Achievements: mechanics or aesthetics?
Trying to help kick off discussion on this thread: Game Design Round Table 2. Most of this is old hat if you’ve read my thoughts on achievements before.
I get in a lot of arguments about achievements, because I like them and I work with a lot of idealists/purists. From a design perspective, achievement structures are typically intangible, superlative, and exploitative—that is, they create game goals not based on good design or good play per se. They also don’t fit neatly in any work ethic or modus—for instance, where would one cram them into the MDA method? Are they really a mechanic, or are they simply an aesthetic after-effect? The distinction seems to be in whether they affect gameplay for a majority of players (or at all). I would argue that this is what lends them design legitimacy, as well as opening them up to be a double-edged sword in the designer’s toolbox.
Let’s start with achievements that simply note player progress through explicit in-game goals. Achievements that mark one’s progress through a game are helpful both for developing gamer cred and for general stats-tracking. It’s a good thing to have a uniform system in place by which to tell how many of your players have completed a certain percentage of your game, right? As a researcher and blogger, I find it very helpful to follow the Gamertags of my colleagues: if they write about a game, and I don’t think their argument is solid, then the first thing I check is whether they’ve actually beaten it (this is by no means an deal-closer, but it helps understand how much of the game their argument is based upon).
That said, can we really claim that an achievement marking level progression or percentage of hidden widgets collected (following Petrie’s example, which I totally agree with) is a design choice? Progression is already a game goal in itself (for most games), and collectibles usually already carry their own in-game reward (if they don’t, shame on the developer—you’ve just artificially extended your game and not thanked your player for allowing you to get away with it). These are cases where the achievements are entirely an aesthetic effect. They may motivate more players to collect or complete, but your game should already be good enough to encourage that behavior in the first place. A game like Oblivion is rare in that every single achievement in the game marks advancement in a faction or the main quest (it was a launch title, I believe, so no surprise); I didn’t pay attention to these things popping onto my screen while I was playing the game, because it was good enough to encourage play without them.
Moving onto functionally superior achievements now, my two favorite 360 games (achievement-wise) are Mass Effect and The Orange Box. Mass Effect did something really novel early on in the console’s life: it provided in-game stats bonuses based on certain achievements, and these bonuses worked on a meta-level that persisted over multiple playthroughs (and in the case of some, such as the experience boost for reaching level 50, building on each other). These achievements vary from marking game progress and completion, tallying kills with specific weapons and against specific types of enemies, rewarding unique in-game actions such as making love or saving a teammate’s life, and for skill-based play such as beating the game on harder difficulties or playing tactically (measured in the damage ratio between your shields and your health). Some of these are not explicit game goals: for instance, an Adept (magic user) can only effectively use a pistol at first; however, after playing as a Soldier and getting a number of kills with a shotgun, an achievement unlocks that allows you to roll an Adept with the ability to effectively use a shotgun. I was more impressed with this achievement system than I was with the rest of the game’s design, to be honest.
The Orange Box is wonderful for numerous reasons (as you all know). Its achievement system isn’t as mechanically refined as Mass Effect‘s; however, it shines in encouraging multiple play styles that the player might not think of themselves. One asks the player to complete Ravenholm using only the gravity gun, which the level has clearly been designed for. The level designers clearly weren’t allowed to take all other weapons away from the player, but they literally littered the ground and walls with sawblades and other nasties to throw at headcrab zombies. It’s not something I would’ve thought to do without encouragement from the achievement system, and it was really thrilling/enlightening to play through the space in that way. There was another one that rewarded you for carrying a garden gnome all the way through Episode 2 (I believe), which was a huge bloody pain but immensely rewarding at the end—the experience could be likened to trying to carry a baby through a battlefield… in a car without a second seat. There are other examples—little touches, like killing certain enemies by feeding them exploding barrels or hitting cops with a crane—but these are the most prominent in my mind.
Finally, as to Trent’s initial and most pressing question: even good achievement structures can be double-edged swords. My third favorite 360 game, achievement-wise, is Halo 3. I loved unlocking armor pieces to customize my avatar with (although I wish they had more-than-aesthetic affects on gameplay)—I was the only person I knew with the Security gear for a really long time, huge nerd cred. As he shares in his anecdote, though, the multiplayer achievements encourage farming and cheating. Bungie has gone through phases where they crack down on this behavior, but in the early days of the game or after a new expansion it’s hard to find a match where half the players aren’t trying to boost in some way. I’ve never played a public Grifball match where one of the eight players didn’t start asking if they could farm for a Killionaire and end up team-killing out of anger. Following Petrie’s comments on Mega Man 9, I prefer in-game MP challenge systems such as CoD4‘s; these allow designers to craft elaborate achievement structures without giving players the incentive to boost just to show off their global Gamerscore to others. I think multiplayer global achievements are possible, but they need to be designed around aspects of play that can’t be easily exploited (win x# matches with x class or weapon, for instance, might work).
I’ll share one last anecdote about the communities that have built up around achievements. There are websites like Achieve 360 and 360 Achievements that basically encourage sharing of game knowledge and cooperative exploration of new games. There’s a downside to this: many people are just there to learn how to cheat, boost, or get an achievement in the easiest way possible. But the work ethic that has arisen out of these communities can approach the amazing—I’ve seen achievement guides with video and photograph accompaniment that rival those produced by companies such as Brady Games. Completionists are an interesting population to study, because they’re kind of like non-professional beta testers—taken in sum, they know everything there is to know about most games (after a few weeks of release).
Again, this is a double-edged sword. When I started becoming interested in achievements, I would check guides just to make sure there weren’t any that I could miss during a particularly long playthrough. This led me to spoil a lot of games for myself, and eventually I had to stop caring about them. As somebody particularly prone to grind addiction, coin-drop, whatever, achievement structures always have the potential to be exploitative, which is why I’m really excited that Trent is kicking off a discussion of how to design them well. Thanks for reading.