If you’ve never watched professional Dota 2, you would quickly learn that some heroes are more picked and banned than others. If you asked someone why, they would tell you that those heroes are good in the current meta. If you asked what a “meta” is, they would tell you that those heroes that are highly contested are such because they are perceived to be good by the members of the professional Dota 2 community. Whether it’s because a hero can teamfight early well, or is able to splitpush with no danger, or any variety of reasons, a hero can become “meta”. The meta, as we call it, is a direct result of a psychological phenomenon known as Groupthink.
What is Groupthink?
Groupthink occurs within a group of people where the desire for conformity within the group results in an irrational or dysfunctional decision making outcome. Group members will minimize conflict and reach a consensus decision without thinking about alternative viewpoints, and at times will actively suppress them. Groupthink is usually characterized by a loss of individual creativity and independent thinking. Because a large amount of people will arrive at this consensus decision, this is referred to as the “illusion of invulnerability”-that is, since so many people have contributed to this idea, it can’t be wrong. It is the best idea, and there are no ifs, ands, or buts about it.
Groupthink and the meta in Dota 2
It’s clear that we have our own ongoing groupthink experiment in the professional Dota 2 scene. As patches are released and the game changes, ideas on how the game should be played change with it. Pros then explore the patch by playing pubs and discover what heroes they think are good or bad and go on to pick them in scrims and officials. When Team X succeeds with a hero and Teams Y and Z see how Team X uses it, they begin using it themselves in their scrims and officials. As a result, a new meta is created around the patch.
Eventually heroes will emerge as the “meta” heroes of the patch, and these meta heroes will be picked and banned by (mostly) everyone. The only people perpetuating what the “meta” is are the players themselves. Icefrog has always been lauded for not adopting League’s style of balance, which some call heavy-handed and Riot-enforced.
The difference between a strong hero and a “strong” hero
An important thing to remember is that while a hero’s strength can be entirely subjective based on everyone’s opinions of it (a “strong” hero), it can still be strong without people thinking it’s good (which we will call a strong hero, without quotations). Of course, what I call a strong hero and a “strong” hero can overlap quite a bit. If you think about 7.00 Ember with his level 10 magic amp talent, he was incredibly strong and as a result, he was fairly perceived as “strong”, and remained as such for a lengthy period of time. Heroes that are strong but aren’t thought of as “strong” tend to be those who are sneakily buffed over a long period of time, rather than one large buff coming out at once. These heroes remain hidden until somebody notices how strong said hero is and begins bringing it into the scene via professional matches.
Crystal Maiden: Strong or “strong”
Crystal Maiden is hands down the best (and most recent) example of a hero being sneakily buffed over a long period of time, as well as how a hero becomes part of the meta. Back in 6.79, CM was the most played hero in professional Dota by a large margin, appearing in 56.45% of games. As with all popular heroes though, her time in the sun came to an end.
With the massive decrease in intelligence, her mana pool suffered drastically, and her high mana costs became too much for the hero to bear. She slowly began the descent to becoming a “dead hero” to most people, and in 6.81 was almost nonexistent, showing up in only 2.25% of games in the patch. Even Bloodseeker was played more. As is the way of Icefrog, he began buffing her. Not by reverting his previous nerfs, but by changing other things.
6.82, released on September 25, 2014 saw the slow buffs begin to Crystal Maiden. Over the span of a little over two years (patches 6.82-6.88), Crystal Maiden saw buffs to all of her abilities, and never once was her Intelligence changed.
Groupthink and Crystal Maiden
Patch 6.88 came out on June 12th, 2016. Surely after all these slow buffs people would have realized how good Crystal Maiden is, right?
She only appeared in 4.07% of games in 6.88, remaining in the dumpster. She didn’t appear in the top 25 hero list until 7.02, six months after her series of buffs finished in 6.88, before becoming the most picked hero in 7.05.
Why did it take over six months for Crystal Maiden to be picked after all those buffs?
Being a good captain and the “good hero” fallacy
The nature of the small buffs that Crystal Maiden got means she never got noticed. Even as they built up over time, nobody thought “CM’s been buffed for six patches in a row, I bet she’s really good now”. Let’s review groupthink quickly:
“Group members will minimize conflict and reach a consensus decision without thinking about alternative viewpoints, and at times will actively suppress them. Groupthink is usually characterized by a loss of individual creativity and independent thinking.”
The group members are the players and captains of the Dota 2 professional community. Here’s a scenario. Player X sees Crystal Maiden being played in a pub and realizes how strong her spells are now, as a result of the consistent buffs. Player X decides to play some Crystal Maiden of his own, and this reinforces his belief that the hero is good. Player X then goes to his captain and tells him he thinks Crystal Maiden is really good. The Captain, who we will call Player Y, reminds player X that CM’s mana costs are still very high, only remembering the old negatives of the hero, not remembering the steady buffs she’s received over the course of two years. Player Y then proceeds to point out that there are better “strong” heroes to pick: That is, heroes that he and most other players think are strong. However, CM could very well be stronger than these other heroes, she just hasn’t been discovered yet.
Is Player Y a bad captain? Absolutely not. As mentioned before, CM was steadily buffed over the course of two years. Two years is a very long time in esports, long enough for careers to rise and fall. Very, very few people would notice these gradual changes.
The other important thing to note with this is that there is often a large amount of money on the line for these teams. How many people do you think would be willing to make the decision that everyone else is wrong: CM is strong, and you believe 100% in the strength of said hero. You’re so confident in the ability of this hero that you would willingly go against every single other drafter in the game that perpetuates the current meta, and say “Screw everything you think about this game: this is what’s good in this game, and you’re all stupid for not recognizing it until now”.
Very, very few people would be willing to act on what they think is right, especially at a LAN. After all, there are heroes that can be picked that everyone else thinks are good. But as players play more and more pub matches and see a hero being repeatedly picked, it eventually makes its way into the pro scene.
Ex-Wings: The masters of breaking groupthink
Wings was one of the most, if not the most unique team in Dota 2 for one reason: they did not adhere to the meta. Wings did not care what other teams thought was good or not, they only cared what they thought was good or not, and their hero pool was incredibly deep. Wings picked heroes that most teams never thought of, even going so far as to pick Techies and Pudge together at TI6… twice. Outside of those Techies plus Pudge matches, they didn’t drop a single game in the main event. But what enabled Wings to completely ignore the meta?
Playing pubs is important
Large patches end up changing the lion’s share of the 113 heroes, as well as several items, interactions, and changing the map/lane equilibrium. With every large patch, the game must be relearned. Analyzing such changes can take upwards of six hours. As a result, the only way to truly get a grasp on the game is to play. A lot.
Wings’ training schedule leading up to TI was absolutely brutal. Training six days a week, upwards of eight hours a day, no arguing, no games other than Dota 2, and no recreation. Some of the players suffered injuries from their intensive training.
"When I was 16, I had a slipped disc, and now when I sit for extended periods of time it hurts very badly. But I can only grit my teeth and endure the pain. Apparently going for a surgery would require me to rest for three months, how do I find that kind of time now?" --Shadow
The sheer amount of Dota Wings played probably outdid every other team at The International, and with their dedication to playing their own game independent of everyone else’s, they were unstoppable. A meta only means something when people care to enforce it, powerful heroes only matter when everyone thinks they’re powerful. If you go into Wings and ban heroes you think are strong, Wings will just pick heroes that are just as good at that you’ve never thought of.
Dota is a game of feel
People often say Dota 2 is a game of feel. When 1437 published his analysis of OG vs. IG at DAC, if you pay attention to his language in the videos he uses the word “feel” in relation to how the game progresses. If a hero gets his stack stolen, that hero “feels really bad about himself”. Dota 2 is also a game of confidence and momentum. If you feel really bad after losing that stack, you won’t have confidence in your ability to fight. “Feel” is very, very important.
In a twitter thread, Scant (a fairly active coach in SEA Dota) makes a lot of points about how ideas in Dota 2 can come to a head. Since Dota is so big, you have to play a lot of pubs. When you play a hero in pubs, you don’t necessarily say “wow this hero’s spell damage ticking every .5s instead of every second really makes this hero powerful”. You simply think the hero is good, and you play more of this hero. As Scant remarks, “you find justification after the fact” and he’s 100% right.
In the original tweet, Scant remarks that “1437 tries jakiro in pubs and likes it, then drafts it four times and wins all those times, then Jakiro becomes meta”. Jakiro was nonexistent before 7.06. Surely he must have gotten huge buffs right?
Wrong. Like CM, Jakiro’s been slowly buffed over a series of patches.
All it takes for a hero to be played is for someone (like Theeban) to notice him in pubs, and begin picking the hero in pro matches. If he wins, as Scant remarked, he becomes “meta” and a “strong hero”. And as a result, Jakiro is now the third most played hero in 7.06 with 32 matches.
As a result of groupthink, a meta can be influenced by a handful, if well-respected, people. 1437 is widely respected due to the credit that Team Secret gave him as their coach, when they were in dominant form. However, a meta is only a construct of the scene itself-and it means little to nothing if teams choose to disregard it. Only one team has been able to break free, and one has to wonder: will we ever see another team like it?