Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
November 1, 2014
arrowPress Releases
November 1, 2014
PR Newswire
View All
View All     Submit Event

If you enjoy reading this site, you might also want to check out these UBM Tech sites:

Designing Achievements/How Good I Should Be At My Own Game
by Andy Wallace on 08/26/14 12:40:00 pm   Featured Blogs

The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community.
The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company.


[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
  • Milestones
  • Skill
  • Inevitabilities



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 guys 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 And you can follow the game at @particlemace on twitter.

Achievement images from

*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.


Related Jobs

Rensselaer Polytechnic Institute
Rensselaer Polytechnic Institute — Troy, New York, United States

Assistant Professor in Music and Media
The College of New Jersey
The College of New Jersey — Ewing, New Jersey, United States

Assistant Professor - Interactive Multi Media - Tenure Track
Next Games
Next Games — Helsinki, Finland

Senior Level Designer
Magic Leap, Inc.
Magic Leap, Inc. — Wellington, New Zealand

Level Designer


Nathan Mates
profile image
The "inevitable" achievements are usually done as a way to get feedback as to aggregate user choices. With a large sample size, you can see what difficulty people really play on, as opposed to reading forums where a vocal minority might be distorting the picture. Designers can also see what mission(s) have the most dropoff in players, and tweak things if a mission is causing ragequits.

Hugo Clarke
profile image
I think the achievement system as we know it came into being around the time PS3 was released, as I had never been exposed to such a system before 'trophies' (although something similar to trophies may have existed prior to PS3.)

As a gamer, I have always despised the trophy and similar achievement based systems, for exactly the same reasons Mr. Wallace brought up in his criticism of it.

My philosophy on games (which admittedly does not represent the feelings of the average gamer) is, *never* remind me that I am playing a game. Even if a game isn't particularly realistic, I want to be absorbed into the experience without annoying gimmicks and superfluous awards designed to give the developer information.

Instead, I want unlockable content, which is attractive (and fun when it has been unlocked) that will drive me to complete even more challenging aspects of the game. This could be costumes, characters, different game modes, weapons, pretty much anything.

In short, proof of my achievement. Even a different title screen after beating the game would suffice. And NEVER award me for watching a cutscene!

John Flush
profile image
Interesting take on the types of achievements. I would disagree on the 'inevitable achievements' though. I expect my games to have about 50% of the points go to this. If I have played your game end to end and saw the first and the end I deserve more than 50% of the points of said game. I shouldn't have to jump through hoops like a dog to feel like I completed anything in your game. I paid for it. I saw the start to the end more or less, reward me for it.

Maybe it is because they were named 'Achievements' and not 'Reward Points' so developers get the wrong idea here, but I think they should be viewed as the later more. Achievements total up to a meta score that shows that I'm rewarded for what I play. I think 50% of the points for a game is saying we appreciate you for buying and playing our game. When I see less than that I usually feel screwed over and not rewarded for even playing the game. That is if I don't see achievements as a huge waste of time in the first place... but here they are and everyone has to use them. Reward standard play with standard points and I think 50% is a good measure.

Then spread the rest of the 50% into the other categories. Milestone's are definitely not my favorite achievement type as they don't express mastery in the least, they describe time sink. What if someone is good in a game and finds a way to win without doing 5000 shotgun kills? maybe they hate the shotgun? Well, now you just reminded your userbase that to 100% your game you have to suffer through 5000 repetitions that you might despise. That sure does leave a good taste in your mouth right? Achievement and completion should not be grind. Grind sucks...

My absolute least favorite though is multiplayer achievements. Lets lock some of the game behind a community that might or might not thrive for very long - time sensitive achievements also fall into this. These usually get compounded into lame ass grind too... "Beat 25 people online." - Grind and time sensitive! Achievement unlocked: "horrible achievement design". As a user I have done this on a few games and I always get burned out on the game by doing them... especially if I'm a late comer to the game. Never remind the player that you don't value their time or you want them to suffer to feel 'completion'

Now, in response though to my own previous comments, achievements should match the game you are making. If you are blizzard you have to have grind achievements. That's the kind of games they make. Same thing with subscription games... But as a consumer these are things I actually look at. For example: "Over 75% of the achievements are Online only? hum, won't buy that one I'm not much of an online gamer"

Just my .02

David Paris
profile image
Because Steam (and other platforms, but I spend more time on PC than consoles at the moment) has no consistent guidelines for what achievements should or should not include, there are already undoable achievements out there. There's also horrible event based achievements (play vs one of the game creators, etc... ) and other stuff that the average player just won't ever have the opportunity to acquire.

Although I'm basically a believer that every achievement should be doable, I don't think it would be any great loss if it turns out one of yours wasn't.

You could also always tweak it later if need be, after monitoring the global achievement rates, and bump it down slightly if only the cheaters can ever get it.

David Canela
profile image
I couldn't care less either way about completing achievements, they are extrinsic and not really part of the game for me, most of the time. The worst are the ones that interrupt my sense of immersion and I think there should always be an option to turn off the achievement/trophy pop-ups.

However, "alternative" play achievements are great imho, they motivate you to explore non-obvious ways to play and give you a sense of feedback: the machine aknowledges what you do and recognizes it, that in itself is pleasant and ecourages you to mess around even more, trying to see what else it will recognize. It's like playing The Stanley Parable. "Milestone" achievements offer the same, but in a much weaker form, because they're much more predictable and obviously arbitrary (e.g. why do I get the achievement at 5000 deaths and not at 4000?)

Andreas Ahlborn
profile image
I like your idea with the show-off ship achievement,
and no I don`t think the developer of a game has to be a virtuoso at his own game,
who beats it on the highest level.

There is an anecdote about the austrian composer Franz Schubert that illustrates this point: Schubert was not a very good pianist and said about his most technically demanding composition: "the devil may play it"

In general its very difficult to design achievements to please every fraction of your players. I´m someone who most of the time doesn`t even look at them, my son is a completionist who has to get 100% of them.

After 30+ years of active gaming I remember merely a handfull of achievements that really meant sth. to me at the point where I attempted them.
All of them had to do with the fact to attempt sth. which you believed to be impossible at first
(like this one for example:
nt/69349-Lone-Wolf.html )

Curtiss Murphy
profile image
Why add achievements? You started with, "I'm not a big fan of achievements personally" and then ended up in some twisted conclusion with "... But achievements are here to stay." As both an Indie and professional designer, I think you might have jumped to conclusions. Achievements exist to provide 'goals', to help players get into flow (google for the win), to provide intermediate accomplishments, or to provide end-game content for deeply engaged fans. They aren't 'here to stay', and they aren't 'inevitable', they are just another tool in the designer's toolbox, and it's just as likely your game might be worse for using them.

Kevin Fishburne
profile image
Maybe make a parody of achievements that would compel the player to attempt to discover them. Like "crouched over a corpse for more than five seconds" or "hammered the pause button to attempt slow motion" or "set chair leg on turbo button...cheater" and weird things like that.

Overall I think achievements are pretty lame. General metrics are fine when performance is important (lap times), an announcer stating "head shot" in their baritone best is satisfying, but formalizing accomplishments with icons in a grid just seems out of place. Perhaps an exception could be a sports game in franchise mode, where you can view your trophy rack, as it's grounded in the idea of the game.

Thomas Henshell
profile image
I liked your analysis of achievements and I agree with your conclusion.
For what its worth, I think it is ok to release achievements you personally can't do but know someone could.

Jason Seip
profile image
If even a single playtester is able reach a difficult achievement, I say leave it in.

Hell, even if you can't find one...some player out there will do it and feel all the more special if completing an achievement is rare.

Eric Salmon
profile image
Personally, I think achievements are weak motivation. It sounds like you already incorporated the best ideas in the actual design of the game which is the best course of action. The section of the audience that actually likes achievements primary just wants to earn them all in the shortest possible time, so use your own systems and to motivate things you really want players to try, and the achievements to attract that section of the audience.

Personally, I don't even glance at anything without a tangible, in-game reward. A little popup just isn't worth the effort for me and is really just distracting and immersion breaking. New ships, new weapons, or even something that doesn't need to be balanced like new ship skins or something might be worth it, though.

Oh, and definitely don't worry about the skill issue. If 20% of your playerbase need and want more of a challenge than you do, it's worth putting it in. I should probably have it tiered so that less skilled players have challenges to attempt and improve their skill, though. Also, big pet peeve of mine -- if you have rewards for beating the game on hard, and also on normal and easy... give ALL THREE when you beat it on hard. It drives me crazy when things unlock on easy mode and I have to play through bored to death to get it.

Benjamin McCallister
profile image
I can't get over how frustrating it is to have almost every single article serve as a trojan horse advertising a product.

John Flush
profile image
I find this comment a little toxic. I would prefer to see articles that pull on direct experience myself, and that usually comes with them talking about their own game, studio, or whatever they do for the day job in the article so I see little wrong with the article as it stands.

Benjamin McCallister
profile image
I'm not trying to be toxic, John, so I apologize for that. I guess I'm just becoming more and more aware of indie guerrilla marketing tactics and native advertising, and after a certain number of articles read in one day that are all, at least on some level, shilling a product, I got frustrated.

The fact that you exist and have an interest in these types of content is the reason these articles keep being written and published so I mean, it has a market.

I guess being a part of reddit communities and other places, I've become accustomed to the "Write an article that may have some interesting topics and/or design merit, and use it to announce/promote your game" page of the indiedev playbook. And seeing it in action so often makes me cringe a bit.

I'm sure the author is a decent guy and no one is really harmed, I just tire of seeing so much native advertising.

(in case anyone isn't familiar with the term)

EDIT: I should add, in case a mod should wander over, that I'm in a situation where at this point I can no longer tell what is native advertising and what is not, which I guess would indicate that the layer of trust is already somewhat diminished.

Josh Foreman
profile image
Thanks for laying out these categories. I feel the same way you do, but maybe with this lens I'll be able to incorporate them into my early design more holistically.