作者：Greg Johnson

我能做的只有射击？

在设计一款互动故事游戏时需要面临的一大挑战便是玩家代理。之前我们说过游戏是关于玩家“做”什么。当我们谈到玩家在一个不断展开的故事中所扮演的角色时，这一问题便变成了“我该如何影响故事的发展？”

如今，大多数游戏所设定的基本标准便包括玩家移动，射击以及拳脚相向。当你停下来并开始思考自己该如何只通过这些行动去影响故事时，这便变成了“我是否该使用暴力”或者“我该杀谁”以及“我该救谁”的选择。你可以基于有限代理去创造更高级别的选择，即引出“我该与谁合作”或“我该选择执行或不执行怎样的任务？”而当你手上没有枪支时，你能够影响故事世界的方法便会非常有限。

我们必须承认的是，除了射击外游戏中还存在其它类型的代理。例如潜行和隐藏便是一种行走路径。也有许多游戏将这种选择路径作为实现目标的方式（游戏邦注：如《神偷》，《刺客信条》，《合金装备》）。而最常见的其它类型的代理（玩家行动）当属环境行动了。从根本上来看，你将接近游戏世界中的某些对象，然后你将按压按键去打开一扇门，喝一杯水，浏览一个标识，或者捡起一把钥匙等等。有时候环境行动会出现在与NPC的互动中，有时候玩家也将从有限的行动集中获得一个环境行动选择，如1）亲切地对待一只小猫，2）踢猫。

毫无疑问我们可以找到一些已经使用过玩家代理的创意故事游戏。而在任何故事游戏中，“我可以在何时做出怎样的行动”这一问题都是核心问题。也许有人会认为如果我们的目标是模拟真正的世界，我们便需要尽可能地提供给玩家使用更多行动的机会。实际上，关于为什么这么做是不可行且不合适的存在一些原因。以下便是我们需要考虑到的内容：

1.在角色的控制方面我们受限于输入设备。

2.我们希望能够保持游戏控制足够简单，容易，不会让玩家感到郁闷。

3.我们希望玩家行动足够清楚且够特别，因为游戏全部都是关于行动。

4.许多玩家行动意味着需要更多玩家和NPC动画资产，而这却是不现实的，除非你能够从程序上创建这些内容，而在这里你也将遇到各种不同的问题。

5.基于各种玩家行动，我们也需要处理各种不同的反应以及大量的故事结果（可能还有分支）。

6.所有玩家行动都需要对“游戏系统”产生影响，例如你能否在玩家实现目标的过程中帮助他们或阻碍他们。许多行动都将创造出更加复杂的游戏系统。

7.硬核主机游戏用户总是喜欢更加强烈的内容，如果我们是面向他们创造游戏并希望将游戏卖给他们，我们就要用行动说话。

也许除了这些内容外我们还有其它需要考虑的内容。从根本上来看这并不是一个简单的问题。而解决这一问题的方法并非提供给玩家更多代理，而是应该采取我们在现实生活中所采取的方式。也就是我们应该提供给玩家合适的代理，并在适当时候让他们能够面对适当的选择。如果你在玩家经历游戏的过程中只呈现给他们少量行动，游戏将大大限制玩家影响故事，特别是其他角色的方式。环境选择所面对的问题在于玩家并不习惯故事中的有限集，因为它总是不断改变着。这意味着玩家将需要留意自己的选择是任意的还是有限的。每当玩家希望自己能在游戏中做些什么而又不能这么做时，他们的沉浸感便会被打破。相反地，当游戏提供给玩家固定的行动集时，他们便会愿意接受这里的局限性并且一段时间后便不会再纠结于此。

有些游戏尝试了所谓的“直接控制”，即玩家可以通过直接移动去控制角色（例如当你移动鼠标或控制器时，你的手臂便会移动）。直接控制似乎能够让玩家更直接地“享有”对于行动的所有权并让他们觉得自己真正参与其中，但同时这也具有许多问题与挑战。例如：你该如何直观地在控制器上呈现复杂的行动，或者你该添加怎样基于直接控制的行动到游戏体验中，以及怎样的行动会让玩家感到厌烦？如果你在错误的地方或基于错误的方法使用了直接控制，这便会破坏游戏的沉浸感并让玩家感受到游戏控制。随着VR以及全新红外线传感器或像戒指和触感手套等输入设备的出现，这些输入问题便得到缓解。也许关于直接控制最有趣的挑战便是NPC角色对于玩家意图或者所传达的意义的理解。这便需要一些非常复杂的AI。即使NPC不能“理解”，直接控制也会很有趣。有些游戏甚至使用直接控制去影响多人游戏设置（如《小小大星球》，《MakeOurWay》）。

关于这一主题的最后一点便是射击本身，或者更准确地说是“杀戮”。我们并不需要深入讨论有关道德和电子游戏的热门话题，但必须注意的是，从现实的游戏玩法角度来看，创造一个不包含杀戮的游戏活动是非常具有挑战性的。当一款游戏通过实体行动在讲述一个故事时，总是存在一些比生存更容易传达或更多动态性的内容。我们很容易通过有限的玩家行动去传达杀戮，并且很容易创造基于技能的杀戮。但即便如此还是有许多游戏找到了其它玩家行动的创意解决方法，我们该为他们的创造性鼓掌。但不幸的是，许多拥有有趣的互动故事的游戏却是基于非常黑暗的主题以及非常血腥的内容。这也是许多潜在玩家所难以接受的。

单词很古怪不是吗？

也许现在去定义这一词有点太晚了，因为我们已经使用了许多次了，但或许我们还是应该花些时间去好好定义我们所谓的“游戏结构”。同时我们也需要定义另一个频繁出现的词，游戏机制。

单词真的很有趣。我们可以将其当成实物任意摆弄，我们可以假设我们所想的有关它们的东西也是其他人所想的那样，因为不管怎样猫就是猫，不是吗？而游戏也就是游戏。在我们的日常生活中，我们很难在第二次就猜出所有东西，但这里所存在的秘密是我们的大脑总是不断地跟我们开玩笑，所以我们才可以每天正常运转。如果你不再去考虑它，你便会发现我们所使用的每个单词都拥有一个抽象结构；这也是居住于这个宇宙中的我们所具有的心智模式中的一部分。不管怎样你所拥有的每个想法，每种直觉以及所有理解都只是不断运转的心智模式机器而已。基于我们大脑中的电线的运转速度，我们的心智模式将根据我们的记忆和情感进行一些随机且具有关联的连接，并尝试着将我们所看到，听到与读到的的事物带进一个更大的背景中从而让其变得有意义。当我们将我们所听到或读到的单词转变成“意义”时，我们的心智模式将考虑像环境和意图等元素。而所有的这一切都是发生在一张巨大的神经网中，并以极高的速度连接着我们的大脑，而我们只能真正意识到那些最终跳出来的想法。

这里的重点是，单词只不过是我们添加到拥有各种“模糊”程度的概念上的标签，并且我们还会将其与其它概念相连接。再加上你自己的心智模式和其他人的心智模式具有很大的区别，所以这真的是我们希望能够好好处理的一部分。让我们以一个较为明确的单词为例，如“cat”。这是指狮子还是猫？当你的朋友说自己要养猫时，你应该就不会有这种疑问了。让我们再看看“game-play”或“art”。当我们说“game play”或“video-game”时，我们认为自己知道这意味着什么，但是这些单词却也拥有许多让人困惑的内容，只是我们假设自己的理解便是其真正含义。让我们再看看“art”这一次。当游戏玩家和非游戏玩家花了好几个小时去争论游戏是否是“art”时我真的非常惊讶，这就好像他们所理解是同样的内容。而这点真的很有趣。

所以为什么我们要绕到语言原理中？这只是为了指出游戏结构这一词只是一个任意结构，就像游戏机制一样。

就像学术一样，似乎我们需要花些时间去定义我们的词语，这里存在非常有帮助的用途。了解一个机制是什么并清楚结构是什么，或者主题，故事，奖励系统等等，你便能够更明确地去考虑它们。模糊的思考总是会浪费你更多时间。所以让我们更清楚地进行定义。

“游戏机制”？

“游戏机制”指的是玩家在游戏中所做的事，并伴随着一些让这些事变的更具挑战性，且是有趣的挑战的元素。之前我们曾经讨论过一款成功游戏的目标是如何强化玩家的行动和选择。而在这里玩家所做的事便是游戏的核心，即玩家期待的很大组成部分。考虑游戏设计的一种最有效的方法便是询问一个简单的问题：“大多数情况下玩家会怎么做”。但是让人惊讶的是，即使是最优秀的设计师也经常会忘记问自己这一问题。

当你在设想一个游戏机制或分解现有游戏的游戏机制时，你可以先考虑类别。在现有的游戏机制中可能只存在40或50种类别。简单来说，这些便是玩家真正会“做”的事。以下便是其中的一部分类别：

徒手打斗

射击

赛车

解决谜题

收集

跟上音乐节奏

锻造

操控对话树

闪避与跳跃

隐藏与潜行

攀爬与跑酷

飞翔

简单的快速按键回应

建造和创造

一些特殊的机制：

使用弹弓通过收缩与释放发射小鸟。根据小鸟的重量以及弹弓拉缩的距离小鸟将呈弧形移动。玩家将尝试着敲击结构以撞击小猪。

在敌人攻击阶段快速移动以避免被袭击，然后在敌人休息的时候准确打压敌人的薄弱区域。

尝试着在音符消失前通过按压准确按键去匹配音符。

一些简单的机制：

随着地域的显现朝前移动并尝试着寻找前进的道路。

在不同对象间移动并躲避敌人的探射灯。

在时间范围内按压按键。

一些复杂的机制：

驾驶坦克的同时转动炮口并使用伸缩功能去射击目标。针对目标选择适当的盾牌，并在盔甲无用时从内部或后方攻击敌人，同时利用地形作掩护。

需要注意的是这些行动案例都是将挑战描述作为行动的一部分。“画张图”是一种活动，这并不是真的游戏机制，而“在5秒钟内画张图”则更像是一个游戏机制。我们未提到但却应该作为游戏机制定义的一部分的内容是，评估玩家的表现以及对于自己表现的反馈。但是“在5秒钟内画张图”还不是一个真正的有机制，因为这是一个很难判断的内容。但这并不是说主观的“创造性”活动在游戏中没有立足地—-它们当然有，只是它们本身并非游戏机制而已。

另外一个需要注意的是：我们经常听到的“游戏玩法”这一词。这通常指代的是一款游戏中出现的机制集合，即伴随着玩家对于这些内容的“乐趣”的期待。玩家和游戏评论者经常将游戏玩法当成游戏中“最重要的内容”。有时候他们可能会困惑游戏玩法到底指什么，但如果你能够提供给他们明确的定义，他们便会豁然开朗。

“游戏结构”？

既然我们已经描述了什么是游戏机制，我们便能够将其与游戏结构区分开来。就像我们之前提到的，游戏可以拥有许多游戏机制，但却只能拥有一个游戏结构。

为了着眼于游戏结构，我们必须着眼于游戏整体。在某种意义上，游戏结构便是更高级的游戏形状。它将定义玩家体验流，即他们将前往哪里，他们将基于怎样的顺序做什么？因为大多数游戏都包含玩家在不同场所间的移动，所以游戏结构其实就像游戏世界的地图一般，并带有能够用于各种场所的设计说明。也有些游戏比这个更复杂，通常情况下如果你能够绘制出游戏结构的话事情便会简单许多。通常情况下在玩家访问全新游戏部分前他们需要满足某些特定条件。这些条件可能非常简单，如到达游戏中的某个场所，收集特定的资源，与NPC角色交朋友，达到特定级别，或者获取一个道具等等。所以你如果能在图表中绘制出这些内容便会很有帮助。除此之外这也能让设计师更清楚地了解玩家在游戏中的选择点。这就像在为一部电影编写提纲。

故事游戏的游戏结构通常包含一些故事章节概念。即故事环境发生改变的游戏阶段。有时候基于玩家所处的章节他们可能会出现新的目标或拥有新的能力。在一款线性游戏中，要映射出这些与故事相关的改变非常简单。而在一款包含自然发生的故事的游戏中，绘制结构便会较复杂。在这里这便是关于明确潜在故事线并罗列出玩家需要满足并且能够改变与故事相关的角色状态和世界状态的条件。

（本文为游戏邦/gamerboom.com编译，拒绝任何不保留版权的转发，如需转载请联系：游戏邦）

Designing Interactive Story (PART FIVE)

by Greg Johnson

ALL I CAN DO IS SHOOT?

One of the big challenges to face in designing an Interactive Story game is player agency. Earlier we spoke about games being all about what players “do”. When we’re talking about the role a player plays in an unfolding story, the question becomes “what can I do to affect the story?”

The basic standard set by most of today’s games involves players moving, shooting, and possibly punching and kicking. When you stop and think about how you can impact a story based on having access to only these actions, it pretty much comes down to choices of “do I use violence or do I not” , or perhaps “who do I kill?” and “who do I save”. You can certainly build in higher level choices with this limited agency that result in “who do I align myself with?” or “what quests do I choose to do or not do?” Still there is a pretty limited set of ways one can affect the world when there is a gun strapped irrevocably to one’s hand.

Admittedly there are other types of agency in games besides simply shooting. Sneaking and hiding is one example of an agency that isn’t too far off this well trodden path. Many games will make this an optional path to achieving goals. (Thief, Assassin’s Creed, Metal Gear Solid). Probably the most common other type of agency (player action) is contextual action. Essentially, you come up to some object in the world and you can press your button to open a door, drink from a cup, read a sign, or pick up the key, etc… Contextual actions are sometime used for interactions with NPCs as well, and sometimes players are given a choice of contextual options from a limited set of actions… (a) pet the cat, (b) kick the cat.

We are undoubtedly shortchanging a few particularly innovative story-games out there that have experimented with player agency. This question of “what actions can I do, and when” is a core question in any story game. One might think that if our goal is to mimic the real world, we always want to try and give players access to as many actions as possible. Actually, there are quite a few reasons why this isn’t feasible, and why sometimes it’s not even desirable. Here are a few ‘example’ considerations:

1.We’re limited by our input devices in terms of controlling our avatars

2.We want to keep our game controls simple and accessible and not bog things down

3.We want player actions to be clear and unambiguous, and generally physical, since games are all about action

4.A wider array of player actions means a lot more animation assets for players and NPCs which gets impractical unless you happen to be building procedurally, in which case you have a different set of issues.

5.With a wide array of player actions, we need to deal with a wide array of responses and a larger set of story consequences (possibly branches).

6.All player actions need to have an effect on the ‘game-system’ i.e. do they help or hinder the player in achieving their goals. Many actions make for a much more complex game-system.

7.The core console gaming audience loves to kick butt – so if we’re building a game for them and we want it to sell…. Just sayin’.

There are probably other considerations but these are a few. Basically, it’s not a simple problem. The answer may not be in giving players a lot of agency, the way we have in real life. More likely, it lies in giving them the right agency…. access to good choices at the right time. The problem with having a small set of fixed actions players can do throughout the game is it severely limits how players can impact the story, especially other characters. The problem with contextual choices is that players can’t get used to the limited set as part of the fiction, because it’s always changing. This means that players are continually aware that their choices are arbitrary and limited. Every time the player wishes they could do something in a game and can’t, it breaks them out of the immersive fantasy. In contrast, when players have a fixed set of actions, they tend to adapt to the limitations and stop thinking about them after awhile.

Some games have experimented with something called “direct control” where players can essentially puppet their avatar with direct movement. (i.e., as you move your mouse or controller your arm moves). Direct control seems to offer promise in terms of connecting players more directly to an “ownership” of their actions and allowing them to feel more directly involved but it poses a number of issues and challenges. These are things like: how do you intuitively map complex actions onto a controller, or what actions add to the experience with direct control, and what actions simply become annoying? Using direct control in the wrong places, or in the wrong ways, can actually work against your immersion, and make players too aware of the game controls. (Until Dawn, Octodad, Growing Home, Surgeon Simulator) With the advent of VR, and new infra-red sensing devices, or input devices like rings and haptic gloves, these input-mapping concerns may start to diminish. Perhaps the most intriguing challenge having to do with direct control has to do with NPC characters interpreting player intent, or expressive meaning. This requires some fairly sophisticated AI. Even without being “understood” by NPCs in the game, direct control can be a lot of fun. Some games have used it to great effect in multi-player settings. (Little Big Planet, MakeOurWay)

A last word on this topic has to do with the shooting itself, or more generally put, the “killing”. Without getting too deeply into the hot topic of ethics and video-games it’s worth noting that it can be a challenge, from a practical game-play perspective, to come up with primary player activities that don’t involve killing. When one is telling a story through physical action, there are few things easier to communicate, or more dramatic, than simple survival. Killing is clear and easy to represent with a very limited set of player actions, and easy to make skill-based. That said, there are a ton of great games out there that have found other creative solutions for player action, and their ingenuity should be recognized and applauded. Unfortunately, a large percentage of the games doing interesting things with Interactive Story have rather dark themes, and bloody subject matter. This is quite a turn-off to a large population of potential players. (I find I can’t even get through many of these otherwise amazing games).

WORDS ARE WEIRD, AREN’T THEY?

It may be a little late to be defining this term, since we’ve been using it left and right already, but perhaps we should take a moment and clearly define what we mean by ‘game structure’. For that matter, while we’re at it lets define another term we’re tossing about, play-mechanic (or game-mechanic).

Words are funny things aren’t they? We toss them around as if they were real solid things, assuming that what we think we mean by them is what other people think as well, because after all, a cat is a cat, isn’t it? And a game is a game. In our daily lives it’s generally not productive for us to go around second guessing everything… but the truth behind the curtain is that our brains are playing a continual trick on us, so that we can function on a daily basis. If you stop to think about it, you’ll see that every word we use is an arbitrary construct; part of our mental model of the Universe we live in. After all, every thought you have, every perception, and every bit of understanding is nothing more than the machinery of that mental model working away. At the speed of the electrical wiring in our brains our mental models make causal and associative connections, layering in memories, and emotions, and trying to fit what we see and hear and read, into a bigger picture that makes sense. When we translate the words we hear or read into ‘meaning’ (reading this sentence, for example), our model is also taking things like context, and intention into account. All of this happens within the vast neural network of links and associations in our brainputers at super high speeds, with us only really aware of the thought that pops out at the end, as if by magic.

The point here is that words are nothing more than labels we slap onto concepts that have varying degrees of ‘fuzziness’, by virtue of this myriad of connections to other concepts. Add to this fuzziness the endless differences between your own mental models, and all those other mental models floating around in all those other brains, and it really is a wonder we manage to communicate at all. Take a solid, unambiguous word like “cat”, for example. Is a lion a cat? Well, sort of, though it’s probably not what your friend meant when they said they were going to adopt a cat. Now consider words like “game-play” or “art”. We think we know what we mean when we say “game play” or even “video-game” but these words have an awful lot of fuzz around their boundaries, yet we sling them about left and right assuming our meaning is getting across. And let’s not even get started on disastrously fuzzy words like “art”. I’d be surprised if two people’s definitions of this concept line up, yet gamers and non-gamers spend hours debating the question of whether games are “art”, as if they all meant the same thing…. Ok, so it is entertaining.

So, why this little detour into the philosophy of language? Well it’s really just to point out that the term Game Structure is just an arbitrary construct, as is the term Game Mechanic. (Come to think of it, I suppose I could have just said that to begin with, but then I wouldn’t have been able to use my “is a lion a cat?” question, and I’ve been wanting to ask that one for a long time now.)

As academic as it may seem to spend time defining our terms, or perhaps even boringly pedantic, there is a very practical and useful application to this. Knowing what a mechanic is, and knowing what a structure is, or for that matter, a theme, or a story, or a reward system, or whatever, allow you really zero in and think about it with much greater clarity and efficiency. Fuzzy thinking takes more time. So let’s get to defining.

‘GAME MECHANIC’?

A ‘game mechanic’ (or ‘game-play mechanic’ or just ‘mechanic’), is what a player DOES in a game coupled with some aspect that makes this “doing” a challenge, hopefully an enjoyable challenge. In the first few pages, we talked a little bit about how the goal of a successful game is to empower players though action and choice. This “doing” is at the heart of what makes a game… a game, and it is a HUGE part of player expectation. One of the most productive ways to think about game design is by asking the simple question: “what does the player do, most of the time”. Surprisingly, even the best designers often forget to ask this question enough.

When coming up with mechanics for a game, or breaking down mechanics for an existing game, one can start by thinking in terms of categories. There are probably only about 40 or 50 categories of existing game mechanics. Again, simply put, these are the things players actually “do”. Here are some examples of these categories:

Hand to Hand Fighting

Shooting

Racing

Physical Puzzles

Collecting

Rhythmic Music Matching

Crafting (combining elements)

Navigating Conversational Trees

Dodging and Jumping

Hiding and Sneaking

Climbing and Leaping “Parkour” Movement

Flying

Simple Quick-Button Response

Building and Creating

A few specific mechanics might be things like:

Shoot the bird with the slingshot by pulling back and releasing. Bird moves in an arc based on weight of bird and the distance it was pulled back. Player attempts to hit and knock over structures to pop the pigs inside.

Move constantly during attacking phase of enemy to avoid getting hit, then strike the enemies vulnerable zone accurately during the enemy’s resting phase.

Attempt to match the musical notes by hitting the correct key within a window of time as shown by the notes passing the bar.

Some simple mechanic might be things like:

Move forward as the terrain becomes visible and try to find the path forward.

Move from object to object, hiding from enemy’s searchlight

Press the button within the window of time allowed

A complex mechanic might be something like:

Drive your tank while also turning your turret and using the zoom feature to shoot targets. Select appropriate shell type for target and attempt to hit enemies in side or rear where armor is weakest, while using terrain for cover and to stay hidden.

Notice that these examples of actions all include the description of the challenge as part of the action. “Paint a picture” is an activity, it’s not really a game-mechanic, whereas “paint a picture within a 5 second window” comes much closer to being a mechanic. One thing we didn’t mention, that should also probably be part of our definition of game mechanic, is the idea of being able to evaluate player performance and feedback to them how they did. “Paint a picture within 5 seconds” still has problems as a mechanic, because it is a difficult thing to judge or give player feedback on. (as opposed to say, something like connect the dots correctly to form a picture). This isn’t to say that subjective “creative” activities don’t have a place in games – they certainly do, but in and of themselves they are not game-mechanics.

One other side note here: we often hear the term “game play”. This almost always refers to the collection of mechanics that are found in a single game, coupled with an expectation of these as being ‘fun’. Gamers and game critics will often talk about game play as ‘the most important thing’ in games. They may sometimes have a fuzzy concept of what they mean by this, but if you offered them this definition, most would say “yeah, that’s what I meant”.

‘GAME STRUCTURE’?

Now that we’ve described what a game mechanic is, we can distinguish this from what a game structure is. As we’ve said early on, games can have many mechanics but only one game structure.

To look at a game’s structure we have to step back and look at the game as a whole. The game structure is, in a sense, the higher-level shape of the game. It’s the thing that defines the flow of player experience….where do they go, and what do they do, in what order? Since most games tend to involve movement from one place to another, a game’s structure might be as simple as the map of the game world, with design notes applied to various locations. Many games are more complex than this however, and it’s often useful to diagram out your game’s structure. Often there are conditions that players need to meet before they are given access to new parts of the game. These conditions can be as simple as reaching a location in the game, or it may be collecting certain resources, or making friends with NPC characters, or attaining a certain level, acquiring an item, etc… Mapping all this out in a flowchart can be of great value. Among other things, this allows a designer to know exactly where player choice points are in a game. It’s a bit like writing an outline for a movie.

Game structures for Story-Games generally included some notion of story chapters. These are the phases of the game where the story context has shifted. Players may sometimes have new goals and new abilities based on the chapter they are in. In a linear-path game, mapping these story-related changes is very straightforward. In a game that allows for more emergent (or organic) story, plotting the structure is more complex. Here it becomes a matter of identifying potential story threads and laying out the conditions that need to be met which will change story-related character states and world states. In the section above titled “Creating a Network of Dependent Gates” we talked a little bit about how some of these conditional networked structures can be built.（source：Gamasutra）