【GDC 2014筆記】Level Design in a Day: How We Used Iterative Level Design to Ship Skyrim and Fallout 3

Bethesda Softworks的資深關卡設計師Joel Burgess在今年GDC的Level Design in a Day和大家分享了團隊在關卡量化上的一些流程和經驗。而我覺得最有趣的一個點是他們所使用的Spaced Iteration迭代流程。這個概念的前提是基於Skyrim團隊有以下幾個處境所發展出來的:

﹣大量關卡內容 (150+ Levels)

有別於一般Continuous Iteration一段時間都鎖在一個關卡內去做好所有四個階段:1.概念發想  2.關卡Layout  3.雛形搭建  4.拋光階段,Spaced Iteration則是第A關的概念發想做完,會接著去做第B關的概念發想;然後才回來繼續做第A關的關卡Layout。這樣做有以下幾個好處:






還記得從前玩過的Dust, Dust 2, Aztec, Italy等CS經典地圖嗎?經典地圖到底有哪些經典的元素?從耳熟能詳的layout中一窺CS:GO的六項多人對戰關卡設計原則:

1: 藉由衝突點Choke points來提升以及促進整個gameplay
2: 衝突點必須放置於攻方到達目的地(objective)之前
3: 兩隊到達衝突點的時間必須一致
4: 限制通往衝突點的入口至一到兩條
5: 提供支持狙擊、近戰和潛行等多樣玩法風格
6: 基於人質護送模式的動態衝突點設計(Italy)


Choke points are areas of the map where the attacking team meets resistance from the defending team before reaching the objective. Choke points are also called control points or bottlenecks.
The attacking team (Terrorist in defuse maps and Counter-Terrorist in hostage rescue maps) must fight through the choke point to reach the objective or retreat and try a different route/strategy.
Choke point areas are specifically designed to enhance gameplay. They are used tocontrol flow, pacing and balance within the map. Whether you are playing Dust (Underpass choke point), Office (Side Hall choke point) or Nuke (Outside choke point), each of these areas are manually crafted. The architecture, cover placement and timing are used to channel each team to attack or defend.
In this blog post I will cover 6 principles of choke point design and choke points used in most played official maps.

Choke points must enhance and facilitate gameplay (combat). To do this you must narrow the flow of your map through architecture down to a single (or double) viewable entrance, such as a hallway, doorway, tunnel or alleyway. This area becomes a choke point where two teams meet and battle for control.
Few examples of choke points:
Dust Underpass/Overpass choke point:
Aztec Double Doors and Bridge:
Italy at Long Hall and after Apartments Alleyway rounding the corner to T spawn:
You want to spend time crafting your choke points because the players will be spending a lot of time in this area.

Choke points must be placed before the attacking team reaches the map's objective. The attacking team needs to fight through the choke point in order to reach the objective (rescue hostages or plant the bomb). You must design your map layout to allow for that. The defending team arrives at the objective and has to defend it by not allowing the attacking team to break through the choke point. This is the key to effective map layout in Counter-Strike.
For example look at Inferno B site Banana choke point, CTs must defend the alleyway entrance:
Mirage site A at Palace Tunnel and Palace Interior choke points:

Timing is very important in choke point design. You want both teams to arrive at the choke point almost at the same time. Defending team can reach the choke point few seconds earlier to set up position. The important part is that the attacking team should never be able to rush through the choke point while the defending team is just arriving there.
Timing choke point arrivals are done in BSP block in stage of your map creation.
Here is the process I use. At the beginning of the round I switch to knife so I can travel at maximum speed and run to the choke point. I use the game clock (round timer) to keep track of the time in seconds it takes me to reach the choke point. Once I get there I write down how long it took me. Then, I do the same thing for the other team side. If the times are different I move the choke point or the spawn points for each team, making sure that the arrival is about the same or slightly earlier for defending team.
Sometimes depending on the map, the defending team can arrive at the objective way before the attacking team. Dust 2 site A is a good example:

Limit entrance options to the objective within a choke point down to one or two entrances. The attacking team should not easily take over the objective. Most maps in Counter-Strike have a single pathway choke point.
Dust 2 (Upper Tunnel) single tunnel leading to the objective.
Aztec (Double Doors):
2 pathway entrance through a choke point offer a choice and strategy for the attacking team. It makes what could be a boring choke point and adds strategy.
If you are going to create 2 pathway entrance through a choke point, make sure that each entrance is clearly visible to the defending team within the same point of view. Avoid more than 2 entry locations to the objective within the same choke point that is not visible from same point of view. Control your choke points and you will control gameplay.
Dust (Underpass and Up the Stairs to Overpass) - both are visible and can be seen at the same time, but giving a strategy option for that choke point:
Inferno (Apartments or Long) two options for the attacking team but manageable for the defending team:
You have to effectively narrow down the options to one or two entrances for the attacking team to get through and plant the bomb. Bottle neck the layout design to funnel teams to these positions.

Each map's choke points should support various playing styles - sniping, close quarter combat or stealth. The more gameplay styles you can support within the map, the more it will appeal to wider range of players. You want to design for strategy and skill. To do this effectively, use distance and cover as a design layout element.
Nuke is a good example of using distance to allow variety of gameplay styles. Outside choke point is good for sniping. Hut, Squeaky and Radio allows for medium and close quarter battles as well as some stealth.
Distance options:
  • Long
  • Short
  • Narrow
  • Wide
You can also use distance vertically (height) as design gameplay element.
Aztec: Double Doors (short and long) good for sniping and close quarter battles:
Train: Middle (short, medium) good for medium/close quarter battles and stealth:
Militia: interior of the house (close quarter combat and stealth)

Choke points in hostage rescue map are dynamic. Hostage rescue does not require very precise and timed approach as it does for defuse maps. It is because choke points will change based on if Terrorist team plays aggressively (attacking) or passively (defending). So when designing hostage rescue maps, you will need to think how the defending team may play.
The main thing you want to do is limit pathways to hostage rescue locations down to 2 or 3.
Italy has 3 main pathways routes from which CTs will come from:
Office also has 3:
Let's me explain a bit further about dynamic choke point design of hostage rescue maps.
Italy is great example. If CTs go right through the market there are 2 battle areas where they may encounter Terrorist.
If terrorist rush, Market will become the choke point:
If Terrorist defend and stay near spawn then Long Hall becomes the choke point.
This defend/attack choke point map dynamic is what you want to create in your own hostage maps.
Here is how to come up with dynamic choke points:
  • First choke point if Terrorist team defends and stays near hostages
  • Second choke point if Terrorist attack and rush towards the CT team.
Make sure to read this in-depth guide about multiplayer map layout creation.

Now that you know the principles let's take a look at some ideas for areas. I will keep this section short because this will depend on the theme you have for your map. Hallways and Doorways will seem like a natural choke point in an interior (office) type map, while Alleyway or Streets is the natural choice for urban (Inferno) setting type.
  • Hallways or Tunnels (Office, Train, Italy)
  • Doors/Doorways (Train, Aztec)
  • Alleyway/Street (Dust, Inferno)
  • Battle Areas (Nuke, Mirage)
Battle Areas: these are more open ended choke points with variety of options and cover within a given area. An example of these are Mirage (Middle), Italy (Market) and Train (Site A):



前段時間 IndieCade East 2014 在紐約舉行。三天的時間來自各地的開發者們聚集此地,參與試玩各種新的已發佈或未發布的獨立遊戲,聽業界著名的設計師和遊戲學者演講,以及參加多人遊戲聚會,認識同好。真的是很棒的體驗!

先介紹下 IndieCade,這是一個在獨立遊戲界的設計師和開發者活動,每年一次,與 GDC 類似。同樣有和 IGF 相似的遊戲獎項評選。相比 GDC 而言,IndieCade 以玩爲主。每年都有很多人在展臺裏展示自己的遊戲,可以嘗試最新的遊戲想法,並與作者交流。IndieCade 在舊金山舉行,而 IndieCade East 則是作爲分會在紐約舉行。在 IndieCade East 中,筆者也有幸在 NYU Game Center 的展位上展示了自己參與製作的三款遊戲,Blind, Ulfbert 和 Shadow (未發布). 得到了很多玩家的肯定,也與很多人交流了自己的想法。展示自己的遊戲,能看到有很多不同的玩家來測試自己的遊戲,得到了很寶貴的經驗。

在 IndieCade East 中,Clara Fernandez Vara 的關於遊戲的敘事設計的一篇講座給人印象很深。她討論了在敘事中會遇到的困難,並全面地總結了對應的一些策略。對於那些對敘事着迷的人,是非常有幫助的指引。而本文則是受到這篇講座的啓發而作。

在此之前,先澄清遊戲的敘事 (narrative):我們在這裏說的敘事,主要並不是指在一些毫不相干的遊戲關卡中嵌入一段故事,而更接近講玩家在遊戲中作出的遊戲行爲,蘊含着故事的含義。








1. 重複使用素材。


有時候,利用重複的素材,也能給故事留下線索。例如 Gone Home 四處可見的箱子。重複的物件可以有一致的行爲,使得遊戲更加易於預測。同時也可以串聯起整個故事。

2. 使用文本。

文本永遠是你的好朋友。有一些敘事遊戲就是以文本爲主的。文本的功能很強大,很簡潔的話語可以將故事的大概展現得很清晰。文本也很廉價,也可以做得很藝術,如 Dear Esther,又如很多跳臺遊戲將文本作爲環境的一部分。當需要考慮到工作量時,就試試文本吧。

3. 提供碎片讓玩家去想像整個世界。

遊戲並不需要完全細緻地展示整個故事。一個有想像空間的故事往往比直截了當的故事更加有趣。遊戲可以僅僅提供故事的碎片,用玩家的理解和想像去完善整個故事。這樣可以大量降低遊戲的工作量,同時也並不會讓敘事變差。比如 Gone home 中,玩家通過探索環境得到故事的碎片,進而發掘出遊戲的故事。

4. 讓玩家去完成故事





多人動作對戰遊戲「Be the Ball」──分享NYU Game Center MFA課程Studio I 的訓練

「Be the Ball」是一款本機最多四人對戰的動作platformer。玩家扮演美國荒漠中兇悍的犰狳,唯一的目標就是用你的拳頭稱王﹣打趴其他犰狳並且將他們投進任何長得像籃筐的東西﹣然後得分!

這款遊戲是NYU Game Center遊戲設計研究所第一個學期必修課程Studio I的期末project。老師Bennett Foddy除了希望學生能藉由這堂production課程學習到具備程式、美術、音樂音效所有面向的獨立開發能力與精神之外,更鼓勵學生大膽在遊戲中做出一些新嘗試。這門課學生共會參與製作三款遊戲專案,每位同學在三個project都會先對班上pitch自己的idea並且進行討論。在成品出來後所有人也會被要求相互playtest, critique給予feedback。

第一個是solo project,但限制學生只能用GameMaker透過簡單的幾何圖形獨立做出一款遊戲。第二個也是solo project,唯一的差別是學生有一週的時間可以製作任何自己遊戲、他人遊戲中需要的asset並上傳至SVN一個共享的folder,這週結束後大家就不能再著手任何asset的製作,而之後自己的所有遊戲素材都只能從這個共享folder中選擇。每個人的素材元件只要被選中一次就加一分。這樣的目的在於除了讓學生擁有獨立開發的能力之外,也在個人asset獲得肯定的過程中找到屬於自己的聲音和強項,幫助每個人未來對自己是屬於怎樣的game designer做定位。

最後一個final project則希望學生藉由前兩個project對自己的了解,以及對班上其他同學的了解,去從每位同學的final pitch中選擇加入/組成團隊進行較細的分工。一個特別的規則是,A同學要加入B同學的project,B是不能拒絕A的。這會強迫學生去思考怎麼針對專案性質做出最合適的分工。每個project最後每個人都要針對"每一位同學在這個project中學習了多少"的原則互相評分,做不記名的comment。

Studio I課堂上的期末playtest session
Studio I共分兩班,兩班同學也會互相交流

「Be the Ball」的團隊組成是一位programmer, 一位level designer和一位project manager。由於團隊中除了我之外沒有其他美術和音樂音效設計,我就得身兼關卡設計以外的職務。例如原本提出的三個關卡設計中有不少event scripting的部份,討論過後則大幅減少以讓我去負責視覺和聲音的內容。由於老師希望每個人的評分原則是"學了多少",團隊中的一位manager雖然是律師背景,但也被指派了角色設計和角色動畫,以減輕我的loading,同時他也能從我和程式之間取得協助。

舉這些例子的用意在於,我們是怎麼面對現有資源做出不管是design decision亦或是teamwork分工的調配。這些都屬於Studio I這堂課所要訓練學生的範疇中。當然,過程中最重要的不外乎是讓學生去執行對自己有意義,自己能放入一百分熱情的專案,然後push每一個人去思考自己的定位。

Studio I期末展的demo現場
Studio I期末展的系上一角


不折不扣的商業藝術品﹣Spec Ops:The Line 特種戰線



Game designer亦然,矛盾的是遊戲身為集眾多藝術於一身的第九藝術,卻往往背負著“遊戲必須好玩”的原罪,迷失在ludology和narratology的爭辯中。

Spec Ops:The Line無疑是一款具有商業考量的Triple A遊戲,卻以少數的design approach去達到讓觀眾反思的目的,著實在商業設計與藝術之間取得了一個微妙的平衡;稱之為第九藝術的一個絕佳商業典範實不為過。


Spec Ops:The Line,並不是一款我們常見的標準軍事射擊遊戲,雖然他的遊戲機制非常的"沒創意",幾乎跟過去的戰爭機器與Mass Effect沒有兩樣,第三人稱的掩體射擊遊戲。但是透過它的敘事手段,讓他這樣古老的遊戲機制獲得了另外一種全新的靈魂,一種只屬於遊戲藝術的象徵手法在這裡被表現的淋漓盡至。

Spec Ops並不是個好玩的遊戲,玩完後會讓玩家心情很差,但是給予玩家的訊息卻具有極大的衝擊力。遊戲一直背負著“遊戲必須好玩”的原罪,但當我們看小說時,或著是看電影時,我們很有可能是想看一場悲劇,為了大哭一場而吸收這些媒體。抑或著是經由這些媒體獲得對於現實與社會的反思。但極少人用這樣的角度去思考遊戲設計。曾經我們一直不停的尋找屬於遊戲的藝術,有人說遊戲需要許多美術、音樂、文學等元素來組成,而前三者都可被認為是藝術品,因此遊戲自然是藝術的一種。但我相信一個成熟的藝術品不應該只是眾多它種藝術的集合。Spec Ops很好的證明了這一點,結合遊戲最重要的元素“互動與Agency”來達到訊息與思想的傳遞,遊戲給予玩家扮演主角的機會,這是其他的藝術型態辦不到的事情,因此這才是屬於遊戲的藝術。

原文出處:Spec Ops:The Line 特種戰線對軍事射擊遊戲的反思