Designing Crumble!

Gary Penn

I HEART JAM! BUT...

There was no way I could spare a dedicated day for this. But I love the spirit of the Jam so I had to find a way to take part. So I took apart a 7.5-hour working day and spread it over a week instead, sort of following our usual development approach but on a smaller scale.

Working in such a fragmented way does mean a loss of momentum. Working outside of the Jam environment means missing out on the general vibe, which is a pity.

I started by breaking down the 7.5 hours as follows:

• 30 Minutes Thinking
  
• 30 Minutes Planning
  
• 60 Minutes Sketching
  
• 60 Minutes Refining
  
• 60 Minutes Extrapolating
  
• 60 Minutes Refining
  
• 60 Minutes Finalising
  
• 60 Minutes Final Refining
  
• 30 Minutes Whatever

I was keen to do everything within the timescale, including identifying and designing a concept, partly for the challenge but mainly because I couldn't spare any more time anyway :) I'd even thought about generating the placeholder audio from scratch but that wasn't viable in the end.

I decided to use nothing but Multimedia Fusion 2 to get the job done. I haven't coded like a real man for two decades and at least I'm familiar enough with The Games Factory that I can get something up and running pretty quickly in MMF2.

What follows is a light, flat and dry account of events.

30 Minutes Thinking
Dream a little dream. Think of what "survive" might mean. Get stimulated. Google the word. I want to stick as closely to the theme as possible, as in create a concept that's related to survival even if it's not immediately obvious. But I can't think of any clever variation on the theme and I'm left feeling creatively impotent.

This was the extent of my thinking:

• Survival of the fittest. Last man standing.
  
• Survive against what? Overwhelming numbers? One pernicious predator?
  
• Survive long enough to...
  
• Survive repeated onslaughts.
  
• Live to tell the tale.
  
• Outlive.
  
• Stay alive.
  
• Against the elements.
  
• Eat to live.
  
• Gloria Gaynor. I Will Survive.
  
• Air Crash Survivor. Survive a plane crash. Eat the passengers to survive until rescue comes. Catch them and eat them. Bash them. Stun them. Kill them. Eat them. In a jungle. On a snowy mountain. In the sea. On a desert island.
  
  
• How to survive in the woods"
  
• Hundreds of angry ninjas attack. Relentlessly.
  
• Survival tends to imply relentlessness.
  
• Survive on a desert island.
  
• Survive stress.
  
• Evolution. Survive to evolve.
  
• Famine.
  
• Plague.
  
• Survive a zombie apocalypse.
  
• Survive families.
  
• Survive the winter.
  
• Survive in business.
  
• Survive a beating.
  
• Survive in the sea after a boat goes down. The Life of Pi.
  
• On a raft in turbulent waters.
  
• Plant survival.
  
• The sense of surviving insanity by the skin of your teeth. An avalanche.
• Robotron.
  
• Earthquake. Tidal wave. Volcano.
  
• Fire. Towering Inferno.
  
• Race. Stock cars.
  
• Cutting through growth that slowly grows back. Use a machete.
• Chop up some creatures, too.

30 Minutes Planning
I picked an idea and ran with it. I fleshed it out and noted what needs to be done (and added notes as I worked).

I'd rather make a Robotron-style combat involving survival against swarms of ninja but there's no way that'd happen in the timescale. So I'm going for something straightforward: an increasingly intense rockfall. The gist is: boulders fall and you avoid them; collect something else for points; collect a few accessories (specifically 'powerups') for good measure.

I couldn't think of anything interesting to do with the chosen idea so I decided to focus on making a very simple thing with a view to making it as refined as possible. I was expecting to make something more hardcore than not (there'd only me to test it so I have to make it for me). I also wanted to do something very "Game & Watch" in scale and tone.

This was my thinking:

I won't have time to do a little platform game variation of this.

I won't be using anything other than placeholder audio and visuals. I won't be animating anything if I can possibly help it (too time-consuming). The scope of the concept means that I can go further than basic play and instead throw together a basic game and product, too.

I want a sense of 'jamminess' - escaping by the skin of your teeth, like Robotron at its finest or when the planet explodes in Defender and there are hordes of mutants after your blood. This suggests that the collision field for the player toy should be small as far as boulders are concerned (but large for treasure to make it easier to collect). Effectively the player toy is no larger than a smallish point at its centre. Moreover, the player toy being removed from play isn't just based on collision but how long a boulder is touching the player toy.

The frequency of boulders and their speed are linked, as are the speed of the treasure falling and disappearing and the speed of the player toy. How long the treasure stays in play is based on how fast the player toy can move across the screen.

Smaller boulders fall faster? Larger boulders fall slower? Is there a fixed number of boulders and treasure per level? Everything falls at a constant speed? There's an increase in frequency and speed? The boulders fall at random positions or in a pattern?

So... The treasure falls twice as fast as the boulders. All boulders fall at the same speed. There's half as much treasure as there are boulders. The less treasure there is to collect, the longer the level is likely to last.

I'd prefer playing intense levels that are short. Longer (but not long) levels that are less intense might be tolerable.

• The collision field is a fixed size or related to the quantity of boulders falling?
  
• Tell the player how many boulders to avoid and how much treasure to collect? Or use a timer?
  
• The boulders explode in a shower of chunks when they hit the floor.
  
• The player toy explodes in a shower of blood when it's removed from play.
  
• Collect treasure to stop the boulders falling. Exits open either side?
  
• One life only.
  
• Crumbling bases like in Space Invaders?
  
• Add solid 'shelters' on the earlier levels.
  
• Shake the cavern at the start of the level to suggest instability.
  
• Shake the logo.
  
• Climb to the top? Collect dynamite to blow open the exit?
  
• Each boulder that breaks releases treasure?
  
• Every third boulder that breaks releases an accessory? Accessories just fall in every 10 treasures collected?
  
• Falling fruit (Fruit Crumble). Falling animals, eg: from a mouse to an elephant. Falling vehicles. Falling buildings. Falling anything.
  
• Close! Test to see if the player toy was almost crushed by a boulder and comment.
  
• Close! Test to see if the player toy almost missed the treasure and comment.
  
• Bonus treasure-only chambers.
  
• Obstacles for boulders bounce off, cf: Pachinko.
  
• Faces on everything. Angry boulders. Happy treasure.
  
• Something silly. Make it sillier.
  
• Use instrument sounds so the breaking boulders play a musical rhythm.

Accessories

• Slow boulders.
  
• Freeze boulders. No use.
  
• Stop time. No use now.
  
• Reduce time. No use now.
  
• Use a parasol to protect yourself.
  
• Change boulders into bubbles or balloons to pop.
  
• A smart bomb.
  
• A big gun to shoot boulders.
  
• Invulnerability.

Final Statistics

• Survival Time
  
• Treasure Collected %
  
• Boulders Dodged

60 Minutes Sketching
"Find the fun." I need to get basic play up and running ASAP. I'm building the basic toyset (the player toy, one type of boulder and one type of treasure) and adding the basic rules of play. It's... not exactly promising, although it's exciting when there are too many boulders falling.

I (ahem) got a little carried away and added some particles and audio effects, mainly because I thought they'd contribute to the overall feel (and it wasn't feeling so good by this point). The boulder particles can be distracting but I like them so they stay.

60 Minutes Refining
Time for some focussed playing and tweaking. I'm also addressing anything that interferes with progress.

MMF2's default controls just aren't working for me. Everything's too precise. The player toy stops moving at a pixel resolution. That and having to avoid boulders entering play at the same resolution makes basic play difficult and definitely unfun. So I'm going to have to do my own controls. Something with a chunky granularity.

The boulders and treasure now all fall in columns 32 pixels wide. The player toy now moves fast but aligns with the nearest column when it stops moving, which is much more 'Game & Watch' (only smoother).

60 Minutes Extrapolating
Now to build the basic game and product structures. I'm also adding the rest of the toys (basically variations of the established archetypes plus a few accessories for the player toy). There won't be any selection modes ('menus') or presentation modes (eg: results) just a 'cover' with a logo.

I added a timer for each level and some counters for score, treasure collected and level number. The timer increasing for each level starts to get tedious so now the survival time is constant for each chamber.

There seems to be a bug with camera shake - everything's drifting off to the left slightly for no obvious reason (I never did manage to fix that).

60 Minutes Refining
More focussed playing, tweaking and fixing. Playing against the clock isn't involving enough. The only other option I can think of is collecting the treasure to make the pain go away for a little while.

I can't think of an elegant formula for the levels (such as how often and where boulders enter play) so I'm going to just do a list.

Incredible how quickly an hour melts away, especially when most of it seems to be spent figuring out how to fix stupid problems with such a basic product structure.

60 Minutes Finalising
Time to finalise the repertoire (as in what I have made and become familiar with). I'm going to add any loose ends that spring to mind - but nothing too risky. It's mostly a few ceremonies and audio-visual effects.

Well, that was the plan. I'm still sorting out issues with the game structure. I ended up using homemade placeholder audio from way back.

I don't like the Windows fonts so I'm making my own boxy bitmap one, which won't take too long. Of course, every minute counts...

60 Minutes Final Refining
Now to play with a view to making final tweaks and bug-fixes and adding final polish.

I'm not sure what rules to use to trigger the appearance of accessories so I'm going to use key presses instead. Oh, what a delicious pile of flaky shite I made. Much faffing ensues but precious little useful tweaking. It surprising how time flies when you aren't having fun.

30 Minutes Whatever
I should probably use this time to nail known issues or add a simple Easter egg or a nice special effect or something funny. The game's hard to play so I should really flatten out the difficulty but I just can't be bothered. I don't care enough now.

I wish I'd done a basic shooter instead.

ENDS