objectives on some maps). It might not have been what people expected from a Relic title. It might not be what players, many of whom understandably approach the game through a context directed by Dawn of War 1, really wanted.

I’m not trying to change minds here. The contextual expectation of something more like Dawn of War 1 is understandable. I just wanted to share my perspective on the game, to provide some commentary, which I think and hope is balanced, on the objective design of the game’s systems.

I hope that you have found something in these words that interested you. I hope that I have provided some insight into why Dawn of War III interests me from a design perspective, and why some people do (and don’t) like it. I, personally, hope that Relic has seen enough success with the game to create additional factions, and maps, and perhaps a co-op story. I’m not sure if that’s the case. But the game, for all its flaws, does have some merit. And for all its merit, a fair crop of flaws. It’s not as simple as “it’s different from its predecessors” and not as shallow as “there aren’t sync kills.”

Whatever you’ve thought of this piece, thanks for reading.