Multiplayer gameplay styles rabble topic yay!

[geoo]

Now that I've finished translating the 2P level discussion from idlingspace http://geoo.digibase.ca/lemmings/IS_2P.html" class="bbc_link" target="_blank">http://geoo.digibase.ca/lemmings/IS_2P.html, it's about time to start our own multiplayer discussion topic. I'll start off with a lengthy essay or rabble about what kind of gameplay I like and what I don't, to share some experiences and impressions. I'm not trying to prove a particular point or anything, just discussing a bit, so you're invited to join the rabbling! So, long rabble ahead! Rabble rabble rabble. At least the forum isn't overly truncate-happy. http://www.lemmingsforums.com/Smileys/lemmings/tongue.gif" alt=":P" title="Tongue" class="smiley" />

Note: For simplicity, I'll henceforth use the generic lemmininum instead of a species-neutral language, so whenever I write 'lemming', this can refer to lemmings, lix, clones, etc.

The original 2P levels have many flaws. We noticed that blockers and diggers are very powerful. So in many levels, if the players can send over saboteurs easily, they can block horizonal progress with blockers, and kill off the route or group of lemmings if they're held somewhere with diggers. As the original levels had the tendency to give the same amount of every terrain removing skills as builders, this usually meant that once a saboteur was in your holding pit, your lemmings there are done (this also applies to clones, perhaps even more there, see far below). If you didn't need a holding pit but have a constant stream of lemmings along your path, you're not gonna lose a whole bunch at once, but if you're under constant attack, the game turned into a builder conservation challenge, where you sooner or later ran out of builders, and were glad if you could at least get these builders into your exit. Either way, all you could do was saving individual lemmings rather than your whole group in that case.

Probably because constantly seeing the majority of your lemmings die in digger pits gets tiring after a while, hima from idlingspace tried to counter these level design flaws with additional rules, like "don't dig there". (One the Genesis the game seems to quit once no player has saved anything, so that might be another motivation for that.) Sometimes, they made sense (e.g. no digging at the bottom of level 7, at least if you care about the lemmings at the bottom), sometimes not so much (e.g. not digging the trunks in level 1, with only 5 diggers it's not that much of an issue, and digging in the thick terrain has just as much effect. Prohibiting blockers would probably have been more effective, as it's pretty much anywhere).

With L++/Lix/Clones, rules like this got redundant, as you can fix flaws in the levels yourself now. In the designing and testing process, a lot of guidelines have come up to avoid the above mentioned issues as a lot of experience has piled up, like highest route wins if the level doesn't prevent it, blockers are evil (though not as evil anymore with the incoming of batters), or to put steel under the exits and hatches. At least for classic style levels, these guidelines should usually be followed. There are more advanced levels that solely focus around hatch digging and defending against it (Simon's "Raise the Bar", my level "Jenga"), so the steel under hatches rule has been broken there deliberately.

I mentioned 'classic style levels', so it sounds like there's some categorisation of levels. Well...I think clearly dividing the levels into meaningful categories might be impossible, but there's two styles that stand out a bit and that are pretty frequent. Apart from these two styles, there's also non classic stuff, which I put into the other two categories:

1) The type of level where you constantly sabotage routes and have to fix your own routes, they're the easiest to build. You throw a bunch of random, connected terrain into the levels, copy it a couple of times so each player gets the same, and you're done. But also many more sophisicated maps play like this. Usually there's not too much to be done in regards to route building, apart from a few gaps or walls the route is already laid out for you. So you got a constant stream of lemmings along that route, and other players easily have access to your route to sabotage. So the gameplay here is a constant sequence of breaking and fixing spots in your route, which I've grown a bit tired of, as in most levels I find it pretty monotonous and tedious and they just drag along until there's no more lemmings coming out of the hatches. It's especially bad if you got a huge level that wraps around in both directions, where it's easy to lose track. Then it's only a game about looking all over the place to find broken parts in your route to fix. I'm not too fond of, and also pretty bad at this usually. In 2P, if you got a level with the player's hatches somewhere at the top, the exits somewhere diagonally below, a chunk of terrain in the middle and crossing paths, then this is a pretty good indicator that you got a level of type 1. In >2P, this kind of gameplay is even more frequent, because if you don't mix too well, a player will be primarily be exposed to a certain other player, and that can throw off balancing.

2) The route building maps. Here, you got a bit of a problem to solve that could easily serve as a Tricky or Taxing SP level. At the start you're usually on your own for a while, and have to be creative...or, well, crafty. Usually multiple routes are possible. Especially then they are awesome maps for 2vs2. Getting saboteurs over isn't easy and takes a while, but sabotage tends to be more effective, by outright killing a large bunch, or getting a strategic advantage. So you try to build routes that are hard to reach by saboteurs so you can send your lems over safely once it's done. Frequently your lems are stalled somewhere for a while, and once you release them, it comes to an epic and exciting showdown with frantic attack attempts. Prime examples are Rubix' "Passing Predicament", my multiplayer adaptation "Stepping Stones (2P)", or the original map "May the craftiest player win". I initially thought the same about the latter as hima, despite the epic battle I had with Steve on it the first time I played it, but playing it 2vs2 with Steve against Rubix and Simon while on voice chat (I think with a slightly amended skillset though) made me really appreciate the level and its craftiness aspect. I really love this type of maps as well as they are very strategic, but they're also very demanding, so not so suitable for late-night phase.

3) Experimental design with a specific gameplay idea in mind. For instance, the above mentioned hatch digging levels fall into this category. Another one is "Tower Defense (Part 3)", where each player has a tower, and a trivial, long platforming route within that tower to complete. While that is being done, you try to get a saboteur over to the other tower through a maze of buzzsaws, but also have to defend against attacks of the same type. "Ghetto Wars" is another one of that type where each player has to defend balconies against the opponent's lemmings which can bat your lems out or bomb a hole in the balcony, until after a while you got a landing platform so you can release everyone from the balconies into the exit. I've made a lot of levels of that kind lately, and it's not only playing them which is fun, but also the design process. You got an idea, design a level around it, and then try to find the best strategy and amend the level if necessary. Finding the perfect opening for the initial version of "Ghetto wars" was one of those strategy finds that was pretty surprising.
It's notable that there's been no attempts at asymetric levels yet, save for the one level in Clones called "Unfair", which was pretty interesting. I think we have a high fear of making imbalanced levels, though in >2P it happens naturally to some extent. When I made a level with an explicit attacker and defender role, I mirrored the whole situation and had both players play both roles in parallel. I think here might still be some interesting experimentation to be done with asymetric levels, as then a player can focus more on the role he takes. Every now and then experimental levels suck, but these can just be scrapped.

4) Anything else. Like "Downward Reduction". I wouldn't know where to put it.

Obviously, many levels also are a mix of different types. I really like strategic levels, where each time you try out new things, or test whether your strategy holds against the other player trying out new things. All-or-nothing levels are also a lot of fun, I like short, intense battles like this, where you immediately start again after a round has finished, and rather than counting lemmings, you count the ratio of rounds you win, if anything.

Some might have noticed that I haven't played much Clones lately. That's simply because I simply enjoy playing Lix a lot more, and with a lot of progress with the game going on lately, there's also plenty of opportunity to play and experiment.
My gripe with clones is that most maps play like type 1, and even those that would be considered type 2 play a lot like type 1 as clones is so grinding-heavy, and levels take very long. Once you got a saboteur in, you'll be busy with the usually drill/bash routine for the next 20 minutes that it's pretty much impossible to get out of, because terrain removers stop builders, and these fat bums turn around at every tiny pixel they bump their head on, or any little step they encounter. RAGE!!! These lazy penners are just too inflexible and slow! With lix, you can jump and turn them around at wish, and get out of your opponent's digger hole with a couple of builders (or nowadays a cuber as well) instead of seeing them drop down repeatedly (though now there's at least an option to keep builders in the air in that case), or turn around at the ceiling if you used an atomizer. In that regard I feel it's a bit of a step backwards towards original 2P lemmings and its levels. Clones policy of all skills aplenty doesn't help either (limiting the types of skills is one of the aspects that adds a lot of diversity!), and make most levels feel the same to me.
There's only one single map I really enjoy, and that's "Cage"/"Rage", probably the only type 3 map. (The latter is my modification which eliminates tedious end-games, and resolves things with a tie instead. Gives a bit of a preference to rush strategies over solid defense building strategies though). I don't think I played all experimental custom maps, I played a couple of Haymanizer's, which were conceptually interesting, but still didn't really work better than the normal levels for most part, at least IMHO. So perhaps there's some good levels I missed, but I've tried my hand at level-making myself, and somehow it's a lot harder to make a fun levels than in lix. I stopped after a while, probably not having tried out everything, but the maps I made didn't play well. Like, I totally managed to ruin "Stepping Stones (2P)", that's quite a feat.

Perhaps I'm just too spoilt by the lix gameplay, where the skillset is nicely rounded out with the skills balancing each other, and providing a lot of control over the lix. And where I can just pick the maps I enjoy out of the very diverse set of maps, and I'm lucky enough that Simon seems to share my taste to a good extent.

tl;dr: Read this post in its entirety you lazy bum!

[Clam]

Yes, this thread is long overdue. Multiplayer Lemmings has been around for, what, 20 years, and still hasn't had a fair run on the forums? Ouch. http://www.lemmingsforums.com/Smileys/lemmings/sad.gif" alt=":(" title="Sad" class="smiley" />

For now, I'll mostly just reply to geoo's points (since there are a lot of them) without adding too much more.

I'll just pass over the old 2P levels by saying they were good for their time, but now it's 20 years past their time, and we've moved on.

I'll go into skills more in later posts, but I can't help wanting to put this one to bed right now: Blockers are pretty much a non-issue since the batter was introduced. Back in the Blocker Dark Ages, removing a blocker with a fling bomber meant a lemming sacrifice, either a blocker of your own (wow unfair) or 5 seconds of your full attention, and another 5 seconds of valuable time waiting for the lemmings to get up again. And then the other player could just place another one, conceding a minimal amount of ground, unless the ceiling is low in which case you wouldn't even get that. In prehistoric times (i.e. pre-L++), it was worse still, if you can even imagine that. Now it takes half a second to whack a blocker away. Nice http://www.lemmingsforums.com/Smileys/lemmings/laugh.gif" alt=":D" title="Laugh" class="smiley" />. Their use now is limited to turning only your own lemmings, and those of other players who are too careless to keep track of them, plus turning some lixes walking just behind your own before they can respond. And IMHO, that's about right.


I like geoo's classification of levels, and it seems to me pretty accurate. It looks to me like you could just about replace those categories with:

1. Bad levels
2. Good levels
3. New levels, to be transferred to either 1 or 2 later on.

Type 1 play (constant route destruction/repair) is all about micromanagemnt, actions-per-minute (apm), and mouse/keyboard control. These are the things that every RTS game should do its utmost to avoid. Sadly, some element of this is going to be present in every level; micromanagemnt is and has always been a part of Lemmings. But as for pure Type 1 levels, right now I'm wondering if these have any merit to them at all. I guess they're easy enough for new players to pick up, and maybe learn some of the skills and tricks that will help when playing the 'good' levels. There's some amount of mindless fun to be had as well, though other levels do better at this (see below). But like geoo says, there's minimal strategy, and much repetitiveness. I'd add maps like "Downward Reduction" to this category; there's only one way to the exit and the gameplay revolves around fighting over paths from start to finish.

The levels geoo lists as Type 2 have a common theme: all prevent contact between lemmings of different teams until after significant strategic decisions have been made. You can choose a route and expect to make some progress down it before you encounter resistance. Just like in a regular combat RTS, you have some time to prepare before the enemy invades, and full-scale invasions are only possible later on. By contrast, Type 1 maps force teams together almost immediately, giving no opportunity to defend against attacks.

As for Type 3 levels, "experimental levels" and "new levels not yet classified" are pretty much synonyms for one another. But, the specific maps geoo cites share something entirely different in common: your exit is on your side of the level. Either it's impossible to send a full stream of lemmings over to the opponent's side, or there's no use in doing so. Forays into enemy territory are aimed purely at killing enemy lemmings, and stealing is nigh-on impossible. So Type 3 really ought to refer to defensive maps where you have to save your own lemmings instead of stealing someone else's.

I'd add another category (let's call it, oh, I dunno, Type 4? http://www.lemmingsforums.com/Smileys/lemmings/tongue.gif" alt=":P" title="Tongue" class="smiley" />) for the pure mindless fun levels, like "Superbowl", "Last man standing", or the old L++ favourite "Experimental: Flinging". While not good by the usual standards, these are great for entertaining between serious games, or when everyone is tired at the end of a session.

In summary (aka. tl,dr):
Type 1: Constant route sabotage and repair, from start to finish. (edit: originally I called this "route-building" instead of sabotage, refer to geoo's post above for the definition of route-building)
Type 2: Strategic levels, limited sabotage early on and larger battles later.
Type 3: Defensive levels, raid your opponent while building your own path.
Type 4: Quick, explosive levels, limited strategy but good fun for little effort.


Finally (for now http://www.lemmingsforums.com/Smileys/lemmings/wink.gif" alt=";)" title="Wink" class="smiley" />), I'd like to give credit to Simon, and everyone else who's helped out with Lix (and the late L++). It's a great indie project in the making, and it's giving Lemmings a new lease on life. Long may it continue! http://www.lemmingsforums.com/Smileys/lemmings/thumbsup.gif" alt=":thumbsup:" title="Thumbs Up" class="smiley" />

[Simon]

Great thread with great analyses and stories to read. Also thanks to Clam for the credits!

I will probably also make a "proper" reply later. Right now, I'll post pictures of some of the discussed levels, so people who haven't played can imagine what happens.

http://asdfasdf.ethz.ch/~simon/etc/levels-network-2p-geoo-ghettowars.txt.png" alt="" class="bbc_img" />

This is Ghetto Wars, the batter map. You have four hatches on one side, and your exit is on the same side. Send a climber to platform under your hatches.

Attack with batters and batted exploders. Most incoming attackers will land on the small ledges in front of your balconies, they're easy to dispose there. If a saboteur actually gets inside your balcony, you must probably bat out many of your own "resources" together with the saboteur.

http://asdfasdf.ethz.ch/~simon/etc/levels-network-2p-geoo-jenga.txt.png" alt="" class="bbc_img" />

This is Jenga, a map focused on killing the other person's spawn point. Again, your exit is on your side of the level.

http://asdfasdf.ethz.ch/~simon/etc/levels-network-2p-geoo-tower32.txt.png" alt="" class="bbc_img" />

Tower Defense Part 3. Make your way through your own tower with platformers, but this part is placed just to delay. Attack and defend in the center in the meantime, trying to get a saboteur into the opponent's crowd. This level doesn't give batters to throw successful trespassers out so easily.

http://asdfasdf.ethz.ch/~simon/etc/levels-network-2p-geoo-steppingstones.txt.png" alt="" class="bbc_img" />

This is two-player Stepping Stones, a highly strategical map. Make a holding cell under your hatch, and decide which of the numerous routes you want to build, possibly several of them. Make sure you release your crowd before saboteurs or even the opposing main group arrives.

http://asdfasdf.ethz.ch/~simon/etc/levels-network-2p-rubix-RubCircus12.txt.png" alt="" class="bbc_img" />

This is Passing Predicament. Your exit is on the other side of the level. Again, you can choose routes, the two main routes are the long direct builder staircase and the ceiling route.

If you want to bunker your main group, don't do it at the bottom of your initial area, instead bunker them close to the hatch. It's a higher location and further away from the opponent. Build on the very edges of the huge steps toward the bottom area, so enemy climbers cannot attack the bunkered lems. For a similar reason, you want to build on the opponent-facing edge of the central pillars.

http://asdfasdf.ethz.ch/~simon/etc/levels-network-2p-oldgfx-geoo-Flinging.txt.png" alt="" class="bbc_img" />

Experimental Flinging. You start walking towards the gap, and must cross it with knockback exploders. Blockers can keep the crowd in place temporarily.

http://asdfasdf.ethz.ch/~simon/etc/levels-network-2p-simon-lastman.txt.png" alt="" class="bbc_img" />

Last Man Standing. A mindless batter fest at first, it can be played skillfully with micromanagement when both players have only a few lixes left. You have just one builder, and if you try to use it while the opponent still has lixes, they just bat your builder so you cannot win.

http://asdfasdf.ethz.ch/~simon/etc/levels-network-5p-geoo-dwn5.txt.png" alt="" class="bbc_img" />

Donward Reduction. It's mainly crowd control with miners. This map is surprisingly hard to play well. Little mistakes are absurdly costly in the later stages of this level, and game outcomes vary greatly.

-- Simon

[RubiX]

Quality fun!

(This post is almost as big as your posts.  Glad I could keep the pace up)

[DragonsLover]

Sounds really cool! I'd love to play a game back someday! http://www.lemmingsforums.com/Smileys/lemmings/cool.gif" alt="8)" title="Cool" class="smiley" />

Just wondering, is cooperative mode planned? Would it be possible to do or multiplayer is really reserved for versus play? And would it be still fun as well?

[okmot]

Interesting thread, allow me to add in my point of view.  My introduction to Lemmings was on the SNES, which included a 2P mode that provided countless hours of exciting competitive chaos.  Clones was an attempt to recreate and expand that feeling.
When designing the MP levels i reviewed all of the SNES ones as well as levels from Pingus to try and get an idea for different categories of levels.  Most obvious grouping was symmetric vs. asymmetric.  The level "Buddy" is an asymmetric map based on a lemmings map where both groups must take the same path the exit, differing only at the very end.  There is an inherent advantage for one player based on the start point positions so noting that i created "Unfair" which attempts to balance the position advantage with a morph count disadvantage.  That map was fun to play, for a time, but after becoming skilled in sabotage it became difficult to counter, although you can always adjust the number of clones/morphs to compensate during the match.  This is why we tweaked the gameplay a bit to try and make sabotage clones less powerful: spinning clone can be shattered, molders can build with no land under them, doppels can be shattered.  I find this makes sabotage easier to deal with, and i am open to further tweaks.
But this leads to the next observation for all symmetrical levels, i believe they are all "type 1" (geoo's classification) in that the goal of MP levels is to win by attacking your opponents path, and thus defending your own constantly.  We tried making maps with no interaction such that the player that makes the best route to their exit first wins, but that was boring (genesis of Multiverse Match mode) so sabotage clones were added.  When in the level the attacking will occur can be delayed for taste by forcing route building or a maze but sabotage is the defining element of MP and once it starts then that becomes the object for the rest of the match if you intent is to win.  The guidelines i tried to follow for creating maps were: allow multiple paths to the exit, use swap traps, force paths to cross, and discourage drilling under start/end points.  Of course some levels break these for variety, but typically if the paths don't cross and there is only one path to the exit it's a boring level.  Recently more maps are incorporating dynamic elements such as moving starts/exits, and modified intent with other game modes such as hiding the identity of a star clone in Capture the Clone mode so that it is only revealed if a dark clone is nearby.  For a true type 3 map i think the goal of the level must not be "who can save more" but a new experimental objective.  The categories of symmetrical level design are broadly just running through all permutations: paths cross above/below starts/exits, cross early/late, inside -> out or outside -> in, multiple starts/extis, and in clones the gravities can multiple these, but if they are the "classic" mode (Corral The Clone) then i'd call them type 1.  Arguably of the most fun Clones maps of this mode is "Lurk" which is small with paths crossing below and almost no route building required.
On the topic of co-op there are 3 kinds: 1) Puzzle co-op in which all players are controlling the same group and working together to complete the level, 2) Puzzle co-op in which each player controls a different group with different abilities (Super Synergy / Quantum Loop) and each group must use its morphs at the right time, and 3) Multiplayer co-op in which each opposing group has multiple players within the group.  I found that #1 was much for fun than i thought it would be, and #3 only works if players within a group are of the same skill level and work together.

[ccexplore]

Hmm, too bad I'll be out of the country for 2 weeks, the lix screenshots were interesting enough that I'd definitely want to give them a try, or at least to spectate a little. http://www.lemmingsforums.com/Smileys/lemmings/cool.gif" alt="8)" title="Cool" class="smiley" /> Any chances of being able to capture movies of matches in Lix that are uploadable to youtube (or at least viewable in Lix, hopefully feature already exists)?  I'm willing to try help out with my admitted limited knowledge in that area of programming (not to mention lack of time http://www.lemmingsforums.com/Smileys/lemmings/undecided.gif" alt=":-\" title="Undecided" class="smiley" />).

I think having read everyone's comments so far, I feel Clam's summary best capture the salient design points to consider when creating multiplayer levels:

Type 1: Constant route building and destroying, from start to finish.
Type 2: Strategic levels, limited sabotage early on and larger battles later.
Type 3: Defensive levels, raid your opponent while building your own path.
Type 4: Quick, explosive levels, limited strategy but good fun for little effort.

I would recommend this blurb above as "MP design 101" that everyone should read before even taking one step at designing multiplayer levels.  Ultimately it's a matter of innate skills, personal tastes and mood which types of levels one will enjoy most, but having the above as a guide should hopefully help people create levels most closely matched to their favorite gameplay styles.

I think I will also give Lix a try one of these days, as it seems easier to create levels there.  I can't blame Clone's level editing since clearly there are people there capable to making full use of it to create levels both fun and beautiful (and there's no question that when you have the talent for using it, it can be very powerful indeed), but I think for me, I can really only work well when given mostly straight-edge terrain pieces, like what you're seeing in Simon's screenshots.

[Simon]

Just wondering, is cooperative mode planned? Would it be possible to do or multiplayer is really reserved for versus play? And would it be still fun as well?

Competitive teamplay is in. Larger maps are are better for 2v2 than small maps.

Cooperative play isn't supported. The closest thing implemented to this is that the game allows to start with everyone on the same team. You could then solve a singleplayer map together.

Assymetric skillsets aren't supported either right now. You could roll a cooperation map from that otherwise, by simply having the players ignore the end-of-game rankings and instead compare their tallied saved counts to a target count.

Clones has cooperative maps, but I never really looked into those...

One of the design rules for Lix was KISS, so I didn't want it to overload it with features, but rather do one thing as good as I can implement it. Most extra features came as a by-product, or from heavy public demand:

• The recently implemented skills were all done when geoo was visiting me. He even drew and programmed the cuber himself.
  
• Replays and savestates were implemented to allow history-changing during networked games.
  
• The server daemon (what people connect to when they hit the "central server" button) was made when Mindless and later tarzeau offered me to host such a program on their static-IP machines.
  
• Observer/spectator mode is a typical heavy-demand feature. Bonus points since it doesn't complicate the game.
  
• A possible future popular-demand feature is neutral lixes, along with optional game modes "only neutrals count" and "first save wins", which might also be combined then.
  

Amendments that aren't fundamentally new features go in very frequently. Bugfixes, physics changes, etc. are often discussed on IRC, and I'll then do what most people consider better or more intuitive.

Any chances of being able to capture movies of matches in Lix that are uploadable to youtube (or at least viewable in Lix, hopefully feature already exists)?

Lix can save replays to text files and play them back later. Multiplayer replays always save the level into the replay file as well, so distributing that single file still ensures everyone sees the same thing. Old replays won't be played back properly if the physics have been changed in the meantime. The replays contain version information though.

Especially on Windows or Mac OS, it should be easy to record the screen with third-party programs. Rubix has done that a few times, and we're actually planning a promotional video with excerpts from games to upload on Youtube.

-- Simon

[geoo]

Hmm, I like Clam's type 3 classification, that's indeed something many of these experimental maps have in common. Still, my original category was a bit broader, and not all these levels are defensive maps. For instance, the Clones map "Cage" is a lot about area control, while you don't care how many die, then there's one sports map I once made using some "I am A.T." (however not the one with the large tower in the middle, another one which I can't find right now) which played totally different as well, and "Downward Reduction" which is about controlling the flow and the direction your tunnels go in, to end up in the right spot, from which you have to attack at the right time (depending on your amount of miners left) with speedmining. @Clam: And don't say this is a type 1 map, just because you suck at it. http://www.lemmingsforums.com/Smileys/lemmings/tongue.gif" alt=":P" title="Tongue" class="smiley" /> On the contrary, the fact that your winning average on this map is lower than on the average type 1 map indicates that there's some strategy element you're missing. http://www.lemmingsforums.com/Smileys/lemmings/tongue.gif" alt=":P" title="Tongue" class="smiley" />
Let's introduce type 3b for these, the type 3 everything-else category. http://www.lemmingsforums.com/Smileys/lemmings/wink.gif" alt=";)" title="Wink" class="smiley" />
Thinking of (most) of the original maps, they are probably worse than type 1. They should deserve their own type 0: grindfests where you just spam blockers and diggers and save only very few.
Then there's of course also the sucky latenight maps that are meant to be played only once, or in very deep late-night phase when even type 4 is too demanding (so type 5 would be quite fitting for these).
Ok, done with the over-categorization.

But this leads to the next observation for all symmetrical levels, i believe they are all "type 1" (geoo's classification) in that the goal of MP levels is to win by attacking your opponents path, and thus defending your own constantly.

Can't say I agree with rt at all though. The mere assumption that there's always a path is already a limitation. And non-crossing paths != no interaction. I guess if the levels in Clones have been designed with these rather limiting guidelines in mind, that explains why they're all type 1 and play very similarly. (Again, except for "Cage" save for the possible endgames at the bottom, I don't know how one can classify that one as type 1 unless one misinterprets type 1 as "there's interaction".)

Hmm, too bad I'll be out of the country for 2 weeks, the lix screenshots were interesting enough that I'd definitely want to give them a try, or at least to spectate a little. http://www.lemmingsforums.com/Smileys/lemmings/cool.gif" alt="8)" title="Cool" class="smiley" />

As coincidentally I'll be away for two weeks as well now, that's kinda fortunate for me, as that way chances are good not to miss the once-in-a-lifetime chance of seeing ccexplore actually try multiplayer. http://www.lemmingsforums.com/Smileys/lemmings/tongue.gif" alt=":P" title="Tongue" class="smiley" />
Ok then, comparatively short post as I'm off now.

[Clam]

But this leads to the next observation for all symmetrical levels, i believe they are all "type 1" (geoo's classification) in that the goal of MP levels is to win by attacking your opponents path, and thus defending your own constantly.  [...]  When in the level the attacking will occur can be delayed for taste by forcing route building or a maze but sabotage is the defining element of MP and once it starts then that becomes the object for the rest of the match if you intent is to win.

Like geoo, I strongly disagree with this point. Yes, route sabotage and repair (edit: not 'route building') is going to be a feature in (almost) all levels, but good map design can reduce its prevalence. A map only truly becomes Type 1 when you have steady streams of opposing lemmings on completed, crossing paths to their respective goals. That's when you get the mindless "looking all over the place to find broken parts in your route to fix". Type 2 levels allow a crafty player to gain an advantage, maybe evn a decisive advantage, before the tiresome sabotage game begins. Another notable level in this category is Rubix's "I want to punch your face in!", in which the route cannot be damaged at all (save for nuking) and the only means of interaction is the Batter skill - it's completely free of Type 1 play. Type 3 levels (plus some Type 2 levels early on) only allow players to send a few (sometimes just one) saboteurs at a time, meaning the other player knows exactly where an attack can occur; contrast this to Type 1 where any enemy lemming can attack anywhere along the crossed section of the path. The predictability of sabotage is a key difference, and the focus shifts to the ability to control a saboteur and neutralise an enemy one, rather than the ability to scroll around the map.

Also, I'm curious as to how "Use swap traps" is a guideline for map design. Can you please explain this further?

Hmm, I like Clam's type 3 classification, that's indeed something many of these experimental maps have in common. Still, my original category was a bit broader, and not all these levels are defensive maps.

Having not played most of your 'experimental' maps, I don't know what they play like. My suspicion, though, is that once the maps have been played for a while and strategies developd, most will settle into one of the categories I mentioned above. So, I just decided to consider new/experimental maps as 'awaiting classification', and change Type 3 to something completely different http://www.lemmingsforums.com/Smileys/lemmings/tongue.gif" alt=":P" title="Tongue" class="smiley" />.

As for Downward Reduction, I called it Type 1 initially because it doesn't fit types 2-4 at all. But on further thought, it doesn't match the typical attributes of a Type 1 level either - there isn't an almost-premade route to the exit, and by the time the exits are reached there are few or no lemmings left in the hatches, so the game is immediately over once the exits are reached. This 'all-or-nothing' nature seems to be the defining nature of this map (along with the somewhat-similar "Pave the way"), so I guess there's no choice but to give this its own category.

And just when we thought we were done categorising things, Rubix has just recently dreamed up a completely novel style of map - combat racing. If you're curious, come on IRC and ask for a round of "Marble Run" http://www.lemmingsforums.com/Smileys/lemmings/laugh.gif" alt=":D" title="Laugh" class="smiley" />.

[Simon]

The constant stream from hatch to exit is a major factor in how a map feels. Stream maps are those where the route is done when lems are still spawning, or more importantly, where they don't cluster, even if you neglect crowd control. Strategical maps haves them cluster much more often.

Stream map = no crowd control, no clustering, typically micro-heavy
Macro map = requires ability to plan and execute several routes at once
Swingy map = you often cannot tell who's won until it's almost over
Grind map = low experienced progress per time, little possibility to deal heavy damage at once
Hot map = lots of damage may happen per time not paying attention
Dumb map = terrain or execution contains jokes, it doesn't even have to be fair

Stream usually implies coldness, no swing, and little macro. Grind maps are cold per definition. Hot maps tend to be short and replayable. http://www.lemmingsforums.com/index.php?topic=243.msg8517#msg8517" class="bbc_link" target="_blank">Immediately go back to Fanning was designed to be hot, but it came out a bit grindy and non-swingy: You have to assign 60 floaters while making a route from a different hatch, overtime was 6 seconds, and you had to make a route for another hatch. Missing a floater wasn't instant-fail though, so the floating started to feel grindy.

geoo's type 2 maps are high on macro, low on stream, and low on grind.

Downward Reduction and Tower Defense part 3 are hot and swingy. Spamfest (a very old dumb map) was swingy, but not hot. Raise the bar (hatch digging map) was extremely hot, but not that swingy.

-- Simon

[ccexplore]

Hmm, too bad I'll be out of the country for 2 weeks, the lix screenshots were interesting enough that I'd definitely want to give them a try, or at least to spectate a little. http://www.lemmingsforums.com/Smileys/lemmings/cool.gif" alt="8)" title="Cool" class="smiley" />

As coincidentally I'll be away for two weeks as well now, that's kinda fortunate for me, as that way chances are good not to miss the once-in-a-lifetime chance of seeing ccexplore actually try multiplayer. http://www.lemmingsforums.com/Smileys/lemmings/tongue.gif" alt=":P" title="Tongue" class="smiley" />
Ok then, comparatively short post as I'm off now.

Believe it or not, I have actually participated in the 1st and 2nd (or maybe it was 2nd and 3rd) multiplayer jams over at Clones.  They were the less populous evening sessions though, and I was seriously outmatched with players like rt, tom and jk.  Aside:  in fact, there is one 2-2 session I participated in during one of the jams, that actually got captured in one of the youtube videos rt posted from time to time.  See if anyone finds it.  Hint:  since I sucked so much, even the video pretty much focused on everything else but me. http://www.lemmingsforums.com/Smileys/lemmings/tongue.gif" alt=":P" title="Tongue" class="smiley" />

Especially with the type-1 heavy Clone-style levels like Octogame, having to pay attention in real-time to all over the map is so opposite to the singleplayer level-playing that I'm so used to, it was a complete shock and I utterly sucked at it.  Perhaps I'll fare better with the allegedly more strategic maps available at Lix, but I still kinda doubt it.  No doubt watching me suck will at least be good for entertainment value though. http://www.lemmingsforums.com/Smileys/lemmings/tongue.gif" alt=":P" title="Tongue" class="smiley" />

[ccexplore]

Also, I'm curious as to how "Use swap traps" is a guideline for map design. Can you please explain this further?

I don't know if it's so much a guideline as simply something useful to throw into multiplayer levels.  In Clones a swap trap changes a clone that touches the trap to the color of your opponents, effectively ceding control of affected clones to your opponents.  It's not hard to see how that would spice things up in MP levels.

And just when we thought we were done categorising things, Rubix has just recently dreamed up a completely novel style of map - combat racing. If you're curious, come on IRC and ask for a round of "Marble Run" http://www.lemmingsforums.com/Smileys/lemmings/laugh.gif" alt=":D" title="Laugh" class="smiley" />.

I expect there will always be some levels that defy categorization.  I think the main point is not to really try to come up with a complete categorization, as simply to highlight salient things to consider when designing multiplayer levels.  I think MP levels tend to be trickier to design because the interactions that occur during MP gameplay are just so much richer than in SP levels, and so it's very helpful to have a good conceptual framework for understanding MP gameplay in order to design good MP levels.

I do have to admit I'm slightly curious at what "combat racing" would be like on Lix.  In contrast, with Clones, the object capabilities are so ridiculously rich, you can easily (well, maybe not "easily", but quite feasibly anyway) design levels that are absolutely nothing like what one normally think of as MP Lemmings levels at all.  I have for example actually been able to create a working 2-player Reversi level in Clones.

[okmot]

Can't say I agree with rt at all though. The mere assumption that there's always a path is already a limitation. And non-crossing paths != no interaction

I was not assuming there was a path provided, it could be a path the player creates.  In either case the point is if the goal is to save more creatures than the other players then since it's not possible to directly damage the creatures then the only way to win is to interfere with their path.  Thus if type 1 is "Constant route building and destroying, from start to finish" then isn't that what the optimal strategy for every mp map is?  How else could you win?  The more time you can allocate to building a good route and attacking opponent routes the more advantage you have.  Perhaps an example would help, can you create some Clones maps that illustrate what you mean?

Yes, route building and sabotage is going to be a feature in (almost) all levels, but good map design can reduce its prevalence.

If you minimize route building (forcing the player to make a path to their exit) and sabotage (paths relatively close to each other, or crossing) then what are you left with?  Keeping the seat warm while the creatures walk to the exit by themselves? I've noticed that on levels in which the start point is in the middle and exits on either end (so that paths don't cross) tend to make it very difficult to make a comeback because once a player has the lead they can strictly focus on path repair and effective sabotage becomes virtually impossible forcing a nuke instead of a certain slow death.  So the tendency is to let weaker players lose by their own poor route building (losing a clone or two) and then just waiting for the map to end without ever making a move against the weaker player.  I don't think anyone benefits from that.  I prefer maps in which it's unclear who is really winning until late in the match.

steady streams of opposing lemmings on completed, crossing paths to their respective goals. That's when you get the mindless "looking all over the place to find broken parts in your route to fix"

I don't see that part of a match as mindless.  By carefully choosing where to inflict path damage while altering your own path you can reverse a losing position.  At some point in every map you will have a completed path to your exit, and your next step (although better to do it in parallel) is to sabotage regardless of whether the paths cross.  It seems like this is just a matter of degree and delay we are talking about.  Even with well designed levels that supposedly increase predictability of the location of sabotage i would think that crafty players would overcome that to sabotage in unexpected places. We should have a  match sometime and you can instruct me http://www.lemmingsforums.com/Smileys/lemmings/smiley.gif" alt=":)" title="Smiley" class="smiley" />

As for the swap trap question, ccx provided a good response.  It's the only trap that is specific to MP and is twice as deadly as a trap that just kills a clone so if you can lead your opponents clones into it there is a high reward.  to truly alter gameplay you need to need to alter objective; what defines a win.  In Clones there is a "WinLoss" trap which you can trigger arbitrarily to issue a win or loss to any group, this is used in one custom map in which you win by either killing all opponent clones or by removing some land at a specific location.  In practice it is much more fun to try to get to that bit of land and blow it up http://www.lemmingsforums.com/Smileys/lemmings/smiley.gif" alt=":)" title="Smiley" class="smiley" />

[Simon]

The Clones physics will naturally lead to more stream. The builder is at least 1/2 as fast as a regular walker. Lemmings and Lix builders work at 1/8 the speed of walkers. It's unfeasible to stream while making a route requiring many builders.

Some Lix maps can't be won without clustering the main group while making the path. We found that these maps tend to be good. Some maps build up huge mental tension about when to free the accumulated crowd and how: Sending the crowd will lose to the enemy sending a few to the same place. Sending only a few yourself might just waste skills to stall them around a completed route (this doesn't include sabotage, which is always useful). Sending none will waste your advantage of having a good route earlier than the opponent.

The only Clones map that enforces the player to not stream all the time is... what was it called, the polar-style map with the horizontally moving exits. That is one of the maps I enjoyed most. Otherwise, it's always bad to cluster in Clones.

Route securing and sabotage are a part of nearly every game. Crowd management is a highly interesting and fun aspect. However, in Clones, it correlates too often with route building, since one is nearly always streaming.

-- Simon