thoughts on Lemmings game design

[mobius]

If I ever made my own Lemmings game..... this is how it would be!

Feel free to suggest your own ideas or disagree and argue or whatever!

1) Block based physics/non-pixel based.

This essentially means that the "pixels" or every place a lemming can be assigned a skill will be larger than the actual graphical pixels. The game's definition will be high so as to be beautiful but they'll be no more "pixel precision" in skill assignment (at least to the degree of classic L1). Though I don't think I want anything as extreme as Lemmings 3D or even L3, unsure I don't have much experience with L3.

2) Logic based puzzles instead of "trick" based puzzles

I consider there to be 2 (maybe more but I won't discuss them here) basic types of Lemming puzzles. A "trick" level is one that requires use of some type of features or mechanic of some skill or object or gameplay itself, that may or may not be intuitive. I don't mean to say that these levels are bad (some of my favorites are these type) but they have faults: Often, a level like this may appear very difficult at first, if one does not know the trick involved. However, once one is made aware of the trick it can becomes trivial. They often can be solved quickly and easily the second time around or quickly at first if one knows the trick from the start.
A pure logic level on the other hand relies on no such trick but rather simple reasoning; timing, order of skill assignments etc. These levels may not rely on any intuitive or non-intuitive knowledge of the mechanics. They may remain difficult even after solving as the number of steps required to remember may be large (or not they could be very easy as well).
Now, it's probably fair to say that most levels don't strictly fall into one or another category but are a mix of the two. And based on all the custom levels I've played over the years it seems logical puzzles are more difficult to create. In any case, I'd like to focus on the latter.

This means some core concepts of L1 will be changed such as; builders will not longer always turn around. If a builder hits an obstacle and could keep walking onward; they will. Almost everyone gets stumped at Postcard from Lemmingland during their first playthrough of that game. Why? I think it's fair to say because the fact that a builder always turns around when being stopped prematurely is not an intuitive design. Why should they turn around? It was an arbitrary design choice. It seems natural to us; only because we're used to it. Of course there are solutions that don't require that trick, but most people seem to resort to that one. At least one solution I know is more clever and also more difficult to see and arguably doesn't rely on any such tricks.


3) Ease of Use/intuitiveness

The sign of a good game imo is when the player starts it up, begins playing and feels like it makes sense from the very beginning. It may take a while or may not be easy for example the player may get stumped at something for a while but when they finally figure it out the reaction should be: "Oh! of course, why didn't I think of that! Or that's clever or elegant!" NOT: "how would I have ever known to do that???"

see #2; level design should not come down to obscure mechanics that don't seem to  propagate naturally from the rules.

Graphics are big thing for me here; there will be no "true physics mode" in my game because it won't be necessary. Terrain should always look like terrain, a button should look like a button, steel.. etc. Background images/colors should not be confused with foreground.

(to be continued......)

[Simon]

Interesting idea to distill the essence of non-trick, non-dexterity into something fresh.

Such a design can even benefit from distinguishing itself deliberately from Lemmings. Builders don't turn? Let your world's builder physics look/work different enough from Lemmings to not provoke thoughts of Lemmings builders. Find your own core ideas for the physics, and introduce puzzle elements that support that core.

You don't have to foster a community around tileset creation, then it's easy to keep the graphics clear. It's not strictly necessary to even foster custom level creation either -- but Lemmings Forums attracts level designers and we would certainly like to play around.

New projects are blissfully unburdened by existing physics/levels/... and you get to bloat features and physics elements at your leisure, then cull the set down to what makes most sense. A luxury to enjoy.

-- Simon

[namida]

I would definitely agree with the use of block-based physics, after my experience in making my L3D pack. Not so sure about the point regarding tricks, though - rather than doing away with them, I'd favor introducing them more naturally. For example, if the builder turns around, before any level that requires knowing this to solve, have a level where the player would almost certianly discover this by accident while solving it.

[Silken Healer]

I feel like lemmings "tricks" are kind of a spectrum from purely logic to purely tricks. Someone could argue the point that it goes from being logical to a trick is way further than somebody else would say. For instance, on one end of the spectrum you could have mining into steel to turn lemmings around example: , in the middle you could have the digging and building to turn around one lemming that mobius said (as this is pretty well known these days) and far on the other end you could have some really hard trick it'd take you ages to figure out otherwise.

I can see where both namida and mobius are coming from, though. Levels that seem near impossible but it turns out was trivial due to using a trick can feel cheap and annoying. However once you know how to use a trick it can really add to levels and feel satisfying to pull off.

[mobius]

-tricks: this is quite a difficult decision for me; I agree with both sides of the argument. Another problem is what exactly defines a "trick" in this sense is very grey and how logical is the trick?

For example; the miners and builders turning; seems a fairly arbitrary design choice, while the trick of creating a notch in a wall with a basher or miner seems less arbitrary; should a basher be able to effect the terrain when they're not going to continue? Or should a miner be allowed to be assigned while standing on steel? and if so what happens? Do they waste the skill? My gut feeling as to what's most logical with the latter is the miner can be assigned, wastes a miner but doesn't turn around; but is delayed slightly. Already that's a bunch of factors that have a trickle down effect on how the level plays out.

Pure logic levels (and/or not so pure) really are my favorite; A Spot of Bother is an example, Devil's Right Hand by Nepster (I think), Dazed and Confused, Parking Garage and Pyramid Scheme by Geoff, also The Floodgates Open by Clam; and several of these also happen to be some of the hardest custom levels ever made (while some are rather easy). But the hard ones require a lot of logical reasoning, which anyone can potentially do without lots of experimentation or in depth game knowledge.
Again, don't get me wrong I still like the tricks; I just feel these logical puzzles are underused. Maybe I'm wrong and I'm not familiar with a lot of good ones in existence?

-Regarding decisions Simon made with Lix:
I'm honestly personally really torn on certain things like the RR and time limits. On one hand (time limits); I genuinely agree with how most puzzles games in recent years have moved toward no time limit. The game is simpler with one less factor in the way to worry about and it having been mostly arbitrary in the main game (and custom content often as well) anyway. However, there are the corner cases where a good puzzle  is enforced [backroutes removed] elegantly by use a of a time limit. And the timer is a very simple mechanic everybody gets; doesn't need any explaining or anything to learn. And the player can pause the game at any time, not needing to be under actual pressure. If I were to include a timer; certain other factors would need to exist.
(for instance; no levels with 100 Lemmings and you need to up the RR just to have them come out in time). Every level with hoards of lemmings should naturally flow into the exit right on time, even if it's tight.

RR-- At first I was against removing that in Lix but now it feels likewise superfluous. Here there are other ways, and some very interesting of executing the same puzzle of spawn interval manipulation without it; runners, jumpers, setting up a unique terrain layout... In the end I'm now simply not as big of a fan of spawn interval manipulation as I was in the beginning. And this mechanic definitely is not as intuitive and not as simple as a time limit.

------

I'd take the best (what I deem as the best) aspects of all of the Lemming games, or some that were terrible but could've been better.... that'll be for a later post.

[mobius]

----more notes on my plans for a game---

Block based physics.

I will try this and see how it goes in early testing. My initial thoughts are:

-should blocks be large or small? That is; a Lemming could be about the size of two blocks (two vertically placed). Or they're smaller so a Lemmings takes up more than that horizontally. With the former the position of each lemming is easier(?) to determine. Smaller blocks however may be easier to deal with the destructive skills and would look smoother.

Here's what chat GPT had to say about it (better than I could)

Pixel-Based Physics:
Pros:

    Precise Placement: Pixel-based physics allow for fine-grained control and precise placement of objects and characters. This can be useful for creating intricate levels with detailed designs.
    Smooth Movements: Since the physics are calculated at the pixel level, the movements of the characters can appear smoother and more fluid, resulting in a visually appealing experience.
    Detailed Interactions: Pixel-based physics can enable detailed interactions between objects and characters, allowing for more complex gameplay mechanics and puzzles.

Cons:

    Performance: Calculating physics at the pixel level can be computationally expensive, especially for large numbers of objects or complex interactions. This can impact the game's performance, particularly on lower-end devices.
    Limited Block-Based Mechanics: Pixel-based physics may not lend themselves well to block-based mechanics where objects interact at a higher level, such as pushing blocks or creating structures.

Block-Based Physics:
Pros:

    Simplified Interactions: Block-based physics simplifies the interactions between objects, allowing for easier implementation of mechanics like pushing or stacking blocks. This can lead to more accessible and intuitive gameplay.
    Performance Efficiency: Calculating physics at the block level can be more computationally efficient compared to pixel-based physics, particularly when dealing with larger levels or numerous interacting elements.

Cons:

    Reduced Precision: Block-based physics can lack the precision and fine control of pixel-based physics. This may limit the complexity and intricacy of level design and gameplay mechanics.
    Visual Limitations: Block-based physics can result in less visually detailed movements and interactions, as objects are constrained to block-sized increments.

Ultimately, the choice between pixel-based and block-based physics depends on the desired gameplay experience and the specific mechanics you want to implement in your Lemmings-style game. It's worth considering the trade-offs between precision, performance, and visual fidelity to determine which approach best suits your game's design goals.

@namida I'd love some more thoughts on this; or is there plenty of it in your threads on the matter (regarding L3D)? I haven't followed those closely until now.

--------------

Some other topics:
I agreed with Nester's view, when he was still around, that the game design and integrity should take precedence over "existing content" which can and should change if the need arises.

I decided to completely cut time limits and changeable RR/spawn interval from the game like Simon did with Lix (except maybe in a few edge cases but that remains to be seen).

carrying on from my first post:
The reason I don't believe in "true physics mode" is because I think the game should be designed such that it isn't necessary. Skill shadows are in a similar vein but a bit different but I'm not in a hurry to implement this; Lix doesn't have it after all and I never felt that I missed it terribly.

More generally; I honestly feel that both games (NL and Lix) have too much complication in their "ease-of-use features". This is just my personal preference yes. I think the game (level design in large part) should be simple enough that the magnitude of features we have in NL today aren't really necessary. Of course I want the game to lack the dullness and tedium that L1 had. But a lot of this can be mitigated by level design. For example you wouldn't rely so much on fast forward if your level simply didn't have a pointless long walk to the exit. My game will likely still have a fast forward, fyi. Although playing the game Temporal gave me the idea of having instead a "speed setting"

[WillLem]

Just to weigh in on a few thoughts here:

Trick-based levels. To add to what's already been said, I'd probably suggest that the best kind of trick-based levels are those which can be pulled off in real-time. Any that require you to pause, backstep, find the exact pixel from which to make the assignment, etc, are not particularly enjoyable to play and even less enjoyable to solve. With that said, the latter kind can justify themselves somewhat if the tricks are interwoven with more of a logic-based solution which might require some fiddling to put together but which looks great in a replay.

Builders turning lems. The Postcard From Lemmingland Fun repeat version provides a scenario where, especially if playing in real-time and for the first time, the chances of falling into the narrow pit and having to build your way out are fairly high, and therefore it's possible to find out from as early as Fun 7 that Builders will turn when they hit a wall. I understand your point regarding this seeming arbitrary, but - to offer a counter-point - it's one of the few ways that it's possible to turn a lemming in the early games, so from a purely gameplay point of view it's perhaps not as arbitrary as the mechanic itself might suggest.

Block vs Pixel-based physics. I've never enjoyed the completely block-based Lemmings games such as L3D and the mobile game; it feels too much like paint-by-numbers, and focuses far too heavily on the puzzle rather than the gameplay (which many will probably say is a point in favour of these games!)

Whilst I do also dislike too many overly-precise stop-start assignments generally, the fact that L1 and its various clones and sequels allow pixel-precision means that the game is more versatile and responds in a much more tactile way to player input. It also makes it much more satisfying to master the gameplay in real-time and be able to pull off the more challenging assignments on the fly. I was fairly dismayed when I first came to these forums after years of enjoying the game in this way only to find that most players find this side of Lemmings abhorrent and have aimed to do away with it. Naturally, I've been on a mission to bring it back and promote it ever since!

From what you've said, I imagine that Lemmings Revolution is probably closer to what you have in mind, since skill cancelling is no longer a thing here and the assignment window is generally quite forgiving. Lix also does a similarly good job of removing pixel precision from the destructive skills. Even NeoLemmix/SuperLemmix has skill queuing to mitigate the more brutal assignments (the ending of Save Me, and the opening of Steel Works when going for 100% spring to mind as examples here).

Maybe there's a sweet spot somewhere, where the game still feels tactile and responsive but doesn't make Icho or Simon rage-quit