从清晰度阐述Riot的游戏设计哲学

theninth1985

　　转载自：http://gamerboom.com/archives/67827

作者：GentlemanGustaf

每当有人要我推荐游戏时，我基本上会先考虑两个问题：

*这个人以前玩过很多游戏吗？

*这个人喜欢单机、合作或竞技游戏吗？

无论答案是什么，我要问的问题都是“我推荐的游戏需要故事/剧情吗？”如果回答是否，我就会推荐《英雄联盟》（即《LOL》），如果回答是是，我可能就要提名其他游戏了。这是因为我希望更多人和我一起玩吗？老实说，我没有跟朋友一起游戏的习惯。对我而言，《LOL》是一款竞技游戏，不过我偶尔也跟朋友玩一些“愚蠢的/可笑”的游戏，我玩的大部分游戏都是有排名的。这是因为推荐玩家点数吗？我还没正式向任何好友推荐呢。这是因为我知道的游戏太少？我已经玩了非常多款电脑游戏和电子游戏，如果一个一个地数，接近一百吧。这是因为我懒得想？应该不是。归根到底是因为我认为Riot的游戏设计理念比现在的其他任何公司的都先进。他们以《LOL》奠定了其开发电子竞技类游戏的基础，击败众多老牌格斗类游戏、《光晕》、《使命召唤》和甚至《星际争霸》与《星际争霸2》。Riot把《LOL》当作一种运动，而不是一款游戏。这种理念对他们的设计产生了许多积极影响。

league-of-legends(from top1mmo)

玩法

是什么导致Riot设计的玩法如此特别？那就是他们专注于清晰度。游戏确实很复杂，但这种复杂与选择有关。这些选择可能很难，但困难与清析度无关。

复杂度VS难度

是什么使复杂度如此显著？Riot设计了游戏的易用性。通常来说，电子竞争类游戏强调复杂度。这大概是因为早期的电子游戏实在太简单了。没什么人真正喜欢游戏《Pong》，也没有人爱看乒乓球锦标赛。许多早期电子竞技游戏的开发者很快发现，许多东西可以使游戏变得更复杂。动画取消、叠加技能、反补等都是从游戏引擎中产生的策略，而设计引擎时并没有考虑玩家将如何探索游戏。这就出现了一个倾向：具有复杂元素的游戏、反抗玩家控制企图的游戏，就是优秀的电子竞技游戏。行家会赞扬《星际争霸》中的小单位的微操作的难度，《纽沃斯英雄》和《DOTA2》的玩家批评《LOL》因为缺少“反补”而降低复杂度。

我们就以《LOL》为案例。《LOL》用什么代替反补？它有驱逐。驱逐是一种走位技术，目的是使你的对手就不能发动最后一击（补刀）。假设游戏允许你反补，这时候的你大约也可以轻易地反补掉那些小兵。一定程度上，反补使你可以轻易地压制对手；你只要点击一下鼠标就可以消灭掉他们的小兵。所以，反补可能增加了机制难度（你必须多做一件事），但它其实并没有增加游戏的复杂度。为了说明这一点，我们再举一个极端的例子。假设有一款游戏叫作《LOL2》。《LOL2》与《LOL》相同，除了一件东西不一样。每一次你进攻时，你必须点击一下魔枢,才能再次攻击。那么，《LOL2》比《LOL》更困难吗？是的。但有没有更复杂？当然没有。这就好像说“篮球太简单了：篮球运动员应该在手臂上绑沙袋，这样才能增加难度！”确实，这条规则会让篮球的难度增加，但是并没有让它更复杂，所以也没有让掌握篮球运动或观看篮球比赛变得更有趣？是的，为什么？因为它没有迫使运动员做出任何决定。运动本身没有变化，但所有运动员都觉得很遭罪。

清晰度

Riot真正做得出色的地方就是清析度。首先，他们将不必要的部分控制到最少。这使玩家很容易看出来游戏的状况。第二，他们保持游戏操作简单。例如，与《DOTA》相比，游戏中没有惯性；角色可以随意改变方向。但最重要的是，他们始终以清析度为原则设计角色和物品。Zileas很详细地探讨了他的设计理念，我将把我看到的部分稍作总结。基本上，他的观点涵盖了三大原则：他从来没有明确地归纳出这三条原则，但在他的文章里解释得很清楚。

效果的清晰度

根据这条原则，如果游戏机制设计得好，那么玩家只要观看或玩上一两次就知道是怎么回事。玩家的能力存在细微差别，所以理解到的机制的深度可能有所不同，但都很容易看出基础的部分。例如，Nasus的大招是非常清楚的。你使用它时，你的周围会产生黑云漩涡（黑云使人产生危险的感觉，所以其他玩家会避开）。此外，角色会变大。事实上，体形大也普遍让人产生“这货不好杀”的印象。Renekton也会变大，Lulu的终极技能还是变大， Cho一直就很大；通过“大”，游戏告诉你“这个家伙真的非常难对付”，非常直观。当然，有时候小东西也很难消灭，但往往是因为特殊的效果。我可能有些离题了：

当然还有其他例子，但在我看来，“大”是最简单、最容易想到的一个。

效果的清晰度还可以用于刺激玩家学习游戏。带有明确效果的技能可以清楚地反映技能的成功或失败。例如，当你释放Jayce的Q时，你很容易看出来有没有命中。但附带其他效果的技能会更复杂。那些效果（通常）也很容易理解。例如Annie的Q/Swain的被动技/Cho的被动技（回复生命值和法力值）/Brand的被动技（晕眩）/Ezreal的Q（冷却时间减少）。这些都有加成效果，会引导玩家学习游戏。

概念的清晰度

根据概念的清析度原则，如果游戏机制有统一的概念，那就说明设计成功。以Soraka的W为例。它治疗目标对象并给予一定的护甲。为什么是给护甲而不是给魔抗？这可能有点像随机选择，但在分裂伤害的情况下，这种安排其实是非常有道理的。魔法伤害往往是突发的，而且是用在最前面。而物理伤害通常是连续的（主动攻击），并且贯穿整个战斗过程。因此，Soraka的W就非常合理了。如果有人需要治疗，可能是因为他们已经被法术攻击打到了。如是战斗仍在进行，他们受到的伤害类型基本上是物理伤害。显然有一些例外（Ryze/Cassiopeia/Kog’Maw），但一般来说这个原则是成立的。所以如果Soraka的W给的是魔抗，一定程度上会产生反效果：你可能会忍不住想“我应该在carry承受术士的法术技能以前用这个；这样他受的伤害就更少。”这显然是一种低劣的治疗设计，因为玩家可能不把治疗技能用于治疗。如果游戏中有这种技能，我希望它：

*影响多个目标对象（魔法师一般有很大的AOE）

*冷却时间短，但治疗程度低（这样你就不会只用它缓和魔法伤害）

*是一种延迟治疗

*增加生命值上限（像Lulu的大招），而不是单纯的治愈

再回到Nasus的大招。Nasus的大招不仅给他带来额外的生命加成，还能对周围的敌方英雄造成伤害。它有清楚的概念：不要离这个家伙太近，因为他的庞大和危险。在什么情况下，Nasus的大招会变成糟糕的技能？让AD Carry具有这个技能。这意味着：“看我，我很大很可怕，就要站在你的队伍中间了！”但是，有这个技能的AD Carry绝对绝对绝对不想站在敌人队伍中间。

尽管不一定这么明显。例如，荆棘之甲这个装备的概念就有一些非常糟糕的互动效果。当本体受到普通反攻击时，这件装备必须根据对方释放的伤害来决定反弹比例，而承受了多少伤害。否则，穿上更多装甲（这样你就不会死在AD英雄手下了）反而会让你反弹更少的伤害（那么为什么还穿荆棘之甲？）。然而，闪避却没有这种例外，这意味着高效的闪避使穿上荆棘之甲显得非常愚蠢。幸运的是，我们再也不用处理闪避了，毕竟荆棘之甲是不常买的吧？

虽然让玩家难以选择如何使用，但设计师不应该因此就不敢设计这样的游戏元素。

league_of_legends_2(from geek.com)

选择的清晰度

但是，当选择应该可用时，为了选择而选择不是你希望看到的结果。选择必须有意义，必须让玩家产生满足感。好的选择让玩家想“我要用这种技能做什么？”坏的选择让玩这家想“这个技能是做什么的？”例如， Amumu的绷带牵引可以用于发起攻击，还可以晕眩目标，玩家要选择用这种技能压制敌方的伤害主力还是保护自己的队伍。但是这种选择取决于玩家如何使用这个技能。但这种技能是做什么的（缠住目标，把目标拉到你身边)是很清楚的。我认为，这就是使Lulu的W令人不满的原因。当然，你可以用它辅助AP英雄而不是AD英雄，或带AP率的AD Carry如Tristana，但那种选择效果不好。当你可以压制敌方英雄2.5秒的时间时，为什么还选择给队友提速和提高AP？选择加速你的carry这样他就能逃跑了；或加速你的Amumu这样他就能发起战斗了；或加速你自己这样你就安全了，这些都是诱人的选择，但大多数时候，你只是想压制敌方英雄2.5秒，因为比较酷，对吧？

一致性

一致性也是游戏设计的重要部分。一致性是可以发挥创意的地方，因为玩家知道（大部分）游戏是按照一定方式运行的。如果没有这种预测性，创新就会很困难，因为无法预测。一致性的作用是，让你知道整个游戏是怎么一回事：如果你得到生命值加成、护甲和魔法抗性，你就很难被击杀；如果你的攻击加速、伤害加大且出暴击，那么你的输出就会比较稳定；如果你的AP增加，那么你用于伤害输出、治疗和防御的魔法就会很强。

肉盾的不一致性

但是，仍然有一些东西公然违抗以上惯例，只是并不常见。例如，Vayne就是一个很有意思的角色。她的银弩虽然没有使培养肉盾变得无意义（游戏邦注：她的绝大部分伤害仍然是物理的），但确实使肉盾不那么强了。毕竟，受到物理/魔法伤害时，你有抗性；打击对方抗性时，你有真实伤害或%穿透；对抗真实伤害或%穿透时，你的生命值长；伤害对方生命值时，你造成的%生命伤害。所以生命值长不管用，因为它有%生命伤害；所以抗性很高也不可行，因为它造成真实伤害。此外，攻击加速和翻滚闪避并不太引人注意。所以对于许多肉盾来说，似乎没有什么办法对抗Vayne。如果她的伤害完全来自银弩，那么就是这么一回事了。但甚至在围绕银弩做设计时，对抗一个血厚防高的角色，银弩只会造成25%伤害。在正常情况下，伤害其实更接近15-20%。她的其他伤害是物理伤害，一般的对抗方法是：护甲。

注：即使只有她的15-25%的伤害不能抵消，但仍然比游戏中的其他英雄多了15-25%的伤害。

乍一看，这似乎不太平衡。但是，Vayne必须经历所有AD Carry中最糟的对线阶段之一。所以，从某种意义上说，她已经因为这个优势付出代价了。这是一个良好的设计，在一方面给英雄优势，但又增加其他弱点。所以，Vayne当然（部分地）破坏了我们的一致性规则，即肉盾=难杀，但她付出的代价就是在游戏初期阶段很弱。她不是唯一一个破坏爆伤、输出和肉盾之间的明确分界的英雄。我为什么这么说？随着穿透率的改变（%达到固定以前），丢失接近0护甲或魔抗的英雄的可能性越来越大。

假设你玩的是AD英雄，你想要“最后的轻语”和“黑色切割者”（你甚至不必得到这些装备，如果你有同伴的话）。假设你确实有了黑色切割者，并且你达到固定穿透比率，英雄可以有多少护甲，并仍然被减少到0？只有85。算高吗？好吧，通过提高1.85倍的物理伤害，它减少你的有效生命值比物理伤害的46%。假设你的同伴是Wukong，他能用粉碎打击帮助你？现在所有低于120的都要减少到0。通过提高2.2倍的物理伤害，减少你的有效生命值比物理伤害的55%。所以如果你的护甲提高到100，会怎么样？在正常情况下，220的护甲会给你3.2倍于有效生命值比物理伤害。但减少和穿透会让你的护甲直线下降到29，同时给你1.29倍的有效生命值比物理伤害。这意味着你的100护甲只实现了29%的价值，你仍然要承受78%的物理伤害（虽然不是确切的伤害，但接近了）。这意味着任何可以利用这%生命值的物理伤害的英雄（如Vi）基本上可以造成%生命值上的真实伤害（更别说他们的物理伤害就是真实伤害）。对于任何认为这很疯狂的人，我想问你，返回去看上一次补丁，看看有多少黑色切割者，有多少队伍没有能输出强大魔法伤害的英雄。是的，黑色切割者的伤害本身不再叠加，但仍然可以叠加四次的护甲穿透。

从固定穿透变成%穿透的组合是一个糟糕的选择：意味着当遇到黑色切割者时，试图用护甲抵抗伤害的英雄就会很尴尬。但我想强调的是不受重视的另一个道具：兰德里的折磨。这是一件%生命值伤害的道具，它的危险在于其法术伤害无视魔法抗性。如果这还不足以让你害怕，那么想想道具持有者得到多少固定穿透率：足以让所有无魔抗道具的英雄的抗性减少到0。比如，大多数Carry有差不多48。很容易就能得到48的穿透伤害：只要使用兰德里的折磨（某评论者指出，其实你只需要幽魂面具）、穿透红药和法师之靴。

我不打算再比较穿透伤害（不要忘了现在抗性类道具更贵了），我要提出一般性的论断：在旧系统中，伤害施放者必须选择：要么用固定穿透伤害针对脆皮英雄并促使抗性叠加，要么以%穿透伤害来打击肉盾，并惩罚抗性叠加。现在他们两个选择都做到了，无论是强抗性英雄还是弱抗性英雄的抗性都减少到近乎为0。此外，AP英雄确实能做到%生命值伤害，因为有了兰德里的折磨和冥火之拥的%命值伤害。这就形成了一个系统，你不必做出牺牲就能获得%生命值的真实伤害；确实很可怕。

伤害和肉盾之间的正反馈循环

另一个有趣的设计是，某些英雄模糊了伤害输出和肉盾之间的界线。Skarner的被动技能在他发动普攻击时会减少所有技能的冷却时间。所以攻击速度越快，他能使用的技能就越多。听起来很不错啊。他的技能之一是护盾。所以，他能输出持续性伤害，又是肉盾。因为这个问题，必须通过损失他的Q来减少他的魔法值，之后还要延长他的护盾冷却时间。但他的优势让他付出代价：他很容易被放风筝。Vlad是另一个例子：他的生命值越多，AP就越大，反之亦然。这导致他集生命值、爆伤和持续性输出优点于一体（因为较低的冷却时间）。他以法术为代价换取这种优势，但有时候他仍然是一个相当难对付的英雄。Riven得到与AD对应的护盾，Irelia的攻击不仅猛，而且速度快（甚至不需要任何吸血道具），现在Rengar也能一边搞输出一边当医生了。

我是在说挑战了我们的肉盾与伤害输出并存概念的英雄不应该存在吗？我希望你不要得出这样的结论。一方面，Skarner是我最喜爱的英雄之一（尽管我讨厌Vladimir）。他们只是实现了挑战的平衡。

Riot试图拓展英雄的能力范围，他们确实（很大程度上）成功了。我只希望他们从拿冥火之拥的Eve或带黑色切割者的4个AD队伍中吸取教训，避免使用过度的穿透伤害和%命值伤害的叠加，否则抗性甚至连肉盾都变得没有意义了。另外，我希望他们从过去的错误中学习，不要用攻击属性替代防御属性或相反。这两种东西都说不上很糟糕，但越来越普遍了，与特定劣势的关联也越来越小了，因此也越来越难处理了。

基本上，他们消解了游戏的一致性之一：爆伤、持续性输出和肉盾的明确分类。前一个问题违反了“肉盾难杀”这条准则，后一个问题模糊了三者之间的界线。再者，如果适度使用，这些都不是设计问题；不过一旦滥用就非常糟糕了。（本文为游戏邦/gamerboom.com编译，拒绝任何不保留版权的转载，如需转载请联系：游戏邦）

How Riot excels at game design (and the pitfall I hope they avoid)

By GentlemanGustaf

Whenever somebody asks me for a game recommendation, I’m basically thinking two things:

Has this person played a lot of games before?

Is this person into solo, collaborative, or competitive games?

And regardless of the answers, I’m going to ask the real question ‘does the game need to have a story/narrative?’ and if they say no, I’m going to recommend LoL, and if they say yes, I might anyway. Is this because I want more people to play with me? Honestly, I don’t make a habit of playing with friends. For me, LoL is a competitive game, and while I will play a few ‘silly/fun’ games a week with friends, most of my games are in ranked queue. Is this because of referral points? I have yet to officially refer any friends. Is this because I don’t know games well? I’ve played scores of computer and video games, probably near a hundred if we count individual games in a series. Is this because I’m lazy? No. Maybe. But ultimately, I think it’s because Riot has the best gaming design philosophy of any company at the moment. They’ve secured a foothold for League of Legends as the premiere e-sport, beating out entrenched industries like fighting games, Halo, Call of Duty, and even Starcraft and Starcraft 2. League of Legends treats their game like a sport, not a game. This has affected their design in many positive ways.

Gameplay

What makes Riot’s approach to gameplay so special? It is their dedication to transparency. There is plenty of complexity in the game, but that complexity is related to choice. And those choices may be hard, but their difficulty is not at all related to clarity.

Complexity vs Difficulty

What makes the complexity so apparent? Well, Riot has designed the game to be accessible. In general, e-sports have thrived on complexity. This is presumably because the earliest computer games were incredibly simple. Short of people who really enjoy the game Pong, nobody’s going to watch a tournament of pong. With many early e-sports people soon discovered a number of exploits that made the game more complicated. Animation canceling, denying, muta-stacking, these are all strategies that grew out of a game engine that was designed not knowing how it would be exploited. As such, there has grown to be a mindset that a game which has complicating elements in it, a game which fights the user’s very attempts to control it, is a good e-sport. Connoisseurs praise the difficulty of microing starcraft units, HoN and DotA 2 players critique LoL for having less complexity due to the lack of denying,

So let’s go with that example, since this is ultimately about LoL. What does LoL have in place of denying? It has zoning. Zoning is the skill of positioning yourself such that your opponent cannot get to the last-hit. Presumably, if you are in such a position, you could easily also just deny those minions. In a sense, denying allows you an easy way to zone people; you just kill their minions with a click. So, sure, denying might add mechanical difficulty (one more thing you have to do), but it doesn’t really increase the complexity of the game. To illustrate this, let’s take an extreme example. Imagine a game called LoL2. LoL2 is just like LoL except for one thing. Every time you attack, you have to click on the nexus before you can attack again. Is LoL2 more difficult than LoL? Well, yes. But is it more complex? Not really. It’s like saying ‘basketball’s too easy: basketball players should have to strap weights to their arms to make it harder!’ Sure, that rule would make basketball harder, but would it make it any more complex, thus making it more fun to master or more interesting to watch? No. Why? It doesn’t actually force any decisions. The game itself is the same, in nature; everybody has just suffered a handicap.

Clarity

Where Riot really shines is in clarity. First of all, they keep needless particles to a minimum. This makes it easy to see what’s happening. Second, they keep controls simple. For example, as compared with DotA, there is no inertia; characters can change direction at will. But most importantly of all, they design characters and items with the principle of clarity of strategy in mind. Zileas discusses his design principles at length here, but I will try to sum them up as I see them. Essentially, he covers what I see as three main principles: he never explicitly states these three principles, but they are very evident in his post.

Clarity of Effect

According to the principle of Clarity of Effect, a game mechanic is well created if it is relatively easy to see what it does just from observing it, or using it once or twice. The nuances of the ability may take more in depth exposure, but the basics come across very simply. For example, Nasus’ ultimate is very clear. You use it and dark clouds swirl around you (nobody likes dark clouds, so they should stay out of them). On top of that, you get huge. In fact, size is part of universal code for ‘hard to kill’; Renekton gets huge, Lulu’s ult makes you huge, Cho is always huge; huge is the way a game tells you ‘this guy is fuck-all hard to kill’, and it paints a clear picture. Sure, sometimes small things are hard to kill, but it’s typically for a specific effect. If I may digress for a moment:

Who played Final Fantasy VIII? How did you feel when you encountered your first Tonberry? You probably thought ‘look at this little guy, I’ll kick his ass’. And then you pounded a bunch of damage onto him, and he just stood  wthere. And you put more damage onto him, and finally, he took a few steps towards you. And if he ever got to you, he poked you with his little knife and you died and you just thought ‘HOLY CRAP WHAT WAS THAT?’ And it made the monster memorable because it defied your expectations. But of course, speaking of not defying your expectations, how did you know Tonberry King was the boss Tonberry? Because he was bigger than all of the other Tonberries.

There are, of course, other examples, but the size one is the easiest one, in my opinion, to get right off the bat.

Clarity of Effect can also be used to motivate learning of the game. an ability with a clear effect also indicates clearly success or failure of the skill. When you shoot Jayce’s Q, for example, when it hits, you feel good, and when it misses, you feel bad. But other abilities have other effects that are more complicated. These effects are also (generally) clearly understood. Examples include Annie’s Q/Swain’s passive/Cho’s passive (yay, I killed a minion, sustain back), Brand’s passive (yay, I got the stun!), Ezreal’s Q (yay, cooldowns decreased!). These all give clear bonus effects for intended use, and, as such, guide learning the game.

Clarity of concept

According to the principle of Clarity of Concept, a game mechanic is well created if it has some unifying concept. As an example, let us take Soraka’s W. It heals a target for a moderate amount and gives them a large amount of armor. Why armor and not MR? This may seem like an arbitrary choice, but it actually makes a lot of sense in the context of the way damage is split. Magic damage typically comes in burst, and typically comes up front. Physical damage, on the other hand, is naturally more continuous (auto attacks), and comes throughout the fight. As such, Soraka’s W makes perfect sense. If somebody needs healing, it’s probably because they’ve taken burst, and need healing. And if the fight continues, the damage they will take will most likely be physical damage focused. There are obviously exceptions (Ryze/Cassiopeia/Kog’Maw), but the general principle holds. So if Soraka’s W instead gave MR, it would feel somewhat counter-intuitive: you might be tempted to think ‘I should use this on my carry BEFORE he takes the mage’s burst; that way he’ll take less damage’. This would obviously be poor design for a heal, because you might decide to use the heal not for the heal. If such an ability were to be made, I would expect it to either

Affect multiple targets (mages have a good amount of AoE typically)

Have a low cooldown but be a lower heal (that way you wouldn’t be penalized as much for using it to mitigate burst

Be a delayed heal

Grant additional health (like Lulu’s ult) as opposed to healing

Let’s go back to Nasus’ ult. It makes you hard to kill, and it makes you hurt people around you. It has a clear concept: don’t stand near this guy, because he’s huge and trying to ruin your day. What would make Nasus’ ult not a good ability? Putting it on an AD Carry. All of a sudden, you’ve got an ability that says ‘look at me, I’m big and scary and going to stand in the middle of your team!’ but you’ve got this ability on a champion who never, ever, ever wants to stand in the middle of the enemy team.

It doesn’t have to be that obvious, though. For example, Thornmail had some awkward conceptual interactions. It was necessary for thornmail to reflect based off of incoming, not received damage. Otherwise, building more armor (so you wouldn’t die to AD champions) would cause you to reflect less damage (so why get the thornmail?). However, no such exception was made for dodge, which meant that building dodge effectively made building Thornmail silly. Luckily, we don’t have to deal with dodge anymore, and how often is thornmail bought, anyway? (basically never except in normals and lower Elo games)

This shouldn’t scare game designers away from designing game elements which force hard choices as to HOW they are used.

Clarity of Choice

But while choices should be available, throwing in choice for the sake of choice isn’t what you want in a game. Choices need to be meaningful to be satisfying. Good choices stem from letting the player say ‘what am I going to use this ability for?’ not ‘what does this ability do?’ For example, Amumu’s Bandage Toss can be used to initiate a fight, and it can also be used to stun a key target, whether that shuts down the enemy team’s damage dealer or protects your own. But this is a matter of choice as to how to use the skill. But what the skill is for (locking a target down briefly and letting you stick to them) is clear. This is, I feel, the heart of what makes Lulu’s W unsatisfying. Sure, you could use it to run support with an AP champion instead of an AD champion, or an AD Carry with AP ratios like Tristana, but that choice never feels good. Why give a teammate some movement speed and AP when you could neutralize an enemy champion for 2.5 seconds? It might seem appealing to have the choice of speeding up your carry so he can escape, or speeding up your Amumu so he can initiate, or speeding up yourself so you can get to safety, but most of the time, you’re just going to want to CC an enemy champion for 2.5 seconds because that’s pretty good, right?

Consistency

Consistency is also an important part of game design. Consistency is what allows for innovation, because the player knows that (for the most part) the game will work in a specific way. Without that predictive ablity, innovation becomes hard, because prediction becomes hard. Consistency tells you about the game as a whole: if you build health and armor and magic resistance, you get harder to kill, if you get attack speed and attack damage and crit you get consistent dps, and if you get AP, you get better spells for burst, heals, and shields.

Inconsistency of Tankiness

There should be things which fly in the face of these conventions, so long as they are not too ubiquitous. For example, Vayne is an interesting concept. Her silver bolts don’t make building tanky pointless (the majority of her damage is still physical), but it does make it less strong. After all, against heavy physical/magical damage, you get resistances, against resistances, you get true damage or % Pen, against true damage or % Pen, you get health, and against health, you get % health damage. So health isn’t really the right choice, since it’s % health damage, resistances aren’t the right choice, since it’s true damage. On top of that, attack speed slows can feel underwhelming with tumble in the picture. So to many tanks, it may feel like there is simply no way to itemize against Vayne. And if her damage came entirely from silver bolts, that would be true. But even in a build designed around silver bolts, against a champion with high health and armor, silver bolts only does about 25% of your damage. Under normal circumstances, it does closer to 15-20%. The rest of her damage is physical damage, and itemized against the normal way: armor.

But Gentleman Gustaf, even if only 15-25% of her damage can’t be itemized against, that’s still 15-25% more than every other champion in the game!

In a vacuum, this would be problematic. But Vayne has one of the worst laning phases of any AD Carry. So, in a sense, she pays for this advantage somewhere else. And this is good design, giving a champion something strong one place, but making up for it with a weakness. So, sure, Vayne gets to break (partially) our rule of consistency that tankier = harder to kill, but she pays for it with a weaker early game. And she’s not the only one who breaks our ideas of the dichotomy between burst, dps, and tankiness. Why do I say this? With the changes to penetration (% before flat), it becomes increasingly possible to drop targets to close to 0 Armor or MR.

Let’s say, for example, you’re an AD champion who wants Last Whisper and Black Cleaver (you don’t even have to be the one to get it, if you have a cooperative teammate). Assuming you do have the Black Cleaver, and you take flat Pen Marks, how much Armor can somebody have and still be reduced to 0 Armor? Only 85. Is that a big deal? Well, it reduces your effective health vs physical damage by 46%, by allowing physical damage to do 1.85 times as much damage. Let’s say your cooperative teammate is a Wukong, and uses Crushing Blow to help you out? Now anybody below 120 is reduced to 0. This reduces your effective health vs physical damage by 55%, by allowing physical damage to do about 2.2 times as much damage. So what happens if you pick up 100 armor to counter this? Under normal circumstances, 220 armor would give you 3.2 times your effective health vs physical damage. But reduction and penetration will bring you all the way down to 29 armor, giving you 1.29 times your effective health vs physical damage. This means your 100 armor purchase gets only 29% of its value, and you are still taking 78% of all physical damage (it’s not true damage, but it’s pretty close). This means that anybody who can take advantage of this with % health physical damage (Vi, for example) can essentially be doing % health true damage (not to mention having all of their physical damage be true damage). To anybody who thinks this is crazy, I would ask you to go back in time to the last patch and watch some games to see how many black cleavers you see, and how many teams you see without a champion who dealt significant magic damage. Yes, Black Cleavers no longer stack with themselves, but they still stacks with the many other sources of armor pen.

It’s been covered to death that the combination of flat pen turning into % pen is an awkward choice: it means people who try to counter brutalizer early with armor can expect to get that countered with black cleaver. But I’d like to focus on an underappreciated item: Liandry’s Torment. Liandry’s Torment is a % health item that starts itself on the path to true damage: it comes with free Magic Penetration. If that thought alone doesn’t scare you, just think about how much flat Pen is available to casters: enough to reduce anybody without MR items to 0 and then some. Most Carries will have about 48, for example. It’s easy to get 48 Pen: just pick up Liandry’s Torment (well, actually, you only need Haunting Guise, as pointed out by a commenter), Pen reds, and Sorc Boots.

Rather than go into more math about Penetration (oh, and don’t forget that resistances are more expensive now), I’ll give a general statement: in the old system, damage dealers had to choose: pick up flat pen to target squishy champions and force resistance stacking, or pick up % Pen to hurt tanks and punish resistance stacking. Now they can do both, forcing both heavy resistance and light resistance champions to almost 0 resistances. On top of that, % health damage is readily available to AP champions, from both Liandry’s Torment and Deathfire Grasp. This creates a system where you don’t have to make sacrifices (like Vayne does) to have access to spammable % health true damage; a scary situation indeed.

Positive feedback loops between damage and tankiness

Also interesting are champions who blur the lines between damage and tankiness. Skarner’s passive reduces all of his cooldowns whenever he attacks. So the more attack speed, the more abilities he gets to use. Sounds good so far. One of his abilities is a shield. Now, all of a sudden, building sustained damage also gives him sustained tankiness. This was enough of a problem that it required gutting his mana through the cost of his Q, as well as later nerfing the cooldown on his shield. But he pays for this advantage with a disadvantage: he gets kited easily. Vlad is another example: the more health he gets, the more AP he gets, and vice versa. This lets him end up with health, burst, and sustained damage (due to reasonably low cooldowns). He pays for it with health costs on all of his spells, but he has remained a very difficult champion to balance for some time. Riven gets a shield that scales with AD, Irelia gets sustain with Attack speed (without even building any lifesteal), and now Rengar can heal based off of AD.

Am I saying that interesting champions who challenge our concepts of tankiness and damage should not exist? I hope you do not draw that conclusion. For one, Skarner is one of my favorite champions (although I hate Vladimir). They just pose interesting balance challenges.

Riot has tried to push the bounds of what a champion can do, and they have (largely) succeeded. I just hope they learn from Eve with Deathfire Grasp or 4 AD teams with Black Cleaver, and avoid allowing overly aggressive penetration stacking in combination with % health damage, or there will stop being reasons to take resistances, and maybe tanks, at all. And as well, I hope they learn from past mistakes with offensive stats feeding defensive stats, or vice versa. Neither of these two things is bad, but the more common they become, and the less linked  to specific weaknesses they become, the harder they become to deal with.

Essentially, they take away one of the consistencies of the game: the trichotomy of burst, sustained damage, and tankiness. The former problem violates the maxim of ‘tanky things are hard to kill’, while the latter problem blurs the separation between the three. Again, neither of these is a design problem when used sparingly, but can become very scary in excess.(source:reignofgaming)