原文作者：Michael How 译者：Megan Shieh

在开发游戏的过程中，有许多需要注意的事情你会希望自己能从一开始就知道，比如可以帮助你避免陷阱的关键点。本文将阐述设计游戏时需要注意的7个关键点。

第一：快速失败，不然就永远不失败。

行业中有许多关于“快速失败，经常失败”的言论。当你拥有大量的时间和金钱，并且想制造出完美的产品时，这无疑是可行的。

但对于大多数开发人员而言，这是不可能的。时间和资源的缺乏限制了这种做法的可行性。我宁愿尽早失败，要不然就永远不失败。你可以在付诸实际行动之前，与可信的设计伙伴在纸上或以文档的形式一起深入探讨某些想法的逻辑。有了可靠的初期理念就把它应用到游戏中去，做出雏形。效果不好的话就赶快放弃；如果它在任务和游戏方面触及到了你的核心支柱，就继续投入时间。

Problem（from gamasutra.com）

千万不要小看特性审批的重要性，向尽可能多的利益相关者展示你的特性，让他们充分理解这个特性并挑战它的设计，从而将它进行优化。如果这个特性很有趣，而且符合你的核心支柱，就把它给签了！然后进行迭代与优化。如果sign off失败了，那就对它进行分析从而找出失败的原因。如果你已经从中学到了一些东西，并且不能把它带入下一个设计的迭代中，那么失败了也没什么大不了。不过不要随意地丢弃工作成果，要从中吸取教训，找出出错的地方并分析原因。

第二：不要显示设计师的意图

为玩家创造更多的有机体验。确保玩家无法看穿你意图创建的体验，确保他们沉迷于游戏之中，以至于看不到框架的运作方式。《Journey（风之旅人）》在这方面做得非常出色。当你穿游戏中的世界时，感觉无比广阔，游戏界面不会弹出“您已偏离路线”之类的提示。似乎无需刻意地尝试就能去到你需要去的地方。温和的引导和路标使得游戏体验成为一种乐趣。在关卡设计中，真正精湛的结构会对成功起到很大的作用，并且有助于保持沉浸感。

多数玩家玩游戏的原因都是希望能够逃到一个新的世界，并在那里玩儿得开心。这个世界需要感觉真实才能让玩家保持沉浸感。如果玩家看到一群敌人在任意的触发条件下凭空产生，他们就会跳出这个世界。

敌人应该在受到某种刺激的时候出现，而不是根据设计师决定的特定条件凭空出现。掩盖你设定的条件，让游戏世界感觉更真实，这样的话玩家也会从优化中受益。在适当的时候引入变量来生成逻辑和条件，因此你需要测试所有可能出现的结局。如果你自己没时间做这个，就请有时间的人来帮你。对于这种带有开放性世界的游戏，你需要测试所有可能发生的体验。敌人莫名其妙出现的情况通常很有可能把玩家从沉浸感中拉出来。

在《Killzone 2》里，敌人出现的情形设计得非常出色，是我迄今为止见过最好的。你可以看到敌人从残骸和碎石中爬出来，最终进入游戏场地。关卡设计师创建了一个外围，并确保它在玩家接触范围之外，所以看起来不容易接近。在保持沉浸感的同时，它也会让玩家产生恐惧。

第三：让各个机制以有意义的方式相互作用

如果你创造出一个无法与环境、敌人或其他机制产生多种相互作用的机制，那么它就无法给人留下深刻的印象，人们很有可能将它遗忘掉。围绕一个主要的机制来设计任务和冲突，并引导你的玩家去探索这个机制的多样性。

这并不意味着建立一个以不同方式进行交互的复杂机制。游戏机制不用搞得那么复杂，只是在互动方式上要有所创新。允许玩家用新的、令人兴奋的方式来使用同一个机制。改变这个机制运作的方式，颠覆玩家的期望，然后要求他们适应这个改变。

Naughty Dog的《The Last Of Us》就是一个典型的例子。该作中的推车可以在游戏的过程中以不同的方式使用。当你认为你已经掌握了它的用法时，Naughty Dog又引入一个新的变量来挑战你，颠覆你的期望，要求你用同一辆推车来解决一个新的难题。

让玩家们更深入地了解一个机制，让他们觉得自己很聪明——可以再次用同样的机制来做一些新的、令人兴奋的事情。在《狙击精英 3》中，我们使用火石和易燃物来分散敌人的注意力，引诱他们去调查烟雾，从而找到在黑暗中隐藏着的敌人。而且如果扔到的位置碰巧有炸药，这些炸药就会被引爆。再者，如果你把它扔到发红的油桶或弹药堆旁边，也会引发爆炸。

简而言之，一个机制应该有多种玩法。给玩家一个机会来掌握一个东西，让他们感觉自己很聪明，并引导他们去探索这个东西的其他用途。

第四：问自己，玩家的能力在什么水平？

这是一个非常重要的问题。你必须要清楚的知道当玩家上手游戏的时候，你希望他们是什么级别的玩家，这就是游戏中最后的任务了吗？我们还能不能继续增加难度？DLC的内容能不能留住老玩家，或者是吸引新的潜在玩家？你是否应该警告打算玩DLC的玩家，这些内容要求你有在主游戏玩过的经验？需要到达中级的时候才可以玩DLC，还是它只是漫长教程中的一部分？确保你的内容不会令人沮丧或觉得太有挑战性。让新手们来玩这个功能，跟踪他们游戏的过程并且做笔记。开发人员通常都无法客观地评估自己的作品。你需要和很多新玩家一起回顾，找到你的真实感悟，从而发现所有你可能从未想过的选择。

第五：界定支柱

这是一个非常有效的方法，能让你保持“在轨道上”。偏离轨道的意思是，在游戏开发的过程中分心或迷失方向。你应该设定游戏支柱，它们可以推动所有研发的基础。你也应该建立部门支柱，它们能够驱动不同的部门，比如UI或美术设计。这些支柱可能会根据部门的不同而存在差异，但它们最终应该都反馈到游戏的主要支柱。

大多数开发者会将游戏支柱当做指南使用，以避免偏离轨道或迷失方向。你甚至可以设定任务支柱和任务目标，界定任务支柱会让你将主要精力保持在优先事项上。如果你游戏的核心支柱是跳舞，任务支柱可能是曼波舞——但是如果你的任务中出现击剑，那你就严重偏离轨道了！ 尽可能频繁地回到你的游戏支柱和任务支柱上，以驱动可靠的迭代和决策。

通常一旦迭代开始，这些支柱就很容易被忽略。如果你已经偏离了轨道，但内容是有趣的，问问自己是否需要为任务界定新的支柱。思考之前的支柱出了什么问题，并讨论设计偏离的原因。把这些经验和教训都找出并记下来，以后不要再犯同样的错误。“乐趣”会推动游戏体验，所以保持灵活变通也没什么不好。

第六：倾听

倾听你的社区，你的玩家，你的老板，等所有对游戏有话要说的人。所有的意见都是有用的。不用把全部的建议都加到游戏里，但必要的时候可以大胆地挑战别人的设计。如果有人询问设计方面的问题，设计逻辑会因为得到额外的关注而受益。

如果你正在制作续作，确保你在原作的反馈基础上做一个回顾表格，并与社区保持联系。基本上就是跟踪所有的评论，将它们进行统计——被多次提到的关键问题，人们喜欢/不喜欢的特性，然后回顾统计出的结果。

如果有26个人赞扬你的关卡设计，并且没有任何负面评论，那么这个方程式就是成功的。如果有10个人抱怨你的AI，那么你就需要对它进行复审和优化。如果没有人提到UI，那么它应该没什么问题。反复审查评论统计中被多次强调的观点，因为关键领域的负面反馈是值得参考的。

让社区感觉他们提出的关键问题得到了开发者的重视，这点非常重要，因为他们会形成游戏受众的核心部分。确保你的社区知道你有在听他们的反馈，并且与他们公开对话，这样的话他们就会了解你做决策的原因。

第七： 集中注意力

说起来容易做起来难。每个人都想打造一款伟大的游戏；发行商们想要有大量的好内容来大肆宣传；开发团队想要一个充满多样性内容的全脂游戏。然而，你所做的每一个功能、特性、机制或模式，都有可能淡化你对重要事情的注意力，剥夺核心游戏的优化时间。

参考你界定的支柱，找出对IP和游戏而言最重要的东西。将资源围绕最重要的特性聚集，如果资源分布得太稀薄就赶快调整。

无论你的团队有多棒，一旦团队的精力被分得太散，很多特性做出来的效果就会变得普普通通，然后你就会想要把所有特性的质量都提高，在这个过程中很有可能把人们的耐心磨光。就跟领导一次性给员工分配太多的工作一个道理。让他们专注并出色地完成一件事，然后再做下一件事。一次性在盘子上堆放太多东西的结果就是所有东西做出来都只有一般般的质量。

谢谢你花时间看完了这篇文章，我希望上述的东西能够对你有所帮助。

PS:我所提到的东西并不是针对所有类型的游戏，个别观点针对的是特定类型的游戏，希望你了解这点。

本文由游戏邦编译，转载请注明来源，或咨询微信zhengjintiao

In development, there are a lot of things that you wish were front facing from the get go. Key’s that could avoid pitfalls and help you had you known earlier. Here I seek to offer some of the keys to solid design. 7 in fact.

1. Fail early or not at all. – There is a lot of buzz around “fail early and fail often”. This is certainly more viable when you have lots of time and money and want to make the perfect product. Supercell can delve deeper into that for you here.

For most developers, this simply isn’t possible. Time and resource get in the way. I prefer fail early or not at all. You can do this by chewing into the logic of something with a trusted design compadre before you run off to tools. Explore the logic in paper/doc form first. From that initial vetted brief – get it in game – prototype it. Fail it early and move on, or, dedicate time to it because it hits your core pillars for both the mission and the game.

Never underestimate the importance of a formal feature sign off with lots of people involved. Get as many stakeholders as possible to see your feature, understand it and challenge the design to improve it. If it’s fun and it hits your core pillars, sign it off! Iterate and polish. If it fails the sign off, then ask why and analyze it. Failing is also ok provided you have learned something and cant take that into your iteration of the next design. Don’t throw away work, learn from it. Discuss what went wrong and why.

2. Don’t show the designers hand - Don’t create experiences that show your hand as a designer. It’s almost never OK to do this (unless you are John Romero and realize your own head is in a level and exact revenge.)

Create a more organic experience for the player. Ensure players can’t see the experience you intended to create. Allow them to become so swept up in the gameplay they can’t see the frame work. Journey is utterly brilliant in this regard. You slide through the world and its feels absolutely vast. There is no warning you are “leaving the battlefield”. You always seem to end up exactly where you need to go without even trying. The effortless soft guiding and sign posting makes it a joy to play. The truly masterful composition in level design plays a large part in the success and helps maintain immersion.

Players are often playing games to have fun and escape to a new world. That world needs to feel real in order for players to maintain immersion. If a player literally sees a bunch of enemies spawn out of nowhere based on an arbitrary trigger condition, they will break out of that world.

Enemies should alert based on stimuli not a set conditions decided by the designer. Mask your conditions so the world feels believable and the player will benefit from that polish. Introduce variables to spawn logic and conditions where appropriate. Your spawn logic may not be appropriate if the player approaches from a new direction – make sure you are testing all possible outcomes. If you don’t have the time, enlist some help to. Games with large environments need testers to help you review all the possible experiences. Often the spawn conditions for enemies are a huge immersion breaker.

Killzone 2 had some of the best enemy spawning i’ve seen, even to this day. You could see the enemies clambering in over debris and rubble, eventually dropping down into the gameplay space to engage. The level designers created a fringe to allow for this and ensured it was outside the player metrics so didn’t look accessible. You’d glimpse the realistic spawning at the fringe of gameplay and it created dread in the player whilst maintaining immersion.

3. Allow mechanics to interplay with each other in meaningful ways - If you make a mechanic and it doesn’t have meaningful interplay with the environment, enemies, or other mechanics in more ways than one; chances are it may be quite forgettable. Design your missions and encounters around one solid mechanic and allow your players to discover the diversity of that mechanic.

This doesn’t mean make a complex mechanic that interacts in different ways. Try keep it the same but have the ways of interaction differ. Give players new and exciting ways to use just one thing. Subvert their expectations as it changes from what they expect, then ask the player to adapt.

Naughty Dog’s The Last Of Us is basically a seminar for unique world interaction. The trolley cart is used and adapted in so many ways through-out play. The second you think you have mastered it, Naughty Dog introduce a new variable to challenge you and subvert your expectations, requiring you to solve a new puzzle with the same trolley.

Let the players forge a greater understanding of a mechanic and feel clever having used it again for something new and exciting. In Sniper Elite 3 we used the flint and tinder to distract and lure enemies in to investigate the smoke. If you place dynamite in the same location, it will also ignite it. If you place it near a red barrel or stash of ammo, it will detonate it. Don’t even get me started on the mechanics interplay in Dishonored – It’s too vast to get into and there’s a host of videos discussing the systemic interplay of the mechanics. The rule remains the same though.

So put simply, don’t make it a one and done. Give your players a chance to master something and feel clever and empower them to explore other uses.

4. Ask what the expected player competency is - What is the player competency? This is a very important question. You need an understanding of the expected skill level for the players when they get to your content. Is it the last mission and we can challenge the player more? Is it DLC where you have a delicate balance to hit content for retentive players or potentially new players? Should you warn players for DLC that the content requires prior experience with the main game? Is it middle of the rung or part of a lengthy onboarding? Make sure you take a step back and get perspective of that skill level so the content isn’t frustrating or too challenging. Get ‘noobs in and test, shadow them as they play and write notes. Dev’s cannot objectively assess their own work most of the time. You become too close to it. You need to review with lots of new players to get your true insights and discover all the options you may never have imagined.

5. Create big and small pillars - Creating big and small pillars is a useful way of keeping yourself “on piste”. For non skiers – “off piste” is back country – filled with deep powder and crevasses – “on piste” is the main slope. Off piste is a place where you will get distracted or lost in both game dev and skiing terms. You should have game pillars – these drive the fundamentals of all your development. You should also have department pillars – these drive various departments like UI or Art direction. The pillars may differ from department to department but they should all feed back to the main game pillars.

Most devs will use game pillars like a guideline to stay on piste and keep from getting lost. Digging deeper, you can go all the way down to a mission pillar or a mission goal. Defining mission pillars allows content to hit the priorities for that mission. If your game has a core pillar of dance, the mission pillar might be the mambo – however If you have sword fencing in your mission you have gone off piste! Fall back to your game pillars and mission pillars as often as possible to drive solid iteration and decision making.

Often once iterations occur it’s quite easy to lose track of the pillars. If your pillars no longer make sense but the content is fun, ask yourself if you need to define new pillars for the mission. Why has the experience metamophasised? What was wrong with the old pillars and discuss why the design moved away. Take those learnings forward so you don’t make the same mistakes again. What is ultimately “fun” will drive the experience forward so don’t be afraid to stay agile.

6. Listen - Listen to your community, your players, your bosses – everyone has ownership and buy in and also opinions. No opinion is invalid. Don’t crowd source design, but don’t be afraid to challenge other people’s designs. If people ask questions of the design, the design logic will benefit through extra attention being paid to the caveats.

If you are making a sequel, make sure you do a review matrix based on feedback from the first game and stay in contact with your community. Essentially track all the reviews and do a count on key issues,likes and dislikes then review the amount in any category.

If you have 26 different reviewers raving about your level design an no negative comments, then your formula is working. If you have 10 people blasting your AI, it will need review and polish. If your UI isn’t even mentioned, it’s probably fine. Hit anything hard that highlights in your review matrix multiple times as it’s likely negative feedback on key areas is worth addressing.

It’s so important to ensure the community feel like their key concerns are addressed as they will forge a core part of your audience. It needs to feel like an inclusive process. Ensure the community know you are listening and open a dialogue so they can also know why your decisions are being made.

7. Focus - Focus is easier said than done. Everyone wants to make a great game. Publishers want an abundance of content to rave about. The team want a full fat game with diversity of content. However, every feature, mechanic or mode you make, has the potential to dilute the focus on what is most important and strip away polish time from your core gameplay.

Reference your pillars and look at what is important to the IP and your game. Keep your resource focused around your most important features and scope/claw back bandwidth if it ever looks like you are spread too thin. Steven Masters delivers a great talk on keeping your game focused and hitting Feature Sign Offs with confidence.

No matter how good your team is or your process is, once a team is spread too thin, you will ultimately make a lot of features to an average quality and risk burning people out as you try to get everything to a high quality level. This goes for assigning staff too much at once. Let them focus and deliver well on one thing with excellence and then move onto the next. You you pile too much onto the plate at once and you’ll end up with a bunch of things done to a mediocre level.

Thanks for your time and I hope some of these keys benefit you as a developer. Please feel free to comment any feedback or add further “keys” in comments.

*I’d caveat this article by saying that not all this information will be relevant to all types of games particularly where I reference spawn logic, enemies or even maintaining immersion. For instance presense may be more important, particularly in VR! It’s merely a guide for general development to help up and coming designers or veterans wishing to verify their own process.

Special thanks to Steven Masters and Mateusz Piaskiewicz for the reference.

All Images are from Zelda 2 The Adventure of Link（Source: gamecareerguide.com  ）