游戏邦在:
杂志专栏:
gamerboom.com订阅到鲜果订阅到抓虾google reader订阅到有道订阅到QQ邮箱订阅到帮看

Ben Brode谈游戏开发中经常被忽略的关键领域

发布时间:2018-08-28 09:24:05 Tags:,

Ben Brode谈游戏开发中经常被忽略的关键领域

原作者:Brendan Sinclair 译者:Vivian Xue

Ben Brode在暴雪工作了15年,现已离职创立了工作室Second Dinner,过往的工作经历使他反复注意到一些“很重要的小事情”。

Brode将在这个月的西雅图PAX Dev主题演讲上对此进行详细讨论。这位前《炉石传说》总设计师最近与GamesIndustry.biz分享了演讲的部分话题——一些对玩家影响很大又可能被低估了的开发领域。

游戏操作感

“玩家们一般不会去注意或者谈论紧凑的操作,”Brode说,“但作为游戏设计者,你必须要把精力多放在这方面,才能体会到操作的紧凑和拖拉之间的区别。”

熟悉暴雪的玩家都知道他们对这一点的重视,Brode甚至提到了前创意总监Rob Pardo经常强调的一句话“操作是关键”。我们在暴雪的很多动作繁琐的游戏中很容易看出这一点,但是Brode说它对于像《炉石传说》这样的卡牌游戏也同样重要。

“我们克服的难题之一是如何使玩家能够想多快就多快地出牌,”Brode说,“即使游戏动画还在运行,你也可以不断地打出下一张牌,只要你的头脑和手速够快。”

Hearthstone(from gamesindustry.biz)

Hearthstone(from gamesindustry.biz)

为了实现这一点,他们运用了客户端预测技术来弥补服务器返回数据产生的延迟, Brode认为这是值得的。

“游戏操作必须尽可能让玩家感觉紧凑,决不能让玩家在某一刻感到自己想做某样事情但是系统却没有在执行它,”他说道,“游戏的响应速度必须超级高,这也是我们在暴雪这么多年以来一直致力于实现的东西。”

匹配制度

Brode建议开发者们重新思考一下传统的匹配制度。当谈到游戏匹配时,人们通常会认为一个好的匹配制度应该确保双方胜率五五开,因为如果玩家一直输,他会沮丧甚至彻底放弃游戏。

“当你尝试把胜率低于50%的玩家拉到50%以上时,你也将那些胜率超过50%的玩家强行降低了,”Brode补充说道。“那些原本胜率70%的玩家会因此流失,当然胜率30%的玩家留存率会变好,但总流失率没什么改变。因此我们做所的是创造一个更好的匹配系统——帮助某些玩家把胜率提高到50%将会伤害到同等数量的其它玩家,并且对降低流失率没有任何好处,至少在某些时候。”

在Brode看来,除了浪费时间却不能真正提高用户留存以外,强制的五五开胜率还会对玩家的长期留存产生不良影响。

“当玩家升级了技能和等级后,他们匹配到了更强大的玩家,这使他们无法知道自己到底有没有在游戏中变强,因为匹配机制模糊了这一点。”

培养积极的社区关系

Brode演讲的话题也并非全关于“小事情”。他也将谈论一些显然很重要的话题,这些话题多年来在业界引起了不少讨论,玩家与开发者之间的关系和摩擦就是其中之一。

游戏即服务的思维和社交媒体的影响已经从根本上改变了开发者和玩家的互动方式。玩家可以更直接地接触到他们最喜爱的游戏的开发者,再加上如今的游戏不是几周就能结束的,玩家们可能会玩上好多年,因此开发者们更需要参与到社区互动中。Brode说,提高这些互动质量的一个好方法是同时增加它们的数量。

“我们在《炉石传说》上做得很好的一点是,我会在很早的时候就尝试参与人们的讨论,” Brode说,“互联网的匿名性使人们对彼此的态度有点差。而一旦有人让他们获得了存在感,大家就不会那么针锋相对,并能进行更有成效的讨论。这在很大程度上缓和了社区情绪,并且玩家们希望开发者能够参与其中。所以我规定了社区团队什么时候要发布什么内容,或者我会自己把它发布到我们的社区中,或者我会尝试成为其中的第一批评论者。这几乎每次都能取得效果。”

Brode说,让开发人员像这样参与社区互动的最大原因之一是为了让人们了解开发人员的思维过程。也许终端用户无法真正地理解,但是让他们知道为什么事情是这样的,至少可以为建立适当的对话提供基础。

“我认为我们的工作是解释我们的观念,这样我们就能进行非常有益的讨论,”布罗德说,“我们当然不可能一直正确,但如果你能够与玩家分享你所有的决策和思维,他们就能站在你的角度思考并进行真正的讨论。”

当然,并不是每个问题都能这样愉快地解决,有时候,玩家和开发人员之间的关系可能会恶化,比如最近的ArenaNet解雇事件以及玩家们对《无人深空》开发商Hello Games的持续针对,当被问到这二者是否是玩家与开发者之间摩擦上升的证据时,Brode表示不确定。

“如果人们投入了很长时间在你的游戏上并且对它充满热情,他们会将自身带入到游戏中并且希望游戏是高质量的,”Brode说,“如果他们已经选择了游戏,但是游戏质量却不高,这将给玩家带来很差的印象。我认为玩家们将高度期待他们选择的游戏,并且希望它是非常优秀的游戏,这是很合理的。他们保持很高的标准,他们也应该保持很高的标准。“

Brode认为,与其把这种关系视为一种对立关系,不如把它看作是一种合作关系。

“一种更积极的态度是,玩家和开发者的目标其实是一致的,即做出一款优秀的游戏。我们也是彼此的工具,”Brode说,“我们可以互相帮助,使彼此变得更好。显然,作为开发人员我们最能够取悦玩家,但是玩家可以为我们提供有效的反馈,让我们知道他们对我们所做的一切的看法,这将使每个人都感到愉快。我认为这种合作对于游戏、玩家和开发者来说都大有好处。有时候很难从这个角度来考虑问题,但我认为这是处理玩家与开发者之间关系的正确方式。“

4)关于加班

Brode不仅希望建立开发者与玩家之间的良性关系,同时也希望建立开发者与他们的作品之间的良性关系。他希望开发者们能够通过更严格地控制范围来实现这一目标。当游戏开发超出了开发人员和设计师的控制范围时,他们往往倾向于通过加班来弥补。Brode更希望每个工作室能够把工作时间当成一个固定的常量,而不是一个用来弥补糟糕的范围控制的变量。根据过往经历,他认为这一点是可以实现的。

“开发《炉石传说》的时候,我们从来没有加班过,”他说,“我们团队的文化非常严格,即确保人们不会透支。这是来自高层的决定,希望每个人都能得到良好的照顾。这是非常重要的。我经常听到一些人超长加班,这对我来说太疯狂了。你不能以这种方式长期经营游戏或公司。”

Broad承认有时候他也会加班一两个小时,但主要是因为服务器故障或一些其他紧急的情况需要大家一起解决。然而,他从不在周末工作,加班绝不是团队的常规,并且如果经理发现员工在加班一般会进行干预,希望他们不要太过劳累。

“这实现起来是很难,你得参与竞争,”他说,“你想提高质量、控制成本,又不想过度加班。很多人确实能够兼顾,但根据我的经验,懂得缩小工作范围比为达目标而增加工作时间更为重要。有时候做减法太难了。我认为我们这个行业需要努力做出正确的取舍。“

本文由游戏邦编译,转载请注明来源,或咨询微信zhengjintiao

Second Dinner co-founder Ben Brode spent 15 years working at Blizzard, and in that span he noticed one idea that kept coming back up time and again: “Little things that matter a lot.”

Brode will be going into that theme in greater detail when he delivers the keynote address at PAX Dev in Seattle later this month. The former game director of Hearthstone spoke with GamesIndustry.biz recently to go over some of the subjects of his presentation, areas of development whose impact on the players’ experience may be underestimated.

“Tight controls are a thing that most players do not notice or talk about,” Brode said. “It’s a thing you have to, as a designer, pay a lot of attention to on your own if you want to learn the difference between very tight controls and loose controls.”

Anyone familiar with Blizzard is likely aware of the company’s emphasis on this point, and Brode even mentioned former chief creative officer Rob Pardo’s frequent insistence that “control is king.” And while it’s easy to see that at work in Blizzard’s more action-heavy games, Brode said it was no less important for a card game like Hearthstone.

“One of the things we did that was difficult to do was to let you play cards as fast as you wanted to,” Brode said. “Even while animations were still happening, you could play your next card and your next card and your next card, playing the game as fast as the decisions and your hand could allow.”

This involved writing a bunch of predictive tech into the client so it could cover the gaps before it heard back from the server, but the results were worth it in Brode’s estimation.

“It had to feel as tight as possible, so there was never a moment where you had something you wanted to do but the game wasn’t doing it”

“It had to feel as tight as possible, so there was never a moment where you had something you wanted to do but the game wasn’t doing it,” he said. “It was super responsive, and that was a very important thing at Blizzard that we spent a ton of time focusing on.”

Matchmaking is another area where Brode is encouraging developers to reconsider the conventional wisdom. One common attitude he encounters is the idea that good matchmaking will ensure that players win about 50% of their games, with the idea being that any player who loses too many matches is very likely to get frustrated and stop playing entirely.

“When you are trying to raise win rates to 50% for players who would otherwise lose more than 50% of their games, you’re also actually trying to lower win rates to 50% for players who would normally be winning more than 50% of their games,” Brode said, adding, “It’s actually possible that by squishing everybody to win 50% of their games, the guy who was winning 70% starts churning on average more, and on average the 30% winners churn less, but the total effect on churn rate is nothing. So the work we do to make great matchmaking systems–presumably to help players get up to 50% win rate–actually hurts an equivalent number of players and gives no benefit in churn rate, at least sometimes.”

Besides the potential for wasting a lot of time and effort without actually improving player retention, Brode said another downside to enforcing 50% win rates is that it obfuscates bigger-picture progress that can keep players engaged over the long haul.

“It makes it hard to tell when you’ve increased in skill, or when you level up that you’ve gotten better at the game, because the matchmaking system hides all that from you by finding you harder opponents as fast as possible,” Brode said.

Not every topic in Brode’s talk is going to fit under the description of “little things.” He’s also going to address some clearly important topics that have generated no shortage of discussion in the industry over the years, among them the player-developer relationship and crunch.

“When you can impart all of the decision-making and thinking, then players can level up as designers, understand where you’re coming from, and really have a good discussion”

Games-as-a-service and social media have fundamentally changed the way developers and players interact. Because players have more direct access to the developers behind their favorite games and those games can now be played over the course of decades instead of weeks, developers have a greater need to interact with their communities. Brode said one good way to boost the quality of those interactions is to boost their quantity as well.

“One of the things that worked really well for us on Hearthstone was that I would try to comment very early on in a thread,” Brode said. “There’s this phenomenon online where the anonymity of the internet makes people treat each other a little bit less good. And when somebody made their presence felt, it got everybody into less of a flaming scenario, and into a more productive discussion. That helped the community sentiment a lot, and players like developers being around. So I just made sure whenever the community team was going to post anything, and I would either try and post it myself to our community subreddit, or I would try to be one of the first comments on it. I felt like that worked across the board basically every time.”

Brode said one of the biggest reasons to have developers engaged in the community like that is simply to explain the developers’ thought processes. Those aren’t always instinctively understood by end users, but letting them know why things are the way they are at least provides a foundation from which a proper dialogue can develop.

“I thought our job was to help explain the reasoning, and then we could have really helpful discussions,” Brode said. “Certainly we weren’t right all the time, but when you can impart all of the decision-making and thinking, then players can level up as designers, understand where you’re coming from, and really have a good discussion.”

Of course, not every issue is going to be happily resolved, and at times, the relationship between players and developers can become abusive. When asked if he sees things like the recent ArenaNet firings or the continuing rancor against No Man’s Sky developer Hello Games as evidence of an increasing amount of friction between the audience and creators, Brode was uncertain.

“When you have a game that is a live game people are playing for a long time and are passionate about, players associate themselves with that game and they want the game to be high quality,” Brode said. “If the game isn’t high quality but they’ve chosen to play it, it reflects badly on the players. I think it makes sense that players would be highly passionate about wanting the games they play to very good. They have very high standards, and they should have high standards.”

Rather than think of the relationship as a contentious one, Brode believes it’s better for all involved to treat it as a collaborative one.

“We just never worked overtime when I was on the Hearthstone team. The culture on the team was really rigid about making sure people didn’t burn out. It comes from the top”

“The better attitude is that we are aligned in our goals of making a great game, and we are tools to each other,” Brode said. “We can all help make each other better. Obviously we have the most power as developers to make players happy, but players can help make everyone happy by giving us really helpful, useful feedback about how they feel about what developers are doing. And that’s a collaboration that I think is hyper great for the game, the players and the developers. It’s hard to think about things in that perspective sometimes, but I think that’s the right way to approach the player-developer relationship.”

Brode not only wants a healthy relationship between the developers and the audience, he wants a healthy relationship between the developers and their own work. One of the ways he wants to see developers achieve that is through more rigorous scope control. When developers and designers let the scope of a game get out of hand, it’s too tempting for them to think they can compensate for it with overtime. Brode would rather see “hours worked” treated less like a variable that can be tweaked to compensate for poor scope control, and more like a fixed constant that studios need to work around. He knows it’s possible, because he’s seen it done first hand.

“We just never worked overtime when I was on the Hearthstone team,” Brode said. “The culture on the team was really rigid about making sure people didn’t burn out. It comes from the top. It comes from the desire to make sure people are being taken care of. That’s a really important thing. I hear stories all the time about people working horrible overtime hours, and it just seems crazy to me. You can’t run a long-term game or company that way.”

Brode did acknowledge he might have done an hour or two of overtime here or there, largely because of server outages or other live ops crises requiring all hands on deck. However, he never worked on weekends, the team never relied on overtime as a standard practice, and managers would actually intervene if they noticed developers working overtime, asking them not to burn themselves out.

“It’s tough, because you do have competing things,” he said. “You want to increase quality, control costs, and you want to not work too much overtime. A lot of people are doing a good job at it, but in my experience, it’s more important to cut scope than it is to increase hours worked to get the result you need. It’s sometimes difficult to cut scope. I think we as an industry need to work harder at making the right trade-off there.”

PAX Dev is set for August 28 and 29 in Seattle, Washington. It is produced by ReedPOP, the event organizer which recently acquired GamesIndustry.biz parent company Gamer Network. (source:GamesIndustry.biz  )


上一篇:

下一篇: