★ TouchArcade needs your help. Click here to support us on Patreon.

Level Design Opinion

07-10-2012, 10:49 PM
Joined: Nov 2010
Location: Portland, OR
Posts: 87
Level Design Opinion

Hey guys,

I had a level design opinion question for you.

Do you think it's better to go for quantity in terms of level design, or freshness of content?

For example, when we're designing around a specific concept or mechanic, we use some repetition to demonstrate mastery of the concept.

Do you think it's better to push the envelope of repetition in order to have a higher quantity of levels, or is it better to have a shorter game that is continuously introducing new concepts?
07-10-2012, 11:24 PM
Joined: Aug 2011
Posts: 1,116
As a gamer (not a developer), I think balance is key. You don't want the player to just try a new mechanic once and then move on to the next one, but you also don't want to bore the player by making him do the same thing over and over again. A little repetition is fine, but you definitely want to make sure you introduce different mechanics periodically to keep players interested in your game =]

Now we are cosmic friends forever, ok?

07-10-2012, 11:25 PM
Joined: Mar 2012
Location: California
Posts: 31
This article was probably the best one I ever read regarding level design with regards to puzzles/mechanics.


Twitter: RuizNick
07-11-2012, 02:08 AM
Joined: Dec 2010
Posts: 240
Originally Posted by ruiznick View Post
This article was probably the best one I ever read regarding level design with regards to puzzles/mechanics.

I've just read it, great article.
Thanks for the link.
07-11-2012, 02:20 AM
Joined: Jul 2012
Posts: 31
I agree it's all about balance. Defining balance is the problem :>

As a rule of thumb, I normally introduce more mechanics at the start and less at the end. So the first levels are constantly introducing new mechanics, to keep player interested. The last levels have less new stuff to introduce and here is where repetition and challenge pumps up. So I'd say quality (high frequency of introducing new concepts/enemies/mechanics) at start and quantity at the end. Start/end doesn't necessary means start or end of the game. It can be start of the chapter, start of some new environment etc....

And to get back to balance - I think the best way to find the right balance is to give the game to the testers and see how they play.
07-11-2012, 04:07 AM
Joined: Dec 2009
Posts: 130
Our iPad game Flockwork had 80 levels for the initial release. In the first 10 levels we introduce a lot of game concepts. Each level is designed to 'teach' the user something.
* your finger moves the sheep
* your finger moves MULTIPLE sheep
* you need to collect the stars
* your sheep need to end up on the targets
* sheep can bounce off stone blocks
* sheep can escape off the edge of the screen
* sheep can die in fire

Over the next 20-30 levels the "new" stuff is introduced more slowly: pushable blocks, water, moving blocks. Remember that each new additional element potentially adds to the complexity due to extra interactions between the new elements: for example in Flockwork you first encounter pushable blocks, then water, and then realise you can push a pushable block in the water to make a bridge.

I'd definitely echo Raines in saying that testing is critical to get a good "learning curve" for the levels. We tested with as many of our friends as possible. The idea was to introduce concepts fast enough that players didn't get bored, but slow enough that they don't get confused/stuck. You can never spend enough time watching other people play your levels
07-11-2012, 05:00 AM
Haven't read the article yet but here are my personal 2 cents...

I think any good game should pleasantly surprise the player on a regular basis, but the amount of content doesn't necessarily have to suffer. I feel that Super Mario Brothers 3 (yes, the NES title) is an extremely good example of brilliant design.

Minigames aside, the game featured a fairly small number of mechanics:

1) A protagonist that could run, gaining some momentum if the right button was held, and jump
2) A screen that could scroll in any direction, both with player movement and independent of it
3) Walls/blocks that could move in any direction or rotate
4) Enemies that were more mobile but could be destroyed, and enemies that were less mobile but could not be destroyed (or were very difficult to destroy)

There were other features such as a timer and power-ups, but they were usually not essential to level design. It was possible to beat almost every level without taking either of those things into consideration...

More to the point, the amount of variety that was achieved with these mechanics was staggering. Many stages found some way to use those 4 basic elements to present a completely new challenge. You might have to perform a series of skillful jumps as you race to keep up with a scrolling screen. You might have to nudge the D-Pad with surgical precision so as not to be destroyed by a static enemy while you attempt to keep your plumber out of the lava. You might have to land on a spinning block at the right time, or risk being flung far away, usually to your death. In a couple of particularly memorable levels, you have to escape an angry sun as it chases you, swirling about on screen in a delightfully crazy way. On the other hand, some challenges were not a threat so much as a promise of potential reward: nail a bunch of enemies standing close together with one turtle shell and gain an extra life... Time your jumps just right and grab stars fast enough, and you can spend almost an entire level being invincible... And so on.

In short, it's not about how many different elements or mechanics you implement. It's about figuring out just what you can do with the mechanics you have. SMB3's levels generally weren't designed to teach you anything at all. They were fun, self-contained tests of reflex and critical thinking, and they never failed to deliver new surprises.
07-20-2012, 10:58 PM
Joined: Nov 2010
Location: Portland, OR
Posts: 87
Thanks guys, these are some great perspectives to consider. I'll do a little recap once we finish the game of how we decided to proceed, if you're interested.

Level/challenge design is currently about 85% done, but with minimal testing so far. That should change soon.
07-21-2012, 06:37 PM
Joined: Jan 2010
Posts: 68
I really liked this article, though it's mostly geared towards platformers-

Personally, I'm not big on 'gimmick' levels, where a mechanic is introduced once and never seen again. I'd rather see game elements that work together to create emergent properties when combined.

I'm working on a puzzle game right now and I'm trying to do something similar... like every few levels a new type of puzzle piece is introduced, by itself at first, and then gradually integrated into what the player already knows.

But yeah, I'd agree with Raines and just test everything.
07-22-2012, 02:48 AM
Joined: Jul 2012
Posts: 31
Originally Posted by u2elan View Post
Level/challenge design is currently about 85% done, but with minimal testing so far. That should change soon.
Level design will (should) change a lot after testing, trust me :> Unless you guys somehow nailed it for the first time, which would be exceptional.

Anyway, keep us posted..