Tuesday, December 22, 2015

Between Success and Failure - I Can't Escape: Darkness

Today, I logged into Steam and noticed that sales for my latest game, I Can't Escape: Darkness, reached 666. A very auspicious number for a horror game, and it made me laugh. However, it also made me remember what my original goals for the game were, and how the reality didn't quite live up to my expectations. After thinking about what my expectations were, and how I defined success and failure, I decided to write my first Post-Mortem - both to put these thoughts down, and to share my conclusions with other game developers.

History
But, first a little about my studio, Fancy Fish Games. We've released three commercial games on Steam, and six free games before that (four of which were for One Game A Month). Fancy Fish Games has been around for three years (although I've been making games since I was in middle school), and consists of around five people who work part time for nothing but revenue share (kudos to them giving up their free time so that we can make our dreams reality).

The Beginning
I Can't Escape: Darkness was designed as the sequel, or as I like to think of it, the full version of I Can't Escape, our January “One Game A Month” back in 2013. I Can’t Escape was a simple atmospheric horror game in which the player is trapped in a seemingly (and in fact, literally) endless dungeon where escape consistently eludes them. Thanks to several youtubers and streamers (including Markiplier), I Can't Escape was a great success - where by success, I mean that over 250,000 people played the game and a lot of people commented and seemed to enjoy it (we didn't make a dime on the game).


The sequel was meant to take the simple idea of I Can't Escape, and make a fully fleshed out game without destroying what made the original interesting. And in that respect, I think the sequel was a success - there are certainly flaws and nit-picks that I could talk about, but the game does draw you in and creates a suspenseful, eerie atmosphere without relying on jump scares (which was part of my goal with the original). The sequel was also fleshed out with a lot new features, including overall progression, events, combat and a story.

Development
Development for I Can't Escape: Darkness took a little over a year. The core team was only three people - myself working on the code, Chase Bethea working on music and sound effects (and he did some awesome experiments with dynamic music for this game), and Matthew Poppe working on art and animations. My original plan was for the game to take four months - which as a rule of thumb I doubled to eight months, which wasn't a terrible estimate for the total development time. Working with collaborators over the internet part time is always a little unpredictable - but I had worked with both Matt and Chase before and trusted them a lot - and they didn't let me down.

The development time broke down something like this:
  • Engine Development - First Month: August 2014. Yes, I developed the engine for the game from scratch. You might think I'm crazy, or it's impossible to make an engine in one month, but I wasn't making a general purpose engine. I was modifying some old maze rendering code and developing the lighting algorithms specifically tailored for I Can't Escape: Darkness and it's gameplay needs. I've had a lot of experience developing game engines, so it really wasn't as difficult as it sounds.
    ICEDfinal.png
  • Minimum Viable Product (MVP) - September 2014. The minimum viable product included the basic gameplay mechanics - movement, exploration, items and combat. Again, you might be thinking one month is pretty short, but remember that the entire original game only took one month. At this point we launched the greenlight page with screenshots and videos.
  • Feature Complete - January 2015. This is the milestone I was estimating when I said the game would take four months. I then double that number because I know from experience that feature complete is really only the halfway point. And while it took six months to get here instead of four, it's interesting to note that it was the halfway point in development for us.
    escape.gif
    Now you might be wondering, if the engine work only took one month, and the MVP only took one month, why did it take four months to get to feature complete? The answer is that even though the MVP has the core elements of game play, it is missing content (including the various event areas and puzzles), UI, and balancing. These add up to a lot of work, and they turn what is more like a prototype into an actual game. We also took about a month off during this time between holidays, a short side project (ADventureLib) and getting Deity Quest on Steam (I launched the game May 2014, thinking it’d never get greenlit, but to my surprise it did in late december).
  • Alpha Test - February 2015. After the game was feature complete, we had to playtest, debug, and finish all the higher priority art and music assets. Then we released the game to a closed alpha to get outside feedback for the first time. I wanted to make sure the game was stable and not missing important assets so that the testers could give more focused feedback on the pacing and gameplay itself. It's also interesting to note that the game was greenlit on Steam at this point, after five months.
  • Beta Test - June 2015. We got a lot of good feedback from the Alpha Test, and there were many pacing and balancing problems we had to address (damn rats!). There were also some last minute features that we decided to add - while none of them were big, they do add up *cough*feature bloat*cough*. However, the difference between the alpha and beta test was pretty incredible. The game drew people in a lot more, the average play length quadrupled, and people began wondering about the story. One interesting thing to note here is that as you get closer to the end of development, it seems like you’re moving slower as there aren’t as many spectacular changes, but those four months of work weren’t wasted - as is clear from the difference in the responses from the alpha and beta tests.
    tutorial.png
  • Second Beta Test - July 2015. After another round of tweaks and improvements, plus cross compiling the game to Mac and Linux, we started a second round of beta testing. Meanwhile, the marketing campaign run by Justin Whirledge went into high gear at this point, with gameplay trailers released every other week.
  • Steam Beta Test - August 2015. At this point, the game was mostly done other than the final tweaks and squashing the last bugs. But there was plenty to do as we prepared to release on steam. We had to set up the store page, get the last of the asset changes in (including assets for trading cards and achievements), integrate the Steam API, get the game translated (we released in English, Spanish, Russian and German - friends, fans, and beta testers offered to translate the game for free!). Finally, near the end of August, the store page was live, and the release date of September 17th was announced. We also did a quick beta test on steam to make sure everything was working with the Steam API and the last of the changes.
  • Release! - September 17th, 2015. After a year and nearly two months, doing some final playtests, tweaks, and a marketing push (including early review copies and attack of the gifs), the game was released. It took nearly six months longer than we expected, and it was starting to feel like endless development hell (we probably could have tweaked the balance and pacing indefinitely - getting a procedurally generated game to parcel out events and progression at an interesting pace is a lot harder than it sounds), but we did it!


Three months later, we've had 666 sales, mostly positive feedback on steam, mostly negative feedback from press, some great videos and lets plays, and one amazing fan-written guide. Was the game a success? How do you measure success?

What is Success - Money?
If success is money, or even making enough to work full time on our next game, we definitely failed. The game retailed for $11.99 on steam, so 666 sales is certainly not nothing. But, even underestimating how much time we all spent on the project, we still made less than $3 an hour (after steam cuts and dividing the revenue among the team). So, it wasn't like we made no money, and it was an awesome end of year bonus for us, but it certainly won’t let us quit our day jobs.

For those of you who like charts and graphs, here's I Can't Escape: Darkness’ sales graph (with the labels removed as we're not supposed to release exact numbers). sales.png
It follows the pretty standard long tail with spikes pattern - where sales drop off very quickly after launch, but then there are spikes from sales and updates. The first big spike was our first major update (with several new secrets and improvements). The second, biggest spike was the Halloween sale including another, smaller update. The third wobbly bump of elevated sales was the Thanksgiving sale - which was not thematically fitting like the Halloween sale, didn’t include an update, and was a smaller sale, so was a lot weaker.

But let's face it - if making money was most important to us, we wouldn't be making games. There are plenty of better paying jobs we could get with our skills. So, there are other definitions of success to consider...

What is Success - Fame?
So, if we're not getting much money, how about fame - people always seem to want money and fame. Well, we certainly aren't famous, and I doubt most of you knew what I Can't Escape: Darkness was when you clicked the link. Additionally, critic reviews are almost unanimously terrible with our metacritic score at 40/100. But, on the other side of the coin, our steam reviews are 91% positive with 23 reviews, and the majority of let’s players who played seemed to really enjoy the game.
But if fame is our goal, again, we failed. Even the original didn't make us famous, and that had 250,000 plays, a lot more than 666 sales. However, while it would be nice for more people to know who Fancy Fish Games was and follow our games, we don't really do it for the fame either.

What is Success - Making a Fun Game?
At the end of the day - we make games because we want to make something creative and fun - something that we enjoy playing, and we want others to enjoy as well. And while we might not have a lot of sales, there are a lot of people who really love the game. A group of players even got together and collaborated notes to create this impressively comprehensive guide to the game.

And this is where I can say I Can't Escape: Darkness was a resounding success - and can feel good about our year long development. We might not be rich or famous, but we made something interesting that a lot of people enjoyed, and we can feel proud about that. Sure - the game isn't perfect, and you only have to read one of the critic reviews to see them point out all of the flaws, but despite that there is definitely a loyal fan base. thankyou.png
As a side note, I'll point out that even the critics who hated the game, perhaps grudgingly, had to admit that the game had a good atmosphere - which was what we set out to create.

Final Notes
There are a lot of post-mortems that talk about how successful and unsuccessful their games are only in the financial sense. However, I feel like it’s also important to think about what success means to you - as even a complete flop that no one plays could be a success if you felt it was a valuable learning experience or you gained one die hard fan who will continue to follow your games - and perhaps even beta test them or offer other help - as you continue on your game development path. This was true for Deus Shift, my very first publicly released game, which didn’t make much of a splash but gained one loyal fan who ended up both playtesting I Can’t Escape: Darkness and translating it to Russian.

Oh, and as I finally finished this post mortem and posted it, the sales have gone up to 674! The launch isn't the end, and people will continue to play and enjoy your games even years after their release! That’s quite a special feeling.