'Wow!' Microsoft has learned nothing about usability in twelve years

Ironically, this is the second time I'm writing this entry. The first attempt to write it involved using IE7 on my new Vista laptop. IE7 helpfully crashed on me losing everything I had written, so here goes attempt numero deux, this time using FireFox on Vista. (I also just downloaded Picasa using FireFox, since it was clicking the Picasa download link that crashed IE, and I wanted to give FireFox its fair chance to crash.)

Back in 1995 I was working as a "Usability Architect" and one of the things I would do was try to minimize keystrokes and clicks for users performing common tasks. Also around this time, Microsoft started pushing Windows (95 or 3.x, can't remember) with an ad showing someone typing a copy command into DOS while announcing that now, with Windows, an operation that previously took dozens of keystrokes could be accomplished with a single click.

So, I counted the number of clicks required to perform this operation (using the "a double-click = 1.5 clicks" metric) and it turned out to be 47. More, in fact, than the number of keystrokes required to perform this same operation.



The reason I am reminded of this is the recent remarks made by some Microsoft apologists (although, surprisingly, not Paul Thurrott) arguing that Apple's ad making fun of Vista's permissions dialogs is unfair, since Mac OS X requires the user to type his/her password in the same situation.

Now let me just make this perfectly clear. I am using Vista for the first time today on a brand new Dell Latitude D620 (nice piece of hardware, by the way; no I didn't pay for it so don't tell me what I should have bought ;-) ), and Apple's ads if anything understate just how dumb, annoying, and ineffectual Vista's permissions dialogs are.

1) Unlike in the ad, the question isn't always phrased the same way, and it's not clear what the unsafe choice is. The ad's annoying security guy is actually more usable and consistent than Vista's permissions dialogs. E.g. sometimes you're asked if you want to allow an action (e.g. when running a program), and sometimes you're asked what to do (e.g. when downloading a file).

2) Unlike in the ad, sometimes you get asked to allow or deny something that isn't clearly an action. E.g. when I was installing a piece of open source software (SciTE) which doesn't have an installer, I got asked to confirm the actions of creating, naming, and selecting a folder in the "Program Files" directory.

Here's a step-by-step breakdown of what happens when I create, rename, and then delete a folder (errrr directory) in "Program Files" under Vista:

I go into "Program Files" (which used to be harder to do!) and then right-click and pick "New | Folder".
1) I am told that this action is denied. Would I like to "Continue" "Skip" or "Cancel"? I continue.
2) Windows needs my permission to continue (system modal!). "Continue" or "Cancel". I continue.

I rename the folder "test".
3) Denied. "Continue" "Skip" or "Cancel". I continue.
4) Permission! "Continue" or "Cancel". I continue.

And then to be tidy I select the folder and press DELETE.
5) Am I sure? "Yes" or "No"? Yes.
6) Denied. "Continue" "Skip" or "Cancel". I continue.
7) Permission! "Continue" or "Cancel". I continue.

So, to create and name a folder in "Program Files" I need to click through four confirmation dialogs.

And might I add that every single one of these dialogs still has the major usability issue that has plagued Windows since version 1? When you create a dialog box to confirm an action -- "delete" say -- instead of asking "Yes" or "No" you should say "Delete" or "Keep File" (say) so that the user doesn't have to read the dialog text every freaking time.

Might I add that I have absolutely no idea what happens if I continue and then cancel or skip. (Edit: actually I tried all the permutations and essentially "skip" and "cancel" have the same effect ... so why have skip at all? Indeed, why have two dialogs to ask the same question ... it's like each dialog is produced by a different team somewhere in Microsoft. It's exactly like that. Funny how a bunch of open source developers can provide a better integrated (and more secure) OS than Windows.)

I'll probably go into more detail on just how badly Vista sucks usability-wise later, but here are some little chesnuts to keep you amused:

1) It appears that IE either won't resolve web addresses unless you've typed in the full "http://..." or takes a random (long) time to do so.

2) A lot of dialogs make Windows modal before they appear. E.g. if you save a picture from IE the "Save" dialog (sheet?) doesn't appear immediately, but the window immediately becomes non-responsive (without visually indicating anything). Lovely.

(Again, inobviously locked up Windows have been a blight on Windows since Windows 95 when "multitasking" was

3) Microsoft still hasn't learned to "ask before the long operation". E.g. when I tried to run a Unity demo (requiring a custom ActiveX control) I get the yellow warning bar in IE, when I ask to install it there's a long pause, then it asks me to "Continue or Cancel", then another long pause and it asks if I want to install the plugin. Putting aside the fact that two of these questions didn't need to be asked, they could all have been asked before downloading the file. The most egregious example of this, of course, is when you install Windows -- but thankfully I didn't have to install Vista on this laptop.

Anyway, a more "realistic" ad might go something like this:

The Mac guy talks, does stuff, and isn't interrupted, but when he tries to throw his cell phone away he's asked whether he's sure and if he is, to provide a password.

The PC guy talks, does stuff, but before each action he is asked a series of differently phrased questions such as "Permission to say bad things about Vista is denied? Continue, Skip, or Cancel." "Continue." The lights go off, a spotlight shines in his face. "Saying bad things about Vista is dangerous. Continue or Cancel?" "Continue". Next, he tries to perform a common operation (e.g. breathe) and is asked "You are attempting to breathe. Run or cancel. If you don't want to have to confirm breathing in future, uncheck this box." Nearly blue in the face, PC unchecks the box, and then is immediately asked "You are attempting to breathe. Run or cancel. If you don't want to have to confirm breathing in future, uncheck this box."

That's right, unchecking that box doesn't work.

"You are coming to a sad realization. Abort, Retry, or Fail."

Addendum

Let me just quickly say some nice things about Windows Vista.

1) Sleep mode actually appears to work. That's right, I can close my laptop and it goes to sleep, and open it and it wakes up (into a PRESS CONTROL + ALT + DELETE to log in screen -- why why why? Talk about dumb things that have been in Windows forever. It's like wiring your house so that to turn on your lights you have to stick a fork in a power socket. At least it's not "Start | Shut Down" any more, because the "Start" menu is the "shiny fake Aqua Windows button" now).

2) When you make changes to a directory they appear to more-or-less update automatically semi-instantly.

3) I like the way gadgets get a slab of the desktop to live in (versus the way Apple hides all your widgets in their own private Idaho).

4) Windows now has an Expose rip-off! Sure, it's ugly, less interactive, and less useful, but it's there. Ugly because the "pseudo 3d overlapping fanned deck" layout isn't very well done. Less interactive because when you point at a window it doesn't tell you what app it belongs to. And less useful because it doesn't appear to support drag and drop (but then drag and drop has always been barely half-assed in Windows). Once again Microsoft shows that when you rip off someone else's ideas, it really helps to understand them.

Yes, these are back-handed compliments. What did you expect?

A Brief History of Bloody-Mindedness & Game Design

Once upon a time, someone invented computer games. Whether it's chess, or pong, or asteroids, or whatever you want to pick as your first computer game (but not space war) the one thing all the original computer games (but not space war) had in common was that you played against the computer. In order to keep the game interesting for stronger players, the game would get harder as you "progressed" (either it got better after you completed a level or you could simply pick a difficulty level).

Fast forward a few years and a new concept was introduced to computer games: content. Suddenly, instead of the game's content being essentially fixed and the speed, number of targets, and/or computer player's reaction speed (or in chess games, depth and quality of look ahead) being raised, games had content. Level 2 might look different from level 1.

Content-based games quickly outstripped content-free faster, faster, more, more games in popularity and market penetration. Perhaps the last big gasp of the old-school games was Tetris, while the first mighty flash of content-based games was MYST (a game that millions of people who didn't play games played).

All of this sets aside player vs. player games (such as Spacewar, DOOM, etc.) because human opponents can be endlessly challenging (well, so far). Humans are bloody-minded as players, but I'm discussing bloody-mindedness in game design.

Why Be Bloody-Minded?

It turns out that the more vivid the content and the more responsive the game world, the more people seem to like the game. It also turns out that the more vivid the game world and the more responsive the game, the more expensive the game is to develop. All those graphics, animations, and special case pieces of scripting cost money to think up, design, implement, debug, and then support.

So, therefore, it makes sense to make maximum use of all the vivid, responsive content you developed. Why produce gorgeous graphics and animation that folks will only see once? Solution: stick it in a really difficult part of the game the player will have to repeat again and again and again.

And, why produce vivid, responsive content players won't see at all? Solution: make your content as linear as possible, forcing players down a "tunnel of fun".

And the end result? Extremely vivid and interactive games that are mind-numbingly repetitive and linear. (You may think I'm describing your favorite game in particular, but guess what -- that includes pretty much every triple-A title out there.)

There's actually a third, and far more insidious, reason for bloody-mindedness that I haven't gotten to, but it has its roots in all that boring history I started off with. It's elitism. Almost everyone in the game design business is really good at playing games. And really good players know that what makes them cool, what makes up for their inability to form social relationships outside cyberspace, is their ability to play Q*bert until they get bored, or clock Time Pilot, or beat anyone in Virtua Fighter 5 as anyone. So there's an inherent belief among many game designers that unless you're this tall you don't get to go on the fun rides.

That's right, you may pay Blizzard $500 for a copy of WoW and three years of subscription fees, but you don't get to see the Twin Emperors or Patchwerk or whatever unless you go through a certain (large) amount of tedium and pain.

Enter the MMORPG*

* Massive Money-making Online Role-Playing Game

All three kinds of bloody-mindedness are particularly prevalent in MMORPGs. You can actually see it evolve within the design teams. They start out designing stuff quite idealistically:

"Look, this dungeon is a bit non-linear and entertaining rather than really bloody-minded."

"Oh dear, folks didn't spend six months doing it over and over, we mustn't make that mistake again."

"If we're going to write a really cool fight script, it's vital that everyone has to watch it fifty times."

This is because all three reasons are strongly sharpened. The economic arguments are stronger because unlike with single player games you want to keep people playing for as long as possible and develop as little content as possible. The elitism issues are magnified because a significant and vocal proportion of your player base hates their real lives and seeks solace in the game world where their entire sense of self value is determined by what they've seen and done (and looted) in the game world that most other people have not.

As a final note. A certain amount of bloody-mindedness is a good thing. If it were easy to see all the content in a game, no-one would bother. Difficulty makes a game rewarding -- to a point. Even tedium and repetition can make an achievement rewarding -- to a point. I doubt that any MMORPG player will disagree that MMORPGs have gone beyond this point in the latter department.

Can Anything Be Done? Should Anything Be Done?

The beauty of MMORPGs is that the economics of the game allow designers to spend much more time building content (and gameplay) for those games (from an individual player's perspective) than single player games. Oblivion, say, has a stupendous amount of content, but it pales in comparison to World of Warcraft. Of course, most people play Oblivion for a few weeks and then, happy they've "finished" it, move on to something else.

So, MMORPGs have a lot more content than single player games because they make a lot more money (when they don't fail miserably). But, single players games give you a lot more content per hour played than MMORPGs. One solution is to play single player games. But single player games have plenty of tedium and bloody-mindedness in them too. How much time spent playing, say, a typical Final Fantasy game is spent dealing with interesting new content.

Solution: Competition

Blizzard, right now, is raking in cash. I very much doubt that they increase content development expenditure based on their revenue; they just treat their player base as a cash cow and enjoy themselves. Every new player is (aside from server/bandwidth costs) pure profit.

My view of every MMORPG out there is that it's woefully short of engaging content relative to the amount of time players are spending in it, and that the "hamburger's helper" being used to keep the players playing is a sordid concoction of tedium, repetition, and general bloody-mindedness. Build more content and you can ease off on the tedium, repetition, and bloody-mindedness. But the only way we'll see this happen is if (a) Blizzard suddenly decides, out of a sense of idealism, that they should spend more money building content for players they already have, or (b) credible competitors emerge who don't take their players so much for granted.

Solution: Hire Your Customer

One of the standard "brilliant ideas" when I was in the Management Consulting biz was "hire your customer". You see it everywhere, from "self-service" fuel pumps to "self-service" ticket collection at airports. (If it weren't for 9/11, you'd be checking in your own luggage too.)

It's clear that the ultimate source of engaging content in games is other human beings, but aside from player vs. player combat, and a few games with scenario design tools, no-one has much attempted to leverage the creativity of other players into games. There are obvious problems: the first thing I did with the Neverwinter Nights "Aurora" toolset (which lets you build game content) was create a "dungeon" consisting of a guy who would give you any amount of money or experience and a small dungeon full of superbly equipped enemies to kill and loot. This only spoils my own fun, but imagine if an MMORPG world allowed such things to be added by players.

The big question then is how to allow players to create challenges for each other in a relatively controlled way that allows the universe to constantly grow more interesting without anything getting too out of control.

I won't go into it in detail here, but I think the solution is, in essence, something like Slashdot or Wikipedia. Everyone can produce content. Everyone can moderate. If a game engine like NeverWinter Nights allowed you to jump from server to server you could build a complete MMORPG universe without requiring a single gigantic server farm and incredibly costly infrastructure.

Hey, it's an idea.

Solution: Emergent Behavior

Will Wright's Spore is an attempt to build a dynamic-ish universe around players' decisions, employing a "fractal" combination of random number generators and player design choices to create a complex, dynamic sandbox. Of course, the game never seems to ship, which tends to imply that while all the algorithms and technology are endlessly cool and fascinating, there's not much of a game there yet. (I played this game before, when it was called SimEarth. Sure, it wasn't quite so elaborate, but it was the same concept with 2d graphics ... and it also didn't work well as a game.)

Emergent behavior is something that game designers need to allow for but not surrender to, and be alert for and willing to exploit. For example, the designers of EverQuest added the "Bazaar" in response to behavior they had seen emerge in the general chat channels. (In contrast, Blizzard essentially implemented eBay inside World of Warcraft. This makes for an efficient market, but not so good game atmosphere.) Many of the features of MMORPGs are essentially codified examples of emergent behavior.

Let me give some simple examples of how emergent behavior could be added to WoW.

Imagine that instead of being quite expensive clickable items, horses were quite cheap (say 1g) but they didn't go in your inventory. They were, essentially, NPCs. If you have the skill you can mount a horse, ride it around, and unmount it. If the horse is "owned" by a given player, it becomes "unowned" if left untended and not "tied" to a hitching post. Unowned horses can be attacked by predators, or stolen. After a while they just disappear (starve or wander off). If you pick up an "unowned" horse you are flagged for PvP (and can be attacked by either side). We just got a ton of emergent behavior -- folks may have to leave people to guard their horses outside dungeons. Players can make money watching horses for people. And, just incidentally, horses can no longer be conjured out of thin air after swimming across an ocean or teleporting into town.

Imagine that instead of simply respawning in place some creatures appear logically from some place and move to their "location" along some reasonable path. For example, you might have a series of "trog camps" in Loch Modan, and instead of their just popping into place in each camp, they appear from the Evil Warlock's spawning cave and then wander from camp to camp until all the camps are full. Then the warlock spawns a big monster that rampages about until killed by guards or players. Now, the order in which you kill things matters, folks doing something in place A affects folks doing something else in place B. The whole area becomes a lot more interesting. Oh, and just incidentally, mobs are no longer "popping" out of thin air.

(Actually, I suggested this as feedback to Verant back before it became Sony Online Entertainment. Apparently, MMORPG designers have decided that mob spawning is just the way things are. Ugh.)

Summing Up

Bloody-mindedness manifests itself in games as repetition, tedium, linearity and gratuitous difficulty. It's caused by the economics of game development and elitist attitudes among both game developers and hardcore players to the detriment of pretty much everyone else. Potential solutions exist, but they're either actually difficult (consider Spore), assumed to be difficult, or simply ignored because designers are too lazy and the existing approach is "good enough". In the end, unless players rebel against repetition, tedium, linearity, and gratuitous difficulty, they'll just get more of the same.

Farewell World of Warcraft

As you probably gathered from my earlier post (if you read it), I have become increasingly disillusioned with WoW after the release of Burning Crusade, and my wife and I have decided to quit the game cold turkey. I mailed my valuables to friends in game and we've uninstalled. Buh-bye.

As with all MMORPGs, WoW's chief addictive quality is personal relationships. While the game may become tedious, boring, unrewarding, or repetitive, one's sense of obligation to friends in the game is often the thing that draws one back in. You might stop playing for pleasure, but a friend will plead with you by AIM or email or phone to just come join tonight's raid.

The single worst feature in World of Warcraft is raid caps. A given "instance" has a maximum number of players who can enter it as a group. For most it is five, but for the hard instances -- the ones for which guilds are formed -- it used to be twenty or forty. Almost all the tension in guilds was a consequence of these caps, since you needed to make a guild large enough to handle specific instances (e.g. 40 for BWL or Naxx) which generally meant making it too large and then cherry picking members to fill your raid.

Contrast this with, say, EverQuest, where most raids could be as large as you liked (within the technical limits of servers). Now, Blizzard's designers were afraid large raids "trivialized" content, which leads us to the second-worst flaw in World of Warcraft: bloody-mindedness.

Tedium is to EverQuest as bloody-mindedness is to WoW. It seems that most high end encounters are specifically designed to be possible only for an ideally composed raid with idealized gear performing at 90% or better of optimal output. Furthermore, you can only succeed by knowing the encounter in advance -- i.e. by trying it and failing or by reading spoilers. Preferably both.

It seems to me that Blizzard has made some huge mistakes with Burning Crusade, and I think I'll list a few of them. The first two mistakes are huge because they undermine the personal relationships which are the chief addictive component of the game.

1) They charged for the expansion.

As I understand it, if you sell a game for $40 you get $15 wholesale of which $10 or so is profit (after you subtract production and distribution costs). Blizzard gets $15/month from its players, so every player who quits because the expansion is too expensive is a huge loss, while the gain is trivial. The negative consequences are huge because they can't make the old world more interesting (then folks who haven't bought the expansion might enjoy themselves, and we can't have that) or thread new content into it. The potential cost is huge since I know of many players who baulked at the $40 which made them likely to quit the game. Why risk subscription income for a one off profit equal to a single month's subscription?

The only reason to charge for the expansion was to keep retailers happy, since most of your $40 goes to them (and middlemen) and not Blizzard. This is dumb; it's not like retailers will refuse to sell some new Blizzard console game because Blizzard sold its WoW expansion direct to customers (or gave it away).

I'm guessing charging for BC will, of itself, cost Blizzard 5-25% of its customer base.

2) They changed raid caps.

The proximate cause of our departure from WoW was idiotic guild leadership trying to reshape the guild solely to deal with Karazhan -- the first "end-game" instance -- despite the fact that it will probably be "old hat" within a month or two, and the next set of instances will be 25-man, meaning more wailing and gnashing of teeth.

Many cry "Karazhan isn't like UBRS, because it's hard" forgetting that UBRS (and Strath and Scholo) were very hard way back when. Oh well, those who do not know history -- even having lived it -- and condemned to repeat it.

On our server it seems almost every major guild is self-destructing in one way or another specifically owing to changing raid caps. I conservatively guess this may cost Blizzard 10% of their player base on its own in the short-run, and possibly seriously dent WoW's player base in the long-run.

3) They didn't fundamentally improve their content in any way.

Burning Crusade is literally just new content in the old engine. Or old content with new graphics. Where are the dynamic instances and quests? Where are the dungeons where you can choose your path? (WoW dungeons are, with almost no exceptions, linear.) Perhaps the most promising thing anywhere in BC is that some of your early quests require you to kill "20 Fel Orcs" (say) where any kind of Fel Orc will do. (Versus quests requiring you to kill 15 Fel Orc Sorcerors and 5 Fel Orc Attendants. "No no no, 36 Sorcerors and 4 Attendants is not good enough, go away.") But this promising trend quickly disappears and we go back to the usual slaughters and collects.

If anything, the new content is more repetitive and banal than anything from the old world. I found myself doing a quest where I'm basically killing giant level 66 sheep and thinking "so this is essentially a super tedious copy and paste version of a newbie quest" ... and guess what, it was part of a set of nine essentially identical quests. The quest "line" was a copy and paste version of a near identical quest line (from an NPC with the same name) you do at level 35 in Stranglethorn Vale. Wow, someone gets a credit for "Game Design" for this masterpiece.

Although this last major mistake does not directly attack the relationships which keep people playing an MMORPG, I think that in the long-run this is going to be the most damaging mistake. In the end, BC shows that Blizzard has run out of ideas (or that the folks with ideas have run out of Blizzard).

What might they have done?

In Dark Age of Camelot there are plenty of dynamically generated mini-quests. If you want to work on smithing, you can get a quest which gives you a customer for whom to make something which allows you to make a slight profit by buying or acquiring materials, making the item, and delivering it. Blizzard hates this idea and has designed all tradeskills to be incredibly tedious to improve and intrinsically unprofitable. The only way to make money is kill stuff or get it from other players. If you just want to help the local authorities you can be assigned a random local mission (go kill some bandits).

In Anarchy Online and EverQuest there are dynamic quests and dynamically populated "dungeons". The most "dynamic" thing you get in Blizzard dungeons is some random spawns (e.g. there might be one NPC who only spawns 10% of the time in a given instance, or can appear in one of three places) and loot tables.

In Guild Wars there are instanced outdoor areas, allowing for epic (or seemingly epic) journeys. Every outdoor area in Burning Crusade is teaming with other players eliminating any sense of heroic adventure. "There they stood, in the Valley of Dark Death, facing the Arch Fiend, when three level 70s wandered by and gave them some free buffs and then killed the shadow fiend three times waiting for motes to drop. 'Wanna join up and score the kill credit?' they asked."

Most of the quests in Burning Crusade are ridiculously local, along the lines of understandably dumb newbie quests: "Seek out wolves and kill 8." Hmm, I seem a bunch wandering aimlessly around me right where we're standing... This is probably because quests requiring you to run around in the old world were very unpopular; but Blizzard never figured out this was because the rewards sucked, not because travel is bad. (A quest requiring 15 minutes of travel through dangerous areas generally yields very little money or experience, while a quest requiring you to kill 10 monsters strolling about within a few yards of the quest giver will yield far more. This is why everyone hates "long journey" quests.) In Guild Wars you can be sent on a quest to kill a bear and bring back its skin because the outdoors are big and finding a bear is an interesting task in and of itself. And you don't mind, because the rewards are decent. (Hey, I'm not saying you should play Guild Wars; I'm saying that Blizzard's designers should -- they might learn something.)

Finally, Blizzard could have allowed Horde and Alliance to cooperate and mingle in some or all of the new world. This makes sense with the back story, and it would have afforded huge new possibilities for creating new social glue to hold players into the game. Instead, the exact opposite makes, as I understand it, the PvP realms almost unbearable (imagine trying to complete quests in a city 2/3 full of enemy players). I imagine a lot of PvP realms have become, effectively, PvNo-one realms.

Anyway, I'm done with time-sink games for now, and -- I hope -- forever. When someone brings out an MMORPG whose business model isn't based on keeping people on a repetitive treadmill while charging them for it, I may come back. Meanwhile, I will probably go back to games that I used to consider outrageous time-sinks, like Final Fantasy XII -- my goodness, 50h of gameplay with 25h of repetitious crap? That's ridiculous.