Super Galaxy Squadron pre-post-mortem (or: how a pixel artist pretended to be a developer for a year and things got weird)

As a brief foreword, this is much less of proper post-mortem (or pre-post-mortem, as it were) and much more of a surreal, rambling self-reflective essay. Throughout it I'll have videos, gifs, and even playable builds to show how the game's evolved, which will also be collected at the end for your convenience.

It's currently 3:04 AM CST, on January 14th, 2015. If everything goes according to plan, Super Galaxy Squadron, a game I've techinically been working on for about a year but actually much longer, will release next Friday, just before I drive down to MAGFest to exhibit in the indie showcase. I've spent the last two weeks working literally non-stop to finish it in time. I'm listening to a shitty J-Pop Pandora station to keep myself in the game's faux Japanese, anime influenced mindset. About 24 hours ago I added the last of the things I needed to add, so I spent today sending it to internet friends for QA (and knocked out a lot of huge bugs) as I wait for the final soundtrack and word from the game's publisher that it's going on Steam. I've played through it about a dozen times today to test and get footage, and now, for the first time in 2015, I have nothing to do but write a bunch of words about the beautiful, confusing, ultimately inconsequential trainwreck that led to this moment.

As the title of this post implies, I'm not a game developer. I honestly don't know how I managed to get this far, I'm sure it was a combination of luck and miracles, but no matter what happens now I feel like I've achieved what I set out to do. What did I set out to do? I don't really know that either.

I realize that the idea of a post-mortem is to look back on the whole process, including release and reception, but this has really been a journey-not-the-destination thing for me, so I feel like this still might be valuable in that sense, even if it's not the most useful for anyone looking for actual gamedev wisdom. Since I'm not actually a developer, I couldn't really give it to you either way.

So instead of trying to do that, I'll ask you to join me as I explore just how things ended up like this and where the hell they could possibly go from here (although if you really just want technical stuff, skip to the “What I Did Wrong” section).

A Brief Super Galaxy History

Many years ago, my brother discovered a game called Metroid Prime 2D, a fan demake of Metroid Prime, as the name suggests. He then decided he wanted to make his own Metroid fangame, and since I had always been the artist of the family, he recruited my 11-12 year old self to make sprites for him. I had just gotten my own computer, so I spent most of each day after school making a close approximation of a Jackson Pollock painting and calling it the Wave Beam. Some time (i.e. 7 or 8 years) later, after continuing to do pixel art (even including working on MP2D) and experimenting in Game Maker on small, terrible projects, I thrust myself into the world of freelancing, and at the moment I'm arting for two pretty sizeable games, but that's (mostly) another story.

In late 2012, one of those small, terrible GM projects was a vertical shmup based on Air Force Gator, the award-winning novel by Giant Bomb editor Dan Ryckert.

Video of the very first AFG build

That game sucked. I mean I don't have to tell you that if you watched that video. It was way too complex, for one- the resolution was ridiculously big for pixel art and there was an embarrassing amount of empty space, the plane had four different weapons with branching upgrades plus an interchangeable secondary weapon, all five of which had their own separate ammo pools, and I wanted this ill-advised divide between air and ground targets that you would need different weapons to destroy, which really doesn't work in a fast-paced vertical shmup. Not to mention that everything about it is just mechanically bad.

After a while I realized all that and decided to scale it back, going strictly 8 bit and simplifying the gameplay a small amount.

Playable build of earliest 8 bit version of AFG (Z to select/shoot, X for secondary weapon, Ctrl to switch weapons)

Playable build of the last 8 bit AFG, with an alternate boss and a few improvements

My plan was to make a decent amount of the game, then show it to Dan and see if it could become an actual thing, but around that time he tweeted in reply to some person that an Air Force Gator game couldn't exist because it would be a conflict of interest with Game Informer, his employer at the time. Reading that tweet was the most anticlimactic end the project could've had, and I never touched it again.

Until...

The Rise of the Super Galaxy Phoenix

One day in late 2013 I was looking through some pixel art website and saw a mockup of a sci-fi shmup. “I should turn that shitty Air Force Gator thing into a sci-fi shmup,” I thought.

And thus, Super Galaxy Squadron was born.

Playable build of SGS from May 16th. At this point it was basically just a reskinned AFG, with a few additions (Z to select/shoot, Space for Hyper)

Gfy of the first appearance of the sweet ass boss title, from May 23rd

The idea was simple- fourteen ships, each with unique stats and weapons, including an infinitely rechargeable Hyper weapon. Although I had no actual programming experience, the years I had spent playing around in Game Maker had somehow added up to the skills I needed to keep it going (not that I didn't spend the majority of my time looking things up in the user manual). Over the next few months I slowly built it up, until summer, when I decided it was time to take it to the next level.

Playable build from June 6th. This was when it started to develop its own identity, but the mechanics were still super rough

I committed myself to actually finishing it, instead of letting it quietly die off like the dozens of other Game Maker experiments I'd made. I decided I'd donate all the proceeds to Child's Play, something I'll talk about toward the end of this, and after a bit of research I started planning a Kickstarter for the essential costs of getting it done and out the door.

Gif of ship redesigns

Gfy of the first appearance of Albatross, who's now the stage 3 boss

Playable build from July 17th. Ships started getting unique weapons, and gamepad support was finally added, making it play so much smoother

The Plot Thickens

The first step was to find someone to do music. I spent a week or two casually looking up composers who did both chiptunes and other genres, since SGS's style was turning out less 8 bit than I had planned, and the first couple I contacted never responded. Looking through my bookmarks, I found something I had forgotten about- the Bandcamp page of Random Encounter, who I saw at the Escapist Expo in 2013, for which I had volunteered. Strangely enough, this is the lesser of two career-influencing coincidences involving The Escapist and finding Bandcamp pages in my bookmarks. I emailed them, and to my surprise and delight, they were interested.

The second step was to contact Child's Play to make sure what I was doing was alright according to them, and it was.

I made a website, put together a trailer, submitted to Greenlight, and the time had come.

Kickstarter trailer

The idea, once again, was simple- a $200 base goal for GM Studio Professional, stretchgoals for Mac and Linux versions, and a $2000 total stretchgoal for a soundtrack by Random Encounter. After several delays, mostly as a result of my using the wrong address in my Amazon Payments account, the campaign launched on August 6th. The $200 goal was passed on August 7th. Then it stopped.

Here's the graph so I don't have to explain the whole thing. As you can see, the $2000 goal wasn't met. However, Random Encounter decided to do the soundtrack anyway (thanks guys).

In the middle of the campaign, something else happened. It makes this story several times stranger, more interesting, and oddly poetic but unfortunately I can't tell you the best part. What I can tell you is this- I was hired by New Blood Interactive, a new studio co-founded by Gunnars marketing director and Rise of the Triad producer Dave Oshry. In addition to working on a thing for them, they also offered to publish SGS, including covering any costs left after the Kickstarter.

So at this point, the KS wasn't much of a concern for me.

Final public demo on IndieDB

The Plot Thins Out a Bit

Once the campaign ended and things settled down, I had a lot of questions on my mind, including but not limited to “how the hell do I make a game?”, “why are people trusting me with their money?”, and “I can't handle this kind of responsibility”, which is simultaneously not a question and a colossal overreaction. Paralyzed by 10% existential crises and 90% laziness, I barely even looked at it until winter, after which we arrive at the first paragraph of this post.

Gameplay video from September 26th, which I recorded for my MAGFest submission

Gameplay video from December 5th, when I finally came back to it and it started to become what it is today

And with that, we come to the game itself.

What I Did Wrong

Usually people start with the good things but this way works better narratively, so let's get brutal for a second.

This is partly due to my inexperience with programming, and I'm sure it's a problem that most of you actual developers are very familiar with. Since I built off of the Air Force Gator “engine”, there was already a lot of fat that needed trimming, and for the first few months I kept the same method of working on it for a couple days at a time adding random things and then putting it off for weeks. It wasn't until July that I actually sat down with it and figured out where I wanted it to go, so everything I wanted to add from then on was accompanied by huge changes to the foundation of the game. There are a lot of things I wanted in it that were just too impractical to make because they would've required massive restructuring, so for the sake of my time and sanity I just worked with what I had.

This extends past the development, too- one of the Kickstarter stretchgoals was a Mac version. I accounted for the expense of the GM Studio Mac module, but at the time I hadn't seen that you also need to be a registered Mac developer, which costs $99 a year. I'll establish this better later but I'm phenomenally broke, so unfortunately the Mac version has to be delayed until I can sort that out.

When I said “working literally non-stop” back in the first paragraph, I was using “literally” in the actual, literal, literally sense (I've been sleeping, of course, but aside from that). This isn't entirely my fault, though, since I'm not being compensated for the game in any way, real life has to come first in many situations, but I'd be lying real hard if I said that I haven't had the time to work on it since September. I'm the worst serial procrastinator. In fact it's a miracle that I managed to pull it all together during the last couple of weeks. Do yourself a favor and don't put everything off until the month you're releasing.

I have trouble even replying to tweets with a simple “Thanks!”, because using exclamation marks feels like I'm coming on too strong but not using them looks like I don't care. When I launched the Kickstarter I sent emails to a decent amount of websites, but they read like they came out of an ironically dry Tim and Eric sketch, minus the irony.

Networking and marketing are really as important as everyone says, and because of that, I have 90 followers on Twitter and the SGS Facebook page has a whole 9 likes. The game hasn't gotten any coverage since September, and the only thing keeping it even slightly alive in the public eye is Screenshot Saturday.

This problem isn't exclusive to my professional life, but to keep it short, getting by without knowing people is really, really hard. I got lucky with New Blood, and if that hadn't happened then SGS probably would've been doomed to complete obscurity.

You want to know what happens when your Kickstarter goal is $200? It gets funded in like a day, people see that it's already done with just $200 and think “well I guess they don't need my help” and move on. I can't believe that never occurred to me until I saw people talking about it in my Greenlight comments, halfway into the campaign.

Deciding your KS goal is a lot like pricing your game. If you set it too high for the value people see in the game, obviously no one will think it's worth it. If you set it too low, it confuses people. The idea was that the $200 goal was the absolute minimum I needed to release the game, and then the stretch goals would have the good stuff. As we saw earlier, that didn't work.

To quote one of the Frogatto developers who I also see around Twitter a lot, “$200 makes it a bit awkward, like there's supposed to be some sort of catch.”

When I say that I really mean it (well I guess I have somewhat of an idea, but only barely). I don't get impostor syndrome, but I still worry that at some point I'll run into an obstacle that I can't weasel past with guts and blind luck, and the charade will be over. In fact, Steam integration could very well be that obstacle, since I don't even know what API stands for.

But I don't mean to sound too dramatic. Ultimately thousands of bad games get put out all the time, and even if SGS turns out to be a complete disaster it's not the end of the world. It's my first game, possibly my only game, and whatever that thing is that I set out to do and apparently achieved... well I achieved it. I think.

What I Did Right

To my surprise, the game might actually not be a complete disaster.

The unnecessarily complex mechanics of Air Force Gator were replaced by a simple two button system- Z/face buttons fire a regular shot which can be upgraded in various ways by classic powerups, and Space/shoulder buttons fire the ship's Hyper weapon, which is charged by collecting point items that are dropped by enemies. This creates a simple gameplay loop on top the typical shmup “shoot things and don't die”. Killing enemies gives you point items, which you collect to charge your Hyper, which you use to kill enemies. Because of this, you constantly have a short term goal in addition to the long term goal of finishing the level. The simple combo system adds another level of interaction and consequences, since timing when you kill enemies will get you more points. Having five different weapons that all have their own ammo didn't add anything like that. All it did was make things more complicated with no reward.

I may pretend to be a developer but I do know a thing or two about game design. This could be a 10 page essay of its own so I'll keep it short. I'm going to use the word “consequences” in a very broad sense here, to mean outcomes, implementation, versatility, and other similar ideas.

-Complex mechanics with complex consequences can be good (see: strategy games, hardcore RPGs).

-Simple mechanics with simple consequences can be good (see: Tetris, chess [maybe checkers is a better example], most early arcade games)

-Complex mechanics with simple consequences are usually problematic

-But simple mechanics with complex consequences are where game design really shines. Look at Super Mario Bros. You basically do two things in that game- jump over/on things and shoot fireballs, and you don't really even shoot fireballs that often, but with that formula Nintendo has made several of the best platformers, and arguably some of the best games, ever. Even if you took the frills out of later games like World, they're still masterfully designed, which keeps running and jumping interesting for much longer than it really should be.

I don't mean to compare SGS to Mario, though. In fact I don't mean to imply that SGS is even particularly well designed, but this one aspect turned out pretty okay.

It's literally my job, so if this wasn't here then I would've fucked up real bad. It would just be horribly sad if the title of this was “how a pixel artist made a game and the art wasn't even good”.

But the anime influence was, in hindsight, a pretty good move, even though I'm not really much of an anime person (never even seen any Gundam and I namedropped it in the KS description). People universally love the boss titles, and if you don't care about spoiling the game's final boss (and I'm sure you don't), check this radness out.

This style turned out to be the kind of style that's confident, consistent, and recognizable. For a vastly superior example, Hyper Light Drifter made half a million dollars almost exclusively on this idea, because it had a command on its style that very few games do.

Think of the most fashionable person you know. They know exactly what they're going to wear and they wear every square inch like they fucking own it. If you can do that with your game, rest assured it'll come across (and possibly get you half a million dollars).

Super Galaxy Squadron has two modes- Arcade (six premade levels), and Endless (one randomly generated level). Arcade mode didn't exist in any form until a couple weeks ago.

Despite the many, many things that went wrong during development, at the very least I did them in the right order. I used Endless mode as a sandbox for everything I put in the game, starting with all the basic functions of moving and shooting things, then different enemies, then all the player ships and their unique weapons, so once all that was done, I had all the tools I needed to make the Arcade mode levels in a tiny fraction of the time. I also increased the cast of enemies by a significant amount, and since I already had a variety of enemy types done and polished, it was an unusually simple process of adapting and expanding those.

This is something I'll absolutely keep in mind if I keep making games. Since I already called it a foundation I'll use a house metaphor- a lot of developers get too concerned too early with the interior of the house, and end up having to build around it instead of up and in. They'll do the bare minimum required to hold up a second floor living room, put a sofa and a TV in it, and then have to go back to ground floor to finish the supports.

I'm definitely in no place to give programming advice, but humor me for a second- the meaty parts of the game are really fun and exciting to make, even for someone like me, but it's better from a technical standpoint and a game design standpoint to start at the bottom and work your way up smoothly.

What's Still Up in the Air

God I hope so. I think it is, and not even because of my massive natural bias. Actually probably just because of my massive natural bias.

This is actually a way bigger deal than it seems. With the demise of Cave, one of the most legendary shmup developers of all time, a couple years ago, it's no surprise that the genre has lost whatever insignificant grip it had in the mainstream. I may pretend to be developer but I don't pretend to know how the Japanese arcade market works, so this is all based on assumptions, but I'd bet everything I own (joke's on you I've only got like $50 in the bank) that shmups have declined as much as they have because they've never evolved. Platformers came back with Braid and Limbo and Fez and Rayman, among others, and roguelikes have possibly become the staple of the indie market thanks to games like Rogue Legacy, Isaac, FTL, Risk of Rain and so on, but shmups have always been stuck in the mindset of infinitely one-upping themselves, without actually changing. And as far as indie shmups go, you can play a drinking game called “Return to Arcade”, where you drink every time an indie shmup talks about emulating arcade classics or having roots in the 80s and then die (no offense to those games, though, that's a completely respectable direction to go, even if it's not personally my thing).

It's my firm belief that trying to be arcade games is what's keeping the genre from going anywhere. No one wants to play shmups but the most hardcore enthusiasts, so games are only made for the most hardcore enthusiasts, creating the aforementioned cycle of one-upping. Some of said hardcore enthusiasts have personally told me that aiming for the “casual shmup audience” is pointless because the “casual shmup audience” doesn't exist. I partly disagree, because I consider myself a casual shmup player. I've run through Dodonpachi a couple dozen times at this point but I'm horrible at it and don't care the slightest bit about my score.

But here's the thing- I would've never played (and thoroughly enjoyed) DDP, or Raiden 1-3, or Battle Garegga and Bakraid and Batrider if I hadn't needed to do research for SGS. So here's where I partially agree- the casual shmup audience doesn't exist, because the people who would comprise that audience have no reason to play them. There are like five “big” shmups on Steam, and the only ones that aren't hard as balls are Raiden 3 and Jamestown on lower difficulties. Anyone who might be interested in the genre, but not interested enough to actively seek out the right games for them, has probably only ever played terrible flash games and seen videos of ridiculous Touhou patterns and thought “this isn't for me”. What I'm trying to say is that there need to be “gateway” games, that are accessible and familiar to regular people. It’s 2015, for god’s sake, and the way people play games is considerably different than it was 20/30 years ago. Vlambeer’s studio tagline is “bringing back arcade games since [random date]”, and they and many other developers have been killing it in that sense over the last few years by making things that are simple and addictive but aimed at the modern PC gamer. I feel like shmups could have a part in that if they were willing to adapt.

So what this all it comes down to is whether or not I'm right. I'd say we'll find out soon but chances are no one will play SGS anyway. Still, though, someone who's better than me should make shmups that do cool new things instead of the same thing again. I want to see what a shmup RPG is like, even if it's terrible, I'd rather see that than generic anime girl bullet hell.

After writing that subtitle I realize it sounds like a thinly veiled humblebrag, but I promise I didn't mean it that way. This is actually an interesting dilemma on the business side of things- the charity thing is in part a cynical, manipulative ploy to bring more attention to the game and serve myself in the future, because that's how humans are (although to be fair to me, it's like 10% of why I'm doing it, at the most). In other words, it's an investment- betting on gaining more success/recognition in the future by doing the charity thing than what I would get from selling SGS normally.

But even considering all that, the answer is yes, because even though I'm super broke and will continue to be super broke after releasing the game, the money is still going to charity and that's just a good thing no matter what.

I'm currently a student at a large art school. Since sophomore year of high school I was set on majoring in film/video, and until the end of last semester, that's what I did. But starting next semester I'll be an Interactive Arts major, which is basically just games but also other cool shit like robotics. I guess I've committed to working on games for the foreseeable future.

SGS might actually work out, but what I've done over the last few days is the extent of my programming ability. I really am a pixel artist, not a developer. My marketable skills are extremely limited, and it would be a long time before I have anything else to offer. This is something I'll have to think hard about, probably for the next few years of my life, and I'll spare you this particular crisis.

Gfy from SSS two weeks ago, comparing footage throughout 2014

Do you want to know what the high point of this whole experience is (so far)?

A couple hours ago, some of my New Blood colleagues were playing through the full game for the first time and talking about it as an actual product.

I may pretend to be a developer, but my drive to make things for other people to enjoy is more real than the fishing section at REI. In fall of 2012 I started making Air Force Gator, and now, over two years later, it became an actual game that people played and genuinely liked. Even after the rocky ass road that led to today, there's a real, tangible thing that I made and, very importantly, finished. This point where tons of tedious, really unsexy work transcends and becomes entertainment is the lasagna to my Garfield.

And on top of that, I now have something I can maybe call a career, making more things for people to enjoy.

My favorite clip from the launch trailer I'm editing

But this story isn't finished. Next Friday the game (hopefully) comes out. I'm simultaneously excited and terrified. And I think that just about does it for this whatever-mortem. If you read through this whole thing then I hope it at the very least wasn't a waste of your time. I don't tweet a lot but you can follow me if you want to see my personal experience of how this story ends, and I'll have that be my only shameless plug for the game, although if you want to see more of my art then check out FTL Remastered, a personal project that was put on hiatus for SGS, and Blood Alloy, which I'm doing character/enemy animations for, or just my portfolio. Meanwhile, keep making cool shit, since you guys can. Maybe someday I'll learn how to do this right and I can too. Well this turned out long. Sorry.

Media