I think what we are seeing in the recent decade is a real codification of these concepts. Fez is another example where the twist in the lock-and-key formula is that the key is knowledge. The more we’ve sat around with these concepts, the more games have been molded to meet our expectations on pacing. I think the reason people in this thread are drawn to earlier games is that those games were designed when there wasn’t that expectation. Early games had to teach players how to play them. Now we’ve reached a point of literacy where that’s taken for granted.
When I first played Riven, I spent my first two hours feeling disoriented because it didn’t match my expectations for pacing. There were no gauntlets of small tests like there are in the Witness; however, there were locks that could only be opened once I had access to special knowledge. There’s an important difference where learning in the Witness is methodically contrived, whereas learning in Riven feels spontaneous and organic. There’s still a guiding hand in Riven; the designers want you to solve the game. But as a player, you cannot guess how you’re going to learn. It surprises you and begs you to be more perceptive
The bit i find interesting to the extreme in The Witness is that it’s both, in that it’s got that super formal net of puzzles designed to teach you the puzzle language in a very controlled manner, but it’s actually a setup for teaching you the mindset that’ll allow you to organically discover the other thing which is completely unstructured and nonlinear.
I really want to play Rhem one of these days, which from what I understand is much more of a Riven clone than a Myst clone mechanically.
Yeah. What you’re saying reminds me of the failure of high modernism. It’s popular nowadays to note that cities started as a warren of evolved human relationships, then were gradually forced into a more and more formal frame so that state goals and policies could be imposed (analogous to game designer goals about exploration pacing, for instance). This trend is understood nowadays to have sometimes gone too far and destroyed priceless organic systems instead of improving them.
Well, I wish there was a game that was an allegory for modernism’s failure, that would be extremely cool. But The Witness is something more like the reverse: nostalgia for when modernism and positivism stood triumphant. The allegory Blow seemed to have in mind, from the sciency talk on the audiologs, is of the paradigm shift from classical to modern math and physics. Total mastery at the old way of formal seeing, leads to insight into a new way of formal seeing that is compatible with it, extends it, and also replaces it. The love and wonder here is at the informal and organic human mind’s capacity to mold itself to understand a world fully explained by rules.
That’s why “the other thing” sooner or later turns into an extension of the main thing. You can be a completionist with it because it’s mapped by the obelisks, for example.
i think i tried one of these, briefly. bafflingly huge and complex, with little story to be observed, at least upfront. felt like first person exploring someone’s Factorio setup
I love this thread because of the thoughts that it is creating, but I’d also reject the notion that adventure games are “Metroidvania adjacent”, though they clearly have things to learn from each other. The key difference: Search Action games require action! It is a key feature of SM and SotN that you are fighting enemies nearly constantly. When you completely absent this requirement, you get a recognizably different genre of game.
So like, Elliot Quest and NES Strider and Milon’s and the other games named early really fit the OP’s bill, but Mystlikes are too much of a jump away to qualify I think. Outer Wilds, with its synchronous and slightly skill-based traversal mechanics, is a really useful and interesting edge case.
I was looking for Switch games to buy when I came across the fact that Toki Tori 2(+) is something that exists. I was a fan of the original GBC game, and I was surprised to see the sequel explicitly billed as a Metroidvania in the promotional material. Having played the game, I can see why, even though I’m not sure I’d agree. I do like its sense of progression, though–it’s more satisfying than that of classic Metroidvanias–and while I’m enjoying it loads, I’m left wishing someone would apply its ethos to a purer platformer. Timespinner, which I also bought recently, could have been that game if it had wanted to, but it’s too beholden to the Metroidvania formula to explore its time stop mechanic as much as it should have.
Toki Tori 2 is also a canonical knowledge-troid-vania, and perhaps the purest distillation of the concept. Stomping and chirping are the only things you can do, and the game is about learning how those actions can be applied. A sufficiently observant player can circumvent the naturalistic tutorializing the game funnels you down for the first act, and I think that’s just wonderful.
Lizard is also a fairly interesting case study. The topography of the world is fairly open, but you’re limited to having only one power-up at a time. This, combined with the unusual topology of the world, helps mitigate the feeling of physical mastery these games tend to celebrate — obstacles are always obstacles, and threats are always threats, but the difference is that your knowledge of how to deal with or work around these things increases. Navigation itself is thus a crucial element of the game’s design, which in turn engenders both a healthy curiosity to poke and prod at the edges of the known world and a healthy doubt of whether or not you fully understand the world.
Megaman games are also a fairly interesting point of reference. In my opinion, the best Megaman (imo: 2, X1, X2) games provide a web of interdependencies between stages and bosses, and not just a rote weakness chain — something simple enough to be intuitively reasoned with, but sophisticated enough to allow for interesting tradeoffs and novel approaches. Thus, in a Metrockman game the questions you would ask youself when playing would be less “when am I going to get a powerup to get to the ledge?” on more “hmm… if I do this other area first I could reach this ledge here, but…” I think it’s a very (potentially) fruitful design space, and one that I would love to see explored outside the context of relatively inaccessible hack and fangames and such.
I think the reason I bring those games up is that search action games encourage a similar relationship with the game world. They ask the player to be perceptive and insightful, to internalize the rules of the environment so they can thrive in it. Mystlikes are “something else entirely” but I see a kindred desire to develop from mystery to epiphany.
Another thing it does quite well is de-emphasize the player as the only source of stimulus in the game. Every animal has a pattern that exists independently of the bird; the bird can be harmed, but animals are not out to hurt it. It’s not something you can have in Castlevania (at least not to a large degree) but it is something I’d like to see more in Metroid.
Thinking about it, the Toki Tori approach would make a really good template for an excellent Metroid II remake, given that game’s setting and story.
Additionally, a more dynamically biological SR-388 would then contrast nicely with Metroid Fusion’s replication of that planet’s environment, since in that game the only animals that remain are X-parasite clones who would have every reason to act like traditional vidcon baddies.
I wonder whether Metroid 2 should be considered more on the quiet adventure side than the search action side? At one point I played NES Metroid and Metroid 2 back-to-back, and the immediately obvious difference is that NES Metroid is an experience of being swarmed from all directions, and Metroid 2 just has an occasional solo enemy every three screens. And this ties in to the contemplative lonely qualities much-praised on IC/SB in the past.
Resident Evil. (hear that in RE title screen guy voice)
it’s kind of a search action but what’s different is that it’s on a real small scale. you find keys, but those keys are to doors on the other side of a mansion or museum turned cop shop, not to like a whole new area; you only go to new biomes to progress the story, and they get smaller as the story gets more action-packed. You’re also gated by firepower to beat harder enemies (bosses are basically an ammo check) and knowledge (knowing efficient paths between save points, where you should kill zombies and where you can just feint around em). Classic RE games kinda flow the same way as a search action game but all the principles are different.
i like thinking about other mundane buildings with secret labs hidden underneath a resident evil game could take place in. A library, an old train station, a sewage treatment plant, the sky’s the limit
It’s not just the quantity of enemies at work, it’s also the behavior of the enemies themselves. Most enemies in Metroid 2 follow set looping paths, or shoot at predictable intervals (or sometimes both). The few enemies that follow more complex patterns, such as metroids, pipe bugs, or hoppers are almost always found alone, not paired with other threats. Contrast this with the dynamic movements of Metroid 1’s wavers and swoopers and sidehoppers and skrees and fireballs and rinkas (and the dense combinations they’re put in) and the difference in design philosophy becomes readily apparent. In one game you are under constant attack from alien intelligence, and in the other you are intruding upon a natural habitat.
The Umihara Kawase games have a similar structure, specifically the first 2, but especially Shun. There’s only a few exits that will send you to an earlier stage iirc, most of the alternate exits are shortcuts or lateral movements to other paths. In the first game there is an overall playthrough timer of 30 minutes that will cause all doors to send you to the “bad” ending field.
There are no upgrades or resources outside of lives and continues, but you do collect fields to play in practice mode by visiting them with the allotted credits in a normal playthrough.
Umihara Kawase Fresh is the metroidvania in the series, but it was very disappointing imo.
Genpei Toma Den is also similar, but I haven’t played it myself.
It might also be worth mentioning smb2j, there were at least a few warp zones that sent you back. You collect 8 clears to play worlds A-D and there’s also a world 9 if you don’t use warp zones.
I like that Cave Story starts with an ambiguous structure that seems like it’s going to be a straightforward hub-and-spokes, only to defy that expectation after the third spoke never sends you back to the hub–and then unexpectedly it does, but the spokes have changed, and it resists easy categorization.
What I dislike most about open spaced games is the lack of uncovering what feel like meaningful coincidences or significance unfolding from marginal decisions. Anything unintentional comes off as sterile.
I guess I sort of just rearticulated what was already posted, but I had been thinking about it a bit and I think a lot of what I like about roguelikes are what I find lacking in “open world” & metroidvania games
I think my main problem with modern indie metroidvanias is that after all this time, they tend to seem more rote and conservative than Cave Story. It’s like despite being so seminal, it’s filled with design choices and ideas that no one’s bothered to build on?
I’ve always found the progression structure of Cave Story very unpleasant on a gut level. And the reason is pretty much those details that it does different from everyone else.
Indie games are generally trying to iterate towards something that goes down smoother, feels like it rewards you for time spent and does not ever feel unfair and arbitrary. The unfortunate reality is they need their Steam reviews to be higher than “Mixed” or they won’t be able to pay their rent. I’m not sure design choices like the recent-performance-scaled gun power level, for example, are compatible with that: that mechanic inherently amplifies streaks of success or failure instead of equilibrating them, and denies the possibility of learning a consistent approach to a level that’s resilient to one or two mistakes.
To extend this line of thought, maybe the only real successor to Cave Story was Spelunky. Which has also had many of its choices ignored by all succeeding roguelites, and for the same reasons. I love Spelunky, but I also remember moments when I hated it from the bottom of my heart. And there was not a hint of appreciation in this hate, it did not enrich my life in any way, however I accept it as the unavoidable price that had to be paid for the Spelunky moments that did enrich my life.