dericbindel - Tumblr blog

dericbindel · 11 minutes ago

Text

BMOP: A history of the Abandoned Koro

Now that things have all come out, and the project is well and truly over, I thought I'd take a moment to talk a bit about something I worked on in the game, the "Abandoned Koro" area, now that there's video footage of a playthrough I can easily reference.

I suppose as a quick primer, I was with the BMOP team from around March of 2021 to November of 2022, serving as both 3D art lead and Level design lead for a time. There's a post out there about my experiences with the leadership of the team that I mostly stand by still, I do feel bad for how the project ended, and I have come around a bit to remember the positive times more than the negative, but that doesn't erase what happened.

So to begin, when I came on level design was a fairly inactive sub channel of the game development discord, some ideas got tossed around occasionally, but not much seemed to be happening, before I'd joined there was a rough map drawn up by the then current level design lead

It was very basic stuff, there was a version of this map in engine that had been extruded a bit to give it make it a kind of playable space, one area had some trees on it and some random platforms.

One area that interested me greatly was this part in the south, the so called "Abandoned Koro"

The rough idea being that it was the Le-Koro seen in the GBA game Bionicle: Quest for the Toa, that had been abandoned in favour of the one seen in the browser game Mata Nui Online Game and the cancelled PC game Bionicle: Legend of Mata Nui, as the one in QFTT had a lot of wooden structures while the one seen in the other two was more like a woven nest.

At the time I was just a 3d modeller, and had never really done any real level design, but the concept just really inspired me, so I went off and in blender sketched out a possible level layout for this area.

I made a bunch of renders and wrote out how I thought the level could work, which can be seen in this document HERE.

For a quick overview, the idea was you'd enter the area, the bridge would break, you'd have to clear a river to activate the water wheel on a large mechanical tree that would allow you passage upwards, and then navigate through the Nui-Rama infested ruins of the village before coming to a cave.

The tree was something I poured a lot of time in to, its worthy of its own post at some point in the future.

Now I'll be honest, I was super nervous about releasing this document to the team, I thought I was overstepping my bounds. I was just a relatively new 3d modeller on the team, I didn't have the right to be talking about level design, but to my surprise people were really receptive to it, and I very quickly got added to the level design team¹.

I ended up doing a rough blockout in engine. This was back in the time before Lewa had been chosen to be the main player character, so it was still Tahu.

youtube

This was the first time I'd ever touched Unreal, so it took a while. That's why the sky is black, I didn't know how to add a skybox.

I never got in to level scripting, but as a proof of concept, I think it was successful.

You can see that in this version of the level had already progressed, now there was the idea of using the Pakari to break the dam, so finding it was the first task on the ground.

This version wasn't in the actual demo map, because as I said before the demo wasn't really very fleshed out at the time. One thing my map showed though was that the demo was way too big. The level design lead at the time had decided the size based on how long it took to get around the map as a perfectly flat plane with no terrain, at max run speed, with no obstacles. This is a very flawed methodology to say the least. Having an actual level with things scaled to the character really started to show the holes in this, and eventually, once I became lead, the demo area was massively shrunk.

Here's a version from just a month later. The version is now part of the new, more compact demo area, but its mostly barren at this point.

youtube

This version was actually completely rebuilt from the ground up because it had been decided to try out voxels as a basis². They were a pain to deal with and this was overturned eventually. I'd say in total, including the 3d sketch I did, I re-built this map maybe 4 times.

One benefit of voxels was it was trivial to have caves, so this is where the idea of the pakari being found in a cave started to emerge.

As I continued to play and refine the demo it became clear how much the movement systems would need to be tweaked. There seemed to be this idea in the team that everything could be done separately and bolted together later, so character moment was all handled by someone jumping around on some big blocks in a test map whereas level design was off doing their own thing, but it really wasn't working. So while the movement felt good in a vacuum, actually putting it in context really exposed a lot of issues. You could easily jump over any enemy you could come across, as said earlier maps had to be huge in order to make the world seem big while you were running at full speed, and after about 3 jumps you'd be 20 metres in the air.

At this point the idea still was to have the entire island of Mata Nui be one large open area you could freely explore, so the sheer size of the maps resulting from the over powered movement was a major issue. I lobbied very hard to have things pulled back and eventually they were.

its a bit out of order, but here's the demo area from closer to when I left. Its a fraction of the size of the original map

Here's another video from a month later.

youtube

As much as it pains me to say it, this is basically it in terms of meaningful development. By this point Lewa had been locked in as the playable character, and the movement had been dialed back to a more reasonable degree, so I was able to really start trying to refine the area. This is where I ran in to a fatal issue.

No one else wanted to play the game.

And what was worse barely anything worked.

The level is still in what I would call a grey box state. I used a couple tree trunk assets I'd made here and there, and put in a big canopy asset from one of the many asset packs we had for a bit of ambience, but it was all still very basic geometry, easily changed or modified.

Unfortunately, there's only so much you can do for playtesting your own area. You built it: you know where everything is, what's supposed to happen, where you're supposed to go. You can try pretend to play it as a new player, but that only gets you so far. I was hoping people on the team would play it and provide feedback³, but outside of maybe one or two people a handful of times⁴, trying to get any really feedback was was a futile effort.

The thing was that this area was very complex. If the demo was a vertical slice of the game as a whole, this area was itself a microcosm of the demo⁵. It had platforming, puzzle solving, combat, mask powers, the lot. Now unfortunately for me, barely any of those systems actually worked.

As an example, this long ledge was for a long time a stand in for vine swinging, then rail grinding⁶, then was ultimately just replaced by a platform.

So things had kind of hit a wall. I couldn't properly design areas with combat in them until combat existed in a more stable state, I couldn't design platforming sections until platforming worked in a consistent way. I couldn't even adjust the overall flow of the level because everyone else basically refused to give feedback.

But the unfortunate thing is, in August of that year they'd released a teaser trailer.

And that trailer had gotten hundreds of thousands of views.

This is where the whole development of this game really went off the rails. Now there was this push to get things in a presentable state, start set dressing and making final assets so things could be shown off.

But I refused. Everything was too up in the air to commit to set dressing, this is why block out exists, if the jump changes height its no big issue to grab the couple cubes an area is made out of and shift them up or down, if combat is found to need more space its easy to make things bigger, or add or take away walls. If something is confusing things can be shifted. Once set dressing starts now you're dealing with dozens to hundreds of objects being scattered about, even the smallest tweak can lead to a mess.

Not to mention set dressing raised its own series of issues, from plants triggering the IK on the toa's feet, making their knees go up to their chin when walking through a bunch of ferns, to collision volumes being oversized or offset, meaning that big rock face they just added has now created a massive invisible wall in another area. Once the addition of some plants caused all ledges within a wide radius to no longer work⁷.

It was a miserable state of affairs. My mental state rapidly deteriorated as I fought against this, I became very short tempered and irritable, and eventually near the end of 2022 I was kicked from the team. It was such a relief honestly.

I think the tragic thing is, set dressing is actually quite a fast process. The starting area went from looking like this to something quite like the final in about a month?

But you can't show off the first screenshot on twitter⁸.

I guess I'll spare a quick moment to talk about the final version seen in the video.

youtube

Its at 46 minutes in if the link doesn't work

Its...fine? The assets they've made are all good, though I think they lost the QFTM inspiration along the way. The thing that stands out to me is just how...little its moved on from 2022. Just look how much progress was made in like 3 months, compare that to now, 3 years later. Temporary platforms I placed are still in the exact same spot.

One part I found quite amusing was the final enemy encounter. It was supposed to be this large hut, probably Matau's, that had been completely overtaken by a hive, and Nui-Rama would spawn constantly from it until the encounter was over.

There were a lot of concepts drawn up, but I guess it just...never happened. I also don't know why there's waves of fikou there either, that was after my time I think.

So that's really all I have to say about that version, it looks okay. Its still clearly unfinished. It kinda works more than it did when I left, but its shockingly not that much different, outside of some new assets.

I'll be honest, I did intend this to be more of a happy, reflective post, but when watching part of the developer commentary a quote stood out to me.

"Traditionally you would kind of have really basic blockout for an environment and play test that to see how the level design is working. Because of the the situation we had we were kind of forced to just go ahead with set dressing and prettying it up and everything so if it were in a perfect world I would have loved to have gotten to do some more play testing early on but we did what we could with it. And I mean you basically had to just blindly trust the process because many systems weren't working at the time things were designed." -AN UNKNOWN BMOP DEVELOPER (2025)

And I just fundamentally disagree. This was a fan project, there were no deadlines but those that were self inflicted. This process they blindly trusted just lead to a thing that on the surface looks okay, but is still riddled with bugs that were well known for years. I've seen some people on the team say the game was 90% completed, feature locked, just 3 more months of polish and it would have all been working, but from what I've seen of the game I really doubt it. They say themselves in the commentary that there's bugs they've been fighting for nearly a decade still rearing their head.

So yeah, its not exactly a happy tale. I'm quite proud of the work I did, I learned a lot, and met a lot of people I'm still friends with to this day. Its a shame the project had to end this way, and I'm sad everyone's work has gone to waste, but I'm also not going to pretend this was some amazing project that was struck down right before achieving greatness.

But most of all: Fuck you lego.

If you're interested in seeing a few more of the things I worked on in my time with BMOP, go HERE. I may post more publicly about some of them in the future, who knows. I have a post about the regional Kini temple I designed and built for the game HERE. Also some renders I did for some of my game design reports HERE.

¹-I say team, it was maybe two other people who were barely around. ²-I'll be honest here: Some people wanted to have the entire map destructible so that they could have Bohrok dynamically destroy it. For the post game DLC after we'd finished this Zelda sized free fan game⁹. ³-💯👍👀 isn't helpful feedback at the end of the day. Its supportive I'll give you that. But sometimes you need more. ⁴-One person on the team even outright refused to play the game until final release, to "save their first reactions for their stream"... ⁵-idk if its clear at this point but the "demo" is, was, and now always will be the entirety of the existing game. Oh there were ideas for other things, but nothing concrete ever materialised*. ⁶-You know, like Sonic⁹ ⁷-The ledge's over sensitive detection for something blocking it is one of the most frustrating things about the game's development for me. ⁸-You absolutely can show off real development stuff, there are lots of people who find that fascinating. ⁹-This game was nothing if not ambitious.

You see that kind of mop is called a "Bee Mop", because of its sponge's resemblance to honey comb.

7 notes · View notes