Welcome, Guest. Please login or register.
Did you miss your activation email?

Login with username, password and session length

 
Advanced search

1366018 Posts in 63996 Topics- by 55884 Members - Latest Member: LoLVesparum

September 15, 2019, 10:52:51 AM

Need hosting? Check out Digital Ocean
(more details in this thread)
TIGSource ForumsCommunityDevLogsCellephant - a pathfinding puzzle - continuing my LD39 entry
Pages: [1]
Print
Author Topic: Cellephant - a pathfinding puzzle - continuing my LD39 entry  (Read 1272 times)
bittwyst
Level 1
*


Twitter: @bittwyst


View Profile WWW
« on: August 22, 2017, 10:14:45 AM »

My first devlog, please let me know where I'm going wrong  Cheesy

Cellephant was my entry to the 48hour solo competition of Ludum Dare 39, the theme was "Running out of power". The LD results have just been published, and Cellephant placed 22nd overall in the compo. I'd been getting some really nice feedback but I still wasn't expecting that.


         


"Cellephant is a game about a robot elephant, who has to collect batteries (aka cells) in order to power him to reach the exit. Your current battery charge is shown at the bottom, and that’s how many steps you need to take to reach the next battery (fail to reach a battery = die, reach a battery too soon also = you die, because your charging circuit can’t handle it or something…). It’s a path finding puzzle."

I made it for LD using Pico-8, which is a "fantasy 8-bit console" with built-in dev tools for gfx/maps/sfx/music and lua scripting. Pico-8 is a lot of fun and perfect for making a small game quickly for a game jam or for prototyping. I'm going to continue working on the Pico-8 version of my game for now, while I improve the gameplay etc, and I like the built-in tracker so I'll probably keep working on the music with that and export it later if I need to. I'm thinking of rebuilding the game in Unity (or similar) later, for improved mobile support, more detailed graphics (representing a robot elephant in 8x8 pixels was never going to be ideal...) etc.

So far since the compo I've started working on a couple of things I got the most feedback about - fixed the delay after collecting a battery until you can move again (the higher the number on the battery, the longer the delay was), stopped it from quitting the game if you press both buttons by mistake, maybe something else. I also fixed a graphics bug nobody noticed and started thinking about improving the difficulty curve so it doesn't stay too easy for too long at the start.

To-do list for the moment:
  • Tweak difficulty curve
  • “Generating (please wait)” screen between levels
  • Show a line for the path you've taken, instead of crosses
  • Backgrounds behind the game grid
  • Remix/expand music
  • Instruction screens
  • Touchscreen
  • Remake in Unity
  • Figure out new graphical style - thinking 16-bit era pixel art
  • Animation, juice
  • Start-level selection
  • Other features
  • Other game modes
  • Online scores?
  • Figure out what else should be on to-do list
  • Woop or boop!

More stuff to come, I'd love to hear your feedback if you have any (you can play the LD version of the game from here, and please follow if you're interested!


Updates:


Follow me on twitter to keep updated!
« Last Edit: October 07, 2017, 11:54:03 AM by bittwyst » Logged

bittwyst
Level 1
*


Twitter: @bittwyst


View Profile WWW
« Reply #1 on: August 22, 2017, 10:24:22 PM »

Thinking about Levels / Game Modes

So I've been thinking about levels, as it stands they're procedurally generated with various factors that influence the difficulty, and there's a basic algorithm that modifies those factors to increase difficulty as the level number goes up. The algorithm does the job but it's less than perfect - there were two bits of feedback about it that I got from multiple people:

  • The difficulty curve is too slow to ramp up - too many easy levels before they get harder.
  • The difficulty of the levels doesn't purely increase as you progress through the sequence, you'll get easier levels in between harder ones - this is because of the random nature of the level generation, even as all the influencing factors are increased, that doesn't set an absolute difficulty, it's partly down to chance.

The first of these could really do with fixing, and it won't be a problem, it just means tweaking the algorithm that ups the difficulty, maybe giving it more of a logarithmic curve so it goes up faster to start with. I'll admit I didn't get to spend any time tweaking the difficulty curve for the LD version of the game, I'd been changing the difficulty factors by hand as I was testing and finishing up the other parts of the game, and only threw in the difficulty curve algorithm within the final hour before the competition deadline, so as soon as it was reasonably working that was it. Will be nice to have the time to tweak it gently.

The second thing in the list isn't the worst thing in the world, some variation in difficulty is ok and it can be nice to have an easier level or two to breeze through after you've been stuck on a particularly challenging one for a while. The random nature of it isn't ideal though, I know there's established theories about what difficulty curves should look like and it could be good to try and model those a bit more closely.

So, tonight's idea is.... *drumroll*

Two game modes:

  • The "regular"/default mode could be a curated list of levels that are ready-made rather than generated on the fly, in a chosen order to give a nice difficulty curve.

    Given that the levels to be selected from can be generated for me, and I can work on the curated list over the course of the game's development time, it shouldn't be unrealistic for this to be a satisfying number of levels, some hundreds or whatever, and potential to add more later. Having a consistent set of levels every time you play, and the same as for other players, additionally lets other nice features make more sense, like continuing from the last level you reached, best times per level, and comparison of progress and best times with other players, online scores and the like.
    ..
  • An "endless" mode that uses the current procedural generation so you never run out of new levels to play.

    Maybe this can be unlocked by a certain amount of progress through the "regular" mode. This mode could potentially have in-game options like starting difficulty and rate of increase,
     so you can play through a game with levels that get progressively harder overall (with the random variation already discussed) like the game as it currently stands, but with the rate tweaked to your liking, or you can stick with endless levels of a specific size and approximate difficulty.

I don't think curating a list of procgenned levels is a brand new original idea to the world, but tonight was the first time I thought of doing it for Cellephant, so it feels like progress. There are bound to be other potential game modes beyond these two, but it seems like a good start - maybe people would even want an editor to design their own levels manually, I'd work on that if there was significant support for it, and I'm always open to other suggestions.

End of update! - glad to get tonight's thoughts down, but man I really need to start throwing some gifs etc into the first post and future updates, make this whole thing a bit more eyecatching.
« Last Edit: August 22, 2017, 10:29:56 PM by bittwyst » Logged

afender7
Level 0
**


twitter.com/afender7


View Profile WWW
« Reply #2 on: August 23, 2017, 01:11:47 AM »

I had a little play and it's pretty cool, definitely has potential! The fundamental premise is simple, but effective. I look forward to seeing where you take this.

One thing I would say is maybe don't overload yourself with too much at once: that to-do list is pretty huge! My advice would be to strip away everything other than the base-level design of the puzzles and focus on that as much as possible, then build up from there. Maybe take a look at Alcazar: it's a simple, very well-designed path-finding game with lots and lots of hand-made levels.

Good luck!
Logged

bittwyst
Level 1
*


Twitter: @bittwyst


View Profile WWW
« Reply #3 on: August 23, 2017, 03:37:49 AM »

Thanks for the kind words and helpful feedback!

Yep I appreciate what you're saying about not taking on too much at once. The to-do list was basically everything I could think of that I would like to, um, do, to turn Cellephant into something resembling a finished game. I should break it down into priorities and milestones, more bite-sized chunks with specific goals. I'll add that to the to-do list  Cheesy
Logged

afender7
Level 0
**


twitter.com/afender7


View Profile WWW
« Reply #4 on: August 23, 2017, 07:18:12 AM »

Haha, yeah, fair enough. I guess my theoretical todo list would also be very long, it's just that I'm too scared to actually write it down, because then it's real... I am very bad at handling multiple problems at once!

Look forward to seeing more :-)
Logged

bittwyst
Level 1
*


Twitter: @bittwyst


View Profile WWW
« Reply #5 on: August 23, 2017, 12:22:51 PM »

Thanks!

I DUN MADE A TWITTER! Follow me  Kiss
Logged

bittwyst
Level 1
*


Twitter: @bittwyst


View Profile WWW
« Reply #6 on: September 23, 2017, 05:18:35 AM »

Giving Cellephant a new look! #screenshotsaturday

So, even though I received at least one compliment about how I'd managed to represent a robot elephant in only 8x8 pixels:


(This the sprite editor of the awesome Pico-8)

...I decided when I thought about working on the game further after Ludum Dare that I wanted to change to a more detailed graphical style, as Cellephant deserves all his glory to be shown off. I figured I would probably go for higher-res pixel graphics, maybe in a 16-bit style as mentioned above, because I like pixels and I'm more familiar with working in 2D. I'm no artist tho, I was going to try and do it myself but had no confidence I could produce anything worthwhile, and I haven't reached a point yet where I want to bring anyone else into this project, so not a lot was getting done.

The other graphical styles I had considered were low-poly (which could look pretty cool and could be more straightforward to make than high-poly, but either way I don't have any real 3D modelling experience), or voxels (kinda like building lego, right?  Durr...? ). The other day I decided to give voxels a go, and fired up an app on my phone so I could create in bed... Here's a first draft of what he now looks like in the editor:



Simple, but I think I like it. Still needs to be animated etc. I'd love to hear any feedback you guys might have.
Logged

afender7
Level 0
**


twitter.com/afender7


View Profile WWW
« Reply #7 on: September 23, 2017, 02:43:25 PM »

Looking nice, I think a simple, blocky style suits it well!
Logged

bittwyst
Level 1
*


Twitter: @bittwyst


View Profile WWW
« Reply #8 on: September 23, 2017, 04:09:18 PM »

Thanks! Yeah he seems happy with his new voxels.
Logged

bittwyst
Level 1
*


Twitter: @bittwyst


View Profile WWW
« Reply #9 on: September 24, 2017, 01:48:26 PM »

I've had a fun new idea I'm excited about, I'm gonna keep it as a surprise for now tho! So, um, maybe this post is unnecessary
Logged

bittwyst
Level 1
*


Twitter: @bittwyst


View Profile WWW
« Reply #10 on: October 07, 2017, 11:52:20 AM »

Map Generation #screenshotsaturday

So I've overcome the mental hurdle that's been preventing me from getting started on coding, hopefully things will start happening a bit quicker. Today so far I've converted the map generation code from pico-8 into the Unity rebuild, here's a quick gif if you can excuse the placeholder graphics:



Shouldn't take long to get the rest of the level/path generation in, then gameplay updates will follow...
Logged

Pages: [1]
Print
Jump to:  

Theme orange-lt created by panic