Timesink: Why MMOs Take So Long

Then there is the hardest thing to account for: You, the player. For example, one legend from early testing of Ultima Online is that it was originally intended to have something resembling an ecosystem. If you hunted deer extensively, deer would become scarce. Naturally, players got into the game and slaughtered all the deer and then nobody had deer. Another conversation I had mentioned a player who was continually crashing a game and none of the staff could figure out how he was doing it. He was doing it by stuffing his backpack entirely full of useless items, bags and boxes and crates all full of one particular useless item, which eventually crashed the database. No one can account for human behavior, especially when it comes to being a jerk.

Because, let's face it, “you” are a jerk. Not you, specifically. I'm sure you, specifically, are a wonderful person, but “you”, in general, en masse, tend to behave according to the established parameters of John Gabriel's Greater “Internet Fuckwad” theory, where anonymity makes people behave in less than ideal ways.

A knockback effect designed to be used in combat might also be used to hurl people off high places, which may be hilarious, but is usually annoying to the people being hurled off high places. That means someone will find a very high place, hang out there for days on end, and knockback everyone to their untimely demise, then screencap it, put it on Youtube with Yakity Sax on the soundtrack, and make lots of comments about how the idiot developers let something like that slip by them.

Or take /follow, a boring and otherwise ordinary command used to keep groups together, especially in the pre-fast travel days where hunting in a group meant trying to herd 20 people on connections ranging from T1 to dialup in roughly the same direction. You see the utility. What you may not see is the potential for comedy. For one thing, a subset of people get really, really annoyed when a strange character starts following them everywhere. And most people get really, really annoyed when a strange character starts following them everywhere and uses the /follow command to free both hands for gibbering obscenities.

There's several ways to deal with this. One is to assume your players will never, ever do what you don't want them to do, and this starry-eyed optimism has damned many a project. Optimistic designers are eternally surprised that, yes, if they put in a behavior or action that could even vaguely be annoying, a select few will use it to be incredibly annoying. They may truly, genuinely, have not realized that people can be jerks on the internet or that anyone would devote days and days to harassing a female-avatared player because there is a faint, remote, distant chance that there is a real girl back there and maybe, just maybe, she'll like him or at least talk to him.

The second way is to find the griefer(s) in your midst and appoint them to what might be casually called "Team Jerk." Inevitably, there's the guy who always looks on the dark side of life and while he may play merry hell with the design stages, it does prevent people playing merry hell with actual playing customers, though it leads to lots of meetings where everyone glares at him with pure venom in their eyes because he's demolished their great idea.

Let's say we're planning to let healers use an area of effect spell healing everyone in the area and want to play a nice pretty particle effect to connote "Healing." It turns out that the particle effect might cause a bit of hitching in those playing on the low-end machines.

Team Jerk pipes up, "So you're telling me I can stand there in town spamming healing spells all day and lock up all the poor bastards playing on Grandma's Wal-Mart laptop?"

"Yes, but those people can just turn it off."

"So we're going to put all these nice particle effects in the game and then tell people to turn them off?"

Silent glowering around the table because everyone wants out of this meeting. "Yes."

"Okay, so, how are they going to know they're being healed?"

"We'll put in a 'healing' sound effect."

"So you're telling me I can spam healing spells and annoy people by constantly playing this sound through their speakers?"

"Fine, we'll make it so you can only heal people in your group."

"So I can just invite them into my group by telling them I want to give them some gold, then lock up their computers spamming heals."

At this point, someone usually has their hands wrapped around his throat.

That said, there's an inevitability to discovering griefing in a game, and the starry-eyed developers tend to experience the depths to which humanity will sink to torment each other via video game very quickly, especially if easily exploitable stuff hits live.

Alright, assuming you've built the game from the ground up, it's time to go into testing. Good news! Everything you did in all the preceding years is wrong. No game survives first contact with the playerbase, even if the fundamental systems are solid and usually they aren't, not because the team didn't try, but because it is impossible to simulate the effects of thousands of people crashing headlong throughout the game, attempting to tear it down and min/max their way to victory.

One team I talked to was genuinely astonished to find their timelines blown because they hadn't accounted for the fact that players would literally sit in their chair playing until they passed out, wake up, and keep playing until they passed out again, and they would do this until they "beat" the game, then they'd go complain on the forums about the lack of content. Even when they looked at the character logs and saw that the players in question were logged in and active, save for random 3-5 hour bursts of inactivity after 72 hours or so, it was still hard for them to comprehend.

Another issue that inevitably arises is the never-ending problem of familiarity. What seems crystal clear when you have three years worth of experience with game systems and the interface can be completely opaque to someone approaching it cold. Assumptions made during development--well, this seems easy to understand--come crashing down the first time a tester stares blankly at something "easy to understand." Something that seemed fun to the guy that thought of it might be incredibly annoying to everyone else. How the team copes with this can say a lot, though it usually comes down to whether a designer is willing to work with the feedback or sticks resolutely to The Vision even when The Vision may be doing the wider game a disservice.

And it goes without saying that beta testing now comes with the burden of being, essentially, a game's audition rather than a final tuning phase to get everything right. Go too long and everyone's convinced your game is a buggy mess and they don't buy it. Go too short and they're convinced you have something to hide, which you may or may not actually have. The current dilemma with betas is, basically, your game has to be actually completed and ready to ship for the beta, or your players will go bonkers. Of course, making sure your game is actually complete and ready to ship requires something resembling a large-scale beta, but these are the dilemmas that define us.

So, now the game is out, and the real struggle begins. Why oh why did we not spend more time in production?