Designing Achievements/How Good I Should Be At My Own Game

[This was originally posted on my personal blog]

Development is coming to a close on PARTICLE MACE, and in the past few weeks I finally got around to implementing something that I always put off, and which is far too often an afterthought in game development: achievements. I'm not a big fan of achievements personally; I often find them to be superfluous, and frequently to be condescending trinkets handed out at regular intervals for basically doing nothing beyond playing the game. Every so often I'll encounter one that I find a bit interesting or that I'll spend an hour pursuing, but that's usually about it. Obviously I'm not alone in this feeling as designer as I've seen similar sentiments pop up in blog posts and tweets very frequently. But achievements are here to stay and I do think they can provide at least some additional interest or engagement when they are done thoughtfully. Perhaps implementing my own achievements so late in the process is not the best way to ensure that PARTICLE MACE has truly exciting achievements, but so it goes. They're in now, and overall I'm pleased with them. However, a few of the game's achievements have caused me to question the importance of the intersection of my ability as a game developer and a game creator, but more on my crisis of faith in my own arcade abilities in a bit. First, let's talk design.

In thinking about how to create achievements for the game, I tried to think of the handful of achievements that have stuck out to me over the years, as well as the ones I found especially dull. I found that most achievements fall into one of four categories:

Alternative Play Achievements are achievements that reward you for playing in a way that is divergent to the normal style of the game. A prime example is the Pacifism Achievement in Geometry Wars. That achievement blew my mind when I first saw it. By asking the player to survive for 60 seconds without firing a shot, it encouraged a whole new exploration of the space of the game that a player would probably never take on without some sort of prodding. These are probably my favorite types of achievements because they expand the game by giving the player new challenges that just don't make sense to include as a part of the game. Most of the missions in PARTICLE MACE attempt to emulate this style: offering new perspectives on the game and the underlying system by tweaking rules the player may have taken for granted, such as wanting to shoot your enemies in Geometry Wars.

Milestone Achievements are maybe a little less interesting, but can be a fun way of patting a player on the back for their dedication to the game. These are achievements that use some internal stat tracker to point out and reward the player for having done something a lot. A perfect example of this is the Practice Makes Perfect achievement in N+, which is awarded when the player dies 1000 times (a number I have far exceeded in my own N+ game). These achievements don't expand the game the way Alternative Play achievements do, but they make for a fun notification, and act as a badge of dedication if not exactly skill.

Skill Achievements are what the achievement systems built into so many modern game platforms are really about (or at least what they were intended to be about). These are achievements that reward actual skill. I cannot stress enough how much it frustrates me when there is an achievement for beating a game on easy or for getting a score that most players could hit with a bit of practice. These things are called achievements, and I like when they must actually be achieved. True to the game's demanding difficulty, many of the Spelunky achievements are great examples of this. Despite many hours of play, I have very few of the score or time based achievements, because they are not just a little tricky, they are truly difficult. I will probably never get most of them and that's how it should be, because that means the folks who did unlock them know that they completed no small feat.

Finally, there are Inevitable Achievements, which make up far far too much of the achievement landscape. Maybe we'd all have a higher opinion of achievements as a whole if we weren't so used to getting an achievement for finishing a tutorial. These may act as good enough carrot for some set of players, but I've never been a fan, and I definitely don't see them working in a deliberately difficult and unforgiving arcade game. When designing achievements, I try to avoid anything that cannot be missed while beating the main game. It doesn't feel like anything has been achieved besides just playing the game if any successful play though of the game will inevitably involve earning it.* So no inevitable achievements in PARTICLE MACE.

Overall, I focussed mostly on Skill and Milestone achievements. Most ideas I had for Alternative Play achievements were already worked into the design of the missions, so there are fewer of these. I was able to include a few Alternative Play achievements for co-op mode, which was fun since mission mode is restricted to single player. Mostly it's skill and Milestone, though. PARTICLE MACE is a masocore arcade game, so things like number of deaths and score in each mode made these fairly simple to design. I wound up with 20 achievements, which seemed like a nice round number. Here they are as they stand now; no promises that they won't get tweaked before release.

Unlock All - Earned by unlocking all ships. Definitely a milestone achievement, but one I like since it requires the player to have beaten all missions and scored fairly well on each arcade setting. This is as close as I'll get to a beat-the-game achievement.

Die 5000 times  - A perfectly good milestone.

Kill 10000 Foes - Another milestone.

1000 Deathmatch kills - It's really hard to think of much to say about the milestone achievements, but they are fun to have and pretty easy to implement on my end.

Get killed by the Snitch - The Snitch is a special foe that shows up in a few missions. It always flies away from you, making it hard to hit. Managing to get killed by it is definitely a feat. This one is Alternative Play and Skill and is probably one of my favorite achievements in the game.

Perfect Credits Run - The credits in the game are an interactive sequence where the player is invincible. This achievement rewards playing through the entire credits without crashing at all. I like it as it encourages the player to check out the credits and has a nice Alternative Play effect by making a challenge out of something not necessarily designed that way.

Score 3000 on Easy

Score 3000 on Hard

Score 400 on Asteroid

Score 3001 on Easy with the Show-Off Ship

Score 3001 on Hard with the Show-Off Ship

Score 401 on Asteroid with the Show-Off Ship

Loop the music in Hard Mode - Goddamn this is hard. Nathaniel Chambers did a killer job on the soundtrack, and this particular track is over nine minutes long. He did that on purpose because he wanted to make sure players never got to the point where it would loop. So really, he made this challenge more than I did.

Team Pacifist - My homage to Geometry Wars, this takes the pacifist Alternative Play mechanic that is frequently used in the missions and let's two players have a go at it together.

Revive each other 15 times in one co-op game - This is a weird one, but it's fun to play co-op where one of the players keeps crashing on purpose, so I kept it in.

Score 4000 on Co-op Easy

Score 4000 on Co-op Hard

Score 500 on Co-op Asteroid

Mess with the XML - Is there an extra type of achievement category for easter eggs or jokes? I guess this may fall into the Alternative Play camp, but messing with the files the game uses to store things feel like it goes beyond getting the player to play differently.

Earn This Achievement - I'll let you folks figure out how to get this one.

It is worth noting here, that I absolutely do not believe that most players should expect to get all of the achievements in a game. Although Alternative Play achievements get a little leeway as they function more as experiments than indicators of ability, Milestone and Skill achievements should be difficult, and in fact should not be earned by most players. It's not an achievement if everybody can do it. I deliberately designed the unlocks and missions in PARTICLE MACE to be challenging but doable for most players because I want most players to feel that they have explored the game to their satisfaction. Achievements exist outside of the game, though, and that is something that can make them fun as a designer. Since they are inherently removed from the game proper, I have no qualms about making them punishingly hard. After all, achievements are the thing thing players show off on their profiles, so I want to make sure they reward deeds actually worth bragging about.

And this is where I hit the tricky part of achievement design for this game. I made three achievements for the Show-Off Ship, who you may remember from my last post about the ship designs. This seemed to be a golden achievement design opportunity as the Show-Off ship is something of a joke unlock that is unlikely to be used much by players in arcade mode as it is objectively the worst ship in the game (same stats as the Normal Ship, but only one particle). By introducing an achievement for each arcade mode that requires you to get a very high score with the Show-Off Ship, it not only encourages an unusual style of play, but provides a very difficult challenge. Skill and Alternative Play! And if the normal score achievement is 3000, let's make the Show-Off score 3001. That's a show-off move. Sounds perfect.

So what's the problem? Well, I'm good at PARTICLE MACE--quite good, as one would expect--but I'm not great at PARTICLE MACE. I am far better at making the game than I am at playing it. At the moment, I am probably in the 80th percentile of arcade players, and probably only the 60th percentile of deathmatch players (you should have seen me getting may ass handed to me while showing the game at the lwlvls festival last week). The Show-Off Achievements are deliberately very hard, and I don't know that I am good enough to beat them. I haven't yet, although I haven't spent more than a few hours trying.

As a game maker, I wouldn't feel comfortable releasing a game that I could not beat, but does the same hold true for achievements? I know these achievements are possible; I've gotten close enough to know that some people will be able to do it. And these are easily the most difficult in the game (along with looping the 9 minute song that plays in hard mode), so they will be the true badges of honor for the hard core players that do earn them. I just don't know if I should include achievements that I, as the game's maker, might never earn.

I've been going back and forth on this. I test my games constantly and believe whole heartedly in using iteration based on testing as the driving force in my design. Releasing something into the wild which has never been completed feels contrary to the way I make games. At the same time, though, I know these achievements are possible, and for the players that are good enough to get them, they will be worth bragging about. I can get every other achievement in the game, so most skilled players will be able to as well, but these last few achievements are going to bother me.

Maybe I'll change them. I still have some time to make that decision. But maybe I'll leave them as a challenge to my players and myself.

----------------------------

You can still get in on the PARTICLE MACE alpha and try to show me up at my own game (I'm Andy Makes on the leaderboards!) at particlemace.com. And you can follow the game at @particlemace on twitter.

Achievement images from www.xboxachievements.com

*I will give some leeway for an achievement for completing the game the first time. This feels like it falls more into the Milestone camp anyway.