Design space of skills and graphic sets

[Strato Incendus]

The developers of the trading card game Magic: The Gathering use the term "design space", referring to "how much can you do with the different card types (without completely breaking the game)?" Perhaps this concept exists in game development on a general level already; anyways, I think it's also applicable to Lemmings.

In our case, it can be broken down into two sub-sectors:
1) How much can you do with this graphic set (=terrain and objects)?
2) How much can you do with this skill?

A) Graphic set design space
Have you ever consciously noticed how the first step of picking a graphic set already influences what the level will look like? And I do not mean the mere aesthetics (that's kind of a "well, duh" point ), but the physics that are actually game relevant from a mechanical standpoint:

Some graphic sets are very block-ish and regular, thereby lending themselves well to e.g. climber stuff - Marble, Crystal, Fire, and Pillar come to mind from original Lemmings, Brick from ONML, and pretty much all of the Lemmings 2 graphic sets.

In other tilesets, in contrast, it is hard to even find such a climber-friendly piece of terrain in the first place: Dirt from original Lemmings is one of the worst in this regard, Rock at least has the gems which are straight walls, and Bubble has a lot of walls that look like climbers could go up there, but they actually have tiny gaps in between that prevent this. More often than not, you'll have to insert a steel piece, pretty much giving away at this point that you're supposed to use a climber here.

Finally, there is the limitation of the Lemmings 2 graphic sets:
I've been thinking a lot about why I don't like Lemmings 2 as much as original Lemmings or ONML, and surprise - it's not actually the multitude of skills. In fact, considering how many different skills were added to NeoLemmix and Lix, few of which are actually overlapping, and people are still considering adding more, it almost seems like one can never have enough different skills available (no one is forced to use them all for level design, after all).

The main issue is the shape of the terrain:
1) The levels often consist of many thin platforms of squared or rectangular blocks.
2) Many feature vertical scrolling.
3) The combination of 1) and 2) leads to a look that is more reminiscent of jump and run games like Sonic or Speedy Eggbert (not so much Mario, btw!), i.e. such that include lots of up and down-movement, rather than a Lemmings game, since up to then, these only featured horizontal camera movement.
4) The execution difficulty in Lemmings 2 wouldn't be nearly as horrible if the terrain were shaped more like in traditional Lemmings levels; however, due to this jump-and-run game level design, oftentimes the execution-heavy skills are the only available option to even make it past all the obstacles in a level. (This is probably also a reason for the increased splat height in Lemmings 2.) Try solving pretty much any Lemmings 2 level with classic skills only and 90% of them would turn into builder fests.
5) The levels look pretty "ugly" and unmemorable as a result, compared to the more landscape-like levels from original Lemmings, ONML, or Lemmings 3D.

The latter point is also the main downfall of Lemmings Revolution to me, which theoretically should be perfect for lovers of original Lemmings, but the different tilesets just look way too similar - literally like they're all "carved from the same wood" . I'd rather play Team 17's remake of original Lemmings, for that matter.

Hence, when designing a new graphic set - or, even before that, considering whether a specific new graphic set is actually needed - it may be worth asking yourself:
1) How much does this graphic set contribute mechanically?
2) How much does this graphic set contribute from an aesthetic standpoint?

For example, we have several snow- or castle-themed graphic sets, but mechanic-wise, they're not actually that different. The food graphic set, in contrast, stands pretty much on its own, and is different enough from the candy set while at the same time providing some interesting synergies with it. Also, there are comparatively many graphic sets that feature water, fire, and triggered traps, but only few contain dedicated one-use traps, splitters, or teleporters.

The Lab graphic set is very block-ish, and therefore it's hard to create interesting looking level shapes with it - in contrast to graphic sets consisting of more distinct, irregular shapes, like Purple or Tree (considering both namida's originals as well as GigaLem's remakes here).





B) Skill design space
Kieran already did an overview of all the Lemmings 2 skills and ranked them all individually, according to usefulness and how much each of them can justify its existence.
I'm going to take a slightly different approach and sort them into categories or niches that they might actually fill, related to the currently existing NeoLemmix / Lix skills.


Redundant:
These skills do basically exactly the same thing as some other already existing skill - which is why I can usually just fill these existing skills in behind a = sign .

Attractor (=Blocker)
L2 Bomber (some uses with knockback behaviour and tumbler physics in Lix, but for NeoLemmix, it would just be non-lethal, and that's all the difference)
Club Basher (=Basher)
Diver (=Jumper)
L2 Fencer (=Basher with a slight angle)
Flame Thrower (=a non-lethal bomber, but the Lemming still has to be close to the terrain it wants to alter)
Hang Glider / Magic Carpet (=Glider)
Hopper (=Jumper)
Paraglider (=Floater or Glider, depending on context, but only either of the two without the fan)
Roller
Scooper (=Miner, given that it literally replaces the Miner in Lix; just a different angle, nothing else)
Skater (=ice objects do not exist yet)
Skier
Stomper (=Digger)



Corner-cases:
These skills are slight variations of what we already have - minor differences that can however become crucial in certain situations and thereby create interesting puzzles. Still, some of these can be grouped together - these are separated with a slash - and therefore, if any of them should be introduced in the future, it should only be one of each of these groups; for example, a temporary swimmer could either be called kayaker or surfer, but there certainly wouldn't be a need for both.

Archer / Spearer / Thrower
A non-lethal stoner with range, basically: It can stop Lemmings, break their falls, and build small bridges. There's even a nice trick with them, where you shoot up in the air and then have to keep the archer lemming busy to prevent him from slipping by his own dropping arrow - similarly to the stacker.

Ballooner
Here we have potential for a one-use climber (see below).

Bazooker / Mortar
And this is the non-lethal bomber with range. Even the holes it creates look similar to the bomber.

Filler / Glue-Pourer / Sand-Pourer
Only slightly different in nature: The shape the sand pourer creates remains constant, always resulting in a little ramp which can be used in a similar fashion as a builder. The filler is just a weaker version of the glue-pourer: Both will fill up small gaps, but the filler is completely useless on flat terrain, whereas flat terrain amplifies the glue-pourer by turning it into a platformer.
Of these three, the glue-pourer probably would have the most amount of remaining design space: The only currently available option to fill small gaps without resorting to standard builders is using a stoner / cuber, always resulting in the death of a lemming.
Most of the time however, the platformer as it currently exists in NeoLemmix and Lix (the latter one also being able to cover up trap triggers on flat ground, like in L2), gets the job of the glue-pourer done just fine.

Kayaker / Surfer
A temporary Swimmer, I emulated this one in my 1.43 pack "Lemmicks" using the "non-permanent skills" gimmick.
If one were to ever introduce a Kayaker into NeoLemmix or Lix, it would beg the question why one shouldn't also create such "temporary" equivalents of the other athletic skills - meaning: a climber that can only go over a single wall before becoming a walker again, floaters / gliders that are only save for one drop, a disarmer that can only disarm a single trap.
Especially the latter one might be worth considering, because the main reason the disarmer is getting little "stage time" at the moment is probably the fact that it is over-powered, at least if traps are supposed to be the main hinderance of a level.

Planter
Kieran's main criticism of this skill was its inconsistent behaviour: The plant animation always looks the same, and since its shape is irregular, it can be walked over from the right, but not from the left. If this behaviour were regular, i.e. going along with the direction the lemming is facing, this could be used to create one-way walls - a creative equivalent to the destructive fencer, which can create tunnels only passable in one direction. It would fill the niche between fencer and stacker, so to say - because fencing through a stacker doesn't actually result in the creation of such a one-way wall.

Pole Vaulter
There is one type of obstacle none of the current skills can overcome - and it actually relates back to graphic sets such as Dirt, which often have protruding pieces. Yet, this type of obstacle is very common in levels, and it should be - otherwise, most levels would look as block-ish and generic as those from Lemmings 2:
A high wall with a protuding piece at the top and/or an irregular shape.
Climbers can't go up, Builders need to take a huge detour around such walls, Gliders require updrafts, thereby clearly giving away "use a glider here". The only way to get around such an obstacle from the crowd's side is the fencer. But oftentimes, you want to send a pioneer lemming ahead who later on frees the crowd from the other side. The pole vaulter could do this, without having to shape the terrain in a straight form that immediately screams at you "use a climber here".

Runner
While I personally just consider this skill the epitome of an invitation to execution-based madness, I have come to accept that crowd control or the lack thereof, with one lemming having to get ahead of the crowd to do stuff in time, is actually considered part of the "puzzle" by a major part of the community. (Basically, if one has to fiddle around with pixel-precision for the lemming to finish its job before the arrival of the crowd, it may just be fiddle-y because it's a backroute .) The runner can get ahead, move faster, and also make greater jumps in combination with the jumper. Now, if it were also bashing / mining / digging / building / platforming / disarming faster, including faster movement while being a climber, floater, glider, or swimmer, there might be potential here.
As long as the only interesting interaction is with the jumper, though, I think a well-working jumper could fulfill the main use of the runner - getting a lemming ahead of the crowd - just fine by itself, as it already does frequently in Lemmings 2.

Slider
I've already spoken about how weaker versions of existing skills can be useful from a level designer's perspective before. The slider is a climber in reverse: A floater that only works on straight walls. Add in the fact that it always turns around when doing a slide, this one could provide a lot of interesting opportunities, considering that turning lemmings around just the right number of times is one of the main challenges of many advanced levels.

L2 Stacker
While this skill's behaviour can be somewhat emulated in NeoLemmix by making the stacker a climber simultaneously and putting several stacks on top of each other, the resulting big stack is more like a climber staircase - one which can only be used from one side. The L2 stacker creates perfectly straight walls while lifting itself up with it simultaneously - and without having the option of climbing any further walls it encounters after having completed the stack, in contrast to the NeoLemmix stacker which, if used this way, continues to remain a climber afterwards.


Unique:
These skills do something which is currently absolutely impossible, therefore level design should benefit greatly from having such options available.

Jumper
Get over a blocker, over a trap, over little gaps, even over fire... Lix levels can already exploit the hell out of this skill, and there's a reason the Linux clone of Lemmings, Pingus, featured specifically this skill as the only additional one on top of the classic eight. When NeoLemmix can finally have this skill as well, this is going to blow the doors to new design options wide open!

Laser Blaster
There simply is no skill that can dig from below. The Laser Blaster does so from a range; an adaptation of the Twister that just moves up straight vertically and then move a couple of pixels to the side it's facing to prevent it from falling down again, this one could fill a niche unoccupied by any other of the skills currently existing in NeoLemmix / Lix.

Magno Booter
Harking back to the graphic set design space of "what can I do with this type of terrain / object?", the main new thing Lemmings Revolution added to the game was the gravity reversal object. In NeoLemmix, this would turn Miners into Fencers, Fencers into Miners, Builders could suddenly build downwards (another currently impossible task)...
Just playing through Lemmings Revolution outlines the loads of potential this skill has - as long as it is possible to assign different skills to it while the Lemming is turned on its head.
In contrast, if the magno booter were to behave like in Lemmings 2, meaning it would be a skill like the climber or swimmer which can't do anything else in the meantime, the design space would be severely limited. In that case I'd rather advise for the introduction of the Lemmings Revolution gravity reversal object at some time in the remote future.

Rock Climber
This skill has two main uses:
1) The ability to climb diagonal walls.
2) The ability to switch to a shimmier while hanging from the ceiling.
This one could become the ideal brother for the climber, like the glider for the floater. If one had asked me before I knew NeoLemmix whether I'd consider a skill like the glider necessary, I probably would have said no. But now that I have seen how much difference there actually is between the two skills, I think the same is true for climber and rock climber.
I don't know how far the development of the shimmier currently is, and what the plans are regarding its interaction with climbers. But if the ultimate goal is to make a climber-shimmier transition possible, then the rock climber would be the way to go, while at the same time adding something new. It wouldn't be wise to change the behaviour of a standard skill like the climber, after all.

Shimmier
The ceiling is unexplored territory for Lemmings so far. Should a gravity reversal object be introduced, the dedicated design space for the shimmier would shrink, of course. But having such an ability as an assignable skill is always nice, too. In essence, the gravity reversal object would be to the shimmier what radiation and slowfreeze were to bombers and stoners - just without the messy execution and a lot more puzzle potential .

Roper
As said for the Laser Blaster, the only direction destructive skills can't go currently is straight upwards.
For creative skills, the only way they can't go is downwards (neither straight nor diagonally).
While the idea of a downward stacker creating a wall at the edge of a platform for another crowd at the bottom or a climber to go up certainly sounds interesting - I already see the Medieval "Rapunzel" levels coming - the main niche the Roper could fill is that of a downward builder / a creative version of the miner.
Sending a lemming ahead, turning it around and mining the way free for the crowd is a common feature of many levels; it would be nice to be able to do the equivalent with a creative skill across gaps.
Perhaps the behaviour should be adapted though, so that the Roper can't go higher than a straight horizontal line, or maybe not even that - reserving this use for the platformer. Because the Roper as it is featured in Lemmings 2 is pretty broken, serving as a builder, blocker, platformer, and downward builder all at once.

Superlem / Icarus Wings / Jet Pack
Well, this one is obvious, I guess? A freely moving single lemming that can fly around... no levels are more jump-and-run like, more in danger of backroutes, and less original Lemmings-like than levels featuring such a skill. But it still would be something completely new, of course - even though I wouldn't be a fan of it .


In general, the ability to affect terrain from a distance, as present in the Laser Blaster, Bazooker / Mortar, and Archer / Spearer / Thrower, is something entirely new as well. By extension, one could put the Roper into this category, too, but to me, the Roper still creates terrain from where he is currently standing, whereas the Archer / Spearer / Thrower can go across a gap, with the resulting terrain being separated from the lemming completely.


Anything I missed? Any slumbering potential you see in certain objects or skills that are not currently accessible to level designers in NeoLemmix or Lix?

[ccexplore]

Interesting points regarding Lemmings 2, though not sure I feel quite as strongly as you do on some points.

I'd say Circus, Classic, Egyptian, Shadow, maybe Space, and Sports are the blocky styles.  Highland is half and half depending on what proportion of green hills vs red blocks you get on a level (which I have to admit, is a very strange mix to my eyes).  Medieval is somewhat block-ish but also features some grass, trees, curve surfaces, and irregular ceilings to hide the blocks somewhat better.  I don't personally have a strong preference for or against the more blocky styles when it comes to the visuals.  I think you can find examples of decent looking as well as less visually inspiring levels across any of the styles.

I do think Lemmings 2 sometimes go overboard with the steel, and in some styles the steel just doesn't blend well at all with the other terrain, and also sometimes the steel blocks used just seems too large to me.  Polar is probably the worst offender in that regard.

The vertical span of Lemmings 2 level I don't generally mind, but on reviewing some levels, I can definitely see that some of the ones that are mostly vertical do tend to make higher use of things like balloons, jet packs, etc. that people probably don't like as much.  I can see that the levels that are more horizontal will tend to inhibit use of vertical movement skills and other skills that require more open space, so you may be right that the more horizontal levels may have a more classical feel to them as well.

Lemmings Revolution may partly be a victim of technology, as well as victim to the core design choices.  Given its choice to feature 3D-ish graphics, combined with the technologies of the time, there may be little choice but to render the world in blocky Minecraft-ish graphics with low-resolution texture.  Those kinds of constraints probably help limit the diversity of styles you can achieve.  Also, I think the fundamental choice to have the levels being visually depicted as stuck spanning around a cylindrical surface makes it more difficult to come up with terrain features (particularly more decorative ones) that wouldn't look out of place within such an unusual, almost very artificial and unnatural setup, again potentially limiting what you can achieve visually.

[Proxima]

Have you ever consciously noticed how the first step of picking a graphic set already influences what the level will look like?

No, I haven't because it's never the first step for me. By the time I pick a graphics set, I usually already have a picture in my mind of what the level should look like, and I choose the graphics set to go with that. Some level ideas -- as you've mentioned -- go well with blocky terrain. Others go better with rough terrain: for example, Level 6 of GemLems requires a steel block in the middle of a wall, to prevent mining all the way through. In Lix this would be unproblematic. But the Lemmings sets all have steel that doesn't match the sizes of terrain pieces, so a single steel block in a wall is rather awkward to get to work. I ended up using the Dirt set.