Ahead of the Curve: The SpaceChem Postmortem

[Music] goodra live hello my name is Zack Barth I work at Zack Tronics industries and we made a game called space chem my part in particular was just design in production but today I'm gonna tell you about things that went right during the development of space chem some things that went horribly wrong and if I'm fast enough some questions from you guys so you ready yeah first we're gonna talk about Space Camp so raise your hand if you've heard of Space Camp yeah keep your hand up if you've played Space Camp keep your hand up if you have a love/hate relationship with Space Camp yes exactly so for those of you who have never heard of space came before this is a screenshot of Space Camp and it's pretty self-explanatory I don't really need to say anything else so a question we get asked a lot is what inspired Space Camp and so some of you may be familiar with some of our older games it's kind of a it's a direct descendant of a game called the Codex of alchemical engineering which is a lot like Space Camp but harder if this is kind of hard to believe that that game was a easier version of a game called manufacturer you programmed a factory in Lua not very accessible which is funny because I made space games so when that's actually based off of a television show called how it's made which is awesome Space Camp is also inspired by a park in Seattle called gasworks Park I don't have anybody from Seattle here but if you go here it's like being in a space cam level and it's a really cool park it was a Superfund site super polluted they cleaned it up and made into a park so another question we get a lot that's actually really hard to answer is what kind of game is Space Camp and so we invented a term called design based puzzle games and it leads to a follow-up question of what is a design based puzzle game so we just made that term up and so we're gonna explore that right now the first answer I will give you that question is a definition so a design based puzzle game is a puzzle game where solutions take the form of a design what is the design this is a definition that I found on Wikipedia that I think is really great because it's kind of complicated but it covers all the points so a design is a specification of an object that complices goals in a particular environment using a set of primitive components satisfying a set of requirements subject to constraints or in normal people talk a plan thing that does stuff in a place using parts with constraints so yeah another another way to look at what a design based puzzle game is is that you kind of know it when you see it so traditional puzzle has usually just one solution with very little replay value because most of the gameplay is in trying to either figure out what that state is or figure out how to get there so portal crossword puzzles both have a single solution on the case of a crossword puzzle it's very important that has a single solution that kind of was part of it on the other hand it is a design based puzzle game has many possible solutions because people you're not looking for a state you're looking for aid you're trying to build a design that does something so there's many possible solutions and in the process has replay value because oftentimes that involves a optimization and other sort of fun things that make it you know every play value of puzzles would usually don't have them another game Mainyu factorio is a Turing machine game that's awesome and it's kind of similar and then you have games that are puzzle games that are not remotely close to being puzzle games like bejeweled and Tetris and they actually have a lot of replay value because they're not puzzles and you can't solve them so whatever so so in our process we've made a lot of puzzle games that ended with space cam and most recently and there are four aspects that we found a design based puzzle games that are common emergent tools are one and this sort of fits back into the definition so you know a design is built out of primitive components for us these are emergent tools the difference between a non-emergent tool an emergent tool is that non-emergent tools do something and then it's done and then it did it you know then you can have a bunch of tools but you just use them individually if you know like a hammer I guess I don't have to go there but you know emergent tools like in space cam you can combine them together and you get really crazy results by combining you know the sum of the whole is more than the sum of the parts and that's true in space cam and you've got reactors you've got instructions some of our older games we had steam valves and integrated circuits and all that so emergent tools that have emergent behavior as a result also having precise goals and this fits into a design also that it has a specific purpose of something that's trying to solve so space cam you have to build molecules from specific inputs and outputs Kotick Slocombe engineering how to do a similar thing in a construct door you had to match little circuit diagram timing diagrams so it's all very specific so that what players know exactly what they have to build which makes it they can they can build lots of possible designs because they know what they're trying to do you can also add canvases and strains which matches into the definition of a particular environment with some with constraints in space can those reactors and monsters trying to kill you and Bureau of steam engineering their steam robot so you had to fit everything inside and another another aspect as a design based public gain should have many solutions and this sort of fits into the specification of an object that if you're if you have a design space you can create many designs in it so there's many possible solutions and this is definitely present in space come so we've got these three sort of ingredients and maybe a lofty goal this isn't really a good hierarchy because it doesn't fit neatly but what happens so what's the point what is this so well it turns out that if you add all of these things together you guys so they get we call crack for engineers it turns out there's a huge portion of the population that they're their jobs are solving problems they go to school to learn how to solve problems and then you get to real life and you're filling out TPS reports and stuff you have to go to meetings and that's not problem-solving that's politics but then you play space cam and it's just pure unadulterated problem-solving and a lot of people really love that and that's awesome and another aspect of this is that because everybody's creating their own design as opposed to just finding the one solution is this people have a really strong sense of ownership which makes this you know very likely to be a favorite game of their social something I've heard from a lot of you which is fantastic and one of the other cool side effects of this is that making these kind of puzzles is really easy so we did a the first ever that I know of space cam like in person space cam tournament at PAX and this is our ironclad tactic smooth and but we had a bunch of people there playing space cam and thirty minutes before I realized oh I did not make puzzle we should make a puzzle for this so I want to Wikipedia found some cool hydrocarbons found acetone and kind of mocked that up in space Kevin the level editor and figured out that oh yeah that's that's mostly methane with a little bit so if we subtract out some methane we've got a carbon and oxygen and it turns out that the actually a oxygen is two above carbon in a the periodic table so we could take those extra hydrogen's and up fusion there we go and adding some bonders that's probably enough and bam it's a puzzle is it playable sure maybe I don't know and it's and that's sort of the great thing about space chem is that and that's how we made all of our puzzles for space chem was just making something that looks like it was playable but because there's such flexibility and how these are solved that you can just do whatever you want when you're making the puzzles something that falls out of a lot of these these design based puzzle games are some really awesome community features so when we did codex we put it on a flash game website I can't think of a name it's a big one you guys can probably know what it is and they have forums and congregate is that it yeah a lot of people post it in the comments and they're like this is my solution here's the crazy code for it and somebody else would load it up and they say you know what i shaped two cycles off your solution so he's like oh I shaved another five cycles off and so what's going on wait what is happening here this is kind of weird this doesn't happen in a lot of other games because a lot of games aren't like these and and we realized that players like to compete they like to optimize their solutions and they like to share them with other people so with when it came to Space Camp instead of letting people just kind of Duke it out in the forums of congregate which is not the best place to do anything we started adding these all his first-class features so starting with histograms I'm sure all of you who have played Space Camp know about the histograms and hate them and histograms are awesome because they're basically like an egalitarian leaderboard when you have a game with an actual leaderboard you know you're gonna be like fifty thousand maybe like slightly behind like xxx panther fan xxx and and somebody who just like slammed on the keyboard you know and doesn't really tell you like is that did you do well did you not how many people are hacking leaderboards are awful in space chem you've got this litter board kind of but it shows you the distribution it shows you like where the average falls it doesn't have any gas holes name on it you know it's it's very it's very egalitarian it also encourages optimization because you see that when you're to the right of the the the average that you can probably go to the left and it wouldn't be that hard also has the side effect that everyone is a winner and what we mean by this is that when you solve a level when you optimize for cycles you probably are not doing well the symbols but we are logging all of those into the graphs so every time you do really well on one you're doing really bad in another and you're patting out the bad side and it kind of increases the ability of everyone to do better than average in something which is fantastic one problem though is that these are not good for every game and actually the portal there's a lot of fans that like it a couple of fans at valve of space can at least and who wanted to add histograms to portal too so they did and their histograms are the axes are time to beat the puzzle I think a number of portal shot both of which are terrible because you know portal is a performance game space comes a design game if you mess up you have to go back and redo it again I don't think anybody tried searching for pictures of the the histograms on portal 2 and could not find them on the internet which is saying some so it doesn't work in every game but for some games like space scam it is amazing the funny thing though we tried to push her like no we don't want leaderboards so people went and made their own leaderboards if you go to Space Camp net which is probably the domain we should have bought there's there's a space cabin leaderboard where they reverse engineered our file format so that people can upload their save files and put it into a leaderboard that we didn't make so and honestly that's probably better than us making our own because if it gets gamed and whatever it's not ours you know that's their problem so but these are like awesome that one of the great things about a hardcore game like space chem is you've got hardcore awesome fans and like who make their own leaderboards and reverse engineer file formats it's awesome another cool committee feature we did were videos we realized that people were sharing the little codes for savings and then that was actually only because we didn't even know how to make games that save stuff in flash but people were sharing these codes and sharing the designs so we realize you know what we're gonna do we're gonna make it so you can press a button in space cam and it'll export like it'll take you to a website where it uploads like images of your solution you can click on reactors in view inside and really quickly realize wow this is a terrible feature because if you look at that is that a good solution is that bad is that clever is it awful I have no idea so we realized really quickly that is as cool as this feature is it really needs to be a video because otherwise it's useless well then we realized oh god where are we gonna put those videos that's gonna take up so much more space it's gonna just ah what I'm on YouTube that's where you put them because they are glad to host your videos for you so we did so we built a recording uploads to YouTube it goes on there and they're dime all of it links back there eleven thousand space cam videos on YouTube well that's kind of Awesome this is normally where I would talk about how to do this but it's gonna take way too long to talk about now so I'm not gonna talk about that but you guys can talk to me later about that if you care about that so we did in a cross-platform way super cool super simple super reliable and potentially open source to anybody who's interested another another community feature we had were custom puzzles this is our kind of gaudy looking level editors base game that's it's ugly but highly functional and it allowed us to do something really cool this base parts kind of conventional a lot of games have level editors what a lot of games don't do is have level editors that feed into a thing that the designer picks the levels that get published instead of just having them go in so every I'll get rated and there's terrible stuff and like I think super meatboy they have a bunch of user levels but users had to filter through them so instead we did something called the journal so it looks like a scientific journal and kind of fits in the universe of the and I went through four months and hated it but it was cool of picking out all the puzzles that looked like they were good because I didn't play them was I don't have to because it's pay scam and and then published ones and volumes that look cool so and there's a lot of actual pros to this there are no trash puzzles like anybody who made up like and my first puzzle asdf like did not see the light of day there are no puzzles that broke the spirit of the game anybody who didn't know or like I think about kind of making fake chemistry not going in themed issues made it feel cohesive that you know here the first two the one on the top issue nine that you see is actual science these are all reactions that are based on actual reactions from real life so we could do that kind of theming stuff like that it's super awesome I'm getting published is really cool and having a game kind of you you do something for the game and it recognizes you and put your name up there when I got really lazy at the end I just started doing research anthologies that were just based on like the common theme was that a person made him so and that's cool like it was easy for me and it was cool for them because the game recognized how awesome they were yeah the downside of this is that it's a pain in the ass so surprise surprise so those are all some good things from Space Camp now we're gonna talk about the bad things that we did terribly terribly poorly bad job cautionary tale number one is appeal everybody loves science right know everybody loves portal you would not believe how many people played today they took a look at space and they're like oh I'm not good at chemistry I'm not gonna play your game and the funny thing you guys all know is that it's not even about chemistry it's a game about programming odd so but what happened I mean is science really that unattractive no but space chem is here's a screenshot from space cam and Brian if you're here we love you and you did a great job but this is me making a game that is like oh yeah there's gonna be lines and circles and we're gonna have to make this look as pretty as possible on an indie budget like this is space chem battlefield 3 also came out that year it looks like this like which one do you want to play like if somebody shows you screenshots don't like you could buy this you can buy this and given this one's a little bit cheaper but this one has guns in it so but the truth of it is that space chem you know like these kind of games can be attractive because a game called splice which looks beautiful and is a puzzle game and is kind of programmer II it's about like binary trees and stuff and it looks really good and it even looks sciency and people are like wow that looks like awesome science as opposed to that science so appeals appeal is important and we sort of didn't realize that until it was far too late but fortunately for us space cam is addictive and like all addictive substances it is up to your questionable friends to get your hooked on it and I think that's really why space can was successful and I'm allowed to be here today it's because all of you who played space chem and everybody who played space chem told their friends about it and then like you should you got to try this game and they're like but it looks like yeah I don't think so and I'm like no it's really fun I will buy you a copy and because of pushy people who really enjoyed the game it really had an awesome viral reach outbreak whatever you call it that has negative connotation I don't know cautionary tale number two difficulty when we design space chem we were designing for our previous games for all engineers audience we made a game about called construct or about integrated circuit design you're placing silicon and it's yeah so it was these are people who really liked hard games and you know played our old games are like where's the challenge these are easy it turns out that is not the same thing as the games for normal people audience who like something like portal that are very approachable and palatable and well you know they they're easy to play it's nice even in that when they're challenging they're still easy to play in space chem you have to be over 50 puzzles to complete the story it takes up to like an hour if not more for a puzzle I talked to somebody who played the last puzzle for like a week not straight but you know in pieces and and then on top of that there are seven defense missions which are super vague and are like harder than average ok so so this is the difficulty percentages of space chem levels and you can see that some of them are like 50 percent or lower people who boot them up complete them which comes out to a bottom rating of like 2% and and if you compare this to other games it's actually it's more on par cluster to meatboy then games like World of Goo or Bastion which have like a 15 percent completion rate which it kind of leaves end up in question is this a bad thing is this acceptable space chem is really hard it's kind of cool being able say you've made one the hardest puzzle games ever but I think next time wouldn't make it so hard because then more people can play it more people can like it collection erielle number three and we're running out time someone having a blast of this so the tutorials in space chem suck like so you guys know why I don't really the the most important rule of making games is you have to ship it so which in this case is finishing the presentation so so our theory about tutorials because we all know why the space chemist Orioles suck is that when we did this face cream tutorials and put a lot of text we tried to explain it kind of like so you're in school you're reading stuff and everybody in school reads everything and thinks about it obviously now people learn through trial and error they're gonna click stuff and they're gonna click it again they're gonna click it like five times just in case two clicks aren't enough people learn through trial and error so in order to make a good tutorial just make a little like bumper padded room where people can just bounce around and click on everything and ultimately find their way to the door and like move on to the next one so that's how all your tutorial should be more simply you could just play test your game until people get it right and then just keep tweaking stuff and then we didn't do enough play testing with space camp you know okay we're gonna finish up rapid-fire post-launch secrets these are the parts that you guys are gonna love Steam great platform or greatest platform greatest over ninety percent of our sales come from Steam Mac and Linux is it worth it this is a question at the end of the last talk it kind of like we got a 10% of our sales came from Mac and Linux and the split was probably like 90% 8% on Mac and then 2% on Linux so and then again let you go on the Humble Bundle which is amazing what about iOS kind of even space come did okay on the iOS although I think it was mostly PC customers does buying it again on their tablet what about Android the Android so that iOS sales were pretty low compared to PC and Dre sales were 1/5 of that not worth it unless you count the Android Humble Bundle which is amazing and did way better than iOS ever did for us so something to consider funding how did we make space can we work nights we worked weekends we did profit sharing cost very little out of pocket and we all profited wildly from it so yeah speaking of how much yes so we made like seven hundred thousand dollars on space chem and that includes humble bundles steam sales private store sales in the longtail we still sell a little bits Kickstarter's for physical copies just kind of stretching it out as much you know I'm looking as much again but that's that's the name of the game right and how do we handle support and that is I do that personally and this is awesome because when people email me they buy the get anything they buy copy Space Camp we 50% off the next day and and they asked for a poetry or something I don't know if you guys read that quickly enough but he wanted me he bought the game on sale here but the game before it went on sale and then it went on sale and he felt dumb and he asked for a poem so I wrote him a poem and I will just I think I got it off my head okay so there once was a bloke on the net who bought space come but then was upset his game was on sale but I'm full of fail so I think that it's time to make threats I've been I've been wronged and plagued with frustrations and demand poetry as reparations although I thought twice and the end I was nice you can't beat my customer relations so the future what are we working on now ironclad tactics check it out online we'll talk about later thank you so much for coming out thank you see is for being awesome thank you GDC for having me so how much time do we have left I'm kind of exhausted how many do we got like five minutes okay question time rise ask questions okay the talk as fast as me if we're gonna get through this there's microphone there that's a microphone you gotta get to the mic oh it's a my from there yes go okay you had an article at one point about how you thought the demo was a failure oh god yes but but I personally found that the demo completely convinced me to play the game it was half as long would it have also convinced you to play it I don't know if that's true because I liked that it was longer than most demos I think you oughta bought it anyway you think so okay so it's space cat yeah so so so quick so your next game yeah are you going to spend more time refining the demo first rather than like the final thing or so I've been told that a game with a compelling gameplay trailer and no demo will do better than the demo or at least people please demo your game I'll just give you a free copy how about that you can be in the beta thank you Oh next how do you know that a puzzle is solvable and also do you know play tester you can also put space can there are two stages of solvability thus kochiyama tree phase where you make sure that there's like the right atoms and the right combinations and then there's the does it fit without losing your mind inside the little box of the reactor phase so I did the first part on paper and I saw I've never beaten the last level of space camp I can be honest okay no space cam I think I deserve a pass I never beat the last level but I designed it I've spent like a week proving on paper that it was possible passed it off some playtesters they beat it or they didn't beat it they said it was too hard went back to find it to make some of it simple instead of hydrogen it used to be methane that powered everything at the end which is way harder to break apart so and then they're gonna redo the stoichiometry pass it back some people beat it not enough people obviously but that's we've been through that so so you proved it on paper I proved it on paper and then gave it to playtesters okay thank you I'm not saying that's the way you should do it I'm just saying that's the way we did it yeah this is not advice these cautionary tales are indeed cautionary tales I actually played through the game and beat it and thought it was awesome be something you didn't talk about was the story was that like an afterthought or is that something that you had conceived it from the beginning like it was the writing was actually really good so we we start all of our and should be careful with my wife writes the story is so we start every game with the story and there was a talk at pax dev last year about board games where a lot of German board games they make a game and then they slap a story on the back of the box and like hope it made it into the printing of the game and we start every game with a like a well thought-out story because eventually we got to a point say like what do we add now or how should this work and you can turn to the story and say how would it work in the story and that's a big part all of the levels in space chem were inspired through matically by the story and where they fit in like the sort of like the career path of the the protagonist and so for us story is very very important production-wise the story in space can was very simple and very limited because everything about space chem was honestly so ironclad tactics again we're working on now for contrast is a 70 page graphic novel that we have way too many people working on and but it's gonna be amazing because it weaves in with the game and it's it's not quite the same as like a world that has the story in it but we're still pushing that way because it's very very important to us the YouTube feature was super cool thank you how long did you spend getting that into the game just roughly it's one or two programs right it's yeah yeah it's like three I guess depends on how much of a programmer I am but it's it's pretty it's surprisingly not hard once you figure out how to do it yeah yeah cross-platform we use a c-sharp for everything exclusively so we have a way to do it if i'll tell you i later how it works I'll give you the code I know for a little while you gave out the game to schools for free don't Oh still is any anybody here who has a school wants base camp it's yours oh maybe I can't get that are they are there any cool like anecdotes that you've gotten from like professors or anything about okay so the one anecdote I don't know if this is cool or depressing but so so we've given out to schools all over the world and that makes it something for lots really more like 100 but um so a typical school in America will say oh yeah we want to copy space because we want to teach kids chemistry and that's where you guys all laugh because yeah the schools in the UK however there's apparently some sort of strong computing in schools thing where they say oh we want to use space come to teach kids programming familiarity which is makes more sense I give it to schools in Eastern Europe and they say oh we want to teach kids how to be like problem solvers and how to just like approach a problem break it down you know and use problem solving logic and I think that I think that shows differences you know like the different ways on how they use the tool reflects their or their culture but those are also just total like anecdotes idea and anecdotal evidence is the best kind of yeah wait time for more questions one more ah great talk by the way but going off of the subject about bringing the game to schools some games in the past in the game of the schools have also had bundled with like lesson plans or things of that sort when you give the game to schools yeah usually give any kind of suggestions of how you've seen it effectively used in the classroom we just kind of let them have at it or have you had any ideas yourself how you would want to we have we have like a two-page document on like ways you could possibly use it but that's really why we're giving away for free right no school is gonna buy Space Camp like it's it's not happening well I only buy it it doesn't what it really doesn't fit into a curriculum and that's the thing like and it has like there's like people's heads exploding and like we describe the nostril sliding down his cheek you know and little touches like that that that's not exactly what schools are into and and it definitely doesn't fit into curriculum because it's nothing like real life and that's sort of I mean we can talk later like we do a little bit education games on the side and and that really fits in our philosophy of how we make educational games is that like try to make it is like weird and different as you can write Duke okay I think that's all the time we have so thank you guys so much for coming out [Applause]

Channel: GDC

Views: 17,039

Keywords: gdc, talk, panel, game, games, gaming, development, hd, design, spacechem

Id: lH7gL3ivgFA

Length: 24min 14sec (1454 seconds)

Published: Fri Jun 29 2018