Header art by Gareth Damian Martin.
///
For the amount of physicality and athleticism we typically associate with sports, it’s easy to forget how artificial, if not arbitrary, they can be. The first olympic competition, held in 776 BC, had exactly one event, and it was about as “organic” as physical competition can get: a foot race.
Smash cut to 2016, and it’s clear that regardless of what anyone perceives to be the platonic ideal of “running,” the various forms running takes as a professional “sport”—and the skill it takes to get good at these forms—are determined not by some naturally-occurring standard of “strength” or “speed,” but by whatever skills society wants to test for in a particular event. Race 100 meters, and you’ll get the tall, toned explosivity of Usain Bolt. Race 26.2 miles, and you get streamlined, endurance-built runners like 2016 Summer Olympic marathon gold medalist Eliud Kipchoge.
Sports take shape when there’s an established, agreed-upon set of skills that a society wants to test for. After the Golden State Warriors’ record-breaking 2015-16 NBA season, some pundits (mostly aging ones) called for the league to alter the basketball court by moving back the 3-point line. To them, the long-range basketball of Stephen Curry and Klay Thompson is inherently less interesting to watch than that of midrange masters like Michael Jordan and Kobe Bryant. That response turned out to be a bit of an overreaction (congrats to the Cavaliers), but the principle holds: the playing field isn’t an objective testing ground, it’s a canvas for displaying whatever skills we value most.
Since physical sports are limited by the laws of physics and limits of human performance, it’s easy to understand why a soccer field isn’t played on a mile-long field, or why a basketball hoop isn’t 30 feet high. It can be much more difficult, however, to parse the geographical makeup of a Multiplayer Online Battle Arena (MOBA), arguably the most popular genre of esports today. Unlike football’s gridiron or MMA’s octagon or basketball’s court, the MOBA map is a complex labyrinth of textures and abstract features set in (and built of) digital space. The possibilities here are virtually limitless. And yet, the MOBA map has calcified into an established form that spans multiple games, including Riot Games’ League of Legends, Valve’s Dota 2, and Blizzard Entertainment’s Heroes of the Storm. For all the supposed rivalries between these titles, mechanically speaking, they’re more theme than variation on a single formula.
Why three lanes? Why the jungle? Why turrets and creeps?
The standard MOBA map comprises two team bases, each housing a main objective. In League, the objective is called the Nexus, in Dota 2 it’s the Ancient, in HotS it’s the Core; destroy it and your team wins. Between the bases lies a “jungle” of complex terrain inhabited by non-player-controlled creatures, and three “lanes” that span the length of the map and spawn creeps (alternately, minions). Along with the jungle monsters, creeps serve as resources, providing players with the gold and/or experience they need to make their avatars stronger over the course of a game. Each lane is also lined with ultra-strong towers, which prevent teams from simply walking into the enemy base and destroying the main objective. By killing enemy minions, players “push” deeper into the lanes and use their own minions to help them destroy the turrets.
To the average viewer, the MOBA field can seem almost too complex. At first blush, most of its features seem inorganic or artificial, especially when placed side-by-side with a sport like soccer, which can be played in almost any open space. Why three lanes? Why the jungle? Why turrets and creeps? The origins of these playable landscapes are obscure, and so they can feel arbitrary. But, in fact, most of the the geography in today’s MOBA maps is the direct result of tweaking and years’ worth of experimentation that slowly developed the standards of the genre.
///
Although it’s often forgotten, the modern MOBA map has roots in a 1997 Starcraft mod called Aeon of Strife. AoS was rudimentary, with three lanes and a primordial resource system that bent the game engine’s rules to lay the groundwork for future MOBA economies. While the typical Starcraft game rewarded players with minerals only after they’d mined the resources with a building, Aeon of Strife rewarded minerals for killing AI-controlled characters. The minerals could only be used to upgrade each hero’s armor or attack damage, and there were no specific items to be purchased the way there are in today’s MOBAs. With no jungle and only eight heroes, the game was heavily stripped-down, but it laid the base for the RPG-esque, team-based, territory-oriented MOBA model. With the frame in place, everything was set, and since the mod community was so prolific, the process of iteration moved along at an extremely fast clip.
Just four years after the launch of Starcraft came Warcraft 3: Reign of Chaos. Even though WC3 was a real-time strategy game with entirely different objectives and mechanics than today’s MOBAs, its robust “World Editor” program gave third parties unprecedented power to modify the game. Using Aeon of Strife as his guide, a mapmaker called Eul, later immortalized in the Dota 2 item Eul’s Scepter of Divinity, used the World Editor program to build a new game mode called Defense of the Ancients. This was the blueprint for the modern MOBA, and like its predecessor Aeon of Strife, DotA’s progeny mutated immensely over a short period of time, branching out into a dense family tree of spin-offs—each with its own perspective on what MOBA functions were the most “fun.”
In its earliest days, DotA emphasized player-vs-environment (PVE) gameplay, with the act of pushing lanes dictating the tempo and feel of each game. This quickly changed when two modders—Guinsoo and Icefrog—got their hands on the game and started tweaking details to emphasize what they saw as the most compelling elements of gameplay by incentivizing team skirmishes over PvE-focused gameplay. Their version of DotA, dubbed DotA Allstars, makes up the core of Dota 2, but it was created in parallel with other progressions and experiments from independent modders, who wanted to test new paradigms and mechanics in this genre which still had yet to prove its mettle in the multiplayer gaming space.
One of these modders, a Brit who calls himself Softmints, spent 15 years putting his own spin on DotA. As the mastermind behind a DotA map called Rise of Winterchill, he’s been experimenting on the DotA map for long enough to understand every aspect of the mechanics behind it. Speaking with The Meta over Mumble, he recalls the early days of DotA modding as a chaotic period where hundreds of people ran independent experiments on the same game—each, in their own way, helping to make the game what it is today: “For almost everyone involved in any kind of lane pushing game, the map that you’re making is very much taking bits and pieces from other people. It was a very primordial soup type of thing.”
Softmints now runs a site called LanePushingGames.com, where he ruminates on the origins of MOBA games and has boiled almost every aspect of the map down to a theoretical science.
The lanes are the arteries of the MOBA map
At the core of Softmints’ science is a set of atomistic design choices he calls “buckets.” Each bucket represents a place to allocate players’ time and attention so that the game balances solo and team play in the most pleasurable way for both players and spectators. Too few buckets and the players get too bunched up and the game ends up becoming a never-ending teamfight. Too many buckets, and the players become too spread apart: their attention wanders to disparate objectives around the map and the game becomes player-vs.-map rather than player-vs.-player.
The way Softmints sees it, the current MOBA structure comprises five buckets: the three lanes, the paths that lead into those lanes, and the jungle. These are fairly consistent across each MOBA game, and while each may vary in lane length or jungle structure, the base concepts stay the same.
The lanes are the arteries of the MOBA map, and they serve multiple functions. The first function is economic: since lane creeps reward gold to the player who kills them, both teams are incentivized to stick around in the areas where the creeps clash—especially in the early stages of the game, where everything else on the map is too powerful to take down. In most MOBA games, a majority of the early game occurs in the lanes, where players “farm” for creeps while trying to get kills on the opposing enemy laners (or, failing that, then at least contesting their farm).
Lane minions also give players constant vision of what’s happening in that area, which is a crucial component in every MOBA’s strategic infrastructure. If a player knows what’s happening in the mid, top, or bottom lane, they can make conjectures about enemy locations in the rest of the map and plan their team strategies to counter. If there are four players in the top lane, a player can either make the call to push the mid and bottom lanes, take another major map objective, or send five players to the top lane to initiate a 5v4 team fight. The combination of incentive and strategic information makes the lanes a crucial bucket in the MOBA structure.
For spectators, lanes are also helpful in illustrating the narrative of a given match. Since MOBAs often revolve around the idea of map control, it follows that a viewer should be able to observe which team controls more of the map as quickly and easily as possible. Lane equilibrium and the status of towers give spectators at-a-glance readability into the state of the game. When the game state is easier to read, so is the narrative drama. If ¾ of the map is controlled by one team, there’s an apparent setup for either a comeback or a closeout. When the map is controlled by both teams equally, each teamfight becomes monumentally important.
The final MOBA “bucket” is the jungle, an area that separates the lanes and provides an extra resource for a team’s “jungler” character to farm. Beyond that, the jungle also represents the informational void of the map. With its branching pathways and myriad hiding spots, the jungle fills the map with potential threats: when a team lacks vision of the jungle, every cross-map movement is filled with the threat of an enemy ambush.
The jungle also makes interactions between lanes more difficult to read. Without a jungle, you get a more static cross-map interactions like the ones in Aeon of Strife: since the only way to get to another lane was by walking back to the point where they intersect, it was easy to deduce enemy movements based on minimal information.
///
Basketball was not always the elegant chaos that it is today. Originally, there was no limit on the number of players each team could have, the first game of basketball ever played was a 9-versus-9 abomination. The final score was 1 to 0. The baskets used were actual peach baskets, so the referee literally had to climb a ladder to remove the ball after one of the teams scored. Only over time did the sport begin to evolve into the game we know it as today. Backboards didn’t exist until five years after the game was invented, and it wasn’t until 1900—almost a decade after the sport’s introduction—that a five-player team limit was settled upon.
The early days of basketball, like the early days of the MOBA, were filled with strange, impractical idiosyncrasies that didn’t stick around, but that helped evolve the sport into what it would eventually become. For the MOBA, most of these ideas would effervesce for only a brief moment, making their way into a single patch and then disappearing into the void forever.
Since the modding community mostly went by aliases, the credit for most of the innovation in MOBA games has been lost to time. According to Softmints, it wasn’t uncommon for people to snag up other modders’ maps and put their own spin on them, remixing mechanics without giving credit to the original creator.
This led to a hyperactive form of evolution for the genre. For his DotA Outland map, Softmints experimented with a few mechanics that don’t often show their faces in today’s MOBAs, like bridges that could be raised or lowered when a player interacted with a switch. He says that terrain manipulation was one idea that always fascinated him, but which was difficult to implement in a balanced way.
The DotA Outland map also featured special “Challenge” runes that, when picked up, would task players with quests like killing a certain target in a three-minute time frame. If the player completed the challenge, they were rewarded with gold. Failure would result in the loss of a level, which was a significant punishment in Outland.
When it comes to the basic form factor of the map—the length of the lanes, the symmetry of each half, the paths that flow through the jungle—most details are the product of this rapid experimentation. The map has settled on three lanes because people tried with two and four, and the game didn’t feel as tight. Lanes have been elongated and compressed, but at a certain point, lanes they became too long or too short—throwing off the balance between isolated skirmishes and macro strategy.
A hyperactive form of evolution for the genre
Today’s MOBA games are still experimenting with new mechanics. One upcoming MOBA called Paragon uses the Z-axis to give players higher ground, but Softmints says he’s seen that before, and that he doesn’t actually like the mechanic: “If you use more of the Z-axis, you lose readability from the lower ground. It also makes things harder to read from a minimap perspective—could you imagine dealing with two floors’ worth of minimap?”
Of course, Softmints’ perspective is just a matter of opinion, and everyone has their unique perspective on what makes a MOBA fun to watch or play. But by reading the details and pedigrees of each MOBA’s map, we can get a better idea of what skills and characteristics each one privileges.
///
One aspect of MOBA design that’s distinct to the genre is the extent to which tradition plays a role in the game’s details. You only rarely notice the vestiges of old mechanics in physical sports, but they’re there. In fencing, you can get a black card—a major punishment—for refusing to salute or shake hands with your opponent. While it’s not practical, and not even really something that needs to be instated as a rule, it’s there for its own sake; it acknowledges the sport’s history and gives fencing a sense of character.
Since many MOBAs directly descend from the Warcraft 3 DotA mod, they’ve all inherited the baseline Warcraft 3 mechanics. Even today, most of the core ideas found in the average MOBA—last hitting, aggro pulling (i.e. strategically drawing the attention of AI characters), damage and defense types—originated in Warcraft 3.
But while each MOBA owes something to their common ancestor, no game hews so heavily to the WC3 tradition as Dota 2.
The advantages of this approach tie in with the strengths of the original DotA AllStars map, which Dota 2 directly imitates. By building on the years of development and iteration that had already gone into the original game, DOTA 2 exhibits a level of complexity that’s unparalleled in the MOBA genre–a complexity that Valve Software uses to facilitate a deeper range of strategic options.
Many of Dota 2’s drawbacks (if you want to look at them that way) can also be traced back to the game’s Warcraft 3 origins. While WC3 provided a robust map editor and options for customization, the engine itself had limitations, and those limitations manifested themselves in the DotA metagame. One vestigial element of Warcraft 3’s mechanics that persists into Dota 2 is the way the jungle works. In the original DotA, jungle monsters would spawn once every minute, provided the spawn zone didn’t currently have a monster dwelling inside it. Players would exploit this mechanic by pulling creatures outside of the spawn zone at the one-minute checkpoint, and a second monster minion would spawn there. Using this strategy, teams could dedicate a team member to roaming through the jungle and “stacking” creatures so that their teammates could come in and farm each stack in one fell swoop, gathering up experience and gold quickly and with great efficiency.
These arcane mechanics can be found all over Dota 2, and the added effect is that they shroud the entire game in a sense of near-impenetrable traditions and strategies that make the skill ceiling arguably the highest among MOBAs. Compared with League of Legends, Dota 2 has a larger, more complex map, filled with more jungle monsters and ambush routes, more locations that afford economic incentives. Overall, Dota 2’s distinguishing characteristic is that it values player knowledge almost as much as it values player skill.
One of the most apparent ways that this manifests itself is easily visible at a glance. While the lanes in both League of Legends and Heroes of the Storm are the same length for both teams—Dota 2 opts for mild asymmetry that sacrifices some accessibility while adding depth to the game’s overarching strategies. On each side of the map, one team’s lane stretches on for just a bit longer than the other’s. Since the discrepancy is consistent for both teams, it keeps the game balanced, but it makes for less rigid roles for each character. Not only are teams forced to construct their compositions to account for enemy teamfighting capability, but they also need to correct for enemy lane advantages depending on which side of the map they’re on. For many months, teams on the Dire-side would instantly pick Shadow Fiend, who could cleave through stacks of jungle creeps from camp that is both safe and near the midlane; Radiant-side had no exact equivalent. Dire side teams also have a modest advantage with Roshan, which is slightly easier for the Dire side to access. It’s a strategic element that exponentially increases the game’s learning curve, but it also keeps the possibilities open, so that multiple combinations of strategies and champions are viable in high-level play. At this year’s primary Dota 2 Tournament, The International 6, the amount of different characters selected breached the 100 mark, making TI6 the most diverse Dota 2 tournament of all time.
///
Compared to Dota 2’s approach of making the map a flexible field open to multiple strategies, Riot Games’ intentions are much more heavily influenced by their perceived ideal metagames. Often, Riot’s method of progressing toward that ideal meta tends to prioritize their own authorial intentions over strategic flexibility.
the delicate threads that weave together game philosophy and map design in every competitive MOBA
Up until a few weeks ago, one of the most prevalent strategies in Riot Games’ League of Legends was the “lane swap,” where the two players on a team traditionally assigned to the bottom lane switch places with their solo top laner. The switch turns both the top and bottom lanes into uneven 2v1 matchups, which has the same effect on each team: it puts the outnumbered laner at a distinct disadvantage, but it also removes a lot of the risk that comes with early game jungle pressure. By outnumbering the enemy and taking the opposing team’s tower early in the game, pro teams essentially buy themselves easy access to the mid game without relinquishing any strategic advantage to the opposing side. While it encourages more macro-level thinking, the lane-swap meta tilts the early game toward a more PvE playstyle, with players prioritizing tower pushing over lane skirmishes.
In a recent League of Legends patch—version 6.15—Riot Games took a blatant step to try and stop the lane swap meta, explaining the decision in their patch notes: “Laneswapping, while difficult to do successfully, is starting to feel pretty formulaic with few strategic tradeoffs. As it’s become more prevalent and teams do it more efficiently, it’s led to passive turret trading and less direct early conflict. When laneswapping becomes a default opener, it creates a non-interactive early game.”
To kill the lane swap, Riot made just a couple small tweaks to the map and its objectives: they increased the gold reward on the first tower kill, they made it harder to destroy turrets in the early game except for the ones in the bottom lane, and they made the early game minion waves less powerful.
The changes may seem superficial, but they’ve drastically changed the way League of Legends is played. In the current meta, lane swaps are mostly gone, and matchups stay even pretty much across the board. With more rigid roles for each player, strategies have shifted toward early jungle pressure and heavy teamfighting, and the League champion pool has calcified into a small set of must-pick heroes who do well in early team fights.
If Riot Games has, indeed, inched closer toward its idealized League meta by incentivizing early game skirmishing and undermining PvE-focused macro strategies, they’ll have to commit to that philosophy by balancing the champion pool around it. In a nutshell, this is a testament to the delicate threads that weave together game philosophy and map design in every competitive MOBA. When Riot makes even a subtle change to their League of Legends map, it has a domino effect on every other aspect of the game.
///
If Valve’s approach to achieving their ideal metagame has been to iterate on established traditions and Riot’s approach has been to tweak map details based on whatever they think will work best for the game at that time, Blizzard has taken an entirely different approach with their MOBA Heroes of the Storm: they’ve turned the MOBA map into a variable component of the game’s strategy.
As it’s already shown with games like Hearthstone and Overwatch, Blizzard is a master of boiling genres down into their component parts and refining them for consumption by broader audiences. In the lead-up to the release of Heroes, Blizzard was so concerned with shirking MOBA complexity that they told journalists they didn’t see Heroes as a MOBA.
Three years later, and lead battleground designer John DeShazer has walked that messaging back a bit (he now calls it “a fresh take on the MOBA experience”), but that desire to shed the MOBA fundamentals in favor of something both accessible and dynamic has led to some creative ideas in their map design.
The first thing that should be noted about Heroes of the Storm’s map design is that it’s not uniform. While other MOBAs treat the playing field as something consistent, like a soccer field or a basketball court, Heroes treats it as a dynamic part of the metagame that’s just as flexible as the characters.
DeShazer believes that the multiple maps encourage different strategies and team compositions that aren’t dependent on a static setting: “By making battlegrounds really strong for certain heroes vs. other heroes, it impacts the meta in a very positive way. Certain MOBA metas will stagnate, but in our game you see a little less of that because depending on what battleground you’re seeing, you’ll see different heroes come to the forefront.”
There are a few other factors at play here, though. First, since Heroes has already simplified the MOBA format by cutting item shops and individualized character leveling, the addition of more maps adds strategic breadth without making the game less accessible; although it’s managed to make the mechanical pool bigger, it doesn’t thrust new players into deep water. Second, the presence of the map as a variable in Heroes of the Storm adds an inherent versatility to its metagame. Whereas other MOBAs alter character viability via item and slight objective changes, Heroes can flip the entire competitive landscape on its head with the release of a single map. It’s emergent gameplay via brute force, and while it keeps the metagame fresh and accessible, it also sacrifices the more granular control that other MOBAs have over competitive tone of their own metas.
There will always be a desire to tweak and experiment with the playing field
On average, Heroes of the Storm’s maps are much smaller than the typical MOBA map, which leads to shorter games. And since each map demands a different play style, the battleground selection process is an important part of the professional metagame. On Battlefield of Eternity, for instance, the objective is to deal damage to the opposing team’s high-health non-player-controlled Immortal character. By focusing down the enemy Immortal, teams can secure their own Immortal ally, which will help them destroy the enemy Core without much difficulty. While teamfighting is always an important skill in Heroes, Battlefield of Eternity requires teams to pick characters who can deal high single-target damage, and the ability to protect the team’s damage dealers becomes a main priority.
On another Heroes map called Dragonshire, the geography encourages players to stay in constant motion. Since each lane is separated by only a small jungle area, rotations are common, and without the ability to read enemy movements, it’s difficult to have much success on Dragonshire.
The original intent behind Heroes was for each game to last around 20 minutes, and DeShazer says that the team views the game as having three separate phases, which seem similar to those in other MOBAs: there’s the lane phase, where characters farm for experience until they’ve become powerful enough to try and impose their strength on the enemy, there are the mid game skirmishes, which involve ganking and lane ambushes, and then, finally, the teamfight phase, where the powered-up teams face each other head-on using the resources they’ve gathered up to that point.
With one of Heroes’ latest maps, dubbed Warhead Junction, DeShazer and his team are looking to bring the Heroes metagame more in line with these idealized phases. Since most of the maps were fairly compact prior to Warhead Junction, “teamfighting started happening much earlier than intended.” By making the map larger and increasing the distance between lanes, the Warhead Junction map makes the act of rotation much more difficult than it is on other Heroes maps. If the teammates group up, they’ll miss out on lots of map objectives that they could better secure through divide and conquer tactics.
While DeShazer is happy with the strategic variety afforded by Heroes’ multiple maps, he sees the introduction of new maps like Warhead Junction as a way to have his cake and eat it; testing for new skills while also working towards a more optimal MOBA map.
///
Even though he’s put more than a decade of work into his DotA mods, Softmints says he thinks the genre is “still very young.” While that’s certainly true, Softmints doesn’t account for the fact that no MOBA will ever reach a peak state—not even if the genre lasted for a thousand years. Just like there is no optimal basketball court and no optimal tennis court, there will never be an optimal MOBA map. As long as people have different opinions about what constitutes “fun” gameplay or “interesting” esports viewing, there will always be a desire to tweak and experiment with the playing field.
What makes esports unique, though, at least when placed side-by-side with their physical counterparts, is that their possibility space is so much deeper than anything that exists in the tangible world. In most sports, there are only so many ways we can test for speed or strength or accuracy, and so we rarely see changes to the playing fields that help us test those characteristics.
Over the past decade of refinements, the MOBA map’s rapid progress suggests that there does exist a general structure that works pretty well when it comes to testing a certain set of skills. We like for players to fight each other so we can measure their individual ability; we like for players to exhibit good reaction times and game knowledge; we like it when teams need to make decisive calls as a single unit; we like to watch teams synergize in an effort to decimate one another during the final teamfight phases. But the MOBA map’s progress also suggests something else: that there may still exist skills that we have yet to test for, in the same way that we couldn’t conceive of a Kassadin backdoor or a million-dollar dream coil back in the days of Aeon of Strife. The constant warping of the MOBA map, then, isn’t just a rorschach test for social values; it’s also a place where can establish new skills to test for altogether.
