Saturday, 5 August 2017

Escape From BioStation PostMortem

Howdy y'all!

It's been such a long time, hasn't it? I haven't really done an update since the release for Escape From BioStation so I thought maybe it was an appropriate time to take a look back and write a post mortem of the game's development.

So sit back, and listen ye to the turbulent tale of how it came to exist!

Planning Issues

Escape from BioStation had 2 very distinct incarnations, which essentially became 2 completely different games, only one of which wasn't terrible.

When I started on the first version of the game I was a green nooby game dev. I only had experience in making silly tiny stuff for friends in game maker. I wanted to use a new engine (Unity) for this, since I wanted it to be 3d. Therefore my initial planning was somewhat limited by what I could do in it. I had a key idea, which was small robot in a vast mysterious space station, to evoke the need to explore. The first concept was summed up as follows:

"An adventure platformer (with fps elements). setting: space and shit"

There was a little more planning that that. But those words are literally lifted from the initial planning document.

I went through what I thought was a very robust and thorough design process. I created a rough draft on what the finished game would be: Little levels which each had a physics puzzle. Each puzzle would be solved by getting a squirrel or other random key object.

So there were already two big mistakes:
  • I only had a rough plan, and even that was flawed. I hadn't really thought about what the puzzles were, why you were there, or even what the environments would look or sound like.
  • I was using a new engine and had no experience in what it was capable of, or the means by which I would realise any of the mechanics I was loosely designing

Early Development

So oblivious to the issues that already underpinned the project, I began designing these self contained puzzle levels.

Very early shot of how the game looked

An example of lack of direction was where I created the first level. I wanted it to just be flat out platforms and obstacles, nothing fancy, leaning on super retro feels. The next level then had a puzzle where there was a gate that would open when you tipped a scale that was just out of reach. This would be solved by finding a random squirrel, and using it to lob acorns onto the scale.
I hadn't planned on any acorn lobbing mechanics. I thought it would be fun, so in it went. At the time I was designing it I made a note to myself: Don't make it a shooter, this is only for a puzzle - otherwise people will compare it to other shooters and that's not what this is.
As soon as it was done though I made another puzzle using the squirrel that was basically shooting barrels that would otherwise prevent you getting past. I'd feature creeped a gun into my not-shooter despite my sternest warnings to myself.

Oblivious to this I went ahead and built a level that covered all of the core mechanics, moving, jumping, picking up stuff and lobbing acorns. This, I thought, would serve as a proof of concept.

Making it public

I was excited to finish and ship a thing - just like proper people - so I decided to make the classic error of putting the "game" up on greenlight and kickstarter as soon as I had something I could show. At this point I had only completed a few rooms. There was no polish and looked like utter garbage.  Unsurprisingly it got very little traction. Greenlight votes were few and far between and there were only a handful of kickstarter backers (excluding my kind friends and family, that is).

After a few months of this I asked around for advice on how to boost interest. One suggestion was to get the demo in a bundle. I didn't really understand what I was doing, but I did it anyway.

So a small number of the public got a playable copy of this thing. Someone then did a lets play of it on the youtubes here.

You can see it looks like crap and I'm surprised that this patient soul played it for as long as he did! (I have given him a free copy of the finished article by way of apology). But when I saw his video I was thoroughly ashamed of myself, and realised that I simply had to make this thing better. much better.

Back to the Drawing Board

The above was all completed in about 3 or 4 months, and I was intending to complete in about another 6.

I didn't carry on.

I'm sure if I finished the project as it was, the game would have been an embarrassing failure in every respect.
So instead what I did was discard the old work. I went back to the drawing board and revised fundamentally what the game was, determined to create something that was... well... better. This is where the second incarnation of Escape From BioStation was born.

So step 1 was that I wanted to spec out everything that would be in the game, what the rough story was, the environments, the mechanics and the sound. Essentially do a game design document. Based on this I could create a new rough plan.

The first key step in the plan was writing a good script. I had a rough idea of a story, so I drafted in a talented writer friend of mine and we co-wrote a good quality script. The plot points of the scripts were hung loosely on the main thematic locations I knew would exist, and a couple of the puzzles I had which were serviceable from the old project. But we still had to tackle the fundamental questions: Why is this robot on a space station? What is he doing? Why on earth does he need a squirrel? Who is the antagonist? I had a title at this point already so.... what is this robot escaping from?

Having covered those points I could "join the dots" as it were, from the key plot points, and all the design decisions subsequently revolved around the script. As such it then became possible to write a plan which was detailed enough for me to stick to and end up with a decent product. This plan covered 18 months, but ended up taking over 30.

Here's my list of what went wrong and what went right:

  • Publishing too early
    • The game was made public about 2 years before it should have. This was a disaster for a couple of reasons. The first is that most people assumed the game was utter shit. The second is that those few that didn't had to wait for 2 and a half years till they got to play the game. This meant that had no chance of building any interest.
  • 2 Incarnations
    • In a way this was a bit of a good and a bad thing. Following the feedback I received, I revised what the game would be quite fundamentally. This meant that I wasted time planning and building something that went nowhere, but on the other hand the fruits of that labour would probably have been a disaster anyway - so maybe I dodged a bullet.
  • Scale 
    • The new version of the game was very ambitious. I had been made aware it was a common mistake for inexperienced developers to bite off more than they can chew, but I went for it anyway. At the time I didn't know it was too ambitious, having not done such a thing before. But I now know that in a way I should have stuck to my first idea since I could have completed it far more quickly. However I ended up having to prune vast amounts of content. For example there were going to be another 6 playable levels toward the end, aboard the "verybigship" (the main villain's frigate). It was going to have a climax where you fight him in a giant robot suit. Instead since I came to realise that due to time and fund related constraints I would not be able to have this - so you just fight the ship itself and don't go aboard.

Not the original "David vs Goliath" moment I had in mind, but one that worked 

  • Creating a core premise I found interesting
    • At it's heart the game never deviated from one idea: A childlike robot who explores a unknowable and vast space station accompanied by a friendly squirrel. It meant that 1) I never once lost interest during development and 2) one of the key environmental style choices always remained consistent (even if few others did)
  • Writing
    • The game in it's current incarnation is very story driven, so coming up with the script at an early stage defined everything that followed in terms of level design and pacing. It's also a decent enough story with funnies which made it enjoyable to work with.
  • Consistent art
    • Being a solo dev I didn't have time to create all of the environmental assets so had to use a small number of 3rd party ones from the store. This could have very easily have been a mistake, as in the game would look cobbled together. However I really didn't want to have been responsible for just another asset flip. So I was careful to make all of the models feel as if they belonged in the same world. I built a large amount of models myself, hired a artists, and made sure I worked on all of the textures (both mine and the 3rd party ones) to achieve a single look.
  • Testing
    • I was adamant that I would not release "early access". The game went into closed alpha in February 2017, and then had the ever living bejeezus tested out of it. It continued in this fashion for numerous months since I wanted it perfect before release (following my previous false start). On lots of occasions I sat with people in the same room to play the game so I could spot not only bugs but also where they were having trouble. That allowed me to ensure that the finished product was stable, worked well and did not frustrate the player (any more than necessary!).

Combat got several overhauls during development thanks to testing


Escape From BioStation was officially released on the 2nd of July 2017.

It went without a hitch! There were only 2 bugs that needed fixing after launch thanks to all the testing. I was half expecting there to be more but very happy there weren't.

The reception it got was good. Here's a nice review it got:

It hasn't exactly flown off the shelves in terms of sales.  I could speculate that the reason for this was a number of things, maybe marketing budget that was too lean. Maybe I could have made the text on the steam storefront more sales-y sounding. Maybe it was something else about the page like the screen shots not being appealing or the video not showing off the game as well as it could. Maybe my distribution channel has been too limited. I think that the real cause is these things:

  • I don't think there is mass appeal for a game like this being developed with the budget & staff I had. It's a little old fashioned, and lacks a strong art style. 
  • Going back to what I said before, I went public far too early. This meant that by the time I released it, anyone who was interested has long since forgotten about the game. 
  • Bad timing. Literally a few weeks before I was going to release steam decided to retire greenlight. This meant that they went through every game that had not been greenlit yet and as I understand the majority got approved. As a result my little title was competing with a lot, I think there were 15 new titles on the "new releases" that day.


One of the best things in the process was that I was seeing people play the game from very early in the life cycle and it really made me concentrate on how people would interact with the game, and make sure that it flowed well, that the story made sense, and that it looked presentable.

A lot of the other challenges, such as the scope being too big or fumbling the release and building 2 games really came from inexperience, and not having released anything big before. I am sure that there will always be something to learn in any game development project, but some of my main hurdles with this project could have been easily overcome by not rushing as much and a good round of prototying.

If you liked the sound of that then Escape From BioStation is available to buy on Steam here:

No comments:

Post a Comment