Level Design: What Makes A Level Challenging?

[Nessy]

So I have decided to try my hand at making my very first topic .

What makes a level challenging?

A quick note: I am a firm believer that you don’t have to create a hard level in order for that level to be a good level. Still, I was curious to see the science behind making a level a worthy "speedbump".

One thing I noticed in some levels considered to be challenging is that a terrain is presented to you and you have very limited skills to use in it. I really like these levels because it forces you to really think about the skills you have, how to use them, where to use them, and sometimes how to combine them. In some areas it may look like you have to use a miner to clear a path, but what if another skill would work better so the miner can be used somewhere else? What if instead of bashing you can maybe dig up the path with a digger and use the basher for other parts of the terrain?

For example, in Postcard From Lemmingland you only get one builder, which means that you can’t easily build over the small wall and pit in the level. You have to focus on what you have instead of what you don’t have, so examining how the terrain looks and using what you have to craft a solution through it is where the challenge comes in. You can bash to the pit from the entrance, so how does that help? (In fact, you can beat this level without using the basher). A miner digs down diagonally, so how does that help? Lemmings turn around when they hit a wall even while building, so how does that help? To successfully beat the level all of these things have to be considered. Another example is The Great Lemming Caper. You only had two builders and builders are the only way to reach the exit, but they were not enough for both lemmings in the level. But what else did you have? You did have a lot of diggers, bashers, and miners. How can the digger, miner, basher, and even blocker skills help out so I can use those two builders wisely? Of course, that’s why No Added Colours Or Lemmings is considered by many people to be one of the best levels in the game. The game never told you that a lemming was automatically doomed if they were turned into a blocker, so using all of your skills wisely to reach the exit and save every lemming was the challenge.

Finally, I also think this is why Save Me is more challenging than Rendezvous At The Mountain. For Rendezvous At The Mountain you had a good amount of each skill to craft a solution. Save Me did not have a lot of different skills, and even the builders run out quickly. You also had the challenge of being forced to send more than one lemming to do work on different parts of the level, so that already puts a challenge in making sure you use every skill in a way as to effectively control the crowd enough to prevent too many of them from getting killed along the way.

Anyway, that’s my two cents on the subject manner. I might even be wrong.

[IchoTolot]

I think "What makes a level challenging?" on the pure puzzle side comes down to 2 things:

1) Entropy
2) Trickery


1) "In statistical mechanics, entropy (usual symbol S) is related to the number of microscopic configurations Ω that a thermodynamic system can have when in a state as specified by some macroscopic variables."
Now my definition of "Entropy" for Lemming levels :
Entropy can be described as the pure ammount of possibilities (chaos) you have in a level. Of course a 20 of every skill level is usually not difficult, but I am rather talking about levels with a barely enough skillset that have a high ammount of possible (not actual) ways to solve them.
For example a level with limited skills that has great hints where you HAVE to put the skills has not a high Entropy. Like for example you have a water gap (no swimmers) and only 1 builder --> the builder HAS to be here --> lower Entropy.
If every skill of the level has multiple places it can be used in a manner that gets you really closer to your goal, that's the type of level that is really difficult. (If you have elimminated all backroutes and only the intended way is left of course )

So: Higher Entropy ---> Higher Difficulty      (example would maybe be the 2nd last level of Reunion "Lemtris Version X" or the last levels of NepsterLems with a high entropy and an easy 1-of-everything level with clear usage places for every skill with a low entropy)

2) Of course difficult tricks increse the difficulty of the level, but there is another thing to be aware of here: If a person knows the trick a lot of the difficulty can get lost. Therefore concealing the trick should also be a goal. Sidenote: A level where the place for the trick is clear and maybe even be hinted at would be a good introduction level for it.
Let's take the last level of the first rank of Reunion for example:
In it's first iteration the starting area did not have it's metal plates --> the blocker or a digging skill was the obvious choice for containment.
What I wanted the player to use was a builder block: 3 builders after each other to make a 1-way wall and create a delay for the 1st lemming. This was considered to be too wasteful for most of the players --> good concealment of an easy trick.
I added the metal plates to lower the entropy (to hint at rule 1) and make the trick more clear --> blocker + digging tools were no longer an option and the level was now easier.
As you can see here is that point 1) Entropy can also be a great tool to hide your tricks . The other way would be in the terrain shape. "The Glacier Station" from the 3rd rank of Reunion would be a good example for hiding sth in the terrain without unfair methods.
Be careful with concealment: Don't actively hide things from the player. Hide it in plain sight! It's always visible, but not clear to the player. A bit like a "Where's Waldo" picture, just now you are searching for a solution or a place to use a trick in a helpful manner.
So the trick complexity (this also includes combination of tricks and maybe even multitasking) and trick concealment is a key to more difficulty.


I hope this helps and I got the main points to create difficulty. I would like to maybe hear other peoples definitions. Maybe I missed a point.

[Nessy]

I like your entropy analysis. I was attempting to explain that general concept a little bit but you articulated it much better.

I forgot about trickery and it does make sense, especially when it isn’t obvious but at the same time hidden in plain sight. That reminds me of I Have A Cunning Plan (sorry, haven’t played a lot of packs yet so I can’t give a lot of examples with variety) where a player would logically delay the lemmings with builders to allow the first one to bash through the wall, but at the same time many new players would still have some lemmings slip by, turn around, but then turn around again by hitting the wall created by the builders. The level can now be done more cleanly by using three or four (I think it differs depending on the version) builders as a "blocker". I suppose another example would be Compression Method I, where it was impossible to beat the level without going through the traps at the bottom, so the player had to figure out how to go through it without losing too many lemmings along the way (although on some ports you could backroute it by glitching through the steel on top ).

By the way, for some reason I really liked that trick with creating a wall using builders and tried to create a level introducing it. Similar to your Reunion example, it wasn’t obvious but it did slowly lead you to that conclusion when you consider the nature of the terrain (using steel blocks to avoid containment through destructive skills) and fact that you have much more builders than needed for the level. I do plan on making another level where the same trick is involved but it builds on it by introducing another trick using it.

[grams88]

Hi Nessy

Welcome to the forum Nessy, like your avatar.

I'm a bit behind with a lot of the new things such as the Neolemmix and Lix levels which I notice a lot people talking about. I might get into it soon in the future. My level ideas might be a bit outdated as I've not really tried many of the new levels but I'll give it a go.

One of my things I tend to like is the whole idea that a level can have a lot of different solutions but I do like the more linear type levels where there is usually just the one solution but it is hard to complete.

One idea well I would probably only use this idea once if you were to create a level pack. The idea is the old hunt for the exit where you have a lot of different skills but are unsure where the exit is. I find those levels kind of fun where you are digging all over the place trying to find the exit. Mind you I think this could be a very unpopular idea if it is overly used. (Maybe once per level pack)

What about the idea where you are having to reach the exit, I think I tried this idea with a few of my old lemmini levels where the exit was up very high and you don't have many skills to get up to the top.

Oh that great lemming caper level was a very good level in that there are a few different ways to complete it. I played that level recently and one of the solutions I tried was to block one of the lemmings once he digs for a little bit, I managed to free him up and complete the level.

[IchoTolot]

One idea well I would probably only use this idea once if you were to create a level pack. The idea is the old hunt for the exit where you have a lot of different skills but are unsure where the exit is. I find those levels kind of fun where you are digging all over the place trying to find the exit. Mind you I think this could be a very unpopular idea if it is overly used. (Maybe once per level pack)

Intervention!

Those kind of levels are one of the worst types! In the old engines you basically have to search blindly and without precise frame-stepping tools in the dark --> pure trial and error. Searching randomly for the exit is neither fun or a fair puzzle! Everything has to be visible to the player! This is one of the most holy rules of level design. No hidden exits, no hidden trpas, nothing is hidden! The levels difficulty has to come purely from the puzzle factor of the solution. Until now every new engine focused on making the execution side easier and focus purely on the puzzle itself. Also in NeoLemmix a quick press on the clear physics button would bring up the trigger area anyway making the search pointless.

To sum it up: Levels with hidden stuff are generally not well received and should be avoided.

One of my things I tend to like is the whole idea that a level can have a lot of different solutions but I do like the more linear type levels where there is usually just the one solution but it is hard to complete.

What about the idea where you are having to reach the exit, I think I tried this idea with a few of my old lemmini levels where the exit was up very high and you don't have many skills to get up to the top.

Well this is very general      You always have to reach the exit going in some direction and levels often have multiple solution ... although more difficulty ones tend to have only one with maybe only slight variations.



After a definition of difficulty in general maybe these level creation methods can be helpful:

- Focus and plan on one central trick. Choose the tileset and terrain layout on a trick you want to force in the level. Aditional puzzle elements can be build around it to create more complexity, but beware of backroutes!

- Build the terrain layout first and after the landscape is finished carve out a challenging solution.


Combining the 2 methods and make a drawing on paper first can also be helpful. Don't forget to add some nice decoration and maybe even a sweet music track!

[Colorful Arty]

I would like to respectfully disagree with your take on hidden exits and traps. I personally (that's the key word) think that hidden exit levels can be fun and hidden traps are fair game provided they are used in good places (i.e. not in solid terrain and not at the very end in front of the exit). Hidden exit levels are a different kind of lemmings level, one that focuses a lot more on exploration than puzzle-solving. I know Lemmings is a puzzle game, but these unique levels are welcome provided you have a good amount of skills to mess around with.

Also, if you use my underwater tileset with the giant updraft, you will not be able to use clear physics to instantly find the exit.

[ccexplore]

It's fair to say that hidden exits and traps are controversial and polarizing.  Some may love it and others will really hate it.  I also think exploration of hidden stuff works better in a non-Lemmings settings, such as most traditional games where you have direct control over a character's movements.  In the long run, you are unlikely to be able to sustain a pack of many levels in a widely appreciated way merely by hidden stuff alone, so you're still going to need to find other ways to make things challenging and interesting.

[IchoTolot]

I would like to respectfully disagree with your take on hidden exits and traps. I personally (that's the key word) think that hidden exit levels can be fun and hidden traps are fair game provided they are used in good places (i.e. not in solid terrain and not at the very end in front of the exit). Hidden exit levels are a different kind of lemmings level, one that focuses a lot more on exploration than puzzle-solving. I know Lemmings is a puzzle game, but these unique levels are welcome provided you have a good amount of skills to mess around with.

Also, if you use my underwater tileset with the giant updraft, you will not be able to use clear physics to instantly find the exit.

What kind of "fair game" is a stomper coming out of nowhere?  Even when it's at the start. I respect your personal opinion that they can be fun for you, although I don't understand where the fun comes from when you are getting screwed out of nowhere, but this is not a fair game by any means. This is never a fair game, you directly lie to the player about the path being safe.
It's like if you have a race and someone just trips you in the middle of it.

I you want to explore I agree wth ccexplore --> play a different game. + regarding underwater: I will find the exit, as overlappng triggers are brighter     True physics will always be there to shift the focus to the puzzle part, as it should be!

[Colorful Arty]

Hidden traps can complement the scenery of a level far better than visible ones sometimes. For example: the retractable boulder trap in the dirt tileset is designed as a hidden trap, hence why it looks identical to one of the terrain pieces. Making this trap visible looks horribly ugly to the terrain. Oftentimes traps are better set as invisible traps provided it makes the level looks less ugly or cluttered. For example: Compression Method I would look far less visually appealing if the mashers were hidden, and that level is regarded as one of the most recognizable and good levels from the game judging by how many homages to that level exist.

[IchoTolot]

Functionality and fairness > looks and decoration.

Decoration and fairness are both key to a good level. If your level needs hidden stuff to look good, it shall be thought over again! I consider the retractable boulder trap in the dirt tileset a misdesign as you are right that it can look stupid if not partly hidden. The ten ton block or the bear trap are better traps in general for usage. They are better designed.
I also critizise Compression Method I for making the mashers hidden. Most of the original levels have design flaws. May it be hidden stuff, misplaced levels, or unnessesary timers. If I would re-release them I would change a lot of things in there     I also think we make and have waaaay better levels than the original ones now.

Functionality and fairness > looks and decoration. That's also the reason I changed my mind and therefore I will change quite a bunch of decorational objects with the format change to be either terrain or have a visible animation and also support the option to blend them out completely.

[grams88]

Interesting debate, I think I'm on Colourful Arty's side but I can understand what Ichotolot is getting at. I agree that a lot of the new level packs beat the original ones as Ichotolot was saying.

I like that level the (X marks the spot), I remember having some sort of idea to the hidden flamethrower trap as you could see a part of the flame coming out. Maybe it might be classified as half hidden.

I loved that boulder trap (Rendezvous at the Mountain), I think it was seeing the blood of the lemmings when they get crushed. I sort of think of it as a decoration element to the level. Nice last level

I'm probably of the opinion where maybe one hidden exit level might be good in a level pack but if it is something that is over used it can be maybe off putting. I like the search around types like in the Martin Zurlinden's level pack, (Episode x the phantom exit) I really enjoyed that one.

I know a lot of people hate the walking bombers as well but I think the walking bomber levels are good in that you will be unsure of whereabouts is it going to explode. Sort of like the bombozal level as it felt unique back in the day. Maybe one walking bomb level might be an nice one as long as we don't overuse it as well. I guess the walking bombing thing made (A TOWERING PROBLEM) level hard.

I will admit (A TOWERING PROBLEM) level from oh no was never one of my favourites, I think it might of been one of my least favourite but the (Bomboozal) level is very very cool.

I think this one will be unpopular but quite a lot of the original levels might of used the old hero type levels where one lemmings does most of the work. I like an occasional hero level such the (What an awesome level) from the original one. THe four special levels from the original could be classified as a hero type one.  Can't quite remember what we call hero levels, I'm sure it has another name in the lemmings community.

That Rick dangerous game was very bad for hidden traps .

[Proxima]

Those kind of levels are one of the worst types! In the old engines you basically have to search blindly and without precise frame-stepping tools in the dark --> pure trial and error. Searching randomly for the exit is neither fun or a fair puzzle! Everything has to be visible to the player! This is one of the most holy rules of level design. No hidden exits, no hidden traps, nothing is hidden! The levels difficulty has to come purely from the puzzle factor of the solution.

Original Lemmings isn't just a different game from NeoLemmix; it's a different genre. Judging it by the standards of a pure puzzle game like NeoLemmix is like faulting Star Wars for making it obvious who the villain is, when it isn't trying to be a mystery.

Original Lemmings, by the standards of its genre, is fair. The early levels cue the player as to what to expect on the later ones. Levels like Fun 17 tell the player that traps can be hidden, and they do so in a harmless way -- when you lose a lemming to the trap, you have plenty of tools to avoid the other lemmings following, and the level requires saving less than 50%. (And, of course, that particular level tells you where the traps are on Compression Method 1, which is its repeat.)

That said... the reason this community has stuck together and is still interested in the games all these years later is that, while there are many other games that scratch the same kind of itch original Lemmings does, there are far fewer other places for us to go when we want the same puzzle-solving stimulation we get from NeoLemmix and Lix. So naturally, today's community has a strong preference for levels that are more puzzle-oriented, and I completely agree that hidden traps are best avoided in NeoLemmix levels.

[Nessy]

Seems like I caused a hot debate here

Another thing that interests me are the levels that appear on the second-to-last difficulty (like in Taxing or Wicked). It seems like those levels are the hardest to design for because a lot of times levels considered to be misplaced come from there. They seem to be in a strange spot on the difficulty curve. Sometimes they seem like they belong in an easier rank and sometimes they seem like they belong in a later rank.

Usually the first rank is the easiest and sometimes (not always) teaches you the skills and hazards you will encounter (like the first few levels of Fun). The second difficulty is when some more complex puzzles are introduced that require more thought and planning than the first rank, but it isn't overwhelmingly hard. The last rank is usually the one that really pushes you to the limit in terms of high entropy and tricks. Finally, that second-to-last rank… what makes those levels that belong in a second-to-last rank and not the last rank? Would it be a rank that starts teaching the player some of the more complex tricks? Would it be a rank that is just more difficult than its previous rank and easier than its last rank? The latter might be more true and I'm just overthinking it.

[Nepster]

If every skill of the level has multiple places it can be used in a manner that gets you really closer to your goal, that's the type of level that is really difficult.

More than any other argument or explanation of the discussion above, I feel that this sentence sums up perfectly my take on the title question.

Another thing that interests me are the levels that appear on the second-to-last difficulty (like in Taxing or Wicked). It seems like those levels are the hardest to design for [...] Finally, that second-to-last rank… what makes those levels that belong in a second-to-last rank and not the last rank? [...]

Your descripütion of the ranks is pretty much spot-on for the original lemmings, and holds at least to some degree for ONML, too. But I am not so sure, that there even exists some defining characteristic of second-to-last rank levels.
My own approach to level pack creation is very pragmatic: Make lots of levels without worrying about their difficulty, then sort them into ranks roughly according to their difficulty and finally rearrange them according to player feedback. I know that at least some others here use basically the same approach, so we should ask ourselves: Did the creators of L1 and ONML do basically the same thing, and just didn't do enough rearrangements at the end? The very chaotic order of the levels in which they were stored in the DOS version of L1 seems to suggest exactly that. So looking for defining characteristics of Taxing/Wicked levels might not yield good results.

[grams88]

If every skill of the level has multiple places it can be used in a manner that gets you really closer to your goal, that's the type of level that is really difficult.

I agree with that statement.

Lets see if I did the quote thing right.