Games

BotW was my favorite gaming experience of all time. Hours and hours of joy for me.

There's one thing they did right that most other open world games do wrong: The map starts blank and it fills in as you explore. Others in the genre, you'll Ubisoft that tower and then it fills in with all the icons of things to do here, so now follow the minimap to them all. In BotW, you Ubisoft the tower, you get the topology map, and now it's up to you to find stuff in it, and when you do you get an icon on the map telling you you've done it.

I'm pretty sure the Rito quest wasn't complete in time for launch so they had to rush to throw something together. The Hebra region is distinctly empty, I'm sure we were going to have to go on an arctic adventure to find Teba's favorite cuttlefish bone or something, but they didn't have time to finish it because the Switch was coming out so they said 'Fuck it, build something that runs and ship it.'

There isn't much variety in the enemy types, a lot of encounters are monotonous, a lot of the systems are so basic that they're easy to break, and they were so afraid of telling a story out of order that the game doesn't have a story of its own; "Link fucks around all over Hyrule for awhile then decides to defeat Ganon." Meanwhile it tells you a different, somewhat related story.

Then there's TotK, which they tried to make a sequel to BotW out of BotW's bones, and it didn't work as well.

That's awesome to hear. I'll have to check it out

Fitzsimmons actually got it even better than my kind of brainless appeal to nostalgia. That's what I want

Some of the games I like the most -- some roguelikes fit into this camp -- had very small teams working for a long period of time on a game, spending a long time iteratively refining the gameplay.

I'm not so sure that I prefer the "wide" model of many people for a short period of time versus a few people for a long period of time. Certainly there are things that the "narrow" model can and has done well that the "wide" model hasn't.

I don't know if I can reduce it to the programmers making the game, though. Like...yeah, the most-replayable games I've played rely heavily on the code, and less on the assets.

But there are also games where the art is pretty critical that I enjoy. Like, imagine games in the Myst series without the art and sound. Like, none of the code is particularly impressive. The puzzles are...okay, I guess. You play a game like that for the art and sound.

Or Lumines. I mean, yeah, they had to get the gameplay loop right, but technically, it's a very simple game, just a falling-blocks game. But the audio and to a lesser degree, the appearance, is important.

https://www.youtube.com/watch?v=bboWlUppp-s

Rez is maybe a little fancier codewise, but it's just a rail shooter. It's really about experiencing the art and music.

https://www.youtube.com/watch?v=gZLHB5e90pU

And I definitely did enjoy those. In the case of the latter two, those were not AAA games, didn't have huge teams creating assets, but it was still really the assets that made the game.

I agree that generally the two are linked, but to nitpick, there are some roguelikes -- games that I would call roguelikes, at any rate -- where the world is to greater or lesser degree hand-crafted.

ToME 2, as I recall, did have a procedurally-generated underworld, but a more-or-less hand-crafted overworld. Been a long time since I've played it, though.

Caves of Qud has a most of the underworld being procedurally-generated, and has many individual overworld maps being procedurally-generated, but the overworld map tiles and many individual overworld maps and some underworld maps are static.

But my point is more just that I don't think that it's right to say something like "procedural generation of worlds doesn't work". I mean, there are games where I think that procedural generation of worlds works, and works well. It's just that we haven't hit that Holy Grail where humans don't have to handcraft worlds for games that rely on handcrafted worlds any more, where the procedural generation engine can just make as much handcrafted-like world as one wants for no cost.

I think that for roguelikes, what makes it work is that rely on creating different, well, tactical scenarios via randomization. Like, you have to play differently based on the environment you're in, and so random generation gives you a stream of unknown environments so that each session is mixed up.

For many traditional games, where the point is exploration or story...we don't have procedural generation that can make interesting plot and characters yet. Maybe we could make aesthetically-pretty procedural worlds, but we don't have software that can generate stories with characters and plot events that we emotionally care about even remotely as well as humans can. Maybe LLMs or something are a tool that can help us get there, but we're not there in 2024.

I don't know what exactly makes procedural generation work for Minecraft. I guess it affects how one plays. It's more-or-less irrelevant from a tactical standpoint, the combat doesn't change up much, but it affects what resources one has, what problems present themselves, what constraints are imposed on how one has to build and act. Maybe games like Dwarf Fortress and Oxygen Not Included and to some extent Cataclysm: Dark Days Ahead, all games with procedurally-generated worlds, would fall into that camp. Starbound. Terraria.