Unending Galaxy: Post Mortem

Here is Unending Galaxy‘s post-mortem! The game still sells relatively well given the niche market I was targeting in the first place, but I think this can be an interesting read to any solo developer tempted to make and publish a game. Also, I haven’t posted something interesting here in a while.

I’ll also publish posts about future projects, and what’s in store for Unending Galaxy in the following days.

Enjoy 🙂

Why Unending Galaxy in the first place

Well, it’s no secret that i was one of the “big” modders for the X-Universe game series, not mentioning an avid player of tons of other space sims. I am also a big fan of strategy games in general. Finally, I love to code general artificial life ecosystems in which different agents can interact. The direct conclusion was to write a game marrying all those things together. Interestingly enough, Unending Galaxy started as a “pirate base simulator” to help me write my plugins for X3, but it quickly grew out of control and became a proper space simulation in its own right.

Hard, cold, numbers

All numbers are approximates [dated from July 2016] [update 2/2017: You can add about 400 copies]

  • Total Sales : ~5,400 units (to this day)
    • direct / donations : ~50 [available during alpha and beta phases]
    • itch.io : ~130 [available since beta] (*1)
    • Indie Game Stand : ~60 [available since 1.0]
    • Groupees : ~2,300 (game bundle, lots of sales, little revenue)
    • Steam : ~2,800 [available since 1.2]
  • Total Revenue (after the stores and the IRS took their respective cuts) : ~18,600$
  • Steam Refunds : 8.5% (no DRM)
  • Budget : ~250$ to pay for various tools (bug report tool, video recording)
  • Team Size : 1
  • Time spent working on the game : 2 years (*2)

(*1) itch.io allows customers to pay more than the price you ask for. I am quite happy to say that they often do. The game was at 5$ during beta, and at 10$ since 1.0. Despite that, the average price people spent on the game is slightly above 10$.

(*2) First year wasn’t full time at all, I worked on it on and off during weekends. The second year got interrupted by family issues during a couple months. I could probably get to the same result in a little less than a year now.

Steam represents the lion’s share of the profits (over 80%). All in all, I am quite content with the numbers. For a game sold at a 11.99$ on Steam, made by an unknown entity and targeting a very niche market, it sold well. And more importantly, it’s still selling, at least on Steam. Thanks to occasional articles, and regular content updates, we’re still making (a bit of) money out of it each month.

To give a reference, according to Gamasutra (2014 numbers, but I suspect it hasn’t changed much), 57% of indie game developers, including both solo indies and members of indie teams across all pay ranges, made under $500 in game sales.

Project scope

Solar power plant
Earliest screenshot I managed to find. 8 directional movement.

Let’s be frank. I went against everything that was ever advised to any new game developer ever. Way too big a project, way too long a development cycle, and so on and so forth. That’s something I knew from the start. My justification for doing so was that I am a programmer by trade,  I had been modding games for more than ten years, and loved to build large/complex artificial life simulations. So I assumed to know what I was doing.

If you’re new to programming, don’t do what I have done, it’s plain stupid. Heck, even for a seasoned programmer, it’s borderline insane. Start small, your grandiose MMORPG project can wait.

In retrospect, it was still slightly too big a project for me. Coding a game supposed to work on tons of wildly different systems is a very different task than building a database application for a couple of similar machines. I underestimated how wildly differently the game would work on different PC. Still, I think I managed decently in the end.

The code itself

I decided to roll my own code and discard popular “engines” like Unity or UDK. This decision was driven by the fact that I needed as much performance as possible. The game was supposed to handle thousands upon thousands of game agents in real time without relying on the usual cheats (extrapolation, spawning stuff out of thin air, …), and such engines are very bad at doing that. UG is also a 2D game, making most game engine too bloated for the task. Also I wanted my code to take full advantage of multi-core processors. While I don’t regret that decision, it also caused many compatibility issues with specific software/hardware configurations, not mentioning how hard it’s to debug a multi-threaded application in general.


So, rolling my own code was a double-edged sword. On one hand I managed to get a simulation handling about 20k agents in real time. On the other hand I was swamped with weird and hard to figure out game-breaking bugs. Some of them persisted until late after the official release, leading to slightly worse Steam reviews than initially expected and less press coverage. Also, with bug fixing taking most of my time, marketing and polishing suffered.

Early development and monetization

It initially looked like that
It initially looked like that, and the physics were absolutely terrible 🙂

That’s something I am pretty happy about. While I wouldn’t do it the exact same way now, I think I more or less did it righ. Basically, I began to post about the game on my official website way before anyone sensible would do, banking on my small reputation as a X-Universe modder. It was very confidential at the time, but I got very valuable input from people from my former community in the process. Then, once the alpha phase began, I put a donation system in place. People who donated anything would get the game, forever. A freeware version of the game was also available for anyone who wanted to try the game. Said version was limited in content (less factions, maps and scenarios). Interestingly enough, it kept the piracy rate to zero. Piracy only began once I stopped updating the freeware version. I’ll get back to the piracy topic later down the line.

Anyway, the general idea was to gradually increase the price of the full game alongside the development. People who bought in the early alpha version got it for 5 dollars, people who waited for 1.0 got it for 10, and people who waited for the steam release got it for 12. It worked okay, and led to the arrival of a few (useful) people interested in the project.

Steam Greenlight, game bundles, and other stores

Initial release on itch.io. Clouds and nebulae were a “bit” on the heavy side.

Once what I considered the game was complete enough, I put it game on itch.io, then on IndieGameStand (IGS) and finally on Steam Greenlight. Before that point, the game was confidential, a few dozen people knew about it at best. Itch.io went quite well. Back in the days, it was the new player in town, it got low traffic, and people generally didn’t know about it. Still, it got me a few hundreds quids to pay for my general expenses. Same goes for IGS to a lower extend. I also put a request for getting into GOG, but it got declined, looking back at the beta version I sent, I can see why now.

Steam Greenlight, on the other hand, was more problematic. While I got overwhelmingly good comments, the traffic just wasn’t enough at the time to get the approval. Back in the days, Greenlight approval was more difficult to get, you needed more than a thousand of “yes” votes to have any chance, granted you managed to achieve some top100 in the process. That’s a number I wasn’t able to get without relying on bots or advertisement, two things I am totally against in general. I wasn’t very far from it, but just enough to force me to choose another road. So I decided to get my game bundled.

Bundled version with more modern looking menus
Bundled version with more modern looking menus. We’re getting somewhere.

For those not aware about game bundles: it’s basically a compilation of small games sold at a massive discount for a limited amount of time. Normally, bundling your game early is a terrible idea if you can do without. Yes, you’ll get at least a few hundreds dollars and get some awareness about your game out of the deal, but clever customers will know that the game has been sold for basically nothing in the past and will wait for that thing to happen again. I should also mention that people who buy bundles tend to check all bundle websites, quite obsessively, meaning that once you’ve sold in a first bundle, it’s unlikely your game will do well in further bundles except if it’s paired with decent/new games. Last thing to know about bundles: It’s unlikely it will extend your community much, most of them won’t even try your game, they’ll just let it run the time to collect Steam cards.

That being said, while I was not especially in a hurry to release the game, I still needed to pass Greenlight one way or another. I also knew that my targeted niche audience tends to be slightly older than your average gamer with more income, meaning they aren’t necessarily monitoring or even caring about bundle websites.

So I accepted a bundle offer from Groupees. A specific “Greenlight Bundle” in which people are asked to vote for the game they buy in order to be released on Steam. I cannot enter in the specifics, the contract includes a non disclosure agreement, the bundle sold quite well, UG was the best seller of the collection (it was a “choose X games from this list” for Y dollars” kind of bundle). Still, in fine, with your share divided by the amount of developers present in the bundle, it’s still pocket money. Another advantage of Groupees is it’s chat-room, allowing me to interact with my new customers and get some valuable feedback.

It got the expected effect, pushing the game relatively high in Greenlight’s top100 leading to its approval for a real release.

Promoting the game before the official release on Steam

Actual game promotion really started with the public beta version.

Space-sim communities

Finally released on Steam!
Finally released on Steam!

Getting and, more importantly, keeping people interested in the game was an extremely time consuming and tedious task. There’s basically no decently sized “space sim fans” forum or website you can advertise on. Large ones tend to be specific to a single game, making it near impossible for me to post there without breaking forum rules (kudos to Egosoft for making an exception, though). Thankfully, some members from my community took on the task of talking about the game on most of those forums. Interestingly enough, the biggest forum leading to the most active thread wasn’t in a space-sim forum but on bay12 (dwarf fortress), a community I wish to thank for its extremely valuable and extensive feedback.

But all in all, it was beneficial. I got a lot of feedback, new members and visibility on space sim news websites like Space Game Junkie, leading to many sales and donations.

Social media and Reddit

Probably the most time consuming and less rewarding task of them all. Cross posting on Facebook, reddit, twitter… takes a lot of time and led to dubious results. Getting new followers is a tedious process if you don’t rely on dumb luck or buying fake ones (yes it’s a thing, very common place at that, and perfectly useless). It’s difficult to put an exact number of sales coming from those sources directly but it’s extremely low, with reddit in the lead.

Indie gaming websites

 This category is a mixed bag. I posted the game on IndieDB quite early and got some decent traffic out of it and some sales, but barely any feedback. The trick is to post news updates (relatively long ones with screenshots) often to get to the front-page. This resulted in me cross-posting my updates on both my main website and IndieDB. TIGSource forums were utterly useless, which doesn’t surprise me, it’s great to get quick feedback on small / straightforward games, not for deep simulations.

In general, however, few large indie websites covered Unending Galaxy. I am not particularly surprised by that. Firstly because early press builds of the game were quite unstable, it was hard to figure out what to do and had quite a steep learning curve. Secondly, space sim / 4X / RTS hybrids are usually pretty far away from what they are used to review.

YouTube / Twitch

The bulk of direct sales (out of standard Steam traffic) came from YouTube channels and to a lower extent, Twitch streams. Lots of small channels did let’s play of the game, giving the game a decent visibility on those platforms. Don’t hesitate to contact content creators, especially small ones, they’ll usually be quite happy to be offered a key.

Gaming Press

The game was mentioned 2 times on Rock Paper Shotgun, sadly way too early in its development process to have much influence on our sales. It was during the early alpha phase (unbalanced, ugly and bug ridden), it probably led to a lasting negative image of the game.

Just before the Steam release, I sent press copies to many PC gaming websites to little effect. English speaking press mostly ignored the game. We however got mentioned in a very large German website (4Players), various medium sized European website, and multiple much smaller news websites.

What I did wrong

First and foremost, our trailer was quite dated. It was as professional as I could make it look and to the point with a great original music, but it was made with an old version of the game with dated sprites and poor visuals. I think it played a major part in why the game got little press. Our press release emails were too generic, sent too late. The key gave access to a game that was still quite unstable as the worst bugs and visual problems were fixed barely a day before release.

In general, I haven’t spent nearly enough time dealing with the marketing and community building aspects, but I had my head buried in code and other problems to deal with. If I had to give one advice, it would be to find someone competent to deal with most of the promotion for you. You still should take time to answer to questions yourself, customers tend to like that a lot.

I also released the game on Steam at least a month too early, and it suffered for that., both in press coverage and Steam reviews.

Piracy and scams

Content update 1.2.6 with ability to play other factions
Content update 1.2.6 with ability to play other factions

Interestingly enough, Piracy is rarely discussed in postmortems. The game doesn’t use any DRM system, and a free, basic version of it (which is still available) was kept updated until the 1.0 release. I like to think it actually helped with the game not being released on warez websites until much later.

Of course, once the 1.2 version got finally officially released on Steam, it appeared in a couple of places including a few major bit torrent trackers and multiple warez websites. While it’s difficult to estimate how many copies were distributed that way, it’s quite low. Also got a mail from a pirate who bought the game shortly after (anecdotal evidence, I know, but still heart warming).

What’s way more interesting are the scammers. Firstly, during the pre release promotion phase I put the game on Distribute(), a website to facilitate how to give steam keys to YouTubers and gaming press. At least 75% of the key requests are scams. For YouTube channels it’s usually easy to detect, the email won’t match the one provided in the channel’s contact tab. For websites, if the mail doesn’t end by @gamewebsite.com but by yahoo or gmail or something like that, it’s 99% likely to be a fake. And for twitch streams, it’s near impossible to tell. Also if it looks to be too good to be true, it generally is. This 300k channel making mostly Call of Duty or Minecraft videos doesn’t give a damn about your game, it’s just another imposter.

After the official release came the emails, tons of them. On top of the previous examples (you’ll now get them in email format too), you get new different categories!

We have the ever popular “I’ll post a review on steam for a key“, a quick look at their profiles told me, that in general, no they won’t. If you’re tempted, think that if something goes wrong, they can threaten you with a negative review instead.

We have the foreign ones that will link to some game website in Cyrillic or other foreign language. Some may be real, but it’s hardly worth the bother.

Finally, we have the professional scammers. Those ones are crafty bastards who probably deserve their own article. They will actually buy a domain name and build what looks like a press website as long as you don’t look too closely. They will send a very professional looking email from this domain (instead of some Gmail account) and ask for multiple press keys. The only big issue (beside the multiple keys thing, which is always a red flag) is that you never heard about them. Those people will never review your game, they will sell it on grey market websites like G2A (a “store” in which people can resell steam keys…). As I don’t want to help them getting any rank on search engines, I won’t link them directly but we have “dzgamereviews:com”, “gamepressed:com” or “s-reviews:eu” to name a few.

You may think that it’s not worth the bother, and save time by not checking every email and key request. Don’t forget that those people are preying one you. They won’t provide the service demanded. They will make money out of selling your keys to G2A. As a side effect, you’ll lose a potential customer who’ll buy the game on G2A instead of Steam. Those people are scumbags, plain and simple.

What Now?

Unending Galaxy was written in an outdated language with weird and unreliable tech. The idea now is to make a few very small (and cheap) games with a more modern development environment. Allowing me to convert reusable parts of my code while acquiring enough money to fund a better and larger project. I will post about those plans in a few weeks.