next »

[diegzumillo]

This is something I've been thinking a lot lately. I played, I think, 3 games that fall on this category: cool new mechanics, boring gameplay. And it's not like the mechanics failed, the experiment was a failure. No. I believe the level design failed to properly explore the mechanics. It's as if most levels in Portal were filled with turrets and death pits and you had to use the portal to take down the turrets and place box-bridges to cross the pits. This is exactly what I see in most games that try to innovate on mechanics but lacked the level design to back it up.

I guess this is why any level design tutorial is inherently limited. The same concepts that people used to design FPS levels in the 90s wouldn't work for current FPS, where the gameplay is only slightly different. Speeds are different, weapons are different etc. So what to say about the infinite range of gameplay possibilities that would create an even larger infinite range of possible level design concepts.

This is just that extra layer of challenge people trying to come up with something new didn't need, but they do have. And it's one hell of a challenge, as I am experiencing it first hand. My Apple game is taking a lot longer than I anticipated because I had no idea how hard it would be on a conceptual level. I knew it would take a long time to make sprites, animations, code and levels. But I spend a lot of time just thinking. Thinking why the demo becomes repetitive, what is missing on the levels. Do I add enemies? instant kill? energy bars? pure puzzles? what are the elements that, if added, would contribute to my main game mechanic? which would detract from it?

It feels like a new mechanic requires you to invent an entire new discipline of level design. And it's only applicable to your specific game. That's a lot of work.

[Alec S.]

What I usually do when I'm designing levels, which you might find helpful, is write down an outline.  Basically, for each level, or area, or zone, or whatever, I write a sentence describing what it is or what function it serves.  The idea is not to repeat yourself.  So you can have one area be "Introduce obstacle A" and then the next one be "have the player deal with multiple of obstacle A in a narrow environment" and then "Introduce a tool" and then "have a puzzle that uses the tool" followed by "have a puzzle where the player needs to use the tool to deal with obstacle A while crossing a chasm" and so on. 

[diegzumillo]

It's good to have a guideline like that.

[ProgramGamer]

Huh, I've been following that advice subconsciously for my game, weirdly enough. It's strange to see it written by someone else lol

[TitoOliveira]

This video provides a pretty good insight into your problem. It resembles what Alec suggested as well.

[diegzumillo]

I've seen it, it's pretty good! but doesn't quite get to the challenge I'm talking about. Take the mechanic from Mario where the tiles flip, for example. Before you can think of how to introduce the mechanic to the player and subsequently present him with more challenging situations that require more sophisticated strategies, you have to figure out what exactly constitute these challenges and strategies. What layout patterns are too hard/easy, which grid size is too big, too small, too narrow, and what kind of enemies combine well with them, which don't? These things have to be well defined and are then used to create the levels. It's all based on experimentation, and I find that hard.

I guess I'm not talking about "level design" exacly, but an almost invisible step preceding it. You can tell that when someone analyze a game, like the video, they already have the knowledge of what a good level using the mechanic looks like and take that for granted.

[TitoOliveira]

I've seen it, it's pretty good! but doesn't quite get to the challenge I'm talking about. Take the mechanic from Mario where the tiles flip, for example. Before you can think of how to introduce the mechanic to the player and subsequently present him with more challenging situations that require more sophisticated strategies, you have to figure out what exactly constitute these challenges and strategies. What layout patterns are too hard/easy, which grid size is too big, too small, too narrow, and what kind of enemies combine well with them, which don't? These things have to be well defined and are then used to create the levels. It's all based on experimentation, and I find that hard.

I guess I'm not talking about "level design" exacly, but an almost invisible step preceding it. You can tell that when someone analyze a game, like the video, they already have the knowledge of what a good level using the mechanic looks like and take that for granted.

Yeah, i get what you mean.
Well, i believe it all comes down to what the player does in the most basic level. In the case of mario everything boils down to jumping and running with him. So i would place a bet that the size of the tiles, the spacing, the timing it rotates, and everything else relating to it is firstly compared to the Mario mechanics.
For instance, rewatching the video, it seems to me that the lenght of mario's jump is th length of the tile, so if he jumps from the middle of a tile directly to another tile, it will land in the middle of it. The time it stays still before rotating is exactly the time it takes to jump, land and run two steps, so the player will always have this window of oportunity of two steps to avoid an error or correct a wrong jump, or something like that. But i'm just speculating and can't really be sure of any of it.
In any case yeah, it is somewhat based on simple experimentation, trial and error, and whatnot. But i believe you can filter down how much, or how far you'll have to trial and error if you have in mind the most basic action the player will be doing the whole game.

[G-Factor]

I think that good level design is under-appreciated. Most people can come up with cool gameplay mechanics. If you spend a day thinking hard enough, you can probably come up with something that sounds interesting on paper.

However, the challenge is to make use of that mechanic in various ways. It's easy to introduce the idea in its simplist form (like a tutorial) and perhaps make 4-5 additional puzzles. But can you make 50 puzzles? Can you make a game that lasts 3-4 hours based solely on that mechanic? Can it even support that many variations?

If someone wrote down in detail the mechanics of say...Portal or Braid, I think many people will still struggle to come up with more than a handful of truly different puzzle scenarios, so I really do believe the hard work is content creation.

I made a puzzle game myself (you can check it out in my sig), and I really struggled after making about 25 levels.

[diegzumillo]

Exactly. That's my point. Laying down the ground rules for good level design is something relative to the game mechanics in question. I mean, sure, there are some universal rules you can come up with but that's a starting point at best.

By the way, I'm playing your game now and it's pretty great. Simple and few rules but each level ties my brain into a different shaped knot.

I'm struggling to make my apple game because I don't feel like there's enough meat on that main mechanic alone for a long game. I made a bunch of levels and I'm starting to understand its mechanics a little better though, what kind of puzzles I can make with it, what is fun and what isn't. But my main question is: did I really explore everything that mechanic has to offer in terms of level design? And adding more mechanics to it starts to lose some of the minimalistic charm I like.

[TitoOliveira]

Exactly. That's my point. Laying down the ground rules for good level design is something relative to the game mechanics in question. I mean, sure, there are some universal rules you can come up with but that's a starting point at best.

By the way, I'm playing your game now and it's pretty great. Simple and few rules but each level ties my brain into a different shaped knot.

I'm struggling to make my apple game because I don't feel like there's enough meat on that main mechanic alone for a long game. I made a bunch of levels and I'm starting to understand its mechanics a little better though, what kind of puzzles I can make with it, what is fun and what isn't. But my main question is: did I really explore everything that mechanic has to offer in terms of level design? And adding more mechanics to it starts to lose some of the minimalistic charm I like.

I'll give my two cents on the level design of your game.
I've played the demo on GameJolt and i feel that you drastically increase the challenge and the number of jumps and twists between each level. The first level could very well be the fifth, or tenth. First few levels could have no hazards and could level by level introduce the ways to rotate the world (walking inside, walking outside, using the same corner to rotate over and over...), then introduce the spikes, and reiterate the same last mechanics, but with the spikes making things more difficult. And keep going like that.
The rest of the game is really nice, the visuals are cool, the character controller feels good to play.

[diegzumillo]

Aw thanks :D I know the demo has a steep difficulty curve. I'm aware of that, I just wanted to keep it short while showing some things it can do. Not ideal at all, I know haha. But, you are absolutely right, the full game would have a more human-friendly progression.