Thursday, December 17, 2009

Familiarity Bias

We act preferentially towards familiar phenomena. We find reasons why the familiar is better than the unfamiliar. We defend what is familiar because we know its strengths and weaknesses. We avoid risk through avoiding change—even known annoyances and negatives are OK in comparison to unknowns.

But is the familiar really better? Familiarity and system quality may be entirely independent. Unfortunately, many of our quality judgments are formed with reference to the familiar and not to objective criteria, especially in games. There’s no avoiding it.

“The only intuitive interface is the nipple.” We travel from interface to interface, from game to game, and what we find easier to learn is whatever happens to be closest to that with which we are familiar. The familiar does not depend on what games were of the highest quality, though—we met our first games for reasons entirely separate from the fact that they were good games that we thought were worth meeting.

When I was 8 or 9 and was confronted with Ferrari Grand Prix for the NES, I messed around with it until I understood how to play. I had no concept of game quality, only an experimental feeling of fun or annoyance. It was novel enough to see the graphics move around on the screen when I pressed buttons. I didn’t play that game very much, but I have fond memories of it and remember it as an enjoyable game. Does that mean that the game is good by any objective criteria? Through some brute force process of experimentation and through the dumb luck of my parents buying me that game, I learned how racing games work and that I find them to be somewhat fun.

The universe of games I can like is limited to the universe of games I’m exposed to. There could be an amazing genre of games hiding away in a niche on the other side of the internet and I would never find it except by chance.

I don’t remember how I found the sports management sim (also called “text sim”) community and genre that they so ardently and loyally support, but I’m very happy that I did. Front Office Football 2007 is probably my favorite football game for any platform. I would have had no idea that the game even existed if it weren’t for blind luck on a google search 5 years ago (when the game was on its last iteration, FOF 2004). At first I resisted the game because the interface is not for the faint of heart. The game sports the best (commercially available? game-format?) football simulation engine that exists—for that reason alone I gritted my teeth and blundered my way through hours of pain. Soon I got past the learning curve and now I absolutely love the game. I pick it up several times per year and play through 20 or 30 seasons.

Front Office Football 2007

When I talk with the community that has developed around FOF games, I find that people overwhelmingly support the game’s interface even though it is objectively ugly, a pain to learn, and often annoying to use even for veteran players. It’s clear that people like this interface primarily because they’re used to it.

And people get used to insufficient or broken game mechanics in a similar way. Consider the debate over building and unit selection limits for StarCraft 2 (the third point in this post). The original Starcraft had unit selection limited to 12 units, though the endgame regularly saw more than 100 units on the field for each side. What seems like a technical limitation ends up being a point of severe contention. SC is one of the most popular eSports in the world. Performing hundreds of actions per minute is crucial to achieving victory in the very competitive Korean leagues. Most of these actions involve circumventing the selection cap to maneuver larger numbers of units simultaneously. Instead of three clicks to move 36 units (click, drag to select, click to move), it took 9. This triplication of clicking for a simple maneuver needs to be carried out thousands of times in a SC match in order for one side to be successful. This is artificial difficultly. People defend a lower unit selection cap in SC2 because this arbitrary skill test separates the the wheat of pro players from teh chaff of amateurs, in a certain way. I think this is candidly ridiculous—the skill test of the game should be inventing and enacting successful strategies, not outwitting the interface—and Sirlin agrees.

When what has come before plays such an important role in what we enjoy now, we run into difficulty having anything resembling an objective discussion. Anyone suggesting meaningful innovation (or even a small change to something fundamental) is trampled by thousands of players who can’t understand how anything different could work. And there are the nostalgia addicts who refuse to understand the difference between playing a game as a wide-eyed 10-year-old and playing a game as an adult with fully developed mental faculties.

Designing a great game involves overcoming a surprising amount of familiarity bias. It’s important to also be aware of our biases when we try to hold productive discussions about games. It’s difficult to be aware of our biases and doubly difficult to mitigate them, but if we can succeed, the bounty in meaningful discussion is enormous.


Anonymous said...

"I find that people overwhelmingly support the game’s interface even though it is objectively ugly, a pain to learn, and often annoying to use even for veteran players. It’s clear that people like this interface primarily because they’re used to it."

I've been playing EvE for a couple of years now, and the same thing could be said about that game's interface.

evizaer said...

Here, again, is an issue of familiarity. I'm used to the interfaces of management sim games. EVE's interface is a good interface among management sims, so I am predisposed towards liking it.

Tesh said...

Tangential thought: For a couple of months, I hid the bulk of my somewhat lengthy blog posts behind a "more" link, purely to keep a lot of articles on the "front" page. It turns out that those links just annoy readers who use RSS feeds. I don't use an RSS reader, so I never thought about it. As it happens, several of my more vocal commenters use RSS readers, and were bothered by my nefarious links. I even received a couple of emails about it. Very polite emails, to be sure, but clear requests for a change.

It was bad UI for those readers, driven by a simple oversight on my part. I didn't think about what readers would do, and I didn't know about RSS feeds in the first place. I've since learned, and made my site more user friendly for those who actually use it. I went with what made sense to me, rather than what made sense to those who were on the other side.

I can't help but think there is a lot of that sort of bias in software development. (Case Study #1: Blender... man, that program's UI is atrocious, but it makes sense to someone somewhere. Just not the typical end user who might be coming from Wings, Max or Maya. Pesky users.)

Oh, and I wholly agree with you and Sirlin; to me, game design is about letting players play the game and suss out the high level strategy. The dubious "skill" of learning how to min/max the UI is something I'd like to eliminate as much as possible.

motstandet said...

How do you feel about first-person shooters, where a large part of the "skill" is in the mastery of the input device?

Dblade said...

Anon, it's interesting. I just strated playing EVE and the interface isn't all that bad. It has its quirks, but it doesn't seem to be the terror people make it out. It can be cluttered, but you do a lot more in EVE than in other games,so it balances out.

I think this is a very good article though, and I wish more designers realized that being familiar with an interface varies from person to person.

Unknown said...

"the skill test of the game should be inventing and enacting successful strategies, not outwitting the interface"

Could not agree more. And that's my biggest gripe with WoW style MMOs. You struggle with the interface of 100 keybinds, rather than struggling with outwitting and outplaying your opponent.

By contrast, the minimal interface and simplicity of control in something like Counterstrike is one of the greatest strength of FPS genre. (Probably a reason why Shooters haven't really evolved much in years...)

Longasc said...

Familiarity Bias unfortunately does not seem to be restricted to the interface.

The whole trinity concept and the idea of an endgame consisting of dungeons and raids, plus every MMO needs to have crafting and housing (which makes me wonder why WoW still has no housing - Wolfshead demands it, and while I think it is debatable if WoW is really a good game to feature housing at all, they usually tend to appease everyone with a bit of everything) and of course levels is very similar to familiarity bias.

It is also interesting to see how strong the desire is to have "tank" builds in Guild Wars, so that the rest of the party can sit back and safely nuke from a distance.

I fear the familiarity and nostalgia involved with antiquated mechanics and concepts is one major of the many reasons why the MMO genre is somehow stuck in a routine that is very obvious to veteran MMO gamers. The game part of MMOs is not their strong point, so maybe it does not matter that much... but superficial and cosmetic changes to the system don't really bring the genre ahead.

But there is change - instanced content and micro transactions for example. They were seen as the ANTICHRIST, HELL ON EARTH, all that, years ago.

Nowadays they are embraced (dungeon finder, DDO) and micro transactions are no longer totally demonized.

Raydenuni said...

"How do you feel about first-person shooters, where a large part of the "skill" is in the mastery of the input device?"

The difference here is the core mechanic of the game. In an FPS, I would argue that the core mechanic is "running and shooting." This is what you do over and over, it's what makes the game fun, it is what you learn to improve upon. If you removed everything else, but you left in running and shooting, an FPS would still be fun.

In an RTS like Starcraft, I would argue that the core mechanic is more along the lines of managing resources, economic, military, and temporal resources. None of those are inherently related to interface and controlling units. Controlling your units efficiently via your mouse is relatively unimportant compared to your strategy and tactics.