Dr. Mundolove: Or How I Learned to Stop Worrying and Love the Metagame

Anyone who has spent any amount of time interacting with the League of Legends community has come across the idea of the metagame. Or, to be more specific, he or she has likely run into the endless complaining directed against this metagame. Hatred towards any one individual champ tends to come and go over time (well maybe not Teemo, he does have a global cuteness taunt after all), but the carping against the metagame appears to be a universal constant of the League of Legends community. "Why do I have to play Ashe in bottom lane?" you might hear. "Why do we need to have a jungler?" is another common refrain. Sometimes you might hear someone pine nostalgically for an older, simpler time: "Back in beta, we played champs wherever we wanted all the time, and it was so much more fun." All of these statements are arguing against the notion of pre-defined champion roles and laning assignments, and implying that a more free-form setup would be better. If you visit any League of Legends forum for any length of time, you will see something to this effect appear and get upvoted by the playerbase, often with some kind of quip about the limited mental capacity of the "meta sheep" who simply do what everyone else does.

But there's one problem with all of this: the playerbase is completely wrong. The League of Legends metagame is a natural outgrowth of the gameplay and design choices made by Riot Games when they set the whole thing up in the first place. The current metagame is the most efficient way in which to allocate limited resources on the game's default map (Summoner's Rift). The adoption of standardized roles reflects the growing sophistication of the League of Legends playerbase, as less effective setups are weeded out and proved to be unviable. Rather than railing against the metagame, the players should instead embrace it as the best way to maximize their own chances of success. And indeed, once a player has accepted the framework of the metagame, he or she can counter-pick deliberate "anti-meta" strategies that offer further chances to achieve victory (Talon mid to kill squishy mages, etc.) In the long run, anyone who refuses to accept the dictates of the metagame is essentially relegating themselves to the League of Legends version of the stone age. You aren't being some kind of bold non-conformist, you're simply championing your own ignorance.

What then is the metagame for League of Legends? In order to understand how and why the metagame exists the way it does, it's important to trace its evolution over time. Most people who play the metagame don't understand how or why it exists in the first place; they play a jungler because "you have to have a jungler" without understanding the rationale behind it. Let's start with some basic facts about the gameplay for League of Legends, and specifically for the default map of Summoner's Rift. In League of Legends, teams of five champions are trying to destroy the enemy team's base (Nexus), having to push through a series of towers to do so. Champions start out weak and gain both experience and gold as the game goes on, both of them deriving from killing enemy minions and enemy champions. Gold gets turned into items, which make champions stronger. Champions with items are vastly stronger than champions without items - the game scales exponentially in this regard, not linerally. In the same fashion, it's better to have one champion with 10k gold worth of items than it is to have two champions with 5k gold worth of items each. Amassing the most gold possible in the shortest time is more or less the goal of the League of Legends early game. Farm gold, convert it into items, then murder the enemy team and destroy their base.

Three Lanes + Five Players = Serious Problems

However, there's an immediate problem here for the map of Summoner's Rift. There are five champions but only three lanes with minions. That means that the gold can't be distributed evenly amongst all five players. (Side note: it would be very entertaining to see Riot design a map with four or five lanes and see how that changes the gameplay. I wish this would happen.) Players found a partial solution to this fairly early on. The area between the three lanes, known as the jungle, contains a series of neutral camps of minions that can be farmed for gold and experience. Farming the jungle almost, but not quite, equals the amount of gold and experience that can be achieved from farming a solo lane. Great, that takes care of four players. What to do then with the fifth player? There are still only four sources of income for five different players. Figuring out what to do with that fifth player has been the source of the greatest innovation in the League of Legends metagame, eventually leading to the current situation, which is unlikely to see serious further changes (for reasons to be covered in time). So what did teams end up doing to solve this problem? Let's look at this in a series of stages.

The initial metagame was to have no metagame at all. You could think of this as Stage Zero, in a way. There was no metagame when League of Legends first emerged precisely because it was a new game. Things that are taken for granted now had to be discovered firsthand, through trial and error over time. No one knew from the game's genesis what would prove to be the best strategies or the most powerful item build sequences. I think of Stage Zero as lasting until the end of 2009, covering the League of Legends beta and the first few months of release. There were no Ranked games at the time, no rating systems, and matchmaking was shaky at best in producing an even matchup. Players would literally play any champion in any lane, at any time. Twisted Fate and Sion together in top lane? Sure, why not! They have two stuns, that's a sweet lane dude. Have you seen the harass that Mundo's cleavers and Zilean's time bombs can do together? I even made a video about this one way back when. Just play anything you want, anywhere, in any lane. Many of the old hands from the beta describe this as a golden age, where they had the freedom to do anything that they wanted, which has sadly been lost in the present day.

Perhaps it was a golden age of sorts, but it was also a golden age of ignorance. Sure, people could play whatever they wanted, but only due to the total lack of organized strategies. This series of unorganized and non-optimized laning setups could only function in a period without serious competitive team play. Once the first true professional teams began to emerge, it was inevitable that more efficient setups would be discovered, and anyone who failed to keep up with the new discoveries would be left in the dust. Oh, you could keep running your Rammus/Amumu top lane, but you weren't going to be winning very often against skilled players. The game had changed.

The first real metagame shift took place in the first half of 2010. I would call this Stage One of the metagame: the recognition that running two duo lanes was strictly inferior to running two solo lanes + a jungler. This might seem obvious now, but it was heavily debated at the time. There were many people who insisted that running a 2 vs 1 lane somewhere on the map, and having the chance to kill the solo laner/push down that tower quickly, would be more effective than employing a jungler. Over time, the superior experience and gold gain from farming another source of income (and two sources of income, if the other team didn't have their own jungler) gradually convinced nearly everyone that running a jungler was a good idea. If you had any desire to win, you were far better off making sure that you took someone who could farm the jungle. The unpredictability of the jungler, someone who is visibly off the map and not in any lane, was another powerful argument in its favor. These were enough to shift the metagame away from a 2/1/2 laning setup towards the now-common 1/1/2 + jungler. My point is that something which seems obvious in retrospect was not obvious at the time. It took months of community experience to work this out. Once it took hold, however, it became a permanent part of the League of Legends gameplay. Riot would have to make gigantic changes to the gameplay at this point to see the jungler disappear, and that simply will not happen. The jungler is here to stay.

Of course, not every champion could be a jungler. In the early days of jungling, it took specialized rune setups and particular champions who could jungle without dying to the neutral minions. Warwick's kit made him the first true jungling champion, and other champs with similar tanky bruiser setups followed in his wake. Soon enough, teams had to make sure that they had one of this limited pool of junglers on their team if they wanted to ensure a decent chance of winning. But wait, this meant that there was less freedom in what players could choose to pick, oh no! "What do you mean I have to jungle? Stupid metagame, argh!" The arrival of the jungler signaled the first limitation on team composition, and narrowed the range of total options. The metagame had weeded out an ineffective strategy (two duo lanes) and replaced it with a more efficient setup (two solo lanes and a jungler). Further developments would continue down the same path, limiting potential options to the most optimal team compositions.

World Cyber Games 2010: Garen/Taric versus Kassadin/Soraka Bottom Lane

The next step was recognizing that certain roles performed better in some lanes than in others. Initially, players would send champions to any lane almost at random. There was a recognition that middle lane was used for some kind of "carry" role, but you would see Ashe or Tristana in mid as often as Morgana or Ryze. This wasn't some kind of quirk of low Elo either; even the best teams would run lanes that look completely crazy today. Watch this video from WCG 2010 (held in late 2010) for a prime example of this. Mid lane in that game was Miss Fortune versus Twisted Fate, top lane was Morgana versus Ashe. Even the best players in the world were sending champions to any old lane that they felt like. Over time, the professional teams discovered that certain roles worked better when sent to certain lanes. The central location of the middle lane meant that it was ideal for champions who could roam to top or bottom lanes, great for AP champions with lots of burst who could dominate the midgame when they are at their strongest. The longer top and bottom lanes were more vulnerable to ganks, which made it more logical to put the duo lane into one of them. In mid, a duo lane wouldn't be able to zone very easily if they snowballed ahead, and solo lanes in top and bot could die very easily. With Dragon being the key early game team objective, it made more sense to put the duo lane in bot, where they would be close by to take or defend a Dragon attempt. This turned top lane into the most isolated location on the map, in turn emphasizing the selection of tanky bruiser champs with either natural sustain or an ability to escape ganks. It just didn't make sense to send Ashe solo top any more; with no escape skills, she could be ganked and killed with ease by a coordinated team. And with her weak early game, she could be dominated by all number of stronger laning champions, before her strong lategame utility could be allowed to shine.

These shifts in where champions were played could be termed Stage Two of the metagame. Increasingly champions were given laning assignments based on the role that they were expected to perform. AP champions generally going mid, AD champions heading bottom, tanky bruisers mostly going to top lane, and a jungler off doing their thing with the neutral camps. This was a direct result of the increasingly competitive nature of League of Legends. Ranked play had been introduced, and the player base became divided up by Elo ratings. It was much easier to see who was genuinely skilled, and who was full of hot air. Streaming software became popular in the League of Legends community starting in 2011, coming over from the Starcraft community, allowing everyone to view what the top players were doing in far greater detail. The first unofficial replay client also came into existence in early 2011: buggy, from a single player's perspective, and prone to breaking with every patch update. But it did work, most of the time, and offered another avenue to watch and compare what the top players were doing. The first teams were beginning to attract sponsors and advertising backers: Counter Logic Gaming and Team Solo Mid in North America, SK Gaming and Fnatic in Europe. There was access to infinitely more knowledge about optimal gameplay strategies. League of Legends was increasingly becoming professionalized.

At this point in time, the roles for four of the five champions had largely been set, but there remained the persistent question of what to do about that fifth champion. There weren't enough sources of income for five champs, and the jungle wasn't sufficient to run a double jungle setup (hilarious as it may be in troll games). The thinking in North America had been to run two champs that had crowd control abilities, having them split farm in the lane and run the thing as a kill lane of sorts. You can see this in the WCG 2010 videos, where combinations like Sion and Garen together were thought to be very strong. Sometimes you would see a healer like Soraka or a utility champ like Janna in these lanes, but they were still splitting farm and trying to build AP to be powerful in their own right. The problem here was that these lanes were running up against one of the iron clad limitations of the League of Legends ruleset: only one income source for two champs. As mentioned at the beginning of this piece, it's much better to have one champ with 10k gold than two champs with 5k gold each. These duo lanes in bot were therefore inherently inefficient, with the way they split the minion kills between two champs. These lanes almost had to get kills to avoid being extremely underfarmed and almost useless in the lategame. If only there were a better way...

A partial solution was developed in North America in the first half of 2011. Recognizing that solo lanes were naturally more gold-efficient than duo lanes, the metagame shifted towards running three solo lanes along with a jungler and a "roamer", a champion role that largely no longer exists today. You could think of the roam metagame as Stage 2.5, a false lead that eventually was phased out. The roamer was designed to be a champion who didn't need much farm on their own, since they would largely be out of lane and not taking last hits. Popular roamers were champions like Taric and Alistar, who didn't need much in the way of levels or farm to use their crowd control abilities. Stealth champs like Evelynn and Twitch were also popular roamers, with this being the brief period where Eve was at the height of her powers. Imagine sitting in mid, then suddenly getting stunned by an invisible Evelynn right as the jungler shows up to make the lane 3 vs 1. Fun stuff. Here is a video demonstrating the roam metagame at its height of popularity in 2011.

The problem with the roam metagame was twofold. First of all, the roamer would spend huge amounts of time out of lane, not farming the jungle, not farming anything at all. They would invariably become underleveled over time and ended up becoming almost useless in the lategame, unless they could successfully snowball the team ahead with a lot of early kills. Playing a roamer meant gambling on getting far enough ahead early on to make up for having a dead weight champion later. Secondly, by having the roamer out of lane, this meant that one of the solos would find themselves playing 1 vs 2 against a more standard team composition. Over time, players became better and better at shutting down the gimped solo lane with proper harassment and zoning techniques. This meant the worst of both worlds, with underleveled and underfarmed champions like a duo lane, but without the safety that it provided. In the end, the roam metagame proved to be a failure, as it was beaten by a superior system of team composition.

The solution was eventually provided by the European community, which was far out in front of the North American one. While NA was still fooling around with an inefficient system, the EU players figured out the answer to the problem: the 0 CS support. The idea was to have a champion who simply did not last hit minions at all, ending up with zero minion kills (0 CS = no creep score, no minions killed). Instead of the roamer, they created the support champ instead. This support role would offset their lack of gold income by purchasing items with passive gold generation: the now-famous Philosopher Stones, Heart of Golds, and so on. These items rarely saw use in the early days of League of Legends, and yet have now become one of the cores of the gameplay. The idea behind the 0 CS support was to have a champion who could still be effective and useful to the team without getting much in the way of items. The support offered all of the superior gold generation of the roamer, due to avoiding last hits, while providing vastly more safety to the third solo lane, who would no longer have to 1 vs 2. By directing all of the gold generation in one lane towards a single champ, and relying on passive gold generation for the other champ, the team would be greatly strengthened as a whole. There was even one role that was suited perfectly to be paired with the new support class: the AD carry, the champion role that starts out the game very weak and eventually goes on to deal the most damage with the proper items in lategame. The goal of the support would therefore be to protect their AD carry and help them farm as many minions as possible, so that the AD could take over the role of "carrying" the game later on. It was a match made in heaven, and forms the basis for the current League of Legends gameplay.

Once again, something that made perfect sense in retrospect was a radical gameplay shift at the time. This change to Stage Three of the metagame took place in early 2011 in Europe, and in mid 2011 in North America. The 0 CS support was deeply unpopular in the community at the time of its introduction, for perfectly logical reasons. After all, who would want to play the role of the support? You don't get to take last hits, and your job is solely to help another player succeed. In the me-first world of online gaming, finding people willing to subdue their own egos for the sake of the team is a rare feat indeed. There's a reason why almost no one played the Paladin class in Diablo 2, while you would find Barbarians around every corner.

Team Fnatic winning the Season One Championship at Dreamhack 2011

In the end, the gradual shift to the 0 CS support metagame took place for the same reason every other metagame shift took place: because it worked. In competitions between professional League of Legends teams played out during the summer of 2011, the EU metagame decisively beat out the NA metagame, with Fnatic winning the Season One championship at Dreamhack 2011. The roaming metagame was completely and utterly destroyed by the support metagame, the roamers achieving nothing of note while the supports allowed their AD to farm up in safety while denying the enemy AD. This was the most decisive statement possible that the 0 CS metagame was superior. The top North American teams quickly imitated what Europe was doing, and the idea trickled down to lower tiers of players over time. It's still possible today to see vestiges of the older systems of play at lower Elo levels. These players might as well be frozen in time, either refusing to adopt the newer and more superior methodology or simply being unaware of it. The result has been their relegation to the lowest levels of play. The 0 CS support metagame simply happens to be a better way of dividing up the limited gold and experience on the Summoner's Rift map. While it might not win any one individual match, over the course of many matches it will produce superior results as its use ensures a greater total gold pool for the team. This is why it has been universally adopted by all competitive teams: because it works.

There has not been a true metagame shift in the year since the adoption of the 0 CS metagame. One could say that we're in a more advanced version of Stage Three of the metagame, maybe Stage Three Plus. That's not to say there hasn't been innovation and new ideas, just that nothing has truly broken outside of the framework that's been in place for the last year or so. European teams showed off the power of running a double AP team comp in top and mid lanes, something which is still used very frequently today. Still, that's not a metagame shift, it's simply the use of another AP champ in top (nearly always one with sustain) instead of a tanky bruiser champion. The actual framework for the gameplay remains the same. We've also seen much more aggressive use of counterjungling, made famous by Moscow Five starting in early 2012, the swapping of the top and bottom lanes, with AD and support going top instead of bottom to exploit 2 vs 1 matchups, and the fast-pushing strategies employed by some of the Korean teams like Azubu Blaze and Frost, which place the highest priority on taking towers quickly. The most interesting recent strategies have been derivations of these ideas by the Korean teams: rotating the AD/support lane around the map to different lanes, using their combination as a pushing lane to take towers rapidly. 2 vs 1 in top or mid lane, take tower quickly, then rotate to another lane and do the same, etc. Like aggressive counterjungling with multiple champions, these tactics require considerable team coordination to pull off, and serve as a testament to how much more complex the League of Legends gameplay has become. Rarely do the professional teams give anything away for free. Every Dragon is fought over, every Baron, and most Red and Blue buffs are contested as well. Even the wraith camps often come under attack. Whatever is not defended is stolen away.

But again, these are all more polished and refined versions of the same general metagame. Everyone is still running two solo lanes, an AD/support lane, and a jungler. Derivations from this pattern (running kill lanes in bot for example) have never achieved sustained success, only temporary success before opponents had a chance to respond. There has not been a major shift since the introduction of the 0 CS support, and there is unlikely to be another major shift any time soon. You won't see a shift because the current setup is the most efficient way to divide up gold and experience on Summoner's Rift. The forum goers can rage all they wish, but these are the facts. Unless there are major changes to the core gameplay of League of Legends, there will not be any more significant meta shifts. The current metagame will remain the metagame going forward. The details will change, there might be more aggressive supports instead of sustain supports, or support junglers who build gold/10 items instead of carry junglers who build Wriggles, but the larger framework will remain the same. There are five well established roles, and each champion gets played to fit into one of those five roles. Such is the gameplay for League, love it or hate it.

And you - yes, you dear reader! - should learn to love the metagame! As this piece has tried to demonstrate, the metagame is not the product of mere chance, or some kind of horrible trick designed to force you to play the support role. The metagame is the product of untold thousands of players and millions of game hours. It's the end result of years of tinkering around with different champions, different roles, and different team configurations. It's the accumulated wisdom of the first pioneers who set out into the waters of a new game, and slowly figured out over time how to play it most effectively. This knowledge is the gift to you of earlier generations who had to figure things out the hard way. The metagame is the end result of the growing professionalization of League of Legends as a competitive game. It is no different than any other sport or activity or whatever; if people are playing against one another, over time, the most optimal strategies will eventually be figured out. Where do you think all of those different chess openings came from? They were derived from centuries of experience of people playing against once another, trying to win, and determining what worked the best. Think of them as the "chess metagame", if you will.

We can extend the same comparison further afield. All of the sports that are professionalized today had to start somewhere. When they first began, no one knew what the best tactics were for them either, and they had to be worked out over time. In the early days of American football, teams barely passed the ball at all, and ran the ball on nearly every single play. It took decades of experience to learn that the passing game could be a formidable weapon in its own right, until we reach the modern-day NFL where passing dominates the sport. In similar fashion, I could go on for pages about all the wacky stuff that took place in pre-20th century baseball. Even today you will still see managers bunting in terrible situations, ignoring all of the accumulated wisdom of the sabermetrics community.

Completely Unrelated Don Mattingly Photograph

The point of this is that competitive play eventually produces optimized strategies. This is a universal rule, whether it be in sports, card games, competitive eating, or whatever. Some strategies will prove to be more effective, and they will be embraced as the "metagame", because people generally don't play competitive games unless they are looking to win. The current metagame is the natural byproduct of the increasing professionalization of League of Legends. We should be celebrating the metagame, not criticizing it! It's a demonstration that the old, ineffective stuff that people were using when they had no clue what they were doing has fallen by the wayside. It's no different than the way in which sports are professionalized. When you are a little kid, you can run around on a soccer field doing whatever you want, without worrying about a position or whatever. You're a kid and the games aren't very serious. But as you grow up and get older, you have to learn to play within a scheme, and to be able to cover a formal position, if you expect to be able to win. You can't just trot out six attackers even if you want to do so; that kind of strategy has been tested, and it's going to result in a loss against any decent team playing a standard formation (unless there's a massive skill disparity between players). Similarly, you can't use four centers in basketball or seven infielders in baseball. Not in any kind of competitive play, because you are going to lose. As the sport becomes increasingly professionalized, the range of viable strategies decreases. It's no different than the kids playing soccer; as they grow up and get older, they can't all be strikers. Some of them have to be defenders. You no longer have the option to do whatever you want. That's what it means to be on a team.

It's the same thing in League of Legends. The creation of formal roles and the establishment of the metagame means that League has grown up as a competitive event. Ineffectiveness has been weeded out. Random or poor choices can no longer be tolerated. The forumgoers complaining about this are no different from the teenager who wants to go back to his youth days where everyone on the soccer team could be a goal-scorer. Well tough luck, buddy. As the sport grows and matures, the range of viable options narrows. If you're serious about trying to win, you're going to have to buckle up when you're last pick and play that Soraka support. And you'll do it because it gives your team the best chance of winning, because you want to protect Vayne so that she can farm up her Bloodthirster/Phantom Dancer/Last Whisper and then murder everyone on the other side. If you insist on picking a second AD carry, splitting farm, and trying to steal last hits, you're not being a free spirit - you're simply being a whiny jerk. Adhering to formal champion roles according to a pre-arranged pattern does not turn everyone else into "meta sheep". What is means is that they are taking a rational, intelligent response to the ironclad rules that govern League of Legends play, and doing everything possible to maximize their chances of winning.

In the end, who is the true fool: someone optimizing their chances to win, or the person sabotaging themselves because they "refuse to be a sheep"?

* Of course, this is all predicated on the notion that each player is trying to win. As we all know, this is not always the case. There's plenty of room to have fun and explore alternative gameplay in custom games, ARAMs, draft racing, and so on. You can always create a team of five friends and screw around however you wish without hurting anyone. At the same time, if you are queuing up with total strangers through the standard matchmaking system, I would argue that yes, you do owe them your best effort to try and win whatever game you are playing. Deliberately doing something that you know is suboptimal is little different than trolling in a ranked or even a normal game. I have little sympathy and no patience for these individuals. While no one HAS to play the metagame, it's a really, really good idea to do so. Trying to go against it for kicks will rarely work out well. Mess around on your own time with friends or in custom games.