#boneworks

As much as that, and Corridor Digital's increased doubling down on AI, has heavily contributed to my feelings (they're the same group of folks), I have a lot of problems with how SLZ approaches game design, combined with how they talk about themselves. The entire marketing for both Boneworks and Bonelab essentially consisted of SLZ saying "We have made the Best VR Engine Ever," meanwhile all they have really done is create an extremely limiting sandbox environment that doesn't meaningfully translate to something actually useful for game design.

Their design philosophy when it comes to VR is to make the simulation aspect as accurate as possible, so that (in their eyes), the player's intuition on what SHOULD happen DOES happen. Unfortunately though, in practice, it doesn't work that way at all. This idea of simulating realistic physics will just never work no matter how many hours they pour into it, because there will always be a fundamental disconnect between what the player is doing in-game, and what they are experiencing in real life. It's frustrating for the player to try and swing something like a hammer, but their hands move too quickly in real life, or their IK rig's arm gets caught on their shoulder, so the physics bug out because their player "shouldn't" be able to do what they are already doing in real life. What they're trying to do is tell the player what SHOULD be happening instead of creating systems that can intuit what the player is trying to do and just doing that. It leads to a game that feels like wading through molasses to play. It's the same idea as to why people hated motion controls during the Wii era, not because of the act of physically moving, but because the simulation creates an inconsistency where there could've otherwise been consistency. Take shooting, for example. The body's IK rig affects the position of the player's gun, so placing your IRL arms in the exact same position two separate times could lead to you aiming at two completely different points. It leads to me missing shots in this game constantly whereas I'm a pretty decent shot in other VR games. You might hear me reference "The Best Guns In VR" from time to time, which is a reference to some of the videos they did teasing Boneworks up to its release, which I bring up in jest because I think the guns in these games feel really bad and awkward to use, which is funny when H3VR had been there for 3 entire years when Boneworks came out (Not to say H3VR has the best guns in VR either, but in terms of the simulatory aspect Boneworks was shooting for, H3VR beats them out by a country mile).

Beyond the issues of player interaction, their insistence on making EVERYTHING physics driven also leads to a really limiting design space. There are technically something like 14 enemies in Boneworks, but in practice there are three. There are the crablets, which are essentially headcrabs. They work pretty good, and are probably the best use-case for physics-based enemies in the game. Then there are the nullbodies, which are essentially just human shaped physics objects being puppeted purely through physics impulses, essentially like marionettes on strings. Then there are the Omniprojectors, which are the guys that roll around on balls and shoot at you. Every other enemy is just a variation on these, with the vast majority of them being either Nullbodies with different models, or Nullbodies that occasionally thwip an energy ball at you. Bonelab isn't much better, only adding some more Nullbody clones and a fireball turret enemy. It's not that SLZ couldn't add more enemies, but they're EXTREMELY limited on what they possibly COULD add due to their hardcore commitment to purely physics-driven NPCs. Their reasoning for this makes sense on paper, they want the player to be able to physically interact with the NPCs, but their implementation is frankly really, really bad. There are examples of how they could do this right in other games, for example GTA which does a really masterful job of seamlessly combining physics-driven ragdolls with their otherwise traditionally animated characters. Obviously SLZ isn't Rockstar, they aren't going to match up to what GTA can do, but they're strangely committed to doing things in their roundabout and limiting way. The Omniprojectors are the exception to all this, because doing an enemy with a gun using their marionette style approach really doesn't work very well, so they created these hologram enemies that shoot at you instead, being controlled with a much simpler and less realistic physics rig. The omniprojectors are... fine? But they're also a complete departure from the other enemies in the game, and it just serves as a really clear example of why their system that they're so committed to doesn't work very well when applied more broadly.

Besides their over-commitment to physics, SLZ is also just. Not good at level design. I'm sorry to speak so bluntly, but I really just truly believe this. Bonelab is the obvious example, but Boneworks is often just as bad. Their levels lack any signposting whatsoever and it's easy to get lost in them, and they constantly break a lot of classic game design principles like "Gamers don't look up." I make a joke during this stream "If this were a Valve game, there would be a look-at trigger for that [setpiece] to happen," which was just a bit, but it really does speak to the massive difference between SLZ and Valve when it comes to guiding the player through their levels and setpieces. Again, SLZ isn't Valve-sized, they don't need to match their quality, but it comes back to the marketing issue. Brandon J Laatsch was very clear on Twitter that he wanted Bonelab to be the "Source engine of VR," echoing the modding revolution that happened during the Source engine's hey-day. They have called Boneworks a half-like game many times, and aren't shy in referencing it. Bonelab even has one line that grinds my gears where a voiceover says over a radio "It all started... with a Valve." The problem is that, despite their boasting, they are nothing like Valve. They share almost zero design principles with Valve, and don't seem to understand what it is that they're even imitating. They spend so much time talking up this engine that they've made acting like it is God's Gift To VR, acting like this physics-based approach grants developers a kind of freedom to make any game they want, but in reality it's more limiting than ever.

At the end of the day, I truly just don't think the people at SLZ are very good at game design. I think they're great software engineers, don't get me wrong, and I don't want to overlook the amount of work that went into these games. The rig for Bonelab is legitimately impressively designed. The issue is that the things they're shooting for just don't work in practice, or at least not with how they've executed it, which wouldn't be as much of a problem if they didn't talk about their work the way that they do. Just as a brief list of a few games that came out before Boneworks that do things they they do better... - Hotdogs, Horseshoes, and Hand Grenades - 2016 (significantly more realistically simulated guns in almost every aspect) - Blade & Sorcery - 2018 (similar physics-driven NPC system, but the NPCs are more reactive and functional than the Nullbodies of Boneworks. Still shares a lot of design limitations but frequently executes them much better) - Robo Recall - 2017 (Enemies that are far more industry-standard with how they function and are animated, but also allow for plenty physical interaction such as physically tearing them apart. Admittedly does not have melee combat)