The Meta, Snowballing and DOTA 2
The following is a big damn wall of text. I tried to make it a little less daunting to read by sprinkling champ icons and "paragraph separators", and I've also made it as concise as I can get it to be.
Why does the meta change? Obviously, it’s to obtain the greatest chance of winning, but if both teams fighting each other follow the meta, multiple games average out to 50/50 win loss rate, yet, people continue with the trends, taking up “mirror roles” somewhat often (e.g. assassin mid gets popular and 
vs 
happens, or top lane bully gets popular and we see 
vs 
). Late in season 4, ADC hyper carries became the flavor of the month, with 
getting more plays, 
getting a lot of attention, and I’ve even seen a 
(in normals) here and there. Now, it’s moved on to bursty bot lane, with 
and 
getting special attention. Again, why does this persist?
How about finding out HOW the meta changes? Gathering some inspiration from Phreak’s comments (in particular, comment #3, “That's not quite correct”)…
…we can see that picking supports, and throwing them mid, got popular with pros because they were uncounterable beyond any consistently viable action (sound familiar?). You can call for ganks, but having the jungler babysit you just won’t work. 
mid got nerfed, then reworked, then the hunt to find the most bullshit, over-kitted laner is on again, hence why overloaded champs like 
(at least in practice) are consistently played in pro gaming, and why 
is popular through ranked and normals, particularly for pub-stomping. Another way the meta changes is say, is picked to counter the flavor of the month hyper carries, exploiting their weak early game to blow them off the face of the rift, then snowball onwards to dominate the game. From then on, he sees a higher pick rate, it’s vs all over again, then we see the phenomena I mentioned earlier- mirror roles/niches.
Now, there’s a disconnect here, but bear with me…
(http://forums.na.leagueoflegends.com/board/showthread.php?t=2742764)
…DOTA 2’s differences from League of Legends are analyzed here, in terms of snowballing, and champion strategy and niches. The summary is, if champions (and by extension, teams) are too similar to each other in function, then whoever gets ahead can do everything the competition can, better. For example, earlier this week, playing Vayne against a Graves who got fed (this was during mid game), I got the perfect 1v1 ambush on him (both of us max health), popping Final Hour, a tumble to get into position, condemned him against the wall, pew pewed a bit, tumbled to the side when the stun got off and he fired his ult (dodging it), kept orbiting him to avoid the buckshot and smoke grenade, nope, still died. He survived by a few minion hits worth of damage. Sure, I definitely screwed up somewhere, but I didn’t feel outplayed, just outgunned. For some of you, considering he is fed, this is reasonable, and I somewhat agree, but the point was, Graves outperformed me at my core mechanic- auto attacks. All my mechanical merit equated to fluff. To my knowledge, Riot consistently considers such, a bad thing. If the game was 1v1, me against fed Graves, Graves will win, seeing as he won the SR duel earlier. Maybe ARAM’s money system will even out our wallets, granting me an advantage there, but not because of my champ.
The auto-target nature of ADC lane is typically why it’s so polarizing as well- you can’t dodge auto attacks there, further establishing the concept that their ADC can out ADC you if they get fed. Say you have 2 out-of-mana 
s , one with 2 bf swords and one with only 1. If they get into an auto attack contest, who will win?
Back to the question- why do people play hyper carry vs hyper carry, or bursty kill lane vs bursty kill lane? Because if someone gets ahead, they stay ahead. Occupying the enemy’s strategic niche, then overpowering them at it, is the best way to destroy counter play, a variation of why the pros picked bullshit champs/lanes. Not to be confused with how the meta changes, but why the trend continues to exist before it changes for the next.
DOTA 2’s design gives it a flexible, vague meta, because no meta is formed to capitalize on the snowballing of one powerful champion against a weaker, similar champion. Because every DOTA hero is bullshit OP in their own way, they still have strategic options due to DOTA’s design in things like the absence of spell/ap scaling. Riot, on the other hand, keeps talking about making champs fair, with counter play, but clearly this isn’t working.
I hope to hell Dawngate makes a comeback- I like this game genre, but League’s all I have for this.
So, what are your thoughts?
