作者：李祥
链接：/question/24376247/answer/64035765
来源：知乎
每个做数值的人都有自己的感受，不过总体来说大家还是认同游戏数值并不需要太多很高深的东西。也许某些游戏在几个地方会需要使用高等数学(例如某些复杂经济、计算机图形、运动模拟)，但基本都不会是在“游戏设计”这个方面。然而从技多不压身的角度来说，也确有相当一部分人认为数值策划应该去学习更多复杂的工具和高等数学知识。
怎么说呢，做商业游戏是个大工程，有很多方面都是上不封顶的，只要你愿意去探索。不过商业游戏会很关注性价比，在任何方面如果投入产出比太低则它其实就只能被归为少数人的私藏爱好，而不是可以通行于业界的东西。例如我个人很喜欢AI和手感，但是在实际项目中对AI和战斗机制的运用程度也就是普普通通的水平，我花个人业余时间去研究感兴趣的东西并不能运用到项目中去，这时候这些东西就是我个人爱好，并不能说必须爱好并学习这些才算是合格的系统策划。
游戏数值也是类似的，大部分项目并用不上高深的数学知识，那就说明游戏业界在策划数值时对数学的应用就是这样一种水平。也有不少人觉得不甘心，“这东西上不封顶的值得探索的地方多着呢，我们觉得初等数学够用了万一是井底之蛙呢”？所以搬运一下Quora上Tom Cadwell的回答(其参与过战争阴影、黑暗王朝2、魔兽3冰封王座、魔兽世界、英雄联盟的平衡设计)：
How much in-depth Mathematics knowledge should you have as a Game Designer?
Very little.
You need to be highly fluent in fast analytic thought... But you need not know much math at all. Having a mind that thinks logically is key though in going through complex system decisions, and considering what might be impacted by what. But you need not know set theory or differential equations or whatever to do that.
Lots of design managers insist on "great math skills" and try to interview for this. I just think it's the wrong way to go since very few game design decisions worth solving require 'great math skills' to solve.
Speaking from experience, I have a lot of advanced math training, but it's really never useful, and I've worked on a variety of highly competitive games that need precise play balance.
The areas that math is beneficial are in three places.
Arithmetic:
1) Balancing like game elements in an MMO (e.g. level 50 sword, level 51 sword, level 51 mace, etc). Using spreadsheets or a templating system can give you easier consistency, but is not 'in-depth' mathematics. It's just arithmetic mapped to a pre-generated curve. .
2) Modeling out complex formulas that you the designer have decided to introduce can benefit from excel fluency, and even better arithmetic skills are required. As an example, calculating in-game DPS for an MMO might require this, though honestly, you are better off in some cases testing results empirically and saving yourself time. So, helpful but optional.
Predictive Modeling:
1) If you have a lot of data, and you want to gain insights about what users are engaging with and why, you might want to know some sophisticated predictive modeling. However, no game designer needs to know this -- this can easily be done by a BI group or a specialist of some sort. It doesn't hurt to know, but it's not really 'game design' as it can be 'insourced' to another department without downsides.
You do need to be decent at excel for some tasks, but that's very different.
Probability:
Some answers have mentioned probability is important. I think a basic understanding of elementary probability is useful but hardly mandatory. This is because most 'bad' decisions that are related to poor use of probability will come out in testing. If you are designing paper and pencil and board game products probability is more important both because you can't patch and because of the level of craziness you can place into back-end algorithms to smooth over outcomes...
But for video games, even very complex ones like the ones I've worked on, it is of marginal benefit. Every time I've had to deal with a complex set of branching probabilities, playtesting and judgement on feel ended up trumping initial probability analysis... So I just don't depend upon probability beyond a basic understanding to avoid obvious bad approaches.
大致翻译一下：
游戏设计中只会用到比较浅的数学知识。
最需要的是熟稔于快速分析，不过并不需要了解太多数学。在做复杂系统的设计决策时逻辑化的思维是关键，要考虑哪些东西会被别的东西影响。但做这些并不需要集合论、微分方程等知识。
许多策划经理(design manager没想到对应职位)很执着于面试具有优秀数学技巧的应聘者。我不赞同这种观点，因为在游戏设计决策时这类数学技巧只有很少的地方能派上用场。
就我个人的经历来说，我接受过大量高等数学的训练。我参与过各种需要精准平衡性的竞技类游戏，但所学的高等数学在其中没怎么发挥用处。
在游戏设计领域有3个地方数学最能发挥用处。
1.算术
(1)用于在MMO类游戏中平衡各种元素(例如50级的剑、51级的剑、51级的锤子等)。借助excel表格或模板系统可以很容易地保持一致性，但这不是多高深的数学。只是一种映射预设曲线的算术问题。
(2)娴熟的excel技术有助于根据设计来构造公式，有良好的算术能力更好。例如计算MMO中的DPS就会用到，不过说老实话，用测试替代理论计算会更省时间。所以算术能力在这里只是可选项。
2.预测建模
如果你有一大堆数据需要透析出玩家与这些数据会如何相互作用，你也许需要学一些复杂的预测建模。然而游戏策划不是必须会这些，一个商业智能专家就能轻松搞定它。技多不压身学了也没坏处，不过这并不是游戏设计，完全可以内包给其他部门。
你需要熟练掌握excel来完成一些任务，但和预测建模有很大区别。
3.概率
我认为对基础概率有所了解还是有用的，但并非必备条件。这是因为大多数跟错误使用概率有关的坏设计会在测试中体现出来。如果你设计的是纸笔桌面游戏，那概率相对来说很重要，因为你没办法打补丁没办法调整后台算法。
但对视频游戏来说，即便如我参与的这种复杂程度的，很懂概率也只能带来些许好处。每当我需要处理某类复杂的概率问题时，游戏测试加上实际感受最终胜于单纯的概率分析……所以为了避免类似问题我并不使用超出基础部分的概率。
如果觉得《李祥对数值策划的定义》对你有帮助，请点赞、收藏，并留下你的观点哦！
