Originally posted by S.N.N.A staple of SMWC has always been its Vanilla Level Design Contests (VLDC). They by far garner the most interest of any contest, usually bringing in between 80-120 levels. What I've had in mind for a while now is to have another contest like this, but with some twists…
In case you can't see what I'm getting at here: an SMWC VLDC collaboration hack.
Since
the first VLDC, there’s been talk of taking all of a contest’s entries and compiling them into a single ROM. It feels like the obvious next step; why force people to download every single entry individually when they could play all the levels in one spot? Unfortunately, this idea always seemed to come as an afterthought, and a compilation never materialized for the first six Vanilla contests (aside from Aja’s independent effort for VLDC1 years later). Instead, more effort was placed on An SMWCentral Production 1 & 2, the only two official sitewide SMW collaborations at the time of this contest. (I do have plans to look at the history of these two projects in the future.)
This changed with VLDC7: for the first time, a compilation/collaboration hack would be the plan from the start. Music would be composed, overworlds would be drawn, and the rules and restrictions of the contest would be rewritten to make the compiling process as smooth as possible. The result would be the best collection of levels this series of contests has brought out thus far, all wonderfully packed together in a single patch.
The Planning Phase
Unsurprisingly, there’s more to compiling levels than just inserting the level from one ROM into another. Your typical level is going to be using sublevels, custom Map16 tiles, custom palettes, Message Box text, and – for VLDCs 3, 7 and onward, at least – custom music, just to name a few. If you really want to get into the process, I highly recommend reading
Aja’s thread on compiling VLDC1, but suffice to say that there is a lot of variables, all while making sure nothing overlaps. Then consider the fact that the first two contests had no limits on level exits, the first
three contests had no limits on level length, and
none of the contests had limits on sublevels or Map16 space; even with today’s tools I’m not sure you could fit all of VLDC3 into a single ROM. Thus, I’m not really surprised that, if anybody actually tried to compile the first six contests at the time, they simply decided it wasn’t worth the effort, even if it meant leaving all those levels in obscurity.
VLDC7 took a different approach: before submitting a level, users would have to sign-up to get assigned resources for their level:
Originally posted by S.N.N.-A main level and 6 sublevels.
-8 secondary exits.
-5 rows of Map16 (FG) and 5 rows of Map16 (BG).
-5 ExGFX slots.
While the rules originally limited the number of entrants to 64, it was decided to host two ROMS to allow even more people to join. This was the right call; even though sign-ups lasted only three days, 125 individuals (not counting partners) showed up. Ultimately, 82 entries would get made, so only one ROM was needed for the final product. It does seem unfortunate, though, given 44 of those slots were wasted when they could’ve been filled by others that could have joined the contest later, like in previous contests. (Strangely, allowiscous was allowed to participate without signing up, and I didn’t see anyone address this at the time of the contest.)
Nevertheless, those were the resources allocated to each entrant. A shared Map16 page was also
assembled by the participants, and custom music was on the table one again. Users could either say what remixes they wanted used for their level or insert up to two unique tracks themselves (
although sampling did become a potential issue). Users could also share Map16 tiles with each other if they wanted to, though I couldn’t tell you how many actually did that.
As you’ve probably noticed, custom music and ExGFX (i.e. tileset merging) have returned to the table.
S.N.N. wanted to reuse the rules from the contest with the highest turnout at the time, which turned out to be VLDC3. Thus, the restrictions of the last three contests were gone for good, and so long as the author stayed within their assigned resources, they were free to rearrange the vanilla graphics however they wanted. (Users were even allowed to recolor tiles, so long as they stayed “reasonably vanilla”; see the IRC log included in my archive for more information on that.) Also removed were the restrictions on what type of levels could be made (very good) and any limits on level length (mercifully, not that big an issue this time).
As for rule additions, the most noticeable changes involve Switch Blocks and level themes. Previously, Switch Blocks were usually used as a required exit for the level; you find the Switch Palace within the level, and then use the now active blocks to reach the “actual” exit. (yogui’s Mystery of the Giant Dark Switch from VLDC5 is a textbook example of this.) For this collab, however, four Switch Palaces would be included in the overworld, so this gimmick clearly couldn’t be used. (Wisely, these palaces were taken from entries from the
24-Hour Valley Switch Palace Design Contest from a few months earlier.) Instead, entrants were encouraged to use Switch Blocks to make it worth the player’s time to activate them; in all, 34 entries would make use of the blocks, some even using them for Yoshi Coins or secret exits. It was also strongly encouraged to be creative with level themes, rather than stick to the basic grassland/forest themes. A list of themes was given as examples, and all were ultimately given their own worlds; this did make for a much more diverse group of levels, although some themes got shafted, like Ghost Houses. (Water, in particular, rarely made an extended appearance in levels.)
The submission period began on January 3rd, the first time the contest didn’t begin in February. That said, it was also the longest submission period thus far, going all the way to February 16th, over six weeks long. (Again, the very brief sign-up period seems pretty unfair compared to this.) One final change is worth talking about: for the first time, the contest would have its own subforum. Rather than being contained to just two threads (submission and discussion), users would be allowed to make their own threads to show off their level and receive feedback. This might be my favorite change in the contest format; it goes a long way of bringing the contest back to its more cooperative, jam-like feel that I felt while playing VLDC1. Quite a few people made sure to post feedback in everything, and (save for some unrelated petty fights and one incident I’ll get to later) overall this was a lax and friendly contest.
The Levels
To put it simply, VLDC7 raised the bar for level quality. Once again, the group of participants are mostly veteran users; more than half have been on the site long enough to remember VLDCs 3 and 4, and over 30 of them have participated in at least 3 of these contests already. Meanwhile, the boundaries of the last three contests have been lifted (albeit with new restrictions as required by the collab), allowing for much more creativity from the contestants, even past visual enhancements. Combined with easier avenues for feedback and an extra two weeks of development time, and you have, in my opinion, the best collection of levels thus far, alongside two shining examples of vanilla design.
To illustrate my point, let’s look at ZMann’s entry, Lakitu’s Castle. ZMann is one of a handful of users that has participated in every vanilla contest since VLDC3 (the others being MarioFan22, TriplePat, and yogui). I ranked their debut level, Scaling the Icy Mountain, 9th place out of 116 when I reviewed that contest; looking back, it’s a quaint level, but solid and streamlined in its design, which is more than I can say for a lot of levels of that time. Their level for this contest, Lakitu’s Castle, is much more developed; even though Lakitus aren’t the most versatile enemy to use, the overall design makes for a unique and entertaining castle level. Despite that, I ended up ranking it 31st out of 81, not because the author’s levels have gotten worse; the competition has just gotten so much better over the years. I would honestly say that half the levels from this contest are at least worth checking out, which is much more than I can say for some of the earlier VLDCs.
Some highlights: Nimono, making his sixth(!) appearance in a VLDC, teamed up with Magi for Algid Ascent, one of the most gorgeous levels I’ve played for this series; it’s also got some real nice design reminiscent of Kaeru’s Mountain Heights from VLDC1. Other previous winners (yoshicookiezeus, allowiscous, Agent Q, and Lui) also returned for some entertaining levels: the simple, solid and tricky Lotus Cave Overpass; the even tougher (but satisfying) Spiny Logging Fort, another great entry into the “Industrial Forest” canon; the flashy and memorable Realtor; and the great desert/ice combo Brisk Barren, in that order. And while this is their contest debut, it’s no surprise that the veteran FPI churned out a great Eater Block level, Snake Eater. (This beautifully pairs with Forty2’s VLDC5 entry, Metal Gear Solid 3.)
As far as common themes go, I hope you like line-guide levels. WhiteYoshiEgg’s Touch Fuzzy, Aeon’s Grinder Grotto, Sokobansolver’s Zip-Line Adventure, amhunter’s Devious Devices, and quite a few others all do an excellent job with this style of platforming. I think it makes sense why this type of level was so popular; line-guides are an easy way to control the pacing of the level, so long as the player isn’t stuck on the same platform the whole time. Still, there’s enough unique ideas in each of these levels to keep them from blending together too much, and each one listed above is definitely worth a playthrough.
In the end, three levels stuck out to me as the very best. Mineyl’s Red Hot Ruckus takes third place, and while I don’t love it as much as the top two, it’s a superbly done vertical volcano level. The platforming is tense and varied, while the pacing of the level was pretty much perfect. Second place goes to Morsel, who has been one of the most interesting designers of this contest since their entrance in VLDC4. One Sunday Morning would have been the winner of any other contest; it looks great, the theme is unique, the use of enemies is incredibly creative, and that ending sequence is top-notch. It’s just a wonderful and energetic level and combined with the Disco Train track from DKC2 really fits with that style of design.
And then there’s Cavern of Pandora, by worldpeace. I don’t think I’ve played an entry that’s really gripped my mind like this. The Eater Blocks have long been underutilized, and previous levels designed around them haven’t been the most interesting. But here, the concept is perfectly explained, starting simple without ever feeling patronizing. The opening grassland portion would’ve worked well on its own; once the player enters the titular cavern, and that music hits, that’s when the level really transcends to a new tier. The puzzles and obstacles inside use these blocks to their full potential (without getting into the glitchier side of that sprite) and are just an incredibly fun challenge to overcome. The secret exit steps up these concepts even further, though it admittedly crosses a line of difficulty for me here (though it’s at least signposted well). I truly haven’t played anything like it in my VLDC reviews, and it’s hard to imagine what levels in the future are going to be able to top it. At the very least, it’s clearly the best entry of this contest… right?
Scoring and Judging
Let’s back up a second. Like I said, the rules effectively took a reset back to VLDC3 standards, and that includes going back to numerically scoring levels. Actually, the score spread feels similar to VLDC4, with four main categories, now with an additional 5-point category for following the rules. These categories are “Design and Fun”, “Creativity”, “Difficulty”, and “Aesthetics”; the first three are worth 20 points while Aesthetics is worth 15, putting the total points at 80, much higher than any contest so far. This feels like a big step back from VLDC5; Design is no longer a majority category, nor even a superlative; Aesthetics is nearly 20% of the entire score; and Difficulty remains a fraught category to use for scoring. This already offers some opportunity for ranking oddities, and unfortunately it just goes downhill from there.
Once again, there are only two judges; S.N.N. had to drop due to the heavy workload on the compilation side. That left Counterfeit, marking her third appearance (the most of any judges besides FPzero), and hack moderator cstutor89. Again, this is a big step down from VLDC6, or even VLDC3 for that matter; considering this has happened 4 out of 7 times, I’m starting to wonder if this contest is cursed. On Counterfeit’s side, the comments are decent enough, but already there’s issues on the scoring side. As it turned out, this was another fairly easy contest; only three entries (plus half if counting Cavern of Pandora’s secret exit path) were deserving of my “Brutal” rating. However, this seems to have resulted in Difficulty basically being a free 19-20 points for nearly all the entries, again kind of showing that Difficulty as a category is mainly just used to punish harder levels. She was also awfully generous with aesthetics, and the only level to not get 5 points for “Following the Rules” from
either judge was Spiny Logging Fort; allowiscous apparently messed up while moving their Map16 tiles around, causing the level to break. Thus, most levels effectively had 35/80 points given to them so long as they weren’t totally incompetent.
As for cstutor…. Look, I don’t like to single out judges while doing these reviews, but I have to be honest: this is the most obtuse judging sheet I’ve ever looked at. All four major categories
are split into several sections, then totaled at the end. (Annoyingly, the total scores for each category were not included.) I’m not opposed to doing this, especially when categories start getting a lot of points, but here it’s absolutely ridiculous. (Separate categories for “Spelling” and “Grammar”? Seriously??) Instead of a tab or column just for comments, they’re all added individual as notes, which is incredibly annoying to read through. And, unsurprisingly, all the comments end up being very minor issues, which (a) generally don’t matter in the broad scope of the level (some even talk about elements
outside of the actual level); and (b) only result in minor penalties for the level anyways. This results in cstutor’s range being 60-77.5 points, which is incredibly small, and allows Counterfeit to have much more influence in the contest rankings. (Ironically, she had the opposite problem in VLDC4, where her range was low while tatanga’s was much higher.)
In the past, I have been lenient about these kinds of judging woes; I do want to respect the original opinions, and I acknowledge that this is a volunteer, hobbyist effort. However, I’m starting to get exasperated with all this; after how well judging went for VLDC6, it’s disappointing to see it stumble once again, seven years into this history. This also wouldn’t be so frustrating if it weren’t for the decision to include two worlds designed specifically for the highest- and lowest-ranking entries in the compilation.
Best World and Worst World
Let’s talk about Worst World first. Even before the results came in, there were
some that suggested to not have a worst world entirely. Recently, the idea of a Worst World has fallen out of favor with users here, and I am inclined to agree that it’s a bad idea. Bad levels can either have a lot of effort put into them or little to no effort to put into them, and it feels wrong to dedicate a world to either category. Levels that have zero effort put into them or are deliberately designed to be placed in a Worst World should be excluded outright, and it’s unfair to mock people that genuinely tried their best on their entry. Even a world designed in a more sympathetic light (like the current VLDC11 collab plans on doing) feels in poor taste; it’s ultimately better to just nix the idea entirely.
Now, maybe you disagree with that argument and think it’s fine to have some schadenfreude in a contest collab. I think even in that case, a Worst World isn’t going to work out most of the time, either. Looking at the bottom eight entries I ranked for this contest, there’s really only one I would consider amusingly bad. The rest are either just blandly designed (or rip-off elements from SMW or other hacks) or have confusing design decisions that don’t work out in the final level. Maybe a Worst World would work for something like VLDC3, but again, people’s skills have gotten better, so it’s rare for me to play a level that’s laughably bad now. And of course, this all depends on the judges’ rankings, so the worst eight levels might not even be that bad. Let’s look at the levels that actually made Worst World for VLDC7:
• Nice Ice Adventure, by Confirm SMW Maker, which was the only level to get a Goal Sphere added to the beginning for being Kaizo. For that reason alone, it shouldn’t even have been included in the collab ROM, but it’s actually not even beatable; some goalpost decoration messes with some line-guides and prevents the player from moving past the second section. (Yet somehow this level got a perfect 5/5 in “Following the Rules” from both judges???)
• Green Muncher land, by natnew, and The Comeback, by Shyguyninja90, which are also in my bottom eight. Fair enough, I suppose.
• Nocturnal Freeze, by Cascade, and Bones – Hot Buss, by BlackMageMario. Again, neither are
that bad, just mediocre.
• Cumulus Crag, by Galactaknight, and Vanilla Death Keep, by Sayuri. These are perfectly fine entries even if they’ve got some issues.
• Twilight Lakeside, by Luaxon, which absolutely doesn’t belong here at all. While it is very long, probably too long, and frustrating to complete, I found it well-designed and visually great. I think it’s worth checking out, at least.
• Finally, there’s Purple Plateau by underway, which technically isn’t in Worst World, but is tied with Cumulus Crag. Like Twilight Lakeside, I’ve got absolutely no idea why it’s ranked so low; it’s a pretty nice entry even if it’s somewhat basic in design.
With all that in mind, it’s really hard for me to justify Worst World in any way. If only a fraction of the levels actually deserves being placed there, and most of them aren’t even interestingly bad to begin with, then what’s the point? This is what frustrates me the most about the judging in this contest; the rankings actually have consequences when it comes to the collab ROM. It’s a permanent presentation of what the best and worst levels of a contest are, so you really have to get it right.
As far as Best World goes, I’m mostly satisfied with what we got. Obviously, I have no qualms about a Best World, since you want to give the spotlight to the best entries of the contest. And of the eight chosen here, they’re mostly solid choices. Blazing Peak by Blue Leaf is a fine choice for first place, and the other picks are either levels that made my top eight (i.e. One Sunday Morning, Touch Fuzzy) or had enough creativity to them that I can understand their high placement (namely SilverSwallow’s Inside Intestines and Milk’s A Study in Fulcrums). The only entry I’m a little iffy on is ToxicRave’s Quicksand Inc. It’s a good desert level but compared to the other entries in Best World (and my personal favorite entries) I found it pretty underwhelming.
There is one glaring omission, however: no Cavern of Pandora. Again, I respect people can have different opinions on levels, but it’s almost comically contentious here. Counterfeit ranked the level first place, with a perfect 80/80, whereas cstutor placed it second-to-last, tied with Green Muncher Land… at 62/80 points. So, it only fell to 17th place, which was still enough to knock it out of Best World. I don’t know if a third judge would’ve brought it higher, but maybe I’d feel a little better about it not being included.
Music and Overworlds
Let’s go back to the collaboration side of things. I’ve already mentioned that custom music was allowed, and from the start of the contest people were working on level remixes the contestants could use. The original tracks, aside from the water and ghost themes, were remixed, and tracks based on different environments were added, like desert or abstract. The overworlds would get original compositions, though a couple were remixes of overworld themes from ASMWCP. While a decent number of levels used their own custom music picks, I’d say a majority utilized these custom level themes. (Also, nearly the entire hack uses custom music; the only exceptions are the title and intro screens, the roulette bonus, a flying Yoshi section, and the various jingles.)
I’ve never been good at music critique, so I’ll keep my thoughts brief. The overworld tracks are overall superb. Each one captures its world very well, from the echo-y Cave theme to the triumphant Best theme. I think my personal favorite would be the Abstract theme; there’s a spaciness to it that I really like, and it really fits with the understated design of the map. The level themes are more mixed, but I’d say most of them are good. Once again, the Abstract theme is a memorable standout, and really energetic; I also enjoyed the Fire and Desert themes, and I think the Castle remix does the best job at capturing the original song while making it sound new. Really the only stinker is the “Here We Go!” remix, but to be fair, it was originally made in 2010 for ASMWCP, and just used here for convenience.
While the music was made on a first-come-first-serve basis, a contest would be held to determine the Overworld, with the winner picked via public voting. After the results came in, MrDeePay played through the levels and
determined all the themes. On the main map would be Grass/Forest, Desert, Mountain, and Ice worlds; submaps would contain Cave, Sky, Fortress, Abstract, and the Best and Worst worlds. There were some quirky placements (Grass/Forest world had a lot of haunted levels for some reason), but overall, I’d say each world felt accurate to its theme. Some worlds also felt rather sparce compared to others, but that seems like it has more to do with the Best/Worst level choices than anything.
The submission period lasted from May 10th to June 15th, roughly five months after the VLDC7 began. The next day,
all seven submissions were presented anonymously. In all honestly, it was a rough selection of maps. Putting entry #5 (the winner) aside, there were individually some good ideas at least. I kind of like Daizo Dee Von’s “layer cake” idea, I like that Aja tried to give each Best/Worst level their own unique island, and Fornaxus’ map is actually pretty solid overall; their Abstract map is a really cool idea, for example. Overall, though, the maps are lacking in detail, which makes some spots look really cramped with level placements, and some really dropped the ball on submap design.
That left Ladida’s map as the stand-out entry. Like I said, they key word is detail; each map has so many little things that help the levels stick out, even when there’s a lot of levels on-screen. The Sky world, for example, is just a simple S-path, yet each level has a unique signifier so that it doesn’t just look like a bunch of dots on flat ground. I also really like that each world is a linear path with warps on each end; having walked through some of the other OW entries, that makes much more of a difference than you might think. Although some of the other overworlds were decent, in every regard this map was the best looking of the bunch. Oh, and the Worst and Best world submaps were a Swastika and a naked lady, respectively.
Aside from a couple petty squabbles, the atmosphere of this contest was fairly calm, so it’s unfortunate that this is the most (in)famous incident of the whole event, aside from the collab hack itself. (Although, I wonder if that has more to do with Ladida also being involved in a
much, much nastier incident, ironically foreshadowed by S.N.N.
in this very thread.) In hindsight, that map shouldn’t have been included with the results, or at the very least had those two sublevels removed. I don’t think people (my past self included!) really appreciated the fact that other people besides SMWC users would want to play this collab hack and be uncomfortable with these two submaps, or that maybe the Worst/Best level creators wouldn’t want their levels placed on a Swastika or naked woman. Unsurprisingly, a lot of people ended up voting for Fornaxus’ (#3) and Prizm’s (#7) entries instead, even if ultimately Ladida’s map got the most votes.
And yet, I have to admit, I am a little glad that it ended up as an option anyways. Like I said, the rest of the map is still incredibly good, and I’d almost go as far to say to accept no substitutes. Fornaxus’ map could’ve at least been tweaked into something special, I suppose, but I was really surprised at how many people latched on to Prizm’s map; it’s the “safe” option, sure, but it’s really lacking in interesting decorations, especially for the submaps. (I honestly wasn’t able to tell which ones were the Best or Worst worlds without picking up on the Star warps.) In the end, Ladida’s map won out, and the Best and Worst maps were replaced with much more tasteful renditions; the Swastika was replaced with a Muncher, a long-running site in-joke, while the naked lady was removed in favor of a Mario-themed
Hinamatsuri display. It’s hard to tell if it was worth all the trouble, but nevertheless, VLDC7 had a great Overworld at its disposal now.
Other Collab Notes
Before I wrap this up, I should talk about how the compilation ROM itself works, as I think there’s a lot of interesting quirks worth mentioning. First off, all levels are available from the start, including the Best and Worst worlds. When a level is completed, an “M” marker appears on the top right corner; either exit can be done. (For those wondering, the Red or Blue color of the M means nothing; it just changes to red if the player completes the level as Fire Mario.) Yoshi Coins are not tracked, which is fine by me, but more annoying is the fact that levels with secret exits are not properly marked; I don’t normally check the levels I play in Lunar Magic, so if it doesn’t look like there’s a secret exit, I usually don’t go searching for it.
Another change involves Yoshi and power-ups: these are wiped after a player completes a level. This keeps the feeling of starting every entry fresh, although it does mean you always starts as Small Mario from the Midway entrance as well; you also can no longer re-enter a level with power-ups. The live system is the same as the original game, though progress is saved after every level. (Technically, it’s saved after walking to different level.) It’s also worth mentioning that the 2-player option is not available for this hack, if that’s something you’re interested in.
There is a small amount of post-game content worth mentioning. The hack download comes with a password-protected ZIP file, and near the warp zone is a Yoshi House that wants you to come back after completing all the levels. An anti-cheat mechanism breaks the hack (in rather spectacular fashion, I might add) if any of the levels are altered in any way; I learned this the hard way when trying to add a Goal Sphere to LHB’s level, Infernal Athletic, as I found it too difficult to complete without tools, and I wanted to get 100% on the save on my console. In any case, all this leads the player to a site like where they enter the Hall of Fame. The original link is dead, and based on this
backup here (thanks Tahixham!), it looks like people stopped getting added after #115 on November 15, 2014. Admittedly, I am mostly glossing over this since I didn’t have a chance to actually access all this, but from what I’ve heard, this is basically all there is to the post-game content.
One last thing to note: aside from Nice Ice Adventure, there is one other level that isn’t fully available from the collab ROM. Frost’s Hostel, one of the most unique (and one of the most frustrating) entries, has only about a quarter of its level available to play, I believe due to a level exit error in the MWL file (based on
this post anyway). I really wish this got fixed for the collab’s release, or even as a later update; it honestly looks like it got left as-is because people said the level was too hard to extend even further. (Which is maybe true, but, considering Spiny Logging Fort’s level was fixed for the collab, I don’t see why we couldn’t fix Hostel.) I am mentioning this only to say that, if you haven’t yet, it’s worth checking out the full Hostel; as challenging as it is, it’s definitely one of the more memorable experiences of this contest.
C3 Release and Final Thoughts
On July 11, 2014 – the opening day of the Summer C3 – six and a half months after VLDC7 began,
the collaboration hack was released. While that’s a long period of time for one of these contests to be active, compared to the development period of ASMWCP1 (let alone its sequel) it’s amazing how quickly it was released. Even though it’s not a perfect compilation, I’d still take it a million times over having to sift through over 80 different patches and having to look at the same title screens and intros over and over.
As I am currently writing this, there have been twelve VLDCs total; thus, this marks the beginning of the second half of this series. Fittingly, VLDC7 feels like the beginning of a new era. I could really feel the energy and excitement from everybody participating in the contest and contributing to the collab. Sure, there were problematic elements like the judging or the OW contest, but those debacles are seriously outweighed by the genuine camaraderie found everywhere else in the project’s development. In a way, it reminds me of VLDC1, where this new idea (or at least new for SMWC) generates a creative spark in people that brings out some really great levels, great music, and great landscapes. It’s a landmark for the community for sure, and I can’t wait to see what lies in the future.