游戏关卡设计6大难处编辑本段回目录
关卡设计可能是电子游戏行业中最有难度的设计工作了，尽管许多关卡团队都为整个行业总结出了不少设计经验，但它依然是这一领域最为棘手的开发问题，主要体现在以下几点：
时间紧迫
关卡设计非常耗时，它贯穿从创建初期原型到排除最后一个漏洞这整个开发过程的始末。对设计和美术来说，时间就是最大的敌人。这种设计工作通常要顶着巨大压力，快速创建出可转换成美术效果的模块。
但根据开发技术完善程度的不同，许多初期模块/原型在开发之前，实际上需要先创建开发工具，而不能直接使用现成技术制作产品。
转换模块的时间限制取决于游戏对视觉保真度的要求。美术设计得在关卡设计完工后，利用有限的时间完成这种转换，否则就会导致后期返功从而耗费高额成本。
重复工作
多数关卡设计都需要提前完工，以便玩法机制填充入内。这常会导致返功（游戏邦注：关卡设计一开工就要面临这种痛苦的过程），或者削足适履地更改游戏机制，以避免其内容与关卡设计无法兼容。
大多情况下，玩法机制在模块环境中的运行效果最为理想。假如添加、移除或修改一个机制，就会直接影响美术设计流程，因为美术设计师此时可能已经在根据那些给定机制或功能设计新内容。
模块过于抽象
一个构思成熟的关卡主题和一堆概念图并不总能完整反映出抽象的模块面貌。美术设计师得理清这些碰撞基元、代理和触发量之间错综复杂的关系，让这些元素无缝兼容，并制作出美观的画面。美术设计师一般都会希望能够自由把控这些领域。
不过，一个优秀的模块确实可以反映关于游戏玩法设置的碰撞结果，所以这些领域的内容都不可过度调整。它还会进一步影响到游戏玩法步测的间距，以及游戏摄像头移动过程中的修正值。
我目前在处理的游戏摄像头需要离墙5米的距离，以便其自由围绕角色移动。这就直接影响到玩家可涉足的任何一个区域，例如房间、走廊、平台、凹点、电梯等。
super-mario-bros(from randomville.com)
难以落实
马里奥和索尼克等早期的平台游戏都有一些悬浮在半空的平台（包括静止和移动的平台），这些游戏的关卡设计师需要在这类游戏原有惯例之下，使其设置与屏幕分辨率、内存、关卡大小等硬件设备局限性相协调。
电子游戏在图像、音效和控制系统上都已超越了过去的马里奥时代，但设计师和美工还是需要吸收仍然很有趣的传统玩法元素，而其难点就在于落实这些元素的美术设计：悬浮的平台需要依附于一些事物，移动的平台也同样得融入定义其移动过程的机制。
落实这些元素需要多边形建模、纹理和基板面，所以几乎没有设计师可以在不影响美术设计的前提下，轻松调整游戏关卡。
调整和返功
多数关卡设计师在最后布局之前，至少都得进行两次大规模修改。在达成一致意见的过程中，其中一些或所有布局都有可能因游戏玩法或产品需求变化而彻底返功。
美术设计师很不幸，他们遇到这种情况就得投入大量时间为一些“不好”的布局创建新的资产，然后再消耗一周的时间完善这些“被认可”的布局内容。
工作量庞大
与其他设计工作相比较，关卡设计需经历多种形式的迭代，包括模块、玩家参数、玩法机制、玩家路径、道具总量、敌人总量、战斗设计、玩法过程、教程设置、变换路径、活动策划、碰撞调整、美术优化、质量管理测试等等。其中有些内容的迭代可以独立调整，但有些内容的迭代则需扎堆进行。
出色的关卡设计需具备建筑设计想象力，大量迭代过程，并且在开发早期就得使用正确的工具和功能。关卡设计师需认识到关卡美工的任务是在保留所有出色玩法的前提下，让游戏外观更为养眼。关卡美工需要充的时间完成这些任务，不可向其施加过多压力。那么具体该怎么做呢，我个人想法是：
项目经理应该将关卡设计视为一个贯穿项目发展始终的工作，而不是在开工之初的数周内就完成所有关卡设计。不可过于仓促地将模块转化成美术内容，而应根据玩法机制的实际需求来执行这一步操作。假如某个关卡暂时不具可玩性，那就不要把所有任务都压到美术设计师身上，最好是先提交那些已经具备可玩性的内容。
关卡设计师和美术设计师之间的工作流程应是平行进行，前者在制作模块时，美工就应该设计场景的概念图，以便关卡设计师测试构成和比例。也许一些公司已经采用这种操作，但多数公司仍沿袭传统模式，先等模块完工，然后再将其转换成概念图。我认为这两者的工作不可脱节，这样才更有助于在项目早期就实现各个部门间的配合。（本文为游戏邦/gamerboom.com编译，作者：Jason Weesner）
Opinion: Classic Problems With Level Design
by Jason Weesner
[In this reprinted #altdevblogaday-opinion piece, Sanzaru Games' Jason Weesner outlines the classic problems that level designers face, like planning stages before important features are in place, block mesh issues, and more.]
As glamorous and fun as it can often be, level design is perhaps the most difficult design job in the video game industry. Despite the collective experience of any level team, the industry is still dealing with some pretty typical production problems:
Time Is A Constant Enemy
Level design is incredibly time consuming. It’s one of the few aspects of design that spans the entire development process from early prototyping to closing out final bugs! Time is an enemy to both design and art. Design is under pressure to rapidly create playable block mesh that can be handed off to art as soon as possible.
However, depending on the maturity of the technology, a lot of early block mesh / prototype time may be spent actually getting the tools to work rather than producing production-ready work.
The timeliness of handing off block mesh is all dependent on the level of visual fidelity the game is aiming for. Art has a limited amount of time to reach that target and needs the level design to be final since changes become more expensive the further along the environment art gets.
The Game Is Missing Important Gameplay Features
Most level designs start well before gameplay mechanics are available to populate them. This leads to rework (a very painful process once level art has already begun) or mechanics being shoehorned into areas where they don’t work very well.
In most cases, gameplay mechanics work best in a block mesh environment. If a mechanic is added, removed, or modified, it directly impacts art due to the fact that the environment artist may already be working on an area that’s still going through design iterations in relation to a given mechanic or feature.
Block Mesh Can Be Vague And Too Rigid
Well thought out level themes and stacks of concept art can’t always clarify the unruly, abstract appearance of block mesh. Artists have to take this tangle of collision primitives, proxies, and trigger volumes and make something beautiful and cohesive out of them! Artists will generally want the freedom to push and pull these spaces.
However, a good block mesh is really the best representation of the collision that works for the gameplay, so many of these spaces can’t be changed too radically. This is further compounded by distances necessary for gameplay pacing as well as allowances for the movement of the game camera.
The camera I’m currently dealing with requires a 5 meter distance away from the wall to allow for the gimbal to freely move around the character. This directly affects the scale of any area the player can get into: rooms, hallways, platforms, pits, elevators, etc.
Block Mesh Can Be Hard To Ground
Early platformers like Mario and Sonic had floating platforms (both static and moving) all over the place. The level designers on these games weren’t insane, they were staying true to the established conventions of the genre while dealing with the restrictions of the hardware: screen resolution, memory, level size, etc.
Video games have come a long way since those days in relation to graphics, sound, and control, but designers and artists butt heads over traditional gameplay elements that are still fun, but difficult for the artist to ground. Floating platforms need to be attached to something. Moving platforms need to fit into a mechanism that defines their movement.
All of these grounding elements require polygons, textures, and real estate, so there is seldom a case where a designer can just throw something in a level to improve the gameplay and not impact the artist.
Be Prepared To Throw Away A Lot Of Work
Most level designs go through at least two major revisions (or more) before a final layout is agreed upon. In the process of reaching agreement, some or all of a layout may be completely reworked to adjust to changing requirements based on gameplay or production considerations.
Unfortunately for art, they may have to spend a considerable amount of time creating assets for the “bad” layout and then have to scrap week’s worth of work to make adjustments for the “good” layout!
There’s Just Too Much To Do
Compared to other aspects of design, level design goes through the most varied forms of iteration: block mesh, player-metrics, gameplay mechanics, player path, item population, enemy population, combat design, gameplay progression, tutorial placement, alternate paths, event planning, collision refinement, art refinement, and quality assurance testing just to name a few. Some of these iteration passes happen exclusively while others have to happen at the same time.
Good level design is a mix of architectural successes / failures based on imagination, a ton of iterations, and access to the right tools and features at an early stage. Level designers needs to be cognizant of the fact that level artists want to make the game look beautiful while maintaining all this wonderful gameplay. Level artists need to be given the time to embrace compromise without falling into the trap of being over-scheduled. So, what can be done about any of this? At the risk of sounding like a know-it-all, I have a few thoughts…
Project managers need to look at level design as a process that occurs over the lifetime of a project and not just a few weeks at the start of a level kickoff. The rush to get block mesh into art should be tempered by the availability of the gameplay mechanics necessary to populate it. If the level is not playable in something close to an alpha form, don’t hand the entire thing off to environment artists! Level design is not an all or nothing situation; hand off the parts that are playable.
The workflow between level designers and environmental artists needs to run more in parallel. While designers are working on block mesh, artists should be working on conceptual forms of the environment in order to test out composition and scale. Perhaps some companies already do this, but the majority still seem to fall back on the traditional model of creating concept art, waiting for block mesh hand off, and then straining the process by trying to mold the block mesh into the concept art. Having all three of these processes run in parallel seems like a far more sane solution with collaboration between departments happening at a much earlier stage!（source:gamasutra）
