We're making progress (despite this article coming much later in the week than I had planned), and though we still can't quite talk about the real meat of League of Legends strategy, we can put a layer on the foundation.  If you haven't read last week's article on zoning, read it.  It will be relevant today.

Of course, before we get to the cool theoretical fusion, we're going to talk about something completely different:

Champion Roles

"Come on, this is something everyone knows!"  Of course it is.  That's why the term "bruiser" means something different to every writer on the internet, and why the term "AD caster" gets used at all.  In reality, the champion roles and their high-level functionality are not very well understood at all.  Let me clarify: I'm not saying that champion capabilities or team compositions are not understood; pros get it much better than I do, but they do so on the champion level.  They understand Gragas, they know what he's good for.  What they don't understand is how to compare other champions to Gragas on a large scale, particularly the ones that don't see a lot of play.  

That, by the way, is the entire purpose of this series of articles: I want to create a basis for analyzing League of Legends that can work independent of experience with the game itself.  I want to be able compare Gragas and Syndra based on their kits and stats alone, as opposed to expert experience.

But yes, back to champion roles.  Riot has done a pretty good job of classifying champions, here's what they have:

Fighter

Tank

Support

Assassin

Mage

Marksman

As far as standard classifications go, this one isn't all that bad, but I have two big questions: First, how am I supposed to take this and use it to analyze anything, it's just a bunch of titles (and vague descriptions in the original post).  Second, what the hell is an assassin?  "Talon is an assassin!"  Okay, but he does the exact same thing as Kassadin.  "Sure, Kassadin can be an assassin too."  Okay, so why do Riven, Ahri, and Zed get grouped in that category too?  "Uhh...because they assassinate people?"

I'm seeing two trends in these responses, and they're both annoying.  In the first, we're labeling a champion role based literally on what their kit looks like (Talon, Kassadin, and Akali all jump on people and do a bunch of damage, so they're assassins!), or worse, we're doing it based on what the champion looks like (Akali and Zed are ninjas, so they're assassins) which is intuitively inane.  In the second, we're labeling them based on their capacity for burst damage, a trait shared by mages, fighters, assassins, and even some marksmen.  It's not a stupid label, it's just not a helpful one.

To twist Riot's model into something that makes sense, I'm going to eliminate the Assassin tag outright (they're all mages or fighters anyway).  For the rest, I have a question that I think sorts them pretty well: When a champion gets gold, what do they spend it on?  "Items!"  Yes, but we'll need to be more specific.  Try damage or not damage.  I think those sum up item purchases pretty well.  That lets us group the roles as follows:

         Damage                   Not Damage

Fighter <------------> Tank

Mage <--------------> Support

Marksman

For some brief definitions, fighters and tanks are defined by their durability, mages and supports are defined by their reliance on their abilities, and marksmen are defined by their reliance on ranged autoattacks (I'm going to put my foot down here: There is no such thing as a melee marksman.  If they have the durability to survive being in melee, they are fighters).

Furthermore, each role can be examined based on three parameters that vary by champion.  These are Damage, Mobility, and Utility, and it is from these two layers of distinctions that we can form team compositions.  Before I get into a real discussion of what a team composition is (that's next week's article), though, I want to clarify what's going on at this mid-level of strategic gameplay.

I have written at length on this blog about the significance of multiplicative relationships (see my articles on itemization), like with HP and Armor and whatnot.  Damage, Mobility, and Utility are like that.  When it comes to a straight-up fight, with equivalent skill of execution on both sides, the team with the highest product of these will win.  I'm not saying that they can be distilled into actual quantities that you can multiply, but I am saying that the whole, mixed between them, is greater than the sum of the parts.  In other words, Damage and Utility (CC) is better than double the damage.

See you next week,

Sorry for the recent radio silence; these guys are coming a little slower than I would have hoped.  Regardless, if you haven't read last week's article, you should scroll down and skim it.  Aaaaaand now that you've done that, you probably have an idea of where I'll be going with this article.  To answer TFN's question on the last one, we will eventually be arriving at something of a playbook for the higher level ideas like vision or team composition, but before we do that we need to lay down a foundation, the basics, hence today's title.  We'll be looking at three of the most basic of basics: range, mobility, and, derived from those, zoning.

Zoning

Counterintuitively, we'll talk about this one first, if only because it's the easiest thing to begin a discussion about.  

So here you are, jumping into a game of League of Legends.  You walk into lane, and you see a thing, where a thing can be loosely defined as anything with a health bar.  The enemy fountain turret is the one exception to this, but let's not be obtuse.  In lane, and in the game as a whole, you see a lot of things, and in order to succeed, you need to make decisions about what do with the things.  We'll start with the first one: Where do you stand, with respect to a given thing (and I mean this in terms of space, not in terms of your opinion)?

If you don't know, that's okay, because I don't know either.  In order to talk about proper positioning, you need to know what the thing is.  The obvious first thing you would need to know about it is how threatening it is.  Minions aren't very threatening.  Champions and turrets tend to be.  However, partially because it's so obvious, and partially because there is too much more to say about it, we're going to leave out threat as a characteristic for now.  That leaves two characteristics to talk about with respect to zones.  Just guess what they might be.

Range

The first (surprise!) is range.  From how far away can they interact with you?  Does the degree to which they can interact with you vary non-binarily with range?  For almost every champion, the answer to this question is yes, but it's less so for turrets or minions.  They will either hit you or they won't.  Champions, on the other hand, have abilities with ranges that often don't match up with their autoattack range, which means that their "zone," as can be derived from range, is not just a simple circle, but a circular heat map, where they can interact more meaningfully the farther you venture into their circle.

Mobility

From a mathematics/calculus perspective, it might be safe to consider range a positional quantity, where mobility is the derivative, velocity.  Where range represents your actual capacity to interact with your opponent(s) as a function of your position on the map, mobility (and your opponents' mobility) represents your ability to change your capacity to interact, also as a function of position.  Mobility might obviously be evaluated based on movement speed or the presence of gap closers, but there are other factors that you should take into account as well.

Consider turrets.  They have no mobility.  They have no inherent ability to change their capacity to interact with you.  Meanwhile, consider minions.  They move, yes, but they're probably no better than turrets in terms of their ability to alter their position in life.  This is because their movement is predictable, because we know, at any given moment, where the minions are going next.  Therefore, when we consider a champion's mobility, it is important to take into account the predictability of their movement in addition to its straight-up power.  Twisted Fate moves blindingly fast in terms of distance over time, but the movement is kind of predictable, at least moreso than a random Kha'Zix jump.  Same goes for Nocturne.

Full Circle

And this brings us back to zoning.  Now, taking the heat map concept I talked about in terms of range, let's incorporate mobility and construct a hypothetical zone map of your lane.  This map, like the capacity-to-interact range map, will also be a heat map, but in this case, the measured quantity will be cost.  This map measures how costly it will be for you to kill your opponent as a function of both of your positions, both of your ranges, and both of your mobilities.  The contribution of minions and turrets can be modeled as static quantities alongside your opponents characteristics, largely because their range is so binary, and their movement so predictable.

Of course, you don't have this map in front of you (it would be insanely difficult to generate in a game situation--but it's probably how Riot programs the bots for AI games), but if you did, what would you do?  To defeat your opponent, you would attempt to maximize the time you spend in zones where your capacity to interact with him is greater than his capacity to interact with you.  And as his HP goes down, the amount it will cost you to kill him will decrease in all regions, and eventually, it will be something you're willing to pay.  An enemy has been slain.

Hey guys, sorry for the silence; I ended up taking the second half of last week off for various personal reasons, but I will hopefully be able to furnish you with content uninterrupted this week.  Onto the point, a popular thread of conversation between the Summoner's Raft "staff" (read: Tom and I) lately has been the overall dominance of Korea (and SKT in particular) in the League of Legends scene and what this means for the future of American esports.  The conclusions of those conversations are neither well-defined nor pertinent right now, but a question keeps coming up in the middle of them: Why is League of Legends theory so poorly documented?  My kneejerk response is that no one really knows what's going on, but that probably isn't necessarily true.  At the very least, it's evident that some people know more about what's going on than others.  SKT in particular.  

To that end, unless I get sidetracked by something that happens either in my life or in League news, I will be devoting the space of my next few articles to talk about elements to be employed in an intellectual understanding of the game: What makes some strategies beat others, how much of a role mechanics plays in all of it, matchup dynamics, ward placement, team compositions, bans, all of it.  But we'll start with the basic stuff.  Today, let's just have a discussion of the difference between mechanics and strategy.

The breakdown of League skills will look something like this: 

Mechanics

Intellectual Mechanics

     -Awareness

     -Game Knowledge

     -Decision Making

Execution

Team Play

Strategy

???

If my conspicuous question marks were not enough of a hint, there's something wrong with this dichotomy.  Mechanics has a full breakdown of parts, while strategy is just that: Strategy.  I'll get to that in a minute, but first I just want to explain what I mean by these terms.

Mechanics, as it is traditionally understood, is execution, aka your ability to press buttons.  Don't knock it.  It's hard.  But the word generally doesn't take the various other parts of playing the game into account.  And by the way, that is what mechanics is: your ability to play the game.  In order to play the game, you need to sharpen certain intellectual faculties as well.  You need to know what your champion does, what the other champions do, how the game works in general (game knowledge).  You need to know when it is safe to concentrate and when you should avoid tunneling at all costs.  You need to know when to watch the map (awareness).  And you need to know what to do with the information that you gather from the game at any given moment (decision making).  Moreover you need to be able to work with your team and coordinate your thought patterns so that you will jump to the same conclusions with a minimum of the messy communication that gets in between.

So then what is strategy?  In short, strategy is the plan, the situation(s) you are attempting to create in your game, that you will then use your mechanics to execute.  Team compositions, picks and bans, and jungle routes are all strategies.  These are things that are worked out ahead of time, and the best ones account for multiple situations that might come from the initial actions.

The problem is that between mechanics and strategy, there is virtually nothing written on it (and certainly nothing that is well agreed-upon) on the whole damn Internet.  Who knew that League writers only like to talk about champions and builds...

Tom: Jungle Riven vs. jungle mana Riven in Korea today.  Mana Riven saw more play though.

Me: What does that mean?  Riven has no mana...

Tom: You've never played mana Riven?

Me: ...Irelia?

Tom: Nobody plays that trash.  Not even in top lane.

Me: Who are we talking about, then?

Tom: Mana Riven!  An attack damage melee champion with a gap closer and an execute mechanic.  Also (apparently) viable in the jungle and has mana.

Me: Monkey?

Tom: Wtf?  What is Wukong's execute mechanic?

Me: Didn't read.  Sorry.  Uh, Darius?

Tom: Viable in the jungle?  Gap closer?  All I see is execute mechanic.

Me: He's been a tier-3 jungler since his release, and he has a gap closer.

Tom: Tier-3 is viable?

Me: Yes.

Tom: This is KR we're talking about.  Think of a champion that just screams Riven with mana.  Then think of a champion with very good jungle clear speed.  If these are not the same champion, try again.  

Me: Very good clear speed?

Tom: Well, not very good, but enough that a Korean jungler would be satisfied running it.

Me: Uh...Pantheon?  I got nothing.

Tom: Khazuzu?  This is 3.14 before the Panth rework.  No way they'd run Panth.

Me: Kha isn't exactly mana Riven...

Tom: And Wukong/Darius are?

Me: No, definitely not Darius, but Wukong kinda beats Kha on that one.

Tom: What?  How does Wukong function at all like Riven?

Me: High-damage bruiser that functions as a slippery off-tank in teamfights.  How the hell does Kha?

Tom: Riven is a tank?  Also, what does bruiser mean again?  I thought Riven and Kha were assassins.

Me: No, she's not a tank, she's a damage soak.  A bruiser is a tanky champion that's not a tank or support.  And she does the assassin job, but she functions very differently from Kha in teamfights.

Tom: What's the difference between a tank and a bruiser?  Also, Riven shouldn't be a damage soak.  If you go in first as Riven, you are either God or you're doing it wrong.

Me: Riven goes in first all the time, and she is a damage soak because she is a high threat while her mobility and shield make her very difficult to kill.  A bruiser's job is to damage, a tank's job is to zone with his body.

Tom: ...What games have you been watching where Riven goes first?

Me: I'm not talking about initiation.  I'm talking about going in with the front line.

Tom: Nowadays, everyone but AD and some supports/mids are on the front line.  Are you implying that Kha'Zix is not front line?  Going in first on any high-damage champion is suicide unless the enemy is bad or has no CC.

Me: Kha is not front line, and I think it's objectively clear that Riven can take more damage.

Tom: How does Kha do damage if he's not front line?

Me: How does Katarina do damage?  How about Zed?  There's a difference here between these guys and Riven.

Tom: No, they do damage the same way Riven does.

Me: No, their timing is different.  If they are in front at the wrong time, they die quickly and unforgivingly.  AoE is a problem for them too.  Riven is much more forgiving in the first sense, and her shield lets her care a lot less about incidental damage.  She can stand in places that the others can't, which leads to fundamental differences in her role.

Tom: But the objective is the same?

Me: Duh.  The objective is the same for marksmen too.  And minions, incidentally.

Tom: What???  Assassins kill the most vulnerable enemy.  Marksmen kill the highest threat (where damage is a subset of threat).  Minions afk push to the enemy nexus.  Sometimes the highest threat is the most vulnerable, but not very often.

Me: I disagree.  The objective is ALWAYS to kill the highest priority target, where priority is a function of vulnerability and threat.  Vulnerability is, in turn, a function of many things, including your mobility and tankiness, their mobility and tankiness, both team compositions, and the situation at hand.  It is from these parameters that we derive roles, and not the other way around.  It is not a matter which roles target which champions.  It is a matter of which champions THE TEAM targets based on the champions in the team.  To that end, roles are more a guideline, but more importantly, the outcome of this function is different for Riven than for Kha'Zix.

Tom: Is that why all the popular marksmen have a method of hitting someone outside their shit-tier range?

Me: Probably.

Tom: That doesn't explain why people go for different targets in a teamfight.

Me: Easy: Suboptimality exists due to miscommunication and error.

Tom: Why do some Koreans purposefully split aggro and win teamfights then.

Me: I see what you're saying.  The optimal target is decided per-player.  Consider the team equivalent of an optimal target to be and optimal "battle plan." 

Tom: So marksmen and assassins have different targets.

Me: Sometimes yes, sometimes no, but all else equal, marksmen and assassins contribute differently to the priority profile.

Tom: So Riven is an assassin?

Me: Uh, no.  She's a high-damage bruiser.

Tom: But her goal is to assassinate.

Me: No, her goal is to kill the enemy Nexus, which is achieved by winning teamfights, which is achieved by her dealing damage to a target that would often be different if she were Kha or Zed.

Tom: How are Riven's targets different from Zed or Kha targets?

Me: I explained this before.  She can take a lot more damage, and her mobility profile is different.  Means she can stand in different places, which means different targets are more vulnerable to her, which means she has different priorities.

Tom: How does her positioning in a teamfight change her target?  She's still doing Zed/Kha's job, only with a better approach due to her durability.

I've written about this before, but I wanted to put it out here in terms that are slightly better supported than my previous rant at Old Man Eyebrows.  In other game communities, notably among the fighting games, there's always been this little sect that's held that well-designed games do not have tier lists.  To a certain extent, balance is a prize in fighting games, whose sub-forty character rosters tend to make it a realistic goal, but (and this should be intuitively obvious to all of you) this doesn't really hold for League of Legends.  In League, the tier list exists, because Riot, no matter how good they are at what they do, can't realistically balance a 120 character roster, because a constant stream of new content will always derail a certain amount of balance, and simply because all of the different champions in League of Legends do different things.  But there's a problem.  We know that the tier list exists at any given time because it can't not, but there is no way we know what it looks like.

That Korea Place

Those of you who have been following for some time know who Tom is.  He wrote some of our earlier articles on team play (though he is inexplicably missing from the Writers tab...I'll get right on that), and he and I have regular arguments, usually about the Korean League scene.  These arguments aren't really the point right now, though I'll probably write about some of them in the future.  The point right now is that a point in one of our most recent arguments was the build order on Korean Ezreal.  Completely independent of what that build order was, I felt somewhat vindicated by the fact that there even was a Korean Ezreal.  After all, Ezreal's viability had been a point of contention in my aforementioned rant (sure, solo queue versus arranged play, but I don't care; it's not Yasuo, it's Ezreal), and here he was competing with Lucian in top-tier Korean play.

But I'll get to the point.  By the sheer virtue of the fact that Korea's competitive metagame is different from ours, I know we don't know what's good and what's not.  Same logic holds for the (arguably stranger) Chinese metagame.  Support Annie and jungle Yorick were standard there in Season 3.  There's discrepancy here; someone's wrong.  But here's the real conclusion: Everyone's wrong.  Even Korea.  

Korea boasts what is probably the strongest collection of teams in the world right now, so wouldn't they be closest to real tier list, you ask.  Well, they would, if there was any hard evidence that champion picks were the dominant force in deciding a victory.  Unfortunately, I don't think that's the case.  For all the writing I do about it, knowledge of individual champions is overrated.  Tactical knowledge, mechanics, and team play are all much more important, and Korea has those in spades.  The reason that Korea is wrong about the tier list is the same as the reason everyone else is, and it has nothing to do with winning, and everything to do with practice.

The Professional (Not) Laboratory

We know the pros aren't always right, but because of their positions at the pinnacle of skilled play, we tend to lean on their anecdotal experience much more than we could typically get away with leaning on anyone else's for anything else.  But the reality is that the pros don't have as much experience as we think they do.  Consider the following list, from the NA solo queue scene:

Thresh

Lee Sin

Lucian

Jinx

Gragas

Shyvana

Annie

Vi

Kha'Zix

Leona

These are the most popular champions over the past month in Challenger.  In the North American, solo queue-oriented training regimen, these are the champions that the pros develop real experience with.  These are the champions that they develop opinions on, and they relate their statistically less-significant opinions of other champions to these.  And these days, since the internet has so tightly synced everyone's opinions on all things League of Legends, the things everyone else is playing are derived from the pros' opinions on this list.

But, you say, pros nowadays practice in scrims where they can try new things out freely.  Not so fast, there.  I have a question for you: How many games with and against a champion does it take before you have mastery over that champion?  I once speculated that competence arrives at about ten games, so let's say that mastery shows up around 100, and you will need mastery if you are taking a champion into a game on the pro scene.  Now I'll speculate that the number of games a pro plays in a day is between five and ten, since they also need the time to eat, sleep, and stand up on occasion.  And don't bother: I know you've probably marathoned more than ten games in a day.  Try doing it every day.  

Back to the pros.  At this rate, it will take ten people about half a month to test whether a champion fits the meta.  Sometimes it's worth the time to find out.  Voyboy did his investigation with AP Tryndamere, Faker did his with solo mid Riven, but I ask: Who's going to check to see if support Poppy is the shit?  The possibility of failure places a heavy cost on this experimentation, which means that pros are only going to test the things that they've seen.  Which means that we're stuck on that same list.  Moreover, when game content is expected to change as frequently as it does, there is a chance that even a brilliant discovery (back to AP Tryndamere) will be worthless within the month.  Pro teams are expected to compete, not to develop the game, and that means that, though innovation might be rewarded, there are few situations in competitive League of Legends where a team must innovate or die.

And that's the problem.  Pros don't have an adequate incentive to innovate, and the people below them are either not good enough or don't play enough (and usually both).  Accordingly, some mythical tier list exists, but we don't know what it is, Korea doesn't know what it is, and the sub-pro streamers on Reign of Gaming and Cloth5 definitely don't know what it is.

Hey guys, I know the update schedule for this blog has been kind of scattered lately, but I will explain as best I can.  I still intend to get two articles to you this week, the second happening either later today or on Saturday, but for the foreseeable future, I will only be able to update twice a week.  I currently have some obligations that I would like to have resolved in around a four-month timeline, and in order to do that I will need to devote a significant amount of time for them.  And that means a little less time on Summoner's Raft.  I apologize, but it's the best I can do at the moment.  TFN is currently busy with his own obligations, so that's that.  As always, if you are interested in writing about League of Legends, feel free to ask me.  In all likelihood, I would love to showcase your work here.

Now onto business.

It's a new year and a new season for League of Legends, and even as the preseason is settling in, I think there is a lot we have left to learn about the new season.  Accordingly, I've prepared a list of things that I think we should pay attention to going forward, be it underrated, underplayed, or perhaps even game defining.  Here goes:

New Vision and Things That Break It: Of course here I'm talking about Evelynn and Teemo.  There are a lot of implications to the ward changes, and one of them is that abilities that establish vision or take advantage of the enemies lack of true vision have a lot of potential to be powerful.  To that end, Evelynn, Teemo, and, to a lesser extent, Rengar/Twitch/Shaco are going to be threats.  Evelynn is performing in team environments while Teemo is doing very well in solo queue, so this isn't exactly new information, but I also don't think the trend has come to a head just yet.  Keep a watch on Teemo and the other stealth champions going forward.  I think they are going to be defining the metagame in a larger way.

Perseverance: Everyone is talking up defensive masteries, but in reality, the thing we need to be talking about is just that specific one.  The three points in Perseverance are the single most valuable of all thirty points we get, and champions that can take advantage of it benefit a lot right now.  Obviously, this includes tanks, champions that build large amounts of defense for whom the percent-missing-HP regen scales with gold spent, but champions that just benefit from HP regen (even those who don't spend gold on defenses), might find it worthwhile to put the points in defense necessary for this mastery.  Riven is a good example.  Yasuo is another (even though he does spend some gold on defense).

Zone Control: Zone control mages like Anivia fell out of favor in Season Three.  The reasons for this are not entirely clear, but the fact that what took their place was assassins is telling.  What we are seeing right now is a metagame that is much more favorable to tanky melee champions, which means two things: First, assassins are much less powerful on the whole, and second, without the crutch of range, zone control is much more important.  This is less of a sure thing, in my opinion, but I'm guessing there will be some opportunities this season for Anivia, Malzahar, and AP Zyra to excel.

Unconventional Supports: Let me be clear.  I'm not talking about support Annie.  Hell, I'm not even talking about support Veigar or Anivia.  Support gold changes have changed the game a little bit, and it doesn't seem to be the case anymore that support viability is determined solely by CC, raw utility, and tankiness.  Supports might just get to do damage now, and that's kinda cool.  But I'm not seeing much reason why that should stop with mages.  I would expect to see more tanks in the position, like Nautilus or Amumu.  I would also *like* to see a little more experimentation with marksmen in the role.  Ashe has already proven herself a competent support in the past, and I've done a little testing of it since.  A Tear of the Goddess will let her act as a utility support for the midgame, and the Manamune and IE that she can complete much later will allow her to provide decent damage in teamfights, along with good siege potential.

That's all for today.  Next up, I'll be talking about Korea, so I'll see you then.

Tuesday was Solo Queue.  Now let's look at what's good with just a little more coordination:

Bronze (57.01-53.31%)

Shyvana

Ziggs

Sivir

Lucian

Evelynn

Nasus

Kassadin

Taric

Dr. Mundo

Riven

Silver (56.94-52.90%)

Shyvana

Ziggs

Sivir

Lucian

Kassadin

Dr. Mundo

Evelynn

Kayle

Taric

Riven

Gold (56.72-53.16%)

Shyvana

Ziggs

Sivir

Kassadin

Dr. Mundo

Lucian

Evelynn

Kayle

Taric

Riven

Platinum (56.89-52.96%)

Shyvana

Ziggs

Kassadin

Dr. Mundo

Sivir 

Kayle

Elise

Evelynn

Xerath

Draven

Diamond (57.03-53.40%)

Kassadin

Dr. Mundo

Ziggs

Shyvana

Elise

Kha'Zix

Evelynn

Rengar

Tryndamere

Sona

Despite the fact that this is essentially the same list four and a half times, here's the breakdown:

Shyvana: 5

Kassadin: 5

Ziggs: 5

Dr. Mundo: 5

Evelynn: 5

Sivir: 4 (Platinum and below)

Kayle: 3 (Silver, Gold, and Platinum) 

Taric: 3 (Gold and below)

Riven: 3 (Gold and below)

Lucian: 3 (Gold and below)

Elise: 2 (Platinum and above)

Kha'Zix: 1 (Diamond)

Rengar: 1 (Diamond)

Tryndamere: 1 (Diamond)

Sona: 1 (Diamond)

Xerath: 1 (Platinum)

Draven: 1 (Platinum)

Nasus: 1 (Bronze)

Which translates directly to:

Shyvana

Kassadin

Ziggs

Dr. Mundo

Evelynn

Sivir

Kayle

Taric

Riven

Lucian

Elise

Kha'Zix

Rengar

Tryndamere

Sona

There are a couple of interesting points here.  First off, you have a consistent top five of Shyvana, Kassadin, Ziggs, Mundo, and Evelynn.  These guys are probably all good right now for different reasons (Kassadin and Ziggs are just strong, as evidenced by their solo queue performance; Shyvana and Mundo are reaping the rewards of the Defensive Masteries; Evelynn...well...wards), but they are probably all good in Ranked 5's for the same reason: Their success revolves around ganks.  And that's an interesting point of League of Legends theory by itself: Team coordination isn't necessarily just about objectives.  At lower levels, kills probably dominate much moreso than things like towers and dragons.  Some of this might also be reflective of the preseason changes to inhibitors and inhibitor towers.  When defenses are less frail, the team that can win fights in the late game (ie: the team with more kills) can actually come back from an early deficit.

Also interesting is the relative performance of Lucian and Sivir.  At low levels, their strength shows, but it disappears at higher levels, which suggests that the relative strength of different marksmen might be a bit more homogenized up there.  This also has some relevance to my rant at Old Man Eyebrows from last week.  His argument was that it sucks to be the AD carry in solo queue, but the data reflects otherwise, for Sivir at least.  It turns out that she cleans house in solo queue, but can't quite measure up at higher levels of coordinated play.  I suspect that marksman mechanics are still quite valuable in solo queue, but Ranked 5's is where the role is much less game defining.  And that's probably the real change.

The last thing I want to comment on is actually just my quip from above.  Bronze through Platinum was essentially the same list four times.  But Diamond (other than our top 5...bleh) was pretty different.  Alien and Predator are seeing success for no clearly defined reason, and the same is true for Elise, Tryndamere, and Sona.  Other than Sona, it is true that all of these champions can be considered difficult to play, so perhaps that's something.  Moreover, they are all somewhat mobility-focused, but other than that, they seem to bring very different elements to teams.  They're good, yes, but I'm not sure why.  More importantly, I'm not sure why they're good now.

The wait is over, here's what's been killing it in Solo Queue for the first month of preseason (data, as always, from LoLKing):

Bronze (55.86-52.88%)

Rammus

Amumu

Taric

Leona

Sivir 

Wukong

Ziggs

Sejuani

Shyvana

Brand

Silver (55.61-52.85%)

Sivir

Amumu

Rammus

Taric

Ziggs

Wukong

Leona

Swain

Sejuani

Shyvana

Gold (57.11-52.89%)

Sivir

Amumu

Ziggs

Rammus

Taric

Kayle

Wukong

Swain

Draven

Teemo

Platinum (57.60-52.95%)

Sivir

Amumu

Ziggs

Kayle

Rammus

Taric

Teemo

Kassadin

Warwick

Wukong

Diamond (57.37-52.81%)

Sivir

Rammus

Kayle

Ziggs

Amumu

Kassadin

Taric

Teemo

Fiddlesticks

Akali

And here's the breakdown:

Sivir: 5

Rammus: 5

Ziggs: 5

Amumu: 5

Taric: 5

Wukong: 4 (Platinum and below)

Kayle: 3 (Gold and above)

Teemo: 3 (Gold and above)

Kassadin: 2 (Platinum and Diamond)

Swain: 2 (Silver and Gold)

Leona: 2 (Silver and below)

Sejuani: 2 (Silver and below)

Shyvana: 2 (Silver and below)

Fiddlesticks: 1 (Diamond)

Akali: 1 (Diamond)

Warwick: 1 (Platinum)

Draven: 1 (Gold)

Brand: 1 (Bronze)

Accordingly, the list is as follows:

Sivir

Rammus

Ziggs

Amumu

Taric

Wukong

Kayle

Teemo

Kassadin

Swain

Leona

Sejuani

Shyvana

Fiddlesticks

Akali

There is a good mix of both new and old faces on here, but the interesting thing is that since we've done multiples of these lists like this, we can draw something more from the data than we usually do: We can tell the difference between something being the meta versus general characteristic strength.  For example, Sivir and Amumu are new additions.  For Sivir, this is obviously because of her buff.  For Amumu, it's probably the Mastery changes.  Meanwhile, Teemo and Kayle have been hanging around the Diamond top 10's for awhile now, and they don't really show any sign of moving (though this list has them seeing success at lower levels than ever before).  So, throwing out Sivir, let's talk about each of these.

Rammus has been here forever.  He's been at the top of the list forever, and he's obviously a pubstomper.  

Ziggs is new, and I don't really have it nailed down as to why he's here.  He has strong waveclear and the ability to pressure towers, not to mention a Twisted Fate-range ult and a pseudo escape.  That sounds good (like, ridiculously good) to me, but I'm confused as to why people have let this cash sit on the table for so long.  Meh. 

Amumu has been a low-level pubstomper for awhile, but he hasn't been seeing success at high levels until now.  I can only guess that the Mastery changes are to blame.

Taric got reworked, support got reworked, and apparently a ranged stun, stat auras, and a metric ton of damage is good now.  To be honest, I think the thing that has set Taric apart from Leona at the moment is what he can do with gold.  Now that supports have gold, Taric has found himself in a position where he can actually be dangerous in the late game.  Pit a support Taric who's ahead against the shut down enemy carry and he wins.  No one else (save maybe Zyra) can do that from a 0-cs botlane position.

Wukong is similar to Amumu: He's crushed low levels historically, but this is the highest level we've seen him perform.  On the one hand, the Mastery changes are probably having an effect.  On the other, he is apparently viable in the jungle now, and he brings a lot to the table from there.

Kayle and Teemo are very similar champions.  Both are ranged pseudo carries with high-utility ults, and they both have historically seen success at high levels.  Now though, they are performing as low as Gold, and I'm not sure why (neither champion should really be benefiting from the Mastery changes).

The reasons why Kassadin is good have been bulldozed over by everyone.  The dude's getting something of a rework next patch though.

The remainder of the list has a couple of familiar faces (Fiddles) and a couple of weird entries (Akali?), but I'm not really sure what to think of it.  Sejuani and Shyvana have been hanging about the list a lot lately, and Leona is probably on high-gain side of the support changes, but Draven and Akali don't make a ton of sense (yes, I know Draven is popular, but what changed about him?).

There are changes coming for some of these guys (Sivir, in particular), but I also expect this to change after awhile, completely independent of Riot's balance.  There is still a lot to be explored about this meta.  To which end...

Yasuo

This champion is really weird.  On a team without other knockups, he appears to have a really high skill floor that would probably preclude him from seeing play at any level.  On a team with two other knockups, he's a truck (this is my experience playing with him so far).  And the knockups can come from any one of many historically favored champions (Vi, Zac, Malphite, Janna, Zyra, Thresh, Jarvan IV, Nami, etc.).  But the discrepancy is interesting.  

That his power level is so dependent on his team means that he will never really be oppressive in solo queue or a team environment, despite the fact that if he is picked with a working team, he might be an unstoppable force, something that would need to be nerfed.  What is worrying is that this means that the statistics might be lying to us about Yasuo.  He might be a lot more powerful than a winrate analysis would indicate.  Just saying, it's worth watching out for.  

To respond to dragonswords, my holidays this past week have not been happy (university finals and whatnot), but I do dearly wish the happiest of holidays on the lot of you.  Enjoy them, dammit.  

But anyway.

I have been teasing a new top 10 (or 15) for a long time now, and I apologize for the wait, but before I got to that, I wanted to talk about something related that was timely about a week ago.  Now, of course, it's buried under a mountain of the things that people actually go to Reign of Gaming to read (read: PBE notes).  So yeah, before you continue, you might want to go read this.

I haven't really ever had a high opinion of Old Man Eyebrows.  He kinda just says shit, which is exactly the problem with this article.  He says a lot of things (masteries and the meta favor bruisers and mages; the best AD carries are Sivir [no duh], Lucian, Jinx and Draven--but Ezreal is terrible).  "AD carries can't get fed and carry games by themselves anymore!"  Thanks guy from Season 2, but this news was old a fucking year ago.  Meanwhile, it's all well and good to do a pseudo-meta analysis.  You can do a pretty good job about talking about what's good based on what people are playing--if Sivir is winning, she's probably good--but this will tell you virtually nothing about anything else.  Hell, even extrapolations from this (tanks are good --> Nasus is good --> Lucian is good because he counters Nasus) are suspect because they are compounded part-truths.  Try cubing the number 0.7.  It's like that.

So Old Man Eyebrows comes up with this tier list that is a linear combination of raw observations and literally nothing, and he concludes for various non-reasons that the reason for the lack of observations of certain characters in high-tier play is that they must be bad.  "Ezreal is great at split-pushing, but he has poor wave clear and is weak as an anti-tank carry."  Right.  Because Draven (who is good now precisely because he was never bad) has fantastic waveclear, right?  And Quinn is good now?  Explain that one for me.  Wait, I'll explain it myself: She was never bad.

The lesson to be learned here is that while it's easy to figure out what's good from the data, it is a psychological trap to label things as bad.  Quinn wasn't bad, Aatrox wasn't bad, Zed sure as hell wasn't bad, and I'll toss on an extra for you: Poppy isn't bad.  So the TL;DR is twofold.  Old Man Eyebrows is an idiot (or at least he writes like one), and Yasuo is probably an OP piece of shit.

I finally got to the last Mastery tree, but I can't really say I've saved the best for last.  Here's the data:

You'll notice first that this is definitely not the full list of Utility masteries.  That's the good news.  There are a bunch of masteries in the tree that just can't be valued with any of the methods I've looked at (Phasewalker, Summoner's Insight).  The bad news is that the ones that can be valued are not overwhelmingly compelling.  Meditation is strong for a tier 1 mastery, and the support gold masteries like Greed and Scavenger are worthwhile, but options like Alchemist or Culinary Master seem kind of silly.  Here are my specific notes:

Fleet of Foot: 1/3 the value of an AD quint.

Strength of Spirit: Champions generally have base mana regeneration between 6 and 7.  This scales up to somewhere between 15 and 20.  I used 12 as an average point.

Alchemist: Using 8 potions and an Elixir of Fortitude as average potion expenditure in a game.

Greed: Using an average game length of 40 minutes.

Culinary Master: 13.3% increase in the effect of potions.  Same average potion expenditure as Alchemist (without the fort pot).

Scavenger and Bandit: Ballpark estimates.

Expanded Mind: Useless on almost everyone.  In the ideal case, I use a 20% expenditure on Mana to account for relevance.

Implications

Here's what we can draw from this:

Supports like utility.

Junglers and mages probably don't.  There is much more gold to be gained from the Offense and Defense trees.  This is likely different from previous seasons (though Greed might be accessible to non-supports).

At this point, we have a pretty good idea of what the power masteries are in each tree, and we can make a few statements: First, almost all melee champions get a lot from defensive masteries right now.  Marksmen and mages, meanwhile, get a lot from offense.  Utility has a lot of interesting things in it, but it is generally weak in terms of gold.  Accordingly, it is probably most attractive for ranged supports.  That's all on that topic.  Let me know if you have any questions; I'll see you next time with the tier list.

I apologize for not having this one up by Friday; I've been kept away for an uncomfortably large portion of the past 48 hours, but I'm back now, and I have some interesting data on the Defensive Mastery tree.  Before I begin, I should mention that I've revised my methodology slightly: For masteries that directly affect percent damage done or taken (as opposed to percent stats), I've given a direct increase in value of 100 gold to account for the extra efficacy of the mastery due to base stats and damages.  This means that the value of the offense mastery builds I mentioned back on Wednesday have gone up (from Havoc and Double-Edged Sword) to:

9 Points: 619.9875

21 Points (AD): 2295.3925

21 Points (AP): 2475.88

To that end, here is the data on the defensive masteries:

And here are the notes on my specific methods: 

Block and Unyielding: Pre-mitigation, so this is essentially just taking AD from your enemy.  I multiplied it by 2 to account for its effect against multiple enemies.

Tough Skin and Bladed Armor: Difficult to value, but I put them down as equivalent to Block and Unyielding for a jungler.

Oppression: Multiplied by 15% to account for the relative lack of situations where being damaged by a CC'ed target is relevant.

Perseverance: This one's tricky.  It scales with maximum HP, but since it's only missing HP, I multiplied everything by .5.  From there, I defined everything to be a base of 10 HP/5 (2k max HP) with 1% defensive scaling.  Obviously, it's less valuable early on (a difference of about 180 gold in value), but it's still kind of monstrous in terms of value given.

Swiftness: I valued this strictly in terms of the passive on Boots of Swiftness.

Reinforced Armor: The value is actually 168.35 (error in the table).  Multiplied by .2 to account for relevance.

Evasive: The value is actually 198.35 (error in the table).  Multiplied by .5 to account for relevance.

Second Wind: I have no idea how to value this, but it seems pretty bad regardless.

Tenacious: I used an average of two nearby enemy champions.  The best case is 2.5 times as good (150 gold per point).

Runic Shield: I assume this might be valuable for junglers.  .3 relevance multiplier.

Legendary Guardian: Tenacity is difficult to value, since I suspect the 35 tenacity on Mercury's Treads is actually worth more than the 115 gold that it costs.  To that end, I am valuing this point at 115 gold as a placeholder if nothing else.

The total value of 9 and 21-point builds is: 

9 Points: 976.44 gold

21 Points: 2811.0525 gold

Implications

The first thing that we can glean from this is that perceptions match the truth: Defensive masteries are significantly stronger than offensive masteries.  The staple points (Fury vs. Hardiness, for example) are roughly the same value, but the Defensive tree has some clear outliers that the Offensive tree just doesn't have.  First on this list is Perseverance, weighing in at a whopping 1200 gold value.  Though regeneration isn't generally valued as a combat stat, the sheer amount that these three points give you is enough to make it worthwhile.  Accordingly, Dr. Mundo is back on the competitive scene.  Huh.  

The second thing worth mentioning is that the second most valuable cluster of points in the tree is Enchanted Armor at tier 1, meaning Defense is strong both for 9 and 21 point builds.

When I previewed this series of articles, one of the questions I had in mind was whether it might be valuable for more offense-oriented champions to put 21 points in defense.  I think this data indicates that it is still not cost-effective for mages or marksmen to do so, but it probably makes sense for melee offensive champions to do so (think Riven or Fiora).  As for supports...well, we'll see about utility next time.

I'm starting to make a habit of this, but I bit off more than I can chew.  This series is going to be a three-parter, and the next two parts will be interspersed with the tier lists in the coming week.  For reference, the purpose of this piece is to provide a context for the value of Masteries in terms of gold value.  Today, I'll be talking about Offense tree, but enough with the introduction; let's get to the data:

There are a lot of interesting things going on here, but I want to take a moment to talk about my methodology.  First, for flat stat values, I used the following table, essentially taken straight from the League of Legends Wiki:

Figuring out how to analyze percent damage or stat increases was trickier, but I approximated both in the same way.  The essential effect of both is that they amplify the amount of gold you spend on offense throughout the game.  To estimate the effect, I assumed that the typical offensively-oriented player will spend around half of his gold on offense in a given game, and he will have somewhere between 10,000 and 15,000 gold.  The numbers I used were on the high side, but the difference between the two numbers is usually not more than 50 gold per point.  Here are the implications of this method:

These masteries are worth more the better you are doing and the more of your gold you spend on offense.

Percent damage is worth some amount more than percent stats in a way that is not reflected in the data.  This is because percent damage also amplifies your base damage done that is not affected by items.  

Percent stat masteries are worth significantly less for champions that build less of those stats.

Those are the general points about the analysis, but there were also some specific ones I want to mention too (just so that the table makes sense).

Double-Edged Sword: I ignore the second edge of the sword.  This point will be less valuable the more you value defense, proportional to the amount of gold spent on defense.

Butcher: I count this point as flat AD.  The restriction of only being able to apply the damage to non-champions probably washes with the utility of increased spell damage.

Expose Weakness: I assume an average of one ally helping you hit the bad guys.  Across the full game, this should be about accurate (more for supports, less for split-pushing duelists).

Brute/Mental Force: I use an average of level 6 for these, since you won't necessarily get to the late game.

Feast: I assume you are getting 75% of the minion kills in your lane for this.

Executioner: The table shows the average value of the points.  The first point is actually twice as valuable as the subsequent points (worth 74.175 gold).

Spell/Blade Weaving: I assume full stacks, and I treat this as a 2-point mastery.

Dangerous Game: The table shows nothing for this because I have no idea what to do with it.  It's not worth nothing, but it's not comparable to anything with a gold value.

Frenzy: To account for the dependency on crits, I assume a constant one stack.

Devastating Strikes: I assume that half of the non-dominant type of damage is useful.

The final point I'd like to add is that I am valuing this Masteries solely on the champions they are useful on (these are roughly the maximum value of these Masteries).  Of course, even niche points like Arcane Blade are useful on many champions.

Implications

To no one's surprise, Havoc is the best point in the tree (though Arcane Blade comes close on ranged AP champions).  But let's talk about the things that are surprising.  First off, Double-Edged Sword is really good, subject to the drawback which I largely ignored.  Nonetheless, it is probably a must-take on marksmen and melee assassins.  Meanwhile Feast is a lot better than some people might think.  It's not an offensive point, but it's worth its weight.  Other than those, there isn't really anything shocking to be seen from the just the Offense tree.  That's where the other trees come in, but we won't get to those until Friday or later.  Until then, though, here is the rough value of 9 and 21 point Offensive allocations:

9 Points (AD--AP is slightly less): 619.9875 gold

21 Points (AD): 2120.3925 gold

21 Points (AP): 2300.88 gold

See you all on Friday.

It's going to be a blunt analysis, and I hate to lock myself in, but the reality is that I simply will not have the time to do the requisite amount of math today or tomorrow.  Thus, a preview: This Wednesday/Thursday, I will write an analysis of the gold value of the various mastery trees.  I aim to answer the following questions:

Is the increased relative value of the defense tree the real reason for the perceived increase in the value of bruisers right now?  For that matter, is its value actually significantly larger?

How does Utility measure up in value to Offense?

If Defense really is significantly more powerful than the other trees, is there value to be gained by running more offensively-oriented champions with Defensive masteries?

In other news, we are nearing a point where we will have enough data to give you all our first preseason top 10/15.  Considering how much the champion pool seems to have been shaken up, I think it will be informative.  Expect that starting this Friday or next Monday!

Originally known as orb walking in DotA, the father/mother of all MOBAs, attack move/stutter stepping is an advanced technique that is staple to an attack damage carry’s repertoire. If you’re finding you can’t kite well or you keep dieing due to shotty positioning, assessing whether or not you’re attack moving should give you useful insight. If you’re looking to up your game as an ADC and you don’t already attack move, you will notice an significant increase to your performance if you start to learn and master attack moving.

Attack moving depends on a technique called “animation cancelling”, a technique primarily relevant to ADCs. Each champion in the League has an autoattack animation. For ranged champions, a projectile will be fired at the same time every time during their animation when you auto-attack. There is a way to make the most efficient use of your time autoattacking (by closing the gap between when you attack and move) by cancelling unnecessary animation. To do this, you move immediately after the projectile leaves your champion, effectively cancelling the rest of your autoattacking animation, enabling you to only be locked in place for as long as necessary to get off autoattack damage. After you move, you can follow this up with another auto attack in which you can cancel as well, chaining together this technique and successfully attack moving. As an aside, a similar technique that can be employed for a vast number of champions in the League is “attack resetting”. An attack reset is essentially a skill that cancels autoattack animations to cast than resumes your auto-attacking. Your ability to attack reset on champions such as Renekton or Riven will separate you from optimal to mediocre dps when considering potential.

While some readers may tell themselves that they already attack move, it may not be true that they ALWAYS do, especially during clutch and high intensity moments. Orb walking should be second nature and automatic whenever playing any sort of ADC and be used in all situations where autoattacking is relevant. I’m not saying you, my reader, don’t: I’m just sharing my personal experience.

While in theory it is simple--step one, auto attack, step two, move, step three repeat--timing plays a pivotal role in your ability to attack move with a specific champion, given the inherent differences in champion auto attack animations. Over time you may even start to feel that some animations are easier to perform an attack move with when compared to other champions’. This is normal, so you may need to adjust to different champions overtime when you first start to get a grip on attack moving.

There’s a couple of methods to improve your ability to attack move (with special bindings, etc.) but for your basic execution, I would recommend moving with right-click (or whatever you have yours binded to) and pressing your “A” key then selecting your target. Why “A” key over right-click attack? Remember all those times you accidentally ran at your target instead of attacking them which most likely resulted in your death? Attack moving with your default “A” key will solve that problem. This will also save you some CS.

A Note from the Author:

Hello, everyone.  I expect that TFN will post his own (actually substantial) article sometime today or tomorrow, but in the meantime I just want to talk a bit about this blog and where I see it going.

First, I apologize about not having an article for you all on Wednesday (again).  I was traveling then, and Thursday was Thanksgiving, so I was traveling then too.  So bleh, I claim a holiday exemption.  

Second, this blog has been running for almost six months now, and I'm happy that it has made it this far, but I think it can go farther.  We (Leo and myself) had originally envisioned this as a competitor to the blogs you see on sites like cloth5 and Reign of Gaming, and while I think we are there, content-wise, we could be reaching a whole lot farther.  I suppose we at least have a logo now (dashboard readers might not have realized, so check out the site), so that's kinda cool.

To that end (outreach), I would like to post once again to let you all know that Summoner's Raft is looking for talented writers who are passionate about League of Legends and are looking for a venue where their voices can be heard.  If you think you can commit to writing a brief piece (250-500 words) once a week, send me a message, and I can get you set up.

Otherwise, I'm glad that you all are following, and I hope to bring you much more League of Legends content in the future.

Thank you,

"You say this '2-1-2' thing is good?  Okay, let's try it!"

I really hate this meta.  Thankfully, normal 5v5 queues seem to have enough jungle mains to ignore the sheer power of the quad-Targon strategy, but arranged 5v5 is a different story.  And the gold, Jesus the gold.  I don't think the strategy is that good, but there is money there, and Riot's nerfs are going to be 100% necessary.  Let's talk for a bit about why:

Money

I'm going to just rattle off some numbers here (a friend did the math, and I trust him; if you can find different numbers let me know): Right now Ancient Coin and gold generation Quints generate comparable gold to afk-jungle farming for the first ten minutes of the game, after which, jungle farming would logically begin to outpace the passive gold gain from support swag.  On the one hand, this doesn't take the new jungler itemization into account.  On the other, this is just Ancient Coin, the weakest gold generation item in its weakest form.  Hell, just upgrading it to Nomad's Medallion will about double the gold it gives you.

So then let's look at Relic Shield, which gives you about 60% more gold than Ancient coin (40 gpm).  But if you're in a lane with double Relic Shield (justifiable on a marksman for sustain, justifiable on a bruiser for the HP and CDR), this more than doubles up to 90 gpm, which is almost four times the gold you get from an Ancient Coin.  This means that, all else equal, double Targon is giving your "jungler" a lot more gold than he could get just by farming the jungle.

Sounds like a good deal.  What's wrong with that?  The short answer is not much, and that's precisely the problem.  The long answer is that running a 2-1-2 lane composition gives two people on your team a bunch of gold and sustain while robbing them of experience, which, unfortunately, often ends up being worth it.  I keep saying that this is a problem, so I think I should probably get to the "why" of it:

Counterplay

Think about the implications of not having a jungler.  Well, there are suddenly a bunch of free resources in the jungle...that's cool, and it suddenly makes map control a hell of a lot more important (you don't have a jungler doing it for you anymore).  This means more snowballing, but it's more of a social problem than a mechanical one.  This makes map control more important because now it's suddenly no one's job to control your jungle for you.  Accordingly, control of the enemy's jungle will fall much more easily to the team that's ahead.  This has potential to be frustrating, but even that doesn't hold a candle to the big problem of not having a jungler: Lane matchups matter.  Like, a lot.

As a top laner, I've had my share of experiences being put in a losing lane.  Ever want to have a bad time?  Try laning Wukong against Darius.  You'll want to break something after.  But with a jungler, you might be able to get through the experience with enough momentum to make a difference in the midgame, despite the early unpleasantness.  No jungler, and you get matched up against a lane that outranges you, out-aggros you, or just mechanically screws you over?  You're done.  Unless you are a lot better than your opponents, you are going to lose your tower, feed a couple of kills, and leave lane three to four levels below your opponents.  Then you will probably lose.  Junglers were a healthy thing for the game, and while 2-1-2 is probably a good thing to keep viable for the occasional strategic surprise in a competitive environment, it should NEVER be the default.  That's why Targon is getting nerfed, and that's why this meta will be unhealthy for the game if it continues.  

Agree, disagree, let me know what you think.  We can have a discussion :).  But I, for one, am glad to see this go.