Synopsis: Videogames typically allow players to learn a theory of design, where you are gaining an understanding of a virtual world and how it operates. The Dota series is a disastrous exception to this rule. Because Warcraft III provided no formal support for many of the concepts that were programmed into Defense of the Ancients, the map has been hounded by glitches and interactions that have compromised the theory. And not only is the theory inconsistent, it is regularly modified in the pursuit of game balance. What could have been an aberration in the history of the dota genre has now become the foundation and backbone for Dota 2, which maintains the broken theory. Why keep it? Not only is the broken theory a means to game balance, it is also a means to depth, where knowledge of the obscure interactions can be used against other players. But this compromises the learning process, where you are learning the game by memorizing small, discrete facts which operate independently of each other. The process of learning the theory is heavily predicated on rote memorization, and it runs counter to the interactivity that makes games so much fun.
But the rabbit hole for proper design goes deeper than game balance. At the end of the day, you must present these rulesets in a way that makes them interesting to explore. For this reason, we must wage war on Defense of the Ancients and Dota 2. Even though we are only holding this criticism against the Dota series, it is the figurehead for the genre. And because its lessons provide developers with an enticing but dangerous precedent, it’s worth cutting the head off of the snake.
Traditionally, games of all kinds allow players to learn a theory of design. From a simple puzzle game like Tetris to a virtual world like Grand Theft Auto V, the idea is that you are not learning a series of interconnected rules, but an understanding of a world and how it operates. Anyone who has played a Japanese role-playing game knows that monsters are vulnerable to certain damage types, where fire monsters fall victim to water, air monsters are weak to lightning, and if the game feels remotely clever, the living dead can be crushed with healing magic. These concepts also require a rigorous cohesion between mechanics and art direction, where fans of the Mario and Sonic series immediately understand not to jump on the pointed objects. But if you have designed strong theory—and with it, a high degree of integrity—it creates an enjoyable learning process where deeper levels of understanding are derived from the broad design choices that provide their foundation.
Developers settled on these ideas without much of a thought because it is a design process that feels natural. They were simply transplanting their understanding of the real world into digital theory. The commercial design process has also encouraged this, because developers have a financial impetus to make sure their rulesets can be easily engaged and understood. But with more games emerging from outside of the publishing model and the commercial spectrum, more creators are neglecting these lessons. I don’t think it’s surprising that one of the most acclaimed amateur videogames of the last decade, Tarn Adams’ Dwarf Fortress, is notoriously difficult to learn through the act of play. This is because Adams is focused on making the game he wants to make, instead of making sure that new players can approach it.1 This is also a development process that has also defined the Dota series, and the design of its ruleset is a disaster.
Remember what we said earlier: Defense of the Ancients was not software as we commonly think of software. It was not even a mod, where players are given significant freedom to alter the game concepts. It was a game map, whose rules existed directly on top of an existing game and game engine, programmed with the scripting language that Blizzard gave its players access to. It is crucial to understand that Just Another Scripting Syntax, the foundation for Defense of the Ancients, is a high-level programming language. In layman’s terms, this means that the language makes the process of creating the software easier by automating or simplifying certain aspects of the programming process, but at the cost of how much control you have over the hardware’s finer, lower-level operations.
This significantly limited what could be done with the Warcraft III game engine and game assets. And relative to the efforts of commercial programmers who have significant experience and the deep pockets of a developer or publisher, the range of what Warcraft III mapmakers could accomplish was minimal.2 Without any formal support for modification, the creators of Defense of the Ancients used JASS to “hack” their concepts into the game map, including enhanced spellcasting features and movement tricks like leaping and “knockback”. In using this higher-level language to modify concepts that Warcraft III provides no native support for, Defense of the Ancients has been hounded by bugs and glitches. They became a necessary evil in the pursuit of a complex game map built on top of the Warcraft III game engine.
As a result, the confusing nature of the Defense of the Ancients ruleset is a two-step process. The first issue is that, quite simply, various glitches and interactions compromise the theory outright. The most obvious example is magic immunity. The Warcraft III rules for magic immunity were simple: Units are immune to all abilities that cost mana, so long as it is not a healing spell or the well-defined ultimate abilities.3 The rules are easy to learn, they’re presented in a very obvious manner, use strong mechanical logic, and can be quickly learned in the course of play. And perhaps most importantly, they make sense. But in Defense of the Ancients, the sloppy use of JASS means that spells can have wildly different effects on magic immune units. Some spells will have their damage blocked, but not the stun effect. Some spells will have their stun effect blocked, but not the damage. Some spells will even have different effects depending on when you activate a skill that provides immunity.4 These outcomes come with no strong logic, and have no reason for their existence other than “that’s how they were programmed”.
In addition, Defense of the Ancients muddles the existing Warcraft III damage types, complicating the matter as players scramble to figure out which spells and attacks deal what kind of damage.5 In Warcraft III, any ability that cost mana was a magical ability, any ability that did not cost mana was a physical ability. From there, those separate ability types acted accordingly. But in Defense of the Ancients, some magic abilities are classified as physical and some are classified as magical. You heard that right: These spells all cost mana, but not all of them are magical abilities. The Dwarven Sniper, a character whose attacks are all performed with a rifle, alternate between Magical and Physical damage. The Juggernaut, who performs all of his attacks with a sword, features the same dichotomy. There are no consistent design or aesthetic principles for these choices, and these are merely two of the countless heroes which demonstrate this sort of erratic spell classification.
These systems are just a small sampling of what has been torn asunder by the Defense of the Ancients design philosophy. These confusing interactions also apply to “debuffs”, to “disjointing”, to “bash”, to nearly every core system which was adapted to Defense of the Ancients from Warcraft III.6 Normally, unintended and surprising outcomes could showcase a game with enough depth in its systems that these unexpected outcomes still surprise veteran players. But as we said earlier, these unexpected outcomes are ideally the result of higher-level interactions, where the core rules build complex but logical outcomes. In this case, the interactions are the result of the lower-level programming choices—the way that the design of an individual spell interacts with the Warcraft III game engine—and hardly anything that can be understood in advance.7 Where the fans of other genres would normally view these things as “bugs” or “glitches”, they are an intended aspect of the Defense of the Ancients experience.
From there, we reach the heart of the matter. At some point in the development process, IceFrog decided that these rules could be indiscriminately changed as a means to game balance. Though many of these changes are characterized as bug fixes, a number are explicitly outlined in the Defense of the Ancients changelogs as balance tweaks.8 These changes go a long way in explaining the inconsistent nature of the Defense of the Ancients ruleset: IceFrog doesn’t view his game world as a universe, but a series of interconnected placeholders that can be changed to suit the culture of competition which surrounds his games.9 10 This means that all of these confusing and poorly-designed rulesets—the collection of inconsistent rules associated with magic immunity, debuffs, disjointing, and other systems—can now be changed on a regular basis. An esoteric game system becomes even more esoteric as players have to scramble to memorize the new interactions.
Now, if Defense of the Ancients had merely become the father of the genre and the commercial game industry had left its lessons to the dustbin, we wouldn’t have needed an entire chapter on this topic.11 But when Valve continued their tradition of taking amateur projects and turning them into commercial games, they would go against the lessons of videogame history. While we’ll probably never know the terms of the agreement between Valve, IceFrog, Eul, and all the people who came together to build Dota 2, the finished project was less of a sequel and more an enhanced remake, a game which bundles nearly identical mechanical concepts with improved graphics and various features for ease of use.
In creating this enhanced remake, Valve took the bugs, glitches, and esoteric intricacies that were the result of the Warcraft III game engine and recreated them in their own Source engine. The goal of a one-to-one recreation is so important that the Dota 2 Dev message board lists anything that is not an “Intended Change” as a “Known Bug”.12 Once an unintended but necessary evil for Defense of the Ancients to exist on top of the Warcraft III game engine, the broken design theory was given a vote of confidence by one of the most reputable game developers on the planet. While improvements to the Dota 2 user interface have attempted to reduce some of this confusion,13 concepts such as Immunity and Bash maintain the same broken theory found in the predecessor.
But other than a stubborn insistence that Dota 2 be little more than an upgraded version of Defense of the Ancients, why would anyone carry on this approach to game design? Well, the integrity of the simulation is irrelevant to those who are most concerned with using the inconsistent theory to assert dominance over other players. Not only are these inconsistencies a means to game balance, but they are also seen as a means to depth. By allowing players to hoard knowledge of the obscure interactions, you create greater separation between stronger and weaker players.14 In a series where the player skillset is so limited and restricted, these inconsistencies become important and must be accounted for. The result is that competing games such as Demigod and Heroes of the Storm—which should otherwise earn minimum praise for their consistent game systems—will actually be considered inferior games. This is because they are not using poor design practice as a buffer zone for the general lack of complexity in the dota model.
None of this is to argue that other games do not have glitches, that glitches are not the realm of the elite in other communities, and that these glitches will always break the aesthetic presentation in a game. It is not even to argue that glitches cannot enhance the experience.15 But the developer must maintain the integrity of their game, and there must be a consistency in the choices that create the world you’re peering into. Just as the overuse of the balance update is a means to cheapening your game world, an inconsistent ruleset will achieve the same result. Once your game calls one rule into question—a world where everything is consistent, except for this one time, and another time—players will begin to question every rule. They will become more disconnected from the experience because there are more inconsistent actions revealing the weakness of the simulation. And where a strong theory of design will create elegance, a weak theory of design will encourage bloat in your mechanics. Creating an exception to the rule is to create another rule.
And let’s be clear: When we use the words “logical” and “consistent”, we are not arguing that there is random chance in how these skills interact with each other. The game is simply running computer code and the code works the same way every time. What we are arguing is that the mechanical outcomes are inconsistent with the visual and narrative presentation, thus rendering an illogical game world.16 In the case of Defense of the Ancients, an entire game has been built around these illogical and inconsistent outcomes, and then rendered more inconsistent in the desperate march to keep the game from disintegrating under the weight of its poor ruleset.
But perhaps you’re not convinced. Maybe you enjoy these “unpredictable outcomes”, where you can organize a mental catalog of loosely-connected facts and use them to gain an edge against other players. Maybe you like the idea that you can cut down an enemy team by knowing that a certain spell will respond in a certain way in a certain situation. Well, just as auto-attacking and the free camera are a boring means to depth, the design of this ruleset is a boring means to depth. In-fact, it is probably the worst means to depth that you can come up with in a videogame! You may say, “That’s subjective!” Oh, but it isn’t! If you understand media, you understand that the best media builds on the strengths of its format. Books are defined by the written word, so the best books are defined by the quality of their writing. Movies are defined by moving pictures, so the best movies are defined by their use of moving pictures.17 Consequently, the thing which defines videogames are their interactive nature, and the best videogames are the ones that play to the strengths of that interactivity.
By creating a world in which no consistent logic exists, theory can be breached at the creator’s discretion, and every rule can be called into question, the defining skill in the learning process is rote memorization. The lessons of these inconsistent rules are self-contained and have no relation to each other, resulting in a learning process that is less about interaction and more like memorizing facts out of a book.18 It is the skill which is the most at-odds with the strengths of an interactive medium, and as a result, the least interesting. Defense of the Ancients is the perfect game for a generation raised by an educational system that values the memorization of concepts over the understanding and application of theory. It is the perfect game for a generation told by public schools that history is about the memorization of facts, dates, and events, instead of drawing intelligent conclusions and concepts from the arrangement of the facts and dates. But it is also a horrible way to design your ruleset, and it runs counter to everything that makes the ideal game an awesome one.
Many players will assert that this ruleset shows how deep and complex the Dota series is. In a learning process where everything can be called into question, there will always be something new to learn. And yes, most critics of my argument will believe I am opposing Defense of the Ancients because the game is too complex, that I have a pea-sized brain, and that my pea-sized brain cannot handle the game. But where game journalists and other players often attack the complexity of a ruleset—because they simply don’t understand it—I am attacking the means by which the complexity is organized and then presented to the player. I do not want a game of weak logic where the theory changes because its creator wants to balance a matchup, and I do not want a game where it is more convenient to learn things by reading a guide than it is to play it. I want games which are worlds, and have rulesets which are confident and strong enough to convince me that is absolutely what they are.
So certainly, the learning process created by this shoddy theory may facilitate “depth”, in that stronger players can overcome weaker players through brute knowledge of the game rules. But as we said earlier, any game can be deep and profound if you have enough people trying to hammer away at it. That is why it is crucial to make sure the depth is presented in the most satisfying manner possible. The learning process in the Dota series relies on the least interesting approach to videogame interactivity, and the outcomes of these obscure rules are so pedantic that there is no visual or mechanical satisfaction to be gained from them. So if it is rote memorization that you desire in your learning process, then I suggest picking up one of the numerous academic disciplines. The thoughts presented in those disciplines are going to be far more deep, complex, and competitive than whatever you can find in any videogame, let alone a dota game.