作者：Giordano Bruno Contestabile

长期以来，游戏行业都在多年未曾有大改变的可预测模型基础上运转。游戏经过长时间的开发，然后提交给发行团队，他们的任务是尽可能高效率地发布和营销游戏。随后，所有团队成员将精力放在下款游戏上，重新开始上述过程，在这样的循环中不断提升产品质量。

游戏行业似乎很信赖这样的开发战略，但是随后发生了某些重大的事情。可以玩游戏的设备间开始通过网络相互连接，使游戏和玩家可以进行不断地交流。

网络连接颠覆了传统游戏开发模型，使行业关注的重点从产品转变成服务，所以开发和发行的各个层面也需要进行改变。这里我所要讨论的就是：基于数据的衡量指标。

Intuition_Vs_Metrics(from casualconnect.org)

两种意见

根据词典的解释，指标的含义为“数量化系统状态的分析性衡量”。在游戏中，“指标”指可以从游戏玩法中收集到的数据，显示出用户在游戏中做的事情以及他们如何同游戏本身互动。这本不会引起激烈的讨论，因为数据本身只是客观的量度而已。行业对此划分为两大阵营的关键在于，这些指标应当如何使用或者它们是否应当被使用。

一方是近几年刚刚兴起的社交游戏公司，他们更注重于网络营销而非游戏开发，他们极为推崇指标的力量，认为数据可以为设计过程中产生的所有顾虑提供答案。

另一方是传统游戏开发商，他们认为指标影响了这个行业，将游戏设计直觉和技巧替换为冷冰冰的数据，最终诞生出的是毫无灵魂的产品。哪方是正确的呢？我觉得，两方的看法都有不当之处。

指标的优点

我先从反指标一方说起。对于想要不断演变和发展的游戏来说，将其作为服务来运转是必要之举，游戏需要在发布后不断更新和修改。从这个方面来说，因为数据可以反映出玩家参与游戏的具体行为以及他们对修改和创新有何反应，所以拒绝监测指标数据似乎是个缺乏远见的做法。

事实上，上述做法意味着设计师没有利用完全同游戏表现相符的客观数据来源，从而草率地做出决定。简单地说，这是种反科学的立场，在极为依赖技术和过程的行业中确实不是恰当的做法，设计师应当在了解一定情况的基础上做出决定。

指标的缺点

但是，如果过分地依赖指标和数据驱动决定，那么同样会导致盲目性。设计师似乎将数据凌驾于所有元素之上，完全忽略了其他元素也在成功设计和运营游戏中占有一席之地。

较为明显的危机就是，过分推崇数据会导致所有产品标准化，开发商摒弃创新和风险，不断追求最优化盈利设计。这不仅会影响到游戏行业未来的前景，而且长此以往产品很难激发玩家的兴趣。这确实是个很糟糕的结果，在这个行业中，我们的任务是让玩家感到高兴，而实现这个目标的唯一方法是构建可持续发展的长期价值。

折中方法

通常情况下，设计师往往会采取折中的方式。在基于服务的游戏运营中，如果不分析和考量使用数据和指标，那么可能失去许多改善游戏、为玩家提供他们想要的东西以及提升业务效率和盈利性的机遇，但是完全靠指标来驱动设计决定会导致最终出现毫无创新性设计的平庸游戏产品。

我的意见是，我们不可推崇“指标驱动”的设计方法，而是“以指标为根据”的设计方法，也就是说应当收集和分析数据并将其用来支持设计而不是决定设计方向，不过在设计过程中仍需时刻考虑到数据的作用。

而且，对指标的重视程度要根据游戏设计和运营的阶段不同而加以区别，设计直觉和数据二者间的侧重取决于开发过程进展到何种阶段：原创设计新IP时应当侧重前者，而已上线产品的优化和改善应当侧重于后者，将新内容和功能引进现有产品中应当同时顾及两个方面。

找到指标和直觉之间的平衡点并非易事，但这对基于服务的游戏的开发和运营来说却至关重要。要弄清楚如何最好地实现这种平衡以及如何将其整合到产品的开发过程和企业文化中，这样的公司才能够长期地繁荣发展，引领游戏行业的未来。

游戏邦注：本文发稿于2011年10月31日，所涉时间、事件和数据均以此为准。（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）

Metrics versus intuitive design – FIGHT!

Giordano Bruno Contestabile

For a long time, the game industry functioned on a predictable model that changed little over the years. A game was developed over a long period of time, then went gold and was handed off to a publishing team tasked with launching and marketing it as effectively as possible. Efforts then shifted to the next game, and the process started anew, becoming more refined with every product cycle.

All seemed well and the industry started to metaphorically doze off, lulled into a sense of complacency by this ever-repeating rhythm. But the something momentous happened. The devices on which games are played became connected, allowing for constant communication between the game and the players.

As we all know, connectivity is upending the game development model, transitioning the industry from product to service-based and requiring dramatic shifts in every aspect of development and publishing. I won’t plunge into a description of how this is happening and what’s changing as a consequence, but will focus on one particular aspect enabled by connected devices, one aspect that has the potential to shape the way games are made, and has generated enormous controversy as well as its fair share of knee-jerk rejection by veteran game designers, to the point of becoming a dirty word in some circles. The word I’m referring to is: “metrics”.

Setting up the sides

Metrics, according to the dictionary, are uncontroversially enough, “analytical measurements intended to quantify the state of a system”. With respect to games, “metrics” refers to data that can be gathered from gameplay, providing an aggregated view of what users are doing, and how they interact with the game itself. Again, this shouldn’t cause heated debates, as data by itself is only an objective measurement. The real issue that has divided the industry, unleashing emotions usually reserved for sport and politics, is how those metrics are used, and if they should be used at all.

On one side we have the upstart social gaming companies that have emerged in the last few years, finding their roots in web marketing more than in game development, and armed with a burning belief in the power of metrics to the point that data overrules any other consideration in their design process.

On the other side, veteran developers see metrics as intruding on their turf, replacing intuition and game design skills with number crunching and ultimately resulting in soulless, cookie cutter products. Who’s right? Neither, I’d say. Or rather: both points of view are tragically wrong, and it all depends on how metrics are used and implemented in the process.

Hooray for metrics

In the spirit of bi-partisanship I’d like to annoy both camps, and I’ll start with the anti-metric one. A game that’s operated as a service is necessarily a game that’s forever evolving and changing, requiring constant updates and tweaks. In that sense, refusing to look at data reflecting how actual players engage in the game, and how they react to changes and innovations, seems to me to be extremely short-sighted.

In fact, doing so means designers are failing to take advantage of a source of objective data that’s completely germane to the performance of your game, and therefore leads to decisions that are less informed than they could otherwise be. It is, in short, an anti-scientific stance, and something that’s really surprising from an industry that heavily depends on technology and process, and that should jump at the chance of being able to make more informed decisions.

Down with metrics

On the other side, fanatics of metrics and data-driven decisions suffer from a similar form of blindness, in the sense that they appear ready to put their belief in data before everything else, and to discount the notion that other factors should play a role in the design and operations of a successful game.

The obvious risk there is to get to a point where the data leads to a standardisation of products, and where innovation and risk are sacrificed to constant optimisation – not just a bleak vision of the future of gaming, but something that would presumably give birth to products that would fail to excite players in the long term. This would be bad practice indeed: as an industry, we’re in the business of delighting our players, and doing so it’s the only way to build sustainable long-term value.

The middle way

The truth, as it often happens, lies in the middle. Operating a service-based game without analysing and considering usage data and metrics would lead to many missed opportunities to improve the game itself and to provide players with what they are looking for, as well as to run a more efficient and remunerative business, but having metrics drive design decisions would eventually lead to a mediocre game that shies away from innovative and daring design choices.

Or, to frame it in a different way, refusing to see that the shift in development and delivery model should reflect also the way games are designed is a surefire way to become a dinosaur and rapidly lose relevance, while slavishly relying on data can lead to a severe lack of long term vision.

What this calls for, in my opinion, is an approach that’s not “metrics-driven”, but “metrics-informed”, meaning that data should be collected, analysed and used to support design decisions rather than drive them, but that design should always take data into consideration.

A further distinction has to be made with respect to different phases of game design and operation, as the balance between design intuition and data should be different depending on what phase of the development process they are applied to: original design of new IP should see a marked predominance of the former, while optimisation of a live product should be strongly influenced by the latter, and the introduction of new content and features in an existing product could most likely benefit by an interplay between both.

Finding the right balance between metrics and intuition isn’t easy, but it’s crucial to the development and operation of a great service-based game. Companies that will figure out how to best achieve this balance and how to integrate it into their own development process and culture will have the best chance to thrive in the long term, and to define the future of gaming. (Source: Edge)