What the Web and games have to teach each other
One of the things that I have been thinking about a lot lately is the way that the current Web world (particularly the Web 2.0 stuff) and games seems to be ships passing in the night. It’s led me to say a lot lately that the two groups have a lot to learn from one another. Right now, with big dot-com folks like Joi Ito calling WoW “the new golf,” there’s a bit more awareness crossing the gap, but sometimes I wonder if the right lessons are bring learned.
So I thought I’d post my quick off-the-top-of-my-head list of stuff that each side should learn from the other.
Things that the Web folks can learn from games:
- Interface. Games bring a lot of interface knowledge to the table; consider that most games offer far more complex environments to navigate, with far subtler information, than the typical webpage. And yet, the current trend towards simpler interfaces in webpages has mostly meant simplifying actual capability.
- It’s the content, stupid. Far too many web services are features, not systems, and far too many of them are intended to grow via user content without being seeded with actual content. The most robust user content communities are those built by fandoms, accreting like pearls around initial ideas.
- Entertainment. Games are about fun; far too many web services are simply not fun. All activities can be improved by adding some fun factor: game-like qualities like collecting, ranking, and so on.
- Feedback. Games understand that everything is about feedback. Websites often seem to forget, and I don’t know why. This is getting better with AJAX, but there’s still a heck of a lot of forms of feedback that are missing, particularly persistent feedback.
- Identity. One thing games typically do well is provide identity and context. I am not referring solely to avatars and characters, but also to themes. Avatars and profiles, obviously, rock. Puzzle games don’t have avatars, and yet the context of Bookworm is memorable, and nobody is going to forget Zuma‘s weirdo frog. Context matters.
- Depth. Games often provide something that is simple on the face of it, yet reveals hidden unexpected depths. It’s implicit in the models games provide. Yet often, a given web service (or even a new application or tool) has no hidden depths. It is what it is on the surface. There’s more enthusiasm for continuing to interact with software when it keeps revealing cool stuff to you.
Things that the game guys can learn from the Web:
- Digital distribution. There aren’t any websites that you buy in a store. That’s because stores are an outmoded way to distribute digital data. The prime reason to do it is because you want to retain control of the entire process from generation of data through distribution and onto the playback mechanism. But bits, by and large, do want to be free.
- Platforms. Web services are full of APIs that connect services and apps, magnify utility, and allow mashups and greater user content. Games, even those which are designed for modding, don’t really embrace openness. The industry is pretty determined to be a content creation industry, but games are not content, they are systems content lives in.
- Everything is a database. The web is built on databases; games aren’t databases, they are models, but they are typically models that interact with databases. Embracing your databaseness opens up all sorts of possibilities for how you interact with the data.
- Small pieces loosely joined. The web has figured out that bite-size chunks are what make sense for the largest amount of people. Sure, lots of bite-sized chunks aggregated into a site like Amazon or eBay makes something that’s big overall, but it’s got some advantages over linear structures: easy to jump in anywhere, easy to do things in different orders, easy to search and index, and easy to add to.
- KISS. Games are in love with overcomplication (particularly the “mainstream” games industry, which is anything but, targeted as it is at mostly hardcore gamers and hobbyists). Most websites do something highly targeted and simple, and do it well. Crazy game budgets are a symptom of a problem, not something to emulate.
- Client agnostic. Web guys rely on standards and assume that any damn browser might interact with their content. They provide alternate versions for differing client platforms. Us, we often design completely new games for different platforms, then release them under the same name.
I’d add “process maturity” to the list of things that game guys could learn from the web (or more specifically software product development) folks. Every day I read about challenges faced on the gaming side that we faced developing software products back in the 90s (clients and web stuff). Stuff like change and configuration management, requirements management, and developing iteratively.
Heck the other day when I was reading about your game modeling language and I kicked myself thinking … we did exactly this with UML back at Rational.
On the flip side, I think the web folks tend to think of what they as nothing more than creating technology. I’d like to see the web approach their craft as more of a craft and try to make something “beautiful”. It’s a mindset change that could result in changing how most people experience the web.
You know, I almost put “modern development methodologies” on there — especially forms of agile development — for something games could learn from the web. But games are starting to embrace scrum and agile methods…
UML is notably ueless at architecting game systems, however — great for planning the flow of technological solutions, but the classic game dev problem isn’t solving the tech, it’s solving the fun. 🙂 What I’ve seen happen with use of UML is a commitment to engineering a solution that may be the wrong one; then, since it was so carefully planned, there can be reluctance to remove or change it. You have to be willing to iterate.
There is no doubt that consumer web is undergoing process of “game-ization”. It was well underway for few years at least. habbo, neopets, MySpace is very game-like from inception.
I think there will be more “distillation” of traditional monolithic game model. There is few distinct set of “game genes”, which historically evolved together so far. Let say one gene is “hardcore-ness”. 3D, increasing levels of realism, blow-zombine-with-shotgun, +5-vs-acid-attacks, and install from 5-DVDs types of things. Another is “social knowledgebase”. How guilds forms? How they live? How they breaks? What makes a vibrant game world? What makes it “sticky” and makes player to return? How you balance accomplishment, possessions and jealousy so it propels the community forward? Thats rough outline, yet there is lots of “social” genes encoded in online games now.
So what happening with AJAX/Web 2.0 is that web is becoming more reactive, more real-time, more intensive. More game-like. Web 2.0 companies are actively forming game-like communities and obviously are very interested in the knowledge accumulated in games “social genes”. Yet at same time there is no need or interest to carry on “hardcore” genes with them. We going to see more of web companies grabbing bits and pieces of game knowledge and integrating it. I’m more skeptical regarding the reverse process – multi-million MMOG juggernauts adopting something from the web playbook. What traditional gamedevs are (critically?) missing is not specific list of web technologies to adopt – its whole culture of fast-paced reactivity and flexibility imposed by the web. MMOG culture is built around assumption to ship once per 5y. Web companies “ship” every other day – often on user feedback they just learned last week.
O’Reilly describes Web 2.0 in 8 bullet points; the second one interests me, “Data is the Intel Inside”. It makes me wonder what the “specialized database” he talks to would be for games.
Wipe the slate; I’ll only talk about RPG genre games.
Mechanics aren’t it; that’s the model. That’s really the rules governing how the data gets altered. No… the data itself isn’t gameplay; it seems to be the gameworld. I think that’s interesting, that what drives an RPG is their unique universe, their lore, their narrative, their… character.
To me, that suggests that MMOs should become more hardcore in order to do better. Why? Well, take the analogy of a good fantasy story. The casual reader, who just picks up a book every few years and reads a few chapters, is clearly not going to get into it. Even the reader who reads all the books and loves it isn’t going to really get into it. They’re spectators, not users. You might place virtual webcams inside your game and let people peek in, but they’re not a target audience.
You want the people who write (possibly crappy) fanfics. You want people who read the Dragonlance Chronicles, and then say, “I want to run a campaign in that world.” Who read “The Last Guardian” and say, “Wow, Khadgar was cool! But what if Khadgar had been, say, a Paladin? I’m gonna play WoW and try it out.” Because those are the people who handle O’Reilly’s #3: “Users add value.” You also have to let them, which is #7, and encourage them, which is #5.
That’s just the RPG genre, though. I’d suspect other game genres ought to widen their market, rather than shrink it. *shrugs* But then again, #1, “The Long Tail”, points to niche and hardcore, not mainstream and casual.
Nice timing. I’ve been recently refocusing on similar issues; only involving videogames, Web 2.0 apps, and PLM software. As games reach out to the web (e.g. NBA 2k6 and Nike iD), internet connectivity becomes increasingly integrated into PLM software, and PLM software simulation gravitates towards videogame-like interfaces, there should be some remarkable developments in the next few years.
Some random thoughts:
Interface – Since when do games have good UIs? Okay. Some of them do, but most have pretty horrific UIs.
Platforms – Very important.
Digital distribution – Very important.
Everything is a database – I actually consider this a problem with MMORPGs and most MUDs. You feel like you’re playing the game in a database… Everything in the world is basically the same except for a numerical variation. Text interactive-fiction and older MUDs have the feel of walking through a program… Everything in the world is unique.
Small pieces loosely joined – Yahoo Games (and similar) are small games loosely joined. (I think) the value in a CRPG/MMORPG are the strongly joined sub-games. If you pull them apart into small pieces they aren’t as fun.
Client agnostic – Of course, text MUDs do this already. From the text-MUD example, you can glean the advantages and disadvantages of the approach.
Actually, I think games are generally excellent at displaying complex information; a modern HUD in a game is a marvel of data display in many ways. Consider how many variables must be kept track of, and yet most gamers manage fine.
My general concerns about game UI:
– When is the last time you spent 20 minutes on a tutorial learning how to use a specific web page? (Yes. I know games are more complex concepts than web pages.) Myst (and adventure games in general) are good examples of easy-to-learn UI. MMORPGs are the opposite. (Why do I have to type /p to talk to my party? Why not a checkbox button on the chat interface: [] talk to everyone, [x] talk to party, [] talk to guild?)
To emphasize the point: Uru had WASD style movement. I seems that so many Myst users complained about the difficulty (in moving) that the lastest Myst defaults to nodal movement, and allows WASD as an “advanced” extra. Just think about how many potential MMORPG players are scared away by the difficulties in movement alone?
– Non-standard controls between games (aka: Does space jump, select, or fire for a given game?). Some of this difference is understandable, but not all.
– Hiding functionality amidst eye candy. Sometimes it’s impossible to tell what’s a button and what’s not on the HUD. WoW did a lot of this. I found EQII to be much cleaner.
– Icons instead of text. Oblivion (to pick a current game) shows bizarre icons where text would be more intelligable.
– Non-obvious hotkeys. Oblivion (to pick a current game) is particularly annoying with requiring hotkeys. Not only are they used a lot, but there’s no on-screen display to say that hotkey F2 = function X, so players have to memorize this.
– HUDs in general. It’s far cleanr to have a region of the screen that is “What you see”, and another region for bits and pieces. Modern MMORPGs (HUD crazy) like scattering the bits and pieces everywhere so you feel like you’re driving around with stickers all over your car window and can’t see everything. EQII (as I recall) handled this fairly well by moving the bits and pieces to the bottom of the screen. (Or how about a second monitor for the bits? Many players have a second monitor lying around, and even have dual-monitor support on their video card. The games don’t support it though.)
(Why do I have to type /p to talk to my party? Why not a checkbox button on the chat interface: [] talk to everyone, [x] talk to party, [] talk to guild?)
I smell command-line legacy design… I was actually going to comment with a plausible explanation, but upon returning to the computer four hours later, I couldn’t remember it.
One that I would add to Raph’s list which is at the core of the Web 2.0 trend is “Users Can Make Content Too” A lot of web 2.0 is about harnessing the collective knowledge of users to create a better experience in some domain. wikipedia, flickr, sourceforge are all good examples. This has been so starkly missing from MMORPG’s since I started playing EQ 6 years ago. The pathetic sight of a hardcord MMORPG player going into an instance for the 450th time to try to get some epic item has got to stop. There is no question that a great storyline and a ton of original content have to exist in order for the world to be consuming. However, to make it live and breath, users have to make content, not just consume. The closest we’re getting now in WoW are addon’s which just skirt the edge of the system. What we need is for a game company to make the leap and invest the time and money to build their world from the ground up not just with an incredible storyline and great entertaining content developed in-house, but a great authoring system, security, etc. for users to add to the system in a way that doesn’t break it. Blizzard obviously invested a ton in the former and Linden a ton in the latter, but nobody has put them together yet.
Web interfaces have had a very different path of evolution.
Because the web is far more accessible (client agnostic, free, etc) I can visit a new site in no time at all. Many new sites even. So a given site has ~4 seconds to capture a user. If I can’t figure out the navigation, or get to meaningful content, then bye-bye. accessibility = simple interface, easy access to content.
Games, on the other hand, have a captive audience. I’ve already shelled out $50, or taken the time to download and install your 250MB demo or 1GB client, so I’m going to take some time to figure out your interface. In addition, in MMOs, users are willing to go on very little content for extended periods of time, provided there’s hope/signs that more content does exist. Imagine if Wikipedia took 2 hours to learn to use, and only let you access 5 articles at first, until you read those 5 articles 20 times, then gave you 5 more. captive audience = complex interfaces, obfuscated content for the sake of increasing the amount of content.
(Why do I have to type /p to talk to my party? Why not a checkbox button on the chat interface: [] talk to everyone, [x] talk to party, [] talk to guild?)
I’d say because typing is often faster than clicking and typing. I would be driven insane if a game forced me to move the mouse and click a checkbox in order to change the scope of my text.
World of Warcraft and City of Heroes both offer variations of the pure command line theme that work well enough for most folks. In WoW, the command line remembers the last place you talked, so you only need to /p the first time you talk to your party. You’ll need to /say or /raid or /gu when you want to stop speaking to your party and switch scope. In CoH, you can set the default channel for text to any available channel (local, zone wide, group, guild, your softball team’s private channel) and then are required only to give the slash command if you want to speak in a different scope. If you really want, you can switch the default manually every time, but taking your hands off the keyboard to move the mouse will be slower than just using the slash commands.
CadetUmfer, there’ a lot of convergence now, because web games have very similar constraints — capture the user quickly, and so on. As more games — even big ones — move to the web, the industry will need to learn the same “hooks” used by the small indies.
I agree with Jason on the rapid typing thing; if you’re typing to talk, then typing for many classes of commands makes more sense than shifting UI methods. The downside comes when the commands you need to type are basic ones for gameplay — most novices reach for the mouse these days, not the keyboard. I am personally convinced that a huge part of the large audiences for Asian games is because they have simpler control schemes. If Korean games had not remained isometric and click to move for so long, I think they would not have gotten such a big audience.
“/p” – Several people have listed the historical reasons for the command, as well as the advanced-user short-cut reason… However, “/p” is not discoverable. For such a major feature to not be easily figured out by users is a design flaw. In fact, the only way you know that /p works is by reading the manual (which normal people don’t) or by having another playing saying, “Hey idiot! Use /p!”, which acts an interactive manual with a bit of insult and embarassment added.
Remember, if you want to get casual gamers, you need to continually ask, “Would my (grand)mother (or other non-computer-literate relative) be able to figure this out?” (If you don’t care about casual gamers, then the market will never get larger than it already is.)
The nice thing about the web is that everything is discoverable. If a link can be clicked on it’s blue and underlined (except for web sites that try to be fancy (and confusing)).
jdanner said
Saga of Ryzom is attempting to do just that, with the upcoming addition of Ryzom Ring. It may not have been built from the ground up to include it but player generated content. From what I understand there will be private and public content. Private allows players to run friends and invitees through an instanced version while content that meets cannon requirements will be added to the world for all to experience.
Second Life could be argued to be all about player generated content. And while not really an MMO Neverwinter nights did a wonderful job of providing a toolset that spawned thousands of individual persistant worlds that are very similar to MMO but wholly player created. Oblivion and Morrowind before it also provide a toolset that allow additions to the game world (although that world is single player only the mods can be exchanged to allow similar player experiences).
So its out there, but I think the issue comes back to the UI. A casual gamer cannot easily add unique content using these toolsets (I haven’t seen Ryzom Ring so can’t comment on that) as the learning curve is far too steep. I agree with jdanner, but from the point of view that the game needs to be designed from the ground up with an “easy to use” toolset to add to the game world without being a script/code guru.
However, “/p” is not discoverable.
What I’d recommend is an interface that incorporates Mike’s graphical approach, with mouseover explanations for commands that can act as shortcuts, and further the ability to make changes to the interface such that the extraneous clickings are disabled when no longer needed.
One that I would add to Raph’s list which is at the core of the Web 2.0 trend is “Users Can Make Content Too”
I actually jumped over here after reading Costikyan’s latest post, The Democratization of Entertainment. I immediately railed about it, but I thought it was an extraordinary point.
Consider that the Web is also a democratization of the act of publication, and you have an explosive combination, which is precisely what the virtual phenomenon is.
Damn Raph. Those observations are unusually down-to-earth for your own bad self.
Heh, Stormgaard, I suspect anyone who only knows me from the lengthy essay-type stuff or the idealistic musings, will have a misperception of how down-to-earth and practical I’ll get regarding design stuff. 🙂
Raph, I would add one more thing to things to learn from the web — and I hesitate to bring out an old overused word:
Well defined paradigms
Furthermore, each paradigm has a consistent interpretation. What is it we do on the web? We go to online stores, auction sites, buy tickets, read new articles, find out about a company (and its services/products), discuss on boards, read blogs, and search for all of the above mentioned sites.
Amazon is simple not only because of its design, but because it has all the elements of every other online store. It has a shopping cart, it has an online catalog that you can key word search or browse via categories, it has a “Your Account” page, etc. REI has the same set of things. Each of the paradigms have a consistent interpretation no matter the site. That greatly increases usability.
The well defined paradigms exist in the gaming industry — but the interpretation varies widely from game to game. For example, if I am in an adventure based MMO, I need to group — but the way I form that group varies slightly from game to game. The controls are different. Each game has the same concept of tank, healer, damage dealer, and utility — but a slightly different way to assist.