How to read and tech against an e-sports meta

When your high-school math teacher insisted that math would be useful in your day-to-day life, she probably wasn’t thinking about getting that sweet, sweet e-sports monaaaay. But a little bit of simple math can take your meta-game knowledge to the next level. In many high-level tournaments it’s this detailed knowledge that gives competitors the edge they need to win. Before we get into the numbers, however, we need to have a crash course in e-sport basics.

What Is “The Meta?”

Meta is short for metagame, which is the game that exists outside the games official ruleset. If you trash talk an opponent or refuse to shower so that he gets distracted during a match, you are metagaming.

The e-sports definition of metagame is something of a misnomer, but has nonetheless become the colloquial definition. It usually refers to the process by which you choose characters, decks, or other strategies in order to give you an advantage against other players.

The classic parallel that is cited is the infamous chess example.  If you see a chess opponent win five times in a row with the same strategy, then making moves specifically to counter that strategy is considered metagaming. Your choice of strategy exists outside of the game itself.

However, character/deck/team/weapon selection is not outside the game itself. It’s an in-game mechanic, so there’s not really anything meta about it. Still, most casual players don’t think of this selection process beyond “choose whatever I enjoy playing the most”, which has given the process of picking and counter-picking a metagaming feel. Hence, our colloquial definition of “metagaming” was formed.

“The meta” is similarly a misnomer, since it usually refers to the current game state, i.e. what other players are likely to do if you enter any random match. When you “play the meta” you play according to the current game state – you see whatever the best strategy is and play it. When you “tech against the meta”, you read the current game state and then play in a different way that is specifically meant to disrupt and take advantage of it.

How the Meta Works

No game will be 100% balanced. As long as there is an opportunity to play with different strategies, some will always be better than others. As gamers, we already know this but we don’t necessarily think in these terms. In shooters, for example, we know that the strategy of “camp near choke points and attempt to shoot my enemy in the head” is a better strategy than “run backward around the map throwing grenades at anything I see.” One of these strategies just isn’t very good. Even if both players were playing their strategy at the highest skill level, one would dominate.

And it’s this lack of balance that makes games fun and interesting. Sniping is only really fun because players who are caught in your line of vision at long distances are essentially helpless. Many would argue that sniper rifles are simply balanced versus other guns, but put the best sniper against the best, say, shotgunner, and it’s still unlikely that you’d see a 50/50 kill rate.

Extend this line of reasoning to anything you can do in your e-sport of choice and you’ll find it still applies. Some characters beat other characters. Some decks beat other decks. Everything has its own strengths and weaknesses. But the meta starts solidifying when the vast range of strategies and decisions become narrowed down to just the few that can routinely win you the game, the “best” strategies. These are the strategies that people who want to win will be using.

Using Math to Break Down the Meta

I’m going to use Hearthstone decks as an example here, but this is applicable to basically any e-sport where you have choices to make before gameplay begins. Let’s say the meta has solidified and you have a list of every single deck that’s worth using. Luckily, we do! Thanks Vicious Syndicate.

There are three statistics you are looking for: 1) Global playrate. 2) Global winrate. 3) Matchup winrate.

Let’s break those down real quick. Global playrate is how often people are using any particular deck. Global winrate is how often a deck wins averaging all matchups together. Matchup winrate is how often a particular deck wins against another particular deck.

So, why are these statistics important? Well, they are the primary factors in any developing meta. The deck with the highest global winrate is largely considered the most powerful deck in the meta. In this case, it’s Mid-Range Shaman. (Once again, this can just as easily be applied to characters in a fighting game or guns in a shooter.)

Usually the “best” deck is also the deck that will be used the most often, and the same holds true here. More people are playing Mid-Range Shaman on the ladder than anything else because it gives them the greatest chance to win.

Next, we look at the matchup data for decks that have a good chance against it. In this case, Freeze Mage is favored to win against Mid-Range Shaman by 57 percent. So on first inspection, Freeze Mage is the deck you want to play in order to take advantage against the meta. But you may also notice that Freeze Mage is favored to lose horribly against some other decks, sometimes with a whopping 75 percent loss rate!

This is where global playrate comes in to the picture once more. To successfully tech against a meta, you have to figure out the likelihood of seeing any one particular strategy. If 100 percent of players were using Mid-Range Shaman, then a Freeze Mage player should win 57 percent of his games. But that’s not going to happen. Let’s say, for the sake of example, 60 percent of players are playing Mid-Range Shaman and 40% of players are playing Spell Druid, which Freeze Mage only has a 25 percent win rate against. To figure out your win percentage, you’d take the percentage of people playing Mid-Range Shaman (60 percent) times your win-rate expressed in decimal form (57 percent or .57) plus the percentage of people playing Spell Druid (40 percent) times the win-rate (25 percent or .25) and you get 44.2. In this scenario, Freeze Mage will win less than half of the games it plays, making it a bad pick.

Now let’s take another example where the craze of Mid-Range Shaman has swarmed the meta, and 90% of people are playing it, leaving only 10% to play Spell Druid. In this case, a Freeze Mage is expected to win 53.8 percent of all games it plays.

These two circumstances used the exact same decks and the exact same players, but in one situation Freeze Mage was a horrible choice and in another Freeze Mage was a top-tier pick. Granted, these situations are vastly oversimplified. To get an accurate read of any deck in the meta, you’d have to multiply every single deck’s matchup win-rate by its play rate, which is a whoooole lotta math. Many times this will average out to about the same as a deck’s global win rate, but every so often you’ll find a deck that counters a meta but just isn’t being played very often. Of course, if that deck catches on then the meta will shift and you will have to do your calculations all over again.

Too Much Math… TOO MUCH MATH!

Let’s face it, you aren’t ever going to have all this data for every e-sport you want to play. That’s fine. You can still make an educated read of the meta even without hard statistics.

Most metas are separated into “tiers” with Tier 1 being the best, Tier 2 being kind of good but not as good as Tier 1, Tier 3 being situational, and things just keep getting worse from there.

But these “tiers” are a lot more complex than “good, better, best.” They are built around a natural rock, paper, scissors hierarchy that emerges due to play and counter play.

Tier 1 is always going to be the best characters/decks/strategies in a game. These are the aforementioned strategies with the highest global win rate.

Think of Tier 2 as strategies that beat Tier 1 but have neutral to bad matchups otherwise. In general, Tier 2 strategies have a high chance of winning a tournament because the majority of any tournament will be made up of Tier 1 strategies.

Tier 3 strategies are basically every other viable strategy in the game. They beat Tier 2 strategies but are stomped by Tier 1 strategies. This completes the rock, paper, scissors triangle.

To triumph over a meta, you basically have to think one step ahead. Estimate how many people will be playing Tier 1, 2, and 3, strategies and stagger your strategies to counter them. As many people as you think will be playing Tier 1 strategies, that’s as many times as you play Tier 2 strategies, and so on. Even if your strategies are just randomly distributed, you should, in general, come out ahead in wins this way.

The trick is you always have to be sure you are ahead. If everyone thinks the same way as you, for example, the entire meta will change and your estimates will be off. Eventually, you might find that the meta chases its own tail until you are back playing the Tier 1 strategies you started with. That is what makes them Tier 1 after all.