作者：David Ferriz

1000个房间是整数，并且听起来很吸引人，所以当我们在发布Greelight活动新闻稿时总是会写上像“拥有1000个关卡的游戏”这样的标题。而1000个房间似乎是促进推广的一个有效数字，但事实上却并非如此，就像《King Lucas》的城堡便拥有超过1000个房间，具体说来是1223个房间。

在之前有关《King Lucas》的文章中我曾和你们描述过我与搭档Laura Suarez一起进行关卡的程序设计。那时候我负责一个一个进行设计，绘制每一个tile，衡量每次跳跃并设置每个危险，并确保玩家能够在适当的地方发现它们。设计如此多数的房间看起来像是一项巨大的任务，但多亏了这一付出，我们可以在每个房间向玩家传递许多细节，参考对象，以及让人兴奋或害怕的时刻，而这也是不能通过程序生成的内容。

明确规则

尽管关卡设计不是基于程序，但在城堡网格中设置它们却几乎是随机的，所以当我们在设计这些关卡时必须明确所有关卡都能够匹配的基本规则：

基于同样的规格（40×30tiles）。

它们都拥有4扇出入门。

当玩家进入房间时他必须能够到达至少一个其它房间。

castle(from gamasutra)

除了连通房间的规则外，我们同样也要考虑英雄的速度以及他跳跃的高度和宽度，因为在每一款平台游戏中，让玩家感受到每个tile非常重要。对此我们设计了多个原型并让不同用户对其进行测试，直至我们获得最理想的速度和跳跃（即3个tiles高和4个tiles长）。

我们清楚角色移动的参数和实际平台游戏的设置相差甚远，但因为我们清楚角色移动必须尽可能接近8位体经典动画，所以这一参数也就变得可预测了。

工具

面对超过1000个房间的设计挑战是非常复杂的，我们需要一个能够像基于tiles进行设计那样简单的工具，如此游戏才不需要浪费较长的加载时间。所以我们最终选择了Tiled这个适合2D关卡设计的免费工具。而多亏了Tiled，我们能够以一天设计5至10个房间的速度进行设计，也就意味着我们可以在5个月时间里以全职的状态设计出所有房间。

尽管游戏将作为完全的2D游戏，但当我们去尝试Unity时我们决定突出平台tiles以创造更多深度并达到2.5D的效果。在动态灯光下，这能够让游戏图像更具现代化。

拼图碎片

我们用于创造城堡房间的元素是受到8位体经典动画的启发，是的，可想而知那里并不会缺少熔岩坑，钉鞋，梯子，滑动平台，水，滚筒，秘密过道，敌人和NPC等元素。

当我和Laura开始设计房间时，我们的程序员并不能全职参与游戏的创造，所以大多数元素还是未能设计好，我们也不能将其用于房间中。在第一周我们创造了大约100个只拥有钉鞋和梯子的房间，这看起来还很无聊，但却能够有效适应Tiled，同时我们还创造了一些更安静的房间为游戏故事添加了一些更放松的时刻。程序员也一点一点地执行着剩下的游戏元素，每次当他们递交了全新的内容时我们便会将其最大化呈现出来，这也是你为什么会发现有些房间充满蛞蝓或蜘蛛的原因。

当我们拥有大概700个房间时开发者也完成了所有元素，所以我们仍有时间去创造几百个包含不同元素的房间，我们也将创造出一个更加匀称的城堡。

灵感

设计关卡总是很有趣的，你可以看看像《小小大星球》或《超级马里奥制造》等能让许多人为其花钱的游戏，但事实上如果你每天花8个小时去做这件事，你的灵感便很容易枯竭。所以我们会在每日生活中去寻找灵感，就像你可以在《King Lucas》的房间中发现许多隐藏信息，圣诞树，甚至是我们的家乡Villena的城堡。

habitaciones(from gamasutra)

当一切都贴合时

就像我在其它文章中曾经说过的，DevilishGames的主要收入来源是我们通过广告代理Spheical Pixel所赚取的的广告收入，而像《King Lucas》这样的独立项目则是未拥有其它更迫切项目的团队成员的潜在任务。

通常情况下我们的开发团队总是比美术人员更忙碌，而这也会影响着《King Lucas》的开发，因为通常情况下我们需要在未明确整个城堡，甚至整款游戏的视觉效果的前提下去设计并测试房间。这听起来可能有点不可思议，但在将近4年的开发过程中我们唯一一次测试了完整beta版本是在4或5个月之前，即在我们最终检测所有房间是否符合我们的设想的时候。毫无疑问也正是这一次的测试让我们决定去改变公司策略并投入更多资源于《King Lucas》中，而不再是去寻求外部客户。对于我们这么做到底是对还是错就让Steam的玩家替我们作出回答吧！

（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 a castle with more than 1000 levels

by David Ferriz

1.000 rooms is a round number, it sounds nice and that’s why when we launched the Greenlight campaign press wrote headlines such as “The game of the 1.000 levels”. 1.000 rooms is a great number for promotion but in fact it’s not real, King Lucas’ castle is made of more than 1.000 rooms, exactly 1.223.

In previous posts about King Lucas I told you that we discarded procedural design for the levels so, together with my partner Laura Suárez, I was in charge of designing them one by one, drawing each single tile, gauging each jump and placing each danger so that the player finds them in the right place. To design such a large number of rooms may seem a heavy task (and it is) but thanks to that work we can empathize with the user at every single room in order to deliver a good bunch of details, references and thrilling or chilling moments that an algorithm wouldn’t have been able to design procedurally..

Defining the rules

Even though the level design is not procedural, the disposition of them inside the castle grid is done (almost) randomly, so when we designed them we had to define some basic rules so all of them could fit together:

They’re all the same size (40×30 tiles).
They all have 4 doors to enter/exit.
When a player gets inside a room he has to be able to reach at least one of the other doors.

In addition to the rules for connecting rooms, we also had to take into account the speed of the hero as well as the height and length of his jumps, because in every platformer is very important for the player to feel that any tile is there for one reason (Super Mario games are a good example of this). For this we designed several prototypes and tested with different users until we found the ideal speed and jump for an optimal control of the character (3 tiles high and 4 tiles long).

We know the parameters of the character’s movement are far away from the frenetic canons that actual platformers set (that let the player move faster, double and triple jump or even bounce on the walls), but this is completely premeditated because we always had clear that the character moves had to be as close to the 8 bit classics as possible.

The tools

Facing the challenge of designing more than 1.000 rooms was complicated, we needed a tool to design them as easily as possible as well as design them based on tiles so the game wouldn’t weight several Gb’s (and get minimum loading times). With these factor in mind we chose Tiled (Tiled Map Editor), a free tool great for 2D level design. Thanks to Tiled we could design rooms at a pace of 5 to 10 rooms for day, what meant designing all the rooms in 5 months working full time.

Although the game was going to be completely 2D, when we experimented with the possibilities of Unity we decided to extrude the platform tiles to get some depth and reach a 2.5D style. This, together with the dynamic lights, gives a more modern touch to the graphics of the game.

The pieces of the puzzle

The elements we used to build the rooms of the castle are absolutely inspired by the 8 bit classics and, as you can imagine, it won’t lack lava pits, spikes, ladders, sliding platforms, water, barrels, secret passages, enemies and NPC’s.

When Laura and me began the design of the rooms, programmers couldn’t work full time to this game (they were working on several advergames for external clients), so most of those elements weren’t designed yet and we couldn’t used them in the rooms. Along the first weeks we created aproximately 100 rooms that only had spikes and ladders, something quite boring but it was great to get used to Tiled and, by the way, create some quieter rooms that add some relax moments to the narrative of the game. The programmers implemented the rest of the game elements little by little and each time they delivered a new one we squeezed it to the maximum and this is why you’ll find some rooms full of slugs, or spiders… probably it will be because we had that new element the week when we designed that room and we were looking forward to use it! That’s all!

The developers finish all the elements when we had about 700 rooms so we still had time to create some hundreds of rooms that included a more balanced number of different elements and this way get a more homogeneus castle.

The inspiration

To design levels is funny and the proof is that many people pays for it in games like Little Big Planet or Mario Maker, but it is also true that when you spend 8 hours a day doing it it’s easy to run out of ideas. It’s because of this why we found inspiration in our daily life and you’ll find in King Lucas’ castle rooms to celebrate moments like the day we designed the 700th. room, affection blinks to my partner Laura, a lot of hidden messages, christmas trees or even the castle of my hometown (Villena), where I usually go to walk with my dog, Willy (that is a character in the game too).

When everything fits

As I’ve said in other posts, the main income source of DevilishGames is the advergames we make through our advertising agency, Spherical Pixel, and indie projects such as King Lucas are an undercurrent task for the team members who are not working in other more inmediate projects.

Usually, our development team is busier than our art directors and, in some way, this fact has conditioned the development of King Lucas because for a long time we have had to design and test rooms alone without being able to visualize the whole castle, the whole game. It may sound incredible but in almost 4 years fo development we only have been able to test a complete beta version 4 or 5 months ago, when we finally could check that all the rooms fit magically as we imagined (even better). No doubt this testing was a tipping point that led us to change our company strategy and invest more resources in King Lucas instead of looking for external clients as we do usually. Whether we are right or wrong only time and Steam players will say!（source：gamasutra）