The Flow of Level Design

[Strife]

As I've spent more and more time on my Sonic Worlds project, I've become more conscious about how my levels "flow" - in other words, how smoothly the player transitions between different points of the level. I think this is something that's important for beginning level designers to get used to, because it makes levels feel more polished and makes players feel like they have ample time to decide what action they want to take.

Take, for example, a Sonic game with this stretch of ground:

Sonic's going to build up speed as he runs down that hill, if he wasn't moving fast enough already, and yet all of that high speed will go to waste when he hits that wall. Unless the player instinctively knows that this wall is coming, all of the momentum they've built up will vanish, breaking the flow of the stage thus far.

Here's a more natural alternative:

Instead of stopping dead in his tracks, Sonic will get carried upwards by the curved corner. Both versions have the same effect in that they reduce Sonic's horizontal speed to zero, but this curve has the added benefit of giving the player a high jump if they approach the curve at a high speed.

Now, let's say that you decide to go with the path in the 1st picture, with the solid wall. The player has two choices: run into the wall and then jump over it, or jump over the wall before they hit it. Either choice results in the player continuing to travel to the right - therefore, you can only continue the flow of the stage by continuing to build your level to the right.

If you decide to go with the path in the 2nd picture, the player also has two choices, but they're different: either they jump over the curve and continue right, or run into the curve and receive an upwards boost proportional to their speed. This results in the player travelling in one of two directions - Right or Up - without breaking the flow of the stage. This means that you can safely provide two different paths for the player to take and each of them will feel like a natural choice. Heck, you could even provide three alternate paths, like so:

So if Sonic is moving fast enough when he hits the curve, he'll land smoothly on the middle platform, but if he runs into it at his highest speed, he'd land on the bridge above where a stash of goodies could be awaiting (assuming there's no other way to get up there). This, of course, does not take into account the fact that Tails can fly, and the special abilities of the player's character are also something you should account for.

Anyway, that's my two cents on how flow can improve stages. There are, of course, other things to take into consideration, such as the player's position in the stage (beginning, middle or end), scenery/background transitions, badnik placement, etc, but I could probably blurb about those in the future. Right awn. :3

[Core]

That's interesting. So I'm wondering, how much important is flow in a Sonic game? Do you think the structure in the second picture is always necessarily better than the one in the first picture?

[GSF]

That second screenshot is just like the layout in Ice Cap Act 2's first area. Structures like that are the base of Sonic games' "rollercoaster platformer" element.

It was a bit abused in Sonic Advance 2, where backtracking was almost impossible, but layering paths that cross with each other at certain points is quite important, I think.

In a game where speed is an important element, it's fun to have multiple paths so you can elaborate the best one and use it. Games with just one path (modern Sonic games in many cases) have very isolated parts where there's an alternate path, and usually these are simply a flat platform over an area with enemies, not a whole new section of the stage.

Pure platforming parts are not bad per se, but it all depends on what you want the player to experience in the level.

[OverbounD]

I understand what you are saying Strife and its a very good point. "Flow" is the reason I always say draw your levels on graph paper first. While you draw a level you can look at it and imagine how large portions of a level will flow. If you realize a section of the level doesn't flow right, its a lot easier to erase it than redo a portion of a level in MMF or GM.

On the other hand, gsoft is kind of right that these types of areas are over used in 2D modern Sonic games. Look at Sonic 1, CD, or Lava Reef Zone. There are zones without half pipes. So I guess I tend to like to break the flow in favor of slower platforming at times in most of the levels I've made.

[Strife]

That's interesting. So I'm wondering, how much important is flow in a Sonic game? Do you think the structure in the second picture is always necessarily better than the one in the first picture?

I'd say that it's not really about what's better, but more about what feels right for a particular stage. I like to think of it in an economic way, where speed/momentum is treated as a base resource like score or rings, but more transparent. If the level's layout allows me to build up an enormous amount of speed, it feels awesome when there's a way for me to "spend" my momentum on something, usually in the form of catching some sweet air time. Modern Sonic games trade the invisible resource of momentum for a visible boost meter.

As for general platforming, the second screenshot in my first post would actually be a great way for the player to transition from a high-speed section to a slower platforming section. The player runs through the curve, catches some air, and then lands flat on the ground to the right, at which point their horizontal speed is minimal. From here, you can safely place obstacles that require more precision. There's actually a good example of this in Sonic 3's Angel Island Zone Act 1.

One of the things I often have difficulty with is badnik placement. Badniks can be tricky to handle properly; If you place them in a high-speed section, they can either pose no threat because the player zooms right past them before they execute their attack, or the player has too little time to react to their presence before they slam into the badnik. Again, this is where the infamous curve comes in handy, since it allows the badnik to become visible on-screen for a couple seconds. Platform staircases also work nicely. Failing that, simply providing some sort of reward for the badnik's destruction also helps. (This was actually part of the core gameplay in Sonic 3D Blast - the no-Flickies hack exposes just how much of a dilemma this can be.)

Looking at official Sonic games, I really like how the drilling badniks from Sonic 1's Labyrinth Zone were handled. They appear in areas where the player is expected to be moving slowly, and their attack pattern makes it so that it's often safer to take the effort to destroy them rather than try and run past them entirely.

[TailsSena]

This will probably help me a lot with Leafgreen Act 2, and possibly in expanding Act 1, as it's not very big or difficult yet. Still needs a bit more of a bunch of variation.

[DimensionWarped]

I don't believe in sticking quarter-pipes and natural curves everywhere. Honestly, I don't think emphasis should ever be placed on making the player never have to stop moving if he follows the natural path along the ground. Don't get me wrong, tons of starts and stops everywhere isn't necessarily a good thing either, but as long as there is a way to circumvent the breakage and as long as you are capable of picking it out after a playthrough or two, that attitude is preserved and looking for how to keep that flow going in an optimal fashion is really supposed to be a central part of the challenge in doing timed runs in the first place.

Half pipes to split paths up are fine of course, but I still prefer for there to be slightly more to it than that.

[Serephim]

Just don't get repetitive. The only time stopping, jumping, falling, springs, boosters, trick rings, rails, spindash, homing attack, blah blah ect ect gets annoying is when they player encounters them too frequently. Then it becomes more of a chore or ritual than a luxury.

[luksamuk]

Just don't get repetitive. The only time stopping, jumping, falling, springs, boosters, trick rings, rails, spindash, homing attack, blah blah ect ect gets annoying is when they player encounters them too frequently. Then it becomes more of a chore or ritual than a luxury.

I agree with this ^

That basically points you to make level-specific gimmicks.

[thesuperinvader]

i always hated sonic games, suddenly stopping my momentum

[DimensionWarped]

You know, that random comma splice completely changes the meaning of that sentence.

[TailsSena]

You know, that random comma splice completely changes the meaning of that sentence.

I was going to mention that, but I figured it was meaningless enough to ignore...

[Serephim]

It was pretty meaningless. It used to be fun, but these days i just don't care to point them out anymore.

People sound retarded on the internet even while forming completely coherent sentences

[Strife]

Hey, we all make mistakes I suppose. :3

I've had another thought recently about flow, this time about when exactly you should introduce new gimmicks and challenges in your stages. From my personal experience with platform game creation, I've found that depending on how complex your engine is, sometimes the only thing you need to do to keep your players engaged and the level flowing nicely is to introduce new challenges. If you're testing your level and wondering if your players will start to get bored after a certain point, then stop and think of a unique enemy or gimmick to put there.

The classic Mega Man games thrive on this tactic - at its core, an NES Mega Man game has a fairly unspectacular physics engine; just running, jumping, shooting, and a few other things. Instead of focusing on the manipulation of physics, a Mega Man game relies on providing the player with a steady stream of unique enemies to fight and gimmicks to traverse. Throughout all 8 of the regular stages, only a handful of enemies and gimmicks are ever recycled or seen a second time.

For those of you who remember Ristar, that's a perfect example of how variety is a good thing - every "Act 2" in each planet not only has different scenery, but at times completely different obstacles. Try to imagine what it would have been like if each second act was the same as the first, just with a different level layout and boss. More repetitive, yes? Repetition is good for getting your players used to a specific gameplay challenge, but there should be a point in the game where you expand on it in order to keep it fresh. That's why I loathe MMOG's that require long periods of level grinding - if it weren't for the anticipation of future rewards, boredom would settle in fast.

[Serephim]

Megaman didn't have a very complex physics engine, but it most definitely required you to master it pretty early on in the game.

Yeah, i think the player should constantly be encountering new stuff.

You don't really even have to program a completely new gimmick or script something special just to achieve the same effect. You can simply alter things they've already seen before. Similar to how Megaman goes from spikes on the floor, to spikes on the floor over fall-through platforms, to spikes on the walls, to spikes on the ceiling, to LOW spikes on the ceiling, forcing you to short hop to not kill yourself.

Or a game like Devil May Cry, which will slowly introduce new enemy types as you continue to play the game. But something as simple as placing one enemy type in a spawn with types of enemies that usually don't spawn together will completely change the player's situation up. Or instead of one enemy spawning, two of the same type will. (Im specifically thinking of the enemy Blitz in DMC4, which was notorious to me on first playthrough.)

Or, instead of changing the enemies, you can just change the environment you face them in. Fighting a ramming bull on a platform is a much different situation than fighting a bull on a platform over a pit of lava. Or likewise, fighting a ramming bull on a platform with higher tiers he cannot reach you on.

I find the most annoying thing about MMOs isn't only the fact that grinding spreads your advances apart so far, but also that because the game is grind-based, the enemies are completely one dimensional.

Usually, in most MMOs, stats like Magic Defense barely exist, and damage formulas rarely ever incorporate more than one stat. As a result, all classes are streamlined into a single damage method, and battling enemies is always a combination of two things -- DPS and Tanking.

But in older RPGs, the ones I used to play when younger, boss enemies and some encounters would require you to formulate different approaches to win, or you'd give yourself tons of trouble. Perhaps you had to actually utulize that spell that was mostly useless until then, or actually fall back on your defensive abilities, or abilities that cause status effects. Sometimes, you were encouraged to not do damage at all in some cases, or it'd backfire. ("Attack while it's tail is up....it'll counter with a laser attack!") Or perhaps you were required to use Common Sense to prevail, like attacking the environment instead of the enemy directly.

Of course, that's old RPG talk, but I feel like the concepts directly carry over into all types of videogames. You just have to keep switching it up, however you can, in unpredictable (cleverly unpredictable, not annoying) ways.

[DimensionWarped]

The problem with MMORPGs is that they can't really anticipate what capabilities a group will have when they encounter any given enemy, so they can't very well incorporate things like 'That one boss where you need to get all defensive when it charges its laser' or the 'elemental barrier change' boss.

As for the 'mean counter-attack' thing you mentioned, I think that actually is incorporated into a number of MMORPGs. By the way, all of the things you and I mentioned are really just... gimmicks. Reaffirming my long held stance that gimmicks are at the heart of all great games.

[Serephim]

I don't call them gimmicks because...i dont know, it just doesn't sound right. Probably because i hear it so much around here.

A better word, or phrase rather, would be "something to do".

"this forest stage needs more gimmicks!"

vs.

"this forest stage needs more to do in it!"

I dunno, "gimmicks" just sounds restrictive to me. It really doesn't matter what you call it though, as long as people get the general idea i guess.

[DimensionWarped]

A gimmick is just a feature that sets something apart from its peers. Nothing more, nothing less. We use it a little more specifically to refer to interactive stage elements that aren't enemies, but that's just a term of convenience.

There are other less flattering uses, the most popular that I can think of being a cheap trick to attract an audience. Anyway, there's really no need to get locked into the negative connotations of the wordl

[Djawed]

I think hitting against a wall and stopping isn't nesceseraly bad for the flow. I think it can be rather usefull as a warning that the player should slow down for reasons of alot of hazards and platforming. Also deathpits when used.