When Solving Levels, What Skills Do You Generally Place First?

[kaywhyn]

This is something that just came up in my mind while I was writing my post in the United topic, and I thought it would make a great discussion. The title says it all: What skills do you generally place first when you're solving a level? As an extension, what skills do you find easiest to place? Hardest to place?

In my reply, I stated that I generally first place skills which are obvious in the solution. For the most part, if a level provides platformers and builders, I'm very likely to start with these two skills first. Most of the time, these two skills can obviously be placed right away because they're the easiest to determine where they go in a level, but if given the choice, sometimes it's not obvious which skill you should use if both work equally well for a particular gap. The terrain is usually what decides whether you should use a platformer or a builder, but sometimes you just have to use trial and error to determine which one to use. I guess the quantity of platformers/builders provided can also be a deciding factor, where the skill with the higher count is generally the correct one to use.

Other skills are much harder to place because it's not always obvious where they should go in the level. For me, I say stackers and blockers are generally the most difficult to determine. The former because it's generally not obvious if they should be used to hold back lemmings or can be used to ascend the level in short bursts, or even a combination of both. The latter because they can be assigned almost anywhere in a level, but where they're placed is often very crucial and so they can't always just be placed anywhere you like. Most importantly, since they can both hold back the crowd, this fact alone can make it very difficult to determine whether you should use a stacker or a blocker instead.

I guess cloners can be difficult too. Most of the time, they're used to clone a lemming performing a skill, but sometimes it's not always obvious which performing skill you should clone. Should you clone a miner? Glider? Not to mention you can just clone a walking lemming, but you almost always don't ever use a cloner for that purpose, although it's always available as an option when solving a level. Generally, if provided for a level, cloning a walking lemming is used just to meet the save requirement to replace any losses.

Even though the walker skill is quite powerful, perhaps walkers are also difficult. The most obvious use for walkers is to release blockers, but they can also be used to turn a lemming around or to interrupt a basher/miner midstroke. Most of the time, though, they're generally not used to turn a lemming around and is used more for the latter purpose. Because of the multiple uses, it's not always obvious for what purpose the walker should be used when a level provides them, especially in the case of releasing blockers when destructive skills are also provided to release a blocker.

Along a similar line, do you generally place the skills that have a greater skill count first, or the skills that have a lower skill count first? For example, if a level gives 2 bashers and 5 miners, do you generally place the bashers first because there's less of them, or do you start with the miners because there's more of them? In this case, does skill count take precedence over skills that are generally easiest to place? If bashers are easier to place because of the quantity, even though platformers/builders are generally the easiest to place IMO, do you start with the bashers instead or still prefer placing skills that generally are easiest to place first? Now that I think about it, I'm not exactly sure which I generally start with, the skills that have a lower quantity count, or the skills with a higher quantity count. I think for the most part I generally start with the skills that have a lower skill count, because they're usually easier to place. The ones with a higher quantity are usually skills that you're given way more than you need anyway, and so they're generally not problematic in determining placement in a level. Of course, they can still be surprisingly all be used up in the intended solution and thus aren't in excess. Then again, skills with a lower quantity count might be harder to place, and so maybe the skills with larger quantities are easier to place and one should start with them first. Of course, everyone has different methods of solving levels, and which skills are placed first depending on quantity is no exception.

Thoughts? Looking forward to hear other people's methods of which skills they generally place first. In particular, it should be interesting to hear from the level solving experts, but any input is appreciated from anyone.

[Proxima]

While this is an interesting question, I disagree with the premise I don't start with the assumption that every skill will be used -- firstly because I mostly play easier packs where that usually isn't true; secondly because designers often miss unintended solutions

I generally start by dividing the level into a series of obstacles -- either literal terrain obstacles, or tasks that must be undertaken during the solution, e.g. building a platform under a splat hatch. Then I consider what skills are needed for each obstacle. If there's a choice of different approaches for one particular obstacle, I try to look for the most efficient way, or the way that saves the skills I'm most likely to need elsewhere -- often builders, but if there's a shortage of destructive skills I will notice that and change my approach accordingly. If one of the tasks looks especially onerous, for instance getting the pickup skill on "Lemgate" from Lemminas, then I might start by asking whether it's actually necessary to do the task at all.

Then, if this approach hasn't led to a solution yet, I try to narrow things down by looking over the skills and seeing if there are any that absolutely must go in a certain place. This definitely includes builders, if there are gaps to cross or steel walls to get up, but depending on the level, it's often possible to place some of the destructive skills, and whatever skills are needed for isolating a worker lemming and freeing the crowd at the end. When looking around for which skills I can definitely place, I'm likely to start with the ones I have fewest of, since if I have only one miner (for example) and it must go in one place, that is much more useful information than if I have four bashers and I can definitely place one of them.

[WillLem]

This is as interesting question. It's the reason why LPs are so good to watch: seeing someone's approach to solving a level is fascinating, and you can learn a lot from how other people construct their solutions.

Personally, I tend to prefer a "let's just go for it and see what happens" approach, which works very well for certain levels but fails very quickly with others. I can use what I call the "picture puzzle method" (essentially pausing the game and solving the level in your head first), but this tends to take me a lot longer and require a more cerebral approach, which doesn't come naturally to me. I tend to treat Lemmings as a video game first and a puzzle game second: I generally want to get into the action as quickly as possible and start to see things take shape.

Easy or open-ended levels can be great fun to just play through and carve out your own solution. Difficult levels tend to have very specific actions or tricks which must be used as "keys" to "unlock" the level's intended solution: once discovered, the rest is usually fairly straightforward (extremely difficult levels will often feature several such "keys", with the solution being more layered and complex). Either can be equally satisfying, depending on the player's preference.

I'd like to see more LPs from people on the forum, even if it's just single levels.

[Armani]

I'm with Willem here : "Let just go for it. Go with the flow." Sometimes this approach doesn't work then I start to place given skills one by one. First I place 1)constructive skills.(builder, platformer) Often but not always, it is quite obvious to see where I need to place given builders and platformers. – to cross the gap or connect the destructive skills. And after that, I think of 2)permanent skills(climber, floater, glider, swimmer), 3)movement skills(shimmier, jumper and sometimes walker) and 4)destructive skills(basher, miner, digger, fencer) to send some pioneer lemmings where the constructive skills are needed. Finally, I placed all the remaining skills.

But in most of the cases, I prefer "just go for it" approach since I feel like there are so many levels that this "placing skills one by one" approach is not the best idea.

[kaywhyn]

While this is an interesting question, I disagree with the premise I don't start with the assumption that every skill will be used -- firstly because I mostly play easier packs where that usually isn't true; secondly because designers often miss unintended solutions

I think you read a bit too much into the title, as nowhere does it imply all skills need to be used, I'm just asking what skills do level solvers here generally place first. However, now that you mentioned it, perhaps a better title is needed, maybe something like "what skills do you place first on HARD levels?" or maybe "on hard levels with only a single solution?" Even then, the question can still apply for easy levels, although it's not as meaningful since pretty much almost every skill can be very obviously placed or don't even need to be obvious because there's usually too much degrees of freedom due to excessive amounts of skills.

Oh, I thought you were one of those extremely good puzzle solvers as well. Even then, I take that you still managed to complete all the packs I've listed that I completed in my topic on the Lemmings Main board?

Personally, I tend to prefer a "let's just go for it and see what happens" approach

I'm with Willem here : "Let just go for it. Go with the flow." Sometimes this approach doesn't work then I start to place given skills one by one. First I place 1)constructive skills.(builder, platformer) Often but not always, it is quite obvious to see where I need to place given builders and platformers. – to cross the gap or connect the destructive skills. And after that, I think of 2)permanent skills(climber, floater, glider, swimmer), 3)movement skills(shimmier, jumper and sometimes walker) and 4)destructive skills(basher, miner, digger, fencer) to send some pioneer lemmings where the constructive skills are needed. Finally, I placed all the remaining skills.

But in most of the cases, I prefer "just go for it" approach since I feel like there are so many levels that this "placing skills one by one" approach is not the best idea.

I think like most others here I tend to solve the levels in my mind first. However, I can still advocate for both WillLem's and Armani's preferred approach of "let's go for it and see what happens." Especially on hard levels, it's likely I will end up confusing myself when I'm trying to think and reason through a solution in my mind or I'm not sure exactly what will happen when the solution is actually carried out, and so I also do the "let's go for it and see what happens" approach. In this case, I generally say that I'm "trying out things" just to make sure that what I had in my mind would or wouldn't work. If the former, then I can proceed with it. If the latter, then at least I don't need to waste any more time with an approach in my mind that's not going to work. I'm a visual learner, and it's usually much easier for me to solve levels when I see the actions carried out. Also, the biggest drawback to the "solving in your mind" approach is the disappointment of seeing what you thought was a solution in your mind turns out to not work at all when you do actually carry it out. Nevertheless, I think it should be clear that what I'm saying here is I myself usually use a combination of solving in my mind and "let's go for it and see what happens."

@Armani Interesting order. So you also generally start with platformers/builders like me. For the constructive skills, you forgot the stacker, and I guess stoners can be included too, since they do create terrain but it also kills a lemming and so can be destructive too? Only thing is that they don't destroy terrain, which is what the destructive skills do. Out of platformers, builders, and stackers, stackers are generally the most difficult to place, again because it's not always obvious for what purpose(s) in the level it needs to be used for. Hold back the crowd? Ascend the level in short bursts? Combination of both?

For permanents, I think disarmer also goes there? In the sense that a lemming stays a disarmer for the rest of the level until it either splats, drowns/burns, exits, or explodes/gets nuked. Also, preassigned hatches can only have permanents, and I remember several levels that have preassigned disarmers.

Finally, for destructive, you forgot bomber. Then again, it's similar to the stoner in that it also kills the lemming, so maybe stoners and bombers are in their own category? Looks like I'm going to have to go back and read the discussion on if they're considered in their own category or included in the main categories. In any case, I think we all agree that bombers and stoners are generally difficult to place. Even if 1 of either is provided, it can be very difficult to know exactly where they're needed, and even more so when multiple stoners/bombers are provided. So, I guess most place these last, and in Armani's case both stoners/bombers would be in his "remaining skills" part of his reply.

[Proxima]

I think you read a bit too much into the title, as nowhere does it imply all skills need to be used, I'm just asking what skills do level solvers here generally place first. However, now that you mentioned it, perhaps a better title is needed, maybe something like "what skills do you place first on HARD levels?" or maybe "on hard levels with only a single solution?"

No, my point is just that starting by trying to place skills makes less sense to me because any skill you look at might not be needed at all. I'm not dogmatic about it, though -- for instance, if there is a really obvious placement, such as a gap or a splat hatch that absolutely requires a builder, that might well be the first thing I notice on a level and then when I'm thinking about how to deal with the level's obstacles, I won't consider solutions that would use up all my builders elsewhere.

Oh, I thought you were one of those extremely good puzzle solvers as well. Even then, I take that you still managed to complete all the packs I've listed that I completed in my topic on the Lemmings Main board?

You're joking, right? You can see all the usermade packs I've completed in my review topic. (I have also completed all levels of the original games.)

To be honest, I'm not sure anyone else has beaten all those packs (but hey, it would be interesting to compare notes). Not only would that be a huge time investment, even for players like Icho who are certainly capable of beating them all, but you have listed packs for DOS, Lemmix, NeoLemmix, Lemmini and SuperLemmini, and I think most of the forum regulars play NeoLemmix either primarily or exclusively.

I would certainly like to beat all new-formats packs eventually -- one reason I started the review topic was to help motivate myself to keep making progress -- but I have no interest in content that's exclusive to the older engines. For one thing, I never liked the gimmicks, radiation or slowfreeze and I'm glad NL has left them in the past. For another, NL has a ton of convenience features that I'd miss when playing the older engines. If I ever have time, then I'm happy to help with converting older content so it can be played in NL, but time is always very hard to find.

[kaywhyn]

No, my point is just that starting by trying to place skills makes less sense to me because any skill you look at might not be needed at all. I'm not dogmatic about it, though -- for instance, if there is a really obvious placement, such as a gap or a splat hatch that absolutely requires a builder, that might well be the first thing I notice on a level and then when I'm thinking about how to deal with the level's obstacles, I won't consider solutions that would use up all my builders elsewhere.

I agree with your first sentence, which is especially true for easy, "anyway you want levels," where placing skills being a priority doesn't matter as much since you can just choose what to use most of the time due to these levels often having many degrees of freedom. However, perhaps if I had been more specific in the title by asking for hard levels, then I think this is where you and I see things differently, where I did mention that I try placing obvious skills first, just because the less skills I'm dealing with in my mind the better. For many things, I have a very specific way of doing things, so really this is just MY specific way of doing things when it comes to solving Lemmings levels. Yes, sometimes there may still be skills that you end up not using on such levels, but I generally don't count on that happening. Most of the time, if that happens, I kind of unintentionally did so because unless I'm super stumped I generally consciously don't try to backroute these levels. Even then, as I have recently seen in United, for example, I never knew I was quite capable of backrouting so many levels, but Icho did say that the last two main ranks haven't been tested as much and hence the "backroute potential is higher."

You're joking, right? You can see all the usermade packs I've completed in my review topic. (I have also completed all levels of the original games.)

Not at all. I honestly thought that you must be one of the much better solvers, just from the fact that you did mention in the United topic that you did beat the Pacifism rank. I thought, oh hey, it sounds like Proxima is capable of taking on and handling hard packs. However, it's only the first rank, and there's numerous difficult levels beyond it, so I'm being somewhat hasty in my conclusion. Now that you linked your own review topic, I totally forgot about that.

To be honest, I'm not sure anyone else has beaten all those packs (but hey, it would be interesting to compare notes). Not only would that be a huge time investment, even for players like Icho who are certainly capable of beating them all, but you have listed packs for DOS, Lemmix, NeoLemmix, Lemmini and SuperLemmini, and I think most of the forum regulars play NeoLemmix either primarily or exclusively.

That's easy: After playing packs exclusively on Custlemm for years, I discovered and found out about Lemmini from rtw's introduction video of his playthrough of Dovelems. That's why I started off with Lemmini. The feature I was most drawn to in Lemmini, and one that I think is a huge welcome, was when rtw mentioned that Lemmini has directional select. Once I finished Reunion, I then found out about Superlemmini here on the forums, and so then I started playing packs made for it. Because I prefer the classic way of playing Lemmings, i.e, without features like framestepping, infinite time, timed bombers, etc, I thought I wouldn't ever transition to playing on Neolemmix. Indeed, I stuck with Lemmini/Superlemmini for at least 4-5 years before I finally decided to give NL a try around this time last year. I still prefer playing the game the classic way, but now that I've been spending so much time on NL I'm liking it to the point where I'm especially thankful for the timeskip feature. 

...time is always very hard to find.

Yes, absolutely! As a matter of fact, I wouldn't even be this far in United that I am right now if I was still working a job the last several months, but I've been out of one ever since the pandemic began in March, hence why I have all this free time that I didn't have when I was working.

[IchoTolot]

If the level is big and has many skills then the "just go for it" method can be VERY successful when you try to be as effective as possible. If you come slightly short then often minor adjustments do the trick. --> That's how I solved levels like "Final Frustraition".

Usually though I make plans and go for patterns I can make out. First, I try to place skills that HAVE to be at a cetain place (no order or distinction between skill types here) and make out important areas of the map - Taking a good look at the map can sometimes solve the level already for you. Then look where the learned tricks in your arsenal could be really helpful - which tricks to consider depends on the level.

Also, sometimes I try to split the level into pieces, process them seperately and see how everything could fit together. Getting the connection right could be the secret.

I often don't solve everything in the head first, but there are still instances where I do it usually when I immidiatly can put together the solution. A mixture of trying and thinking works best for me. Things should be tried out and not just assumed as you might miss something here. Also, try to be open and not just assume something have to be a certain way. If it doesn't work try another method while actively avoiding your first setup.

A lot of this comes down to experience and general intuition though, while making every action as effective as possible. I can't really give an exact formular to this, but simply solve and solve and solve and soon, if you remember the things you learned, you can see the patterns and things will automatically come together in your head.
In order to get this intuition/automatism you will need to bash your head against many brick walls again and again. Only if you don't shy away from the brainmelters and don't mind if you get stuck for ages you can get this type of intuition.

[kaywhyn]

Thanks for the reply, Icho! It's great hearing thoughts from a puzzle creator/solving great like yourself.

If the level is big and has many skills then the "just go for it" method can be VERY successful when you try to be as effective as possible. If you come slightly short then often minor adjustments do the trick. --> That's how I solved levels like "Final Frustraition".

It's going to be a while before I even get to the final Black Hole level, as I'm currently stuck on a Nepsterlem level in the Sun rank that both you and Namida were stumped on, and so no you two weren't the only ones with a brain freeze on it. I absolutely agree with you that this level that has also managed to stump me in Nepsterlems is way too hard for its position.

First, I try to place skills that HAVE to be at a cetain place (no order or distinction between skill types here) and make out important areas of the map - Taking a good look at the map can sometimes solve the level already for you. Then look where the learned tricks in your arsenal could be really helpful - which tricks to consider depends on the level.

Then we're in agreement here, as I too pretty much start with exactly that: Place skills that I know for sure have to go at a certain spot. It helps me a lot just because the less skills I have that I need to think about, the better. Of course, this applies more for hard levels than easy ones.

I often don't solve everything in the head first, but there are still instances where I do it usually when I immidiatly can put together the solution. A mixture of trying and thinking works best for me. Things should be tried out and not just assumed as you might miss something here. Also, try to be open and not just assume something have to be a certain way. If it doesn't work try another method while actively avoiding your first setup.

Even though I mostly try to solve in my head first, I'm also in agreement here with the hybrid trying and thinking things approach. I usually start using the "try things" approach once I start getting confused thinking through the solving process, but trying things is especially useful because I'm a visual learner. And yes, I did mention the biggest drawback of solving in your head first is the disappointment that can result when you actually try the solution and what you thought was going to work in your mind is not going to work at all. Sometimes, it's too difficult to tell whether something will work without trying things. Good example is big platformer/builder gaps, where it's often impossible to know exactly how many you need when you're solving. Sure, you can estimate, but the only way to accurately pinpoint an exact number is to try it out.

A lot of this comes down to experience and general intuition though, while making every action as effective as possible. I can't really give an exact formular to this, but simply solve and solve and solve and soon, if you remember the things you learned, you can see the patterns and things will automatically come together in your head.
In order to get this intuition/automatism you will need to bash your head against many brick walls again and again. Only if you don't shy away from the brainmelters and don't mind if you get stuck for ages you can get this type of intuition.

Absolutely agree with all of this! I also know that you mentioned that on hard packs that people generally give up way too early, and I agree with that. This is definitely why I'm able to get far in hard level packs: I generally have a much higher level of patience and perseverance than most. I also like taking my time, and so I'm willing to be stuck, and stay stuck, on a level for however long it takes. For example, it's been more than 3 years and I'm still stuck on Bonus 25 of RotL on Lemmini! I'm aware that the levels in that rank use glitches, and I've read through the Lemmini glitches topic so many times to see what can help me solve it, but I still cannot figure out which one needs to be used to solve the level. It's possible that it might not be listed either. It's very easy to save 1 lemming, but to save both I still have no clue. As much as I hate to, this might be one of those levels where I'm just about to throw in the towel completely, but it's likely that I'm just missing something extremely obvious. I'm well aware there are video solutions of the level on Youtube, but I dare myself to not watch it, as I don't watch solutions until I solve the level myself first. Nevertheless, I'm just going to keep trying despite it already taking this long, and so no spoilers, anyone!

[namida]

Often, what approach I take may depend on the mood I'm in at the time. But overall - while I might look for obvious things (eg. "there's a gap with no way around it, I need a builder / platformer here"), generally, I'm just going to start doing things and refine things until something works. If it really feels like it isn't going anywhere, I might throw the whole thing out and start again, making sure to do things a bit different this time.