Fear Me! – Post-mortem (pun not intended)

Posted by (twitter: @JustSomeNikola)
August 28th, 2015 1:58 pm

In Fear Me! you play as a happy-go-lucky ghost who has recently breached the veil of the damned in pursuit of his favorite leaisure activity – scaring children. A sort of action-puzzle blend, check it out if you dig that sort of stuff.

One of the more action oriented levels.

One of the more action oriented levels.

Short disclaimer: I’ll be writing like I know everything about everything, but in fact I just have approximate knowledge of many things.

How it went:

After being rudely awoken somewhere in the early afternoon by a combination of my dogs yelping and my neighbours drill-and-hammer work, I came to the realisation that Ludum Dare was going strong for about 11 hours. After a couple of sobering slaps, cups of java and some sandwiches I was ready for some gamedev on PC action. It was time for…

Brainstorming! (1-2 hours)

I put some depressing music on – Steve von Till – and pulled out my notebook and pen. A list of wild ideas came torrenting from the weird places of my brain.
It includes a Through the eyes of a dictator – Painting Simulator where you’re a painter. But you’re also Hitler.

Pictured: Game design document.

Pictured: Game design document.

Noticably (or not) – the game I made isn’t on the list. After  I was done laughing at my own jokes, it was time to do some actual work. So I began thinking about what sort of game I would like to make, what design elements would I like to try implementing. I remembered some experiences from my previous (and first) game jam – in which I overzealously attempted to create a procedurally generated party based roguelike dungeoncrawl RPG – this time, instead of intentionally trying to make something impossible to make in the alloted time, I decided to scale the project properly.
It was decided that:

  1. The art would be minimalistic – I’m not an artist, but I’ve got some decent chops in some aspects. The complete art for the game took less than an hour to make. Half of it is unused. I designed 13 tiles, but used only 2. Only the main character has any actual animations.
  2. A less is more approach, gameplaywise. I wanted there to be one main mechanic other than movement that was essential, everything else would be designed with it in mind – that mechanic was phasing through walls. I had this picture of ghosts running around and going back and forth through walls in my head while thinking about monsters – it felt right so I went with that. At this time it was also decided that the game would be a sort of puzzle focused action thing. A pursuit of elegance in design.
  3. It would have music I wrote if there was time – fortunately there was – writing the music and getting the right sfx took about 5-6 hours. Unfortunately, not quite properly mixed and mastered, but good enough. It really brought out the feel and tone of the game I was trying to make.

Now – for the puzzle part – it was alright if the only lose condition was to not solve the puzzle. But I wished for the game to be something a bit different than your ordinary put x into y puzzle. Thus death by illumination was introduced.

Early somewhat functioning prototypes.

Early somewhat functioning prototypes.party hard

At this point there were a couple of mechanics I wished to implement:

  1. Movement with phasing.
  2. Dynamic lighting for 2d tiles using my own light manager that changes the color of the tiles based on the distance from the lamps – there’s a bug related to this that I never got to fixing, but fortunately it was easy to bypass. 😉
  3. Scaring
  4. Timers and switches
  5. Basic kid AI
  6. An enemy patrol AI (Got cut)

As I later found out – just these mechanics were enough to provide a satisfactory amount of depth in the puzzle and level design.

Time for the actual coding and stuff.

Coding and stuff (after brainstorming – 2 hours before submission hour)

Somewhere during june I began work on a Unity Editor extension for tile based level editing. It’s far from complete – but it includes all the important features. Having this was a big timesaver, without it I would most certainly not be able to make this many levels while getting this much sleep. It also brought up some minor complications related to the lighting – thanks to the map files not really being designed for being accessed outside the editor, but a couple of precise hacks later, everything was in working order.

The mechanics themselves took less than 15 hours to implement – they’re pretty simple and straightforward after all. Then came the initial levels and some intense playtesting. I played it a lot, I bothered my friends a lot to play it a lot. I bugged my family to play it (not a lot because they’re terrible at games). This was probably the most important part. Based on their reactions I’d gauge the level of difficulty, level of fun and investment – also stress and rage at some pain in the a** parts like level 7. Not to mention that the whole process was totally awesome and fun for me. I’d observe them and laugh maniacally when they stopped to think in places I wanted them to stop and think, and I stared with fascination when they came up with a different solution than the one I initially expected.

Then came the UI.
I hate doing UI so much.
I completely hate it.
If UI was a person with a face, I’d punch it right in the teeth (kinda like what I did with the UI on my first jam game – it’s terribe).

Fortunately, Unity has a decent UI editor and there wasn’t too much UI to actually make. That took about 3-4 hours.
Then back to testing and designing levels.

The last two hours were filled with frantic compiling and recompiling and fixing minor quirks and bugs and fine tuning the numbers.

Levels 4 and 5

Levels 4 and 5

Levels 6 and 7

Levels 6-8

What I also hoped to acomplish was to fully exploit/demonstrate the variety of emergent problems from such a small set of mechanics. Unfortunately, the levels are probably not ordered properly, ex. level 4 is a big jump in difficulty from level 3, and most of the levels after that one are quite simpler and more direct. In the next section I will talk a bit about each level and what I tried to achieve with them – maybe at a later time I’ll record a video with a detailed explanation.

  • Level 1 – Simple tutorial level, made for familiarising the player with the basic controls
  • Level 2 – Tutorial explaining how lights work with a super simple “puzzle” in the start. In retrospect – the second part of the level with the switch could’ve been less of a pain in the ass and teach that without a high chance of killing the player. But maybe this way it delivers a different and overall more important message in a not too punishing environment – “be careful when you see a switch that appears to be doing nothing”.
  • Level 3 – Getting the player comfortable with the controls and mechanics with a simple and direct level.
  • Level 4 – The level that kind of got out of hand. It was supposed to set the tone of the game, with puzzles blended with mild action. It does this, but ends up being kind of hard and unusual for most players. Super proud of this one, but would be better if it switched places with level 5 or level 8.
  • Level 5 – Showing the player that levels can also be somewhat dexterity based.
  • Level 6 – Initially Level 7 – but I thought it would be cool to give an alternative easy way to solving the puzzle since you don’t really see any kids in the initial screen – giving the player the opportunity to explore the layouts so they would know what to expect in Level 7.
  • Level 7 – trigger fingers abound – introducing the phase through the blocks while timing properly mechanic.
  • Level 8 – A level using the bounceback mechanic.
  • Levels 9 and 10 – mixing the various previous mechanics in an interesting and new way.
  • Level 11 – A sort of anticlimactic congratulation screen. I should improve this.

What went well:

  1. Good and reasonable goals and scale.
  2. Even though the art was meant to be just placeholders, it kind of ended up working well after I was done with the lights.
  3. I had a storm of inspiration for the music, praise the sun!
  4. Tons of testing with a lot of different people – super important, especially with puzzles – you need to know how someone who hasn’t seen the game will react at various different stages.
  5. Reused lots of code.
  6. Cut a feature that would clutter the design with added complexity that simply would not get properly exploited in the timeframe. A patroling enemy AI requires navigation and pathfinding, line of sight programming and various other thingys I believed wouldn’t add as much as it would take. Maybe in some later revisions.
  7. Slept long and well <3

What went bad:

  1. Cut a feature that would add more complexity and potentially a lot of depth (No. 6 on the previous list – the patrol). Could’ve been cool.
  2. I suck at mixing and mastering.
  3. One bug and one quirk went under the radar. A certain wall doesn’t quite work right under a very specific scenario and sometimes players while holding down and dying would leave to the main menu by accident.
  4. The bug with the lighting (not seen ingame)

What I learnt/relearnt:

  1. Push your strong points to the front
  2. Knowing what you want is important
  3. Knowing why is more important
  4. Try to create a comprehensive and coherent experience, every element of your game should stear in the direction you want your game to take
  5. Designing good puzzles is hard

Thanks for reading! Thanks for playing, rating and commenting! Each bit of support means a lot to me <3
I’ll be recording a walktrough for confused people at some point, until then, best of luck on your own!
If you would like to contact me for anything, your best bet is to hit me up on twitter – @JustSomeNikola

Tags: , , , , ,

Leave a Reply

You must be logged in to post a comment.

[cache: storing page]