Revenge of the Lemmings - Level selection and ordering for v5

[WillLem]

10-of-everything levels are not just for beginners. They make good warm-up levels to get the cognitive synapses firing ready for more challenging fare.

Above all, though, the most important thing a pack should have: variety.

So:

Sure, levels with lenient skill sets are beginner friendly as they allow players to explore mechanics. But similarly, levels can have a small set of skills (in particular few different skills) and only use straight-forward mechanics and thus be beginner friendly and yet not easy ... At least these could be interleaved between the 10-of-each levels.

Yes, agreed - mix it up, give people a bit of everything.

[Proxima]

The poll is still open (for another six days, in fact), but in a community this size, 8 votes means we can already draw some conclusions based on current standings. It's pretty evenly split between yes and no, but slightly favouring yes. A strong majority are not bothered by "ghost repeats". Also, I want to take into account that a recurring theme in previous feedback (e.g. Pieuw's spreadsheet and Flopsy's video) has been that previous versions of the pack get tough too quickly.

It's true that the pack is unlikely to be played by anyone as literally their first experience of Lemmings (which is why I didn't bother asking for opinions on skill tutorial levels; we are simply not bringing those back); but it should be approachable for players at all skill levels.

So, my inclination is to keep 10-of-all repeats but reduce their number, and in particular, I think we should add some new ones to help out the very beginning of the pack, while dropping some of the less interesting existing ones. I'll come up with more concrete suggestions for this around the time the poll closes.

geoo, given your stance, we can cut "Canopy Lemmings" (the 10-of-each repeat of "Stroke at Retirement Age") if you prefer, though I would like to keep that one as it's certainly one of the more interesting 10-of-each levels.

[JawaJuice]

10-of-everything levels are not just for beginners. They make good warm-up levels to get the cognitive synapses firing ready for more challenging fare.

Completely agree with this. I consider myself a fairly experienced player at this point, but I still enjoy X-of-everything levels. They can certainly be made more interesting with talismans (the recent Lemmings Faithful pack is a good example of this), but you can also set yourself challenges like always saving 100% wherever possible. It's just nice sometimes imo to not have an extremely limited skillset and a single intended solution that you have to find, as you can get creative with your solution and explore the open-endedness that affords.

[Proxima]

On another note, I've spoken with Will and he's agreed to add the L2 Circus style to RetroLemmini. (It's used by three levels in RotL that we are probably keeping: Cry for me, Another Day at the Office, Getting a Cat Down from a Tree.) There are also some levels using the L2 Egyptian and Sports styles, which are already supported by RL.

As well as those three, RL supports the L2 Shadow style, and Will is considering adding Beach, but at the moment there are no plans to add all twelve. So, for the third pack of the RotL v5 project (the pack with new levels), allowed styles will be all the original and ONML ones, and the five L2 styles just mentioned.

[WillLem]

I think we should add some new ones to help out the very beginning of the pack, while dropping some of the less interesting existing ones.

You're definitely thinking along the right lines as far as pack approachability is concerned; your post above has my full support.

Two additional suggestions that might be worth considering:

1) When deciding which levels to feature as "X-of-each", look to the general layout of the level. The more detailed and featured it is, the more likely an X-of-everything version will work well in its own right. I'm sure you're already aware of this, but just thought it was worth mentioning in-topic. Also, I'm happy to post a list of possible candidates when the time comes (i.e. if you're taking suggestions on specific levels).

2) Consider that not all X-of-everything needs to be 10. For instance, 5, 3 and even 1 of everything can make for great quality accessible levels. I'd suggest adjusting X to suit the level layout being used. Again, I'm sure you're already aware of this anyways

[kaywhyn]

Maybe people are misunderstanding what I mean by "First Draft".

My suggestion is simply that we set up a framework on which to build. Levels can be swapped out, moved around, and we get a clear sense of cohesion and direction right from the very start. I'm not suggesting that the first draft be viewed as something that could potentially be released, which is what people seem to think. On the contrary, the first draft should really be nothing more than a list of placeholders. I mean, it could literally say "to be decided" for every entry at this stage!

Right, it's a first draft. A rough draft, so to speak, not a "final" draft. That's not the issue. Rather, the problem was with you suggesting it be done quickly, which Proxima and I think isn't the approach to take with this. There's a lot of levels to scrutinize here, and that will take a lot of time in general. Some have already put a bit of a list together and posted them, and I intend to do the same by providing mine. I know I said I would post mine at least a weekend ago or so but I haven't gotten around to doing so yet. I'll hopefully have it by this weekend. It's been a really hectic last several days/weeks for me. I'm intentionally ignoring looking at the spreadsheet for now in order to avoid any bias when I come up with my list. Once I have something, then I'll take a look to see where there's agreement on what levels should and shouldn't make the cut for the pack(s).

Of course, this taking time extends to all steps of the project, not just the first draft part of it.

With that said, the spreadsheet is providing real-time progress updates, which is great. And, Proxima's mentioned that he intends to draft the first rank fairly soon anyway. At that point, things will start to take shape and we'll have a clearer idea of what we're dealing with.

Yes, having a great first rank is important for a community pack such as this.

At least personally, I rather do things right the first time around

Impossible. That's why versioning is a thing.

Not unless you're me who tends to be a perfectionist kind of person Semi-joking aside, I do consider myself one, even though I know I'm not perfect. Regarding getting things right the first time, it is possible, but it's extremely difficult, to the point that it might as well be considered impossible anyway, especially with multiple steps here.

To put it in a better way and that I should had said before, I rather do things slowly, carefully, and properly. Just the way I work and which works well for me. Everyone is free to work at their own pace, of course. In the same way everyone has their own ways which work well for them

[WillLem]

the problem was with you suggesting it be done quickly, which Proxima and I think isn't the approach to take with this. There's a lot of levels to scrutinize here, and that will take a lot of time in general.

There has definitely been a misunderstanding about what I mean by 'first draft'. Essentially, what yourself and Proxima want to do is skip what I'm calling the 'first draft' and go straight to sorting through the levels.

And that's totally fine, as I said previously I support whatever approach Proxima wants to take. If it doesn't help you guys to have a 'working playlist' (which is probably a better descriptor for what I'm referring to than 'first draft'), that's not a problem. Do what works best for you.

Anyways, take all the time you need with sorting through the levels. You're right, it's important not to rush that particular part of the project.

[kaywhyn]

There has definitely been a misunderstanding about what I mean by 'first draft'. Essentially, what yourself and Proxima want to do is skip what I'm calling the 'first draft' and go straight to sorting through the levels.

Ok, I re-read the Discord discussion on this and I think I understand what you're getting at here. I'll see if I can interpret what you mean correctly. If you mean what I think you mean, essentially your approach goes like, "I choose this level, that level, etc." without needing to necessarily play the levels beforehand as your list we can analyze as a group to see what we can further cut down. Yea, this way of doing so would likely introduce arbitrary bias as Proxima mentioned. I'm definitely not in favor of randomly picking levels that should and shouldn't make the cut without properly analyzing the levels first, especially as we might miss out on a lot of great levels this way. As a reminder, it's been more than 3 years since I LPed v4.0 of RotL and hence a lot of the levels I have forgotten by now. I would definitely need to go through a lot of the levels myself again, which of course takes time.

Feel free to let me know if I'm still getting or interpreting any details of what you were proposing with what you envisioned with the "working playlist" incorrectly!

To be honest, the approach I likely would had taken had I made an update to RotL v4.0 a while ago would had been where I would mostly fix any backroutes for the current levels of the pack whose known intended solutions aren't enforced yet, swap levels around with one another to make the difficulty curve better, and for any authors who are around include updated versions of their levels that have been sent to me, while also being open to any level suggestions people post for possible inclusion to replace levels that end up being cut for the update. I know Crane sent me some stuff a while ago, but those were mainly levels of his that are currently in the v4.0 RotL pack that he updated for possibly the RotL update that I was going to do a while back. I'll also have to dig them up again, but I'm sure they're still on my hard drive somewhere.

At the same time, this Excel spreadsheet is definitely an excellent idea and I'm fine with going this route as well, particularly for the purposes of this community project.

And that's totally fine, as I said previously I support whatever approach Proxima wants to take. If it doesn't help you guys to have a 'working playlist' (which is probably a better descriptor for what I'm referring to than 'first draft'), that's not a problem. Do what works best for you.

Anyways, take all the time you need with sorting through the levels. You're right, it's important not to rush that particular part of the project.

Thank you. At least we agree there. Also as I previously mentioned taking time extends to all parts of the project so that we hopefully won't need to make many more updates later on once the pack(s) have been posted. I may still do some long-term maintenance in the form of updating the levels of the v5.0 RotL pack where needed if there's still backroutes, but I'll only make updates if they are necessary and after I post asking about the changes I'm thinking of making Again, likely won't be needed once the packs are posted and hence considered "finalized" but I'm fine with doing post-maintenance of the v5.0 NL RotL packs.

[WillLem]

I'll see if I can interpret what you mean correctly. If you mean what I think you mean, essentially your approach goes like, "I choose this level, that level, etc." without needing to necessarily play the levels beforehand as your list we can analyze as a group to see what we can further cut down. Yea, this way of doing so would likely introduce arbitrary bias as Proxima mentioned. I'm definitely not in favor of randomly picking levels that should and shouldn't make the cut without properly analyzing the levels first

I've highlighted in yellow bold the parts that are being misunderstood.

The following is the approach I suggested, as clearly as I can make it. I've spoiler tagged it to make this post smaller as this part of the conversation has taken up enough of this topic already!

Suggested approach, and discussion thereof

(1) Make a list of 30 x 4 levels which will form the "working playlist" (Draft 1) - they can be any levels at this point, and have no priority over anything not on the list. We do NOT decide at this point which levels should and shouldn't make the cut.

(2) Play through every level in the total pool (i.e. all levels on the list, and all levels off the list). During this step, we make notes, update any levels that need updating, and identify which levels should and shouldn't make the cut.

(3) Having completed step 2, we should now have a more firm idea of which levels will actually make it into the pack. We go through the process of swapping, moving, and replacing all levels in Draft 1 until we have Draft 2, which will be much closer to what the pack will actually end up looking like.

(4) Play through Draft 2 as if it's the completed pack, making note of how well the levels flow from one to another, etc. to get Draft 3, which would likely be a release candidate by that point.

All that's happened is, Proxima doesn't think step (1) is necessary and so we're starting from step (2). And that's fine, level selection was delegated to him for a reason and I'm happy to support the approach he wants to take - he has my full confidence.

The only difference it will actually make is to step (3). I imagine that it might it take a bit longer or be slightly more difficult if we don't already have a working list at that point. Then again, it might not. I could be wrong. It's natural for me to want to refer to a method has repeatedly worked for my own projects when suggesting how to go about this particular project - but, of course, I'm always happy to try new ways of doing things!

Anyway - by now, I'm explaining this only because I dislike misunderstandings. I haven't suggested rushing anything; on the contrary, my suggestion was that we add an extra step, which would actually slow the process down initially, but (maybe) make a later part of the process quicker and easier.

Think of it like this: when you take your washing out of the dryer, do you first sort everything into piles by clothing type (shirts, t-shirts, pants, socks, etc) or do you sort through the whole pile one item at a time? Is that first step necessary? I'd argue it is. It makes it much easier to see what you're dealing with, makes the total task feel less daunting, and increases the chances that the task will get completed by > 0%

Again, bottom line: I'm no longer making the case for my suggested approach at this point. We've already decided to do it Proxima's way and that's absolutely fine by me. Happy to leave it there and move on

[WillLem]

OK, so Proxima mentioned on Discord that we can start suggesting levels for Pack 1 Rank 1, particurly those that might make good X-of-each repeats.

These would be my top picks:

9 Ways To Live - perfect start to the pack, an excellent take on tutorial levels in general
Over or Under - provides an interesting challenge, and multiple possible ways to solve it
Snowy Caves - snow levels usually aren't my favourite, but this one is very well put together. 20 of each is excessive though, 10-15 would probably be fine
Pipe Dream - this gets more interesting the more you look at it. I'd probably suggest reducing it to 3 or 4 of each
Lemmington Spa - presents multiple 'get down from there' challenges in a single level, but isn't too difficult for the first rank by any means
Surrealism - probably my favourite level visually, could be good as a 2 of each
Neighbours - rock levels usually put me off, but this one is very well done as a 'use this crowd to save that crowd' challenge
Let's Play Lemmings! - very good pillar level, well put-together and requires use of several different skills to get a basic solution. Shame it doesn't make use of the OWW, but other than that it's a decent candidate for inclusion
Crystal Caves - this one's a classic custom level, and makes for an interesting challenge given only 1 Basher
Lem Dunk - a favourite of mine, just a shame it isn't possible to save 100%!
Minesweeper Lemmings - include it for sure, maybe reduce the X to some amount that makes the level require floaters?
THE FEARSOME FOURSOME - excellent level, a great multi-tasker with many possible solutions, perfect for X of each

In general, my criteria for X-of-all repeats would be: how many different skills are needed to create a basic solution? If the level can be done with just one or two skills, it's perhaps either not worth including, or the X amount should be reduced so that other skills need to be used.

That's it for now, I'll report back more when I've worked through more of the pack.