The Archipelago train continues, this time with me rating all of the APWorlds I have experience with, because I’ve wanted to do that in at least recent times, especially as the quality of them sees massive variance due to how community-driven the project is. I’ve been having so much fun with Archipelago in general, but with some recent snags, I wanted to relay my experiences forward to anyone considering a game I might have experience in for the next one; plus, this gives me an excuse to write about something, which is always welcome.

I’ll be rating APWorlds in three categories, as well as giving an Overall rating: Gameplay covers how the game feels within the context of a multiworld, Logic covers how games interact with the randomizer logic at large and affect other slots as a result, and Customization covers how personalizable a slot can be for the same game. I’ll try to list these in alphabetical order and I’ll put a disclaimer as to what I’ve actually played of the game.

If genAI was used in the development of an APWorld, that’s just a sign to try to stray away from it in the future.

Autopelago#

I haven’t goaled this game yet, but I’ve played a lot of it and seen a lot of it.

Gameplay: 0/10#

It’s a slow release simulator. What do you expect? There’s a reason I ban it from my lobbies; it’s trivial, even if not as trivial as Clique, and it’s not fun to have to sit behind while a rat slowly chugs along the line. This isn’t an experience I would recommend to anyone.

Logic: 3/10#

If you want to treat this like an actual game, the logic can be very annoying; there are a million blockers despite the low sphere count, and it can be very annoying to get anything done, especially in larger lobbies. This game is perfect for simulating what a lobby could look like, and especially what a heavy blocked/blocker game could end up playing like with your game of choice, but it’s not a great choice for many.

Customization: 8/10#

Autopelago was made for APWorld testing, and it shows with how much goes on in it for just being a slow release simulator: Custom chat messages, filtering of buffs/debuffs, multiple goals, and lots of fun little options to make this game a better fit for your lobby.

Overall: 2/10#

Avoid without purpose.

Celeste (Open World)#

I’ve goaled this game before playing to Summit A in a sync that included B-Sides on an older version before the more recent updates. I’ve been playing in an async in a slot made for the most recent update and I have experience with the PopTracker as well for both.

I generally just call this “Celeste” and the other APWorld for this game “Celeste Legacy”, as this is often considered the de facto implementation for the game.

Gameplay: 9/10#

This game plays like it was made for Archipelago. Each check feels rewarding, there are long stretches of good gameplay (especially later into the run; one of the issues I’d have with this is how slow progression can come near the start, but I think that’s on my progression balancing value a bit), and the game tasks you with creative ways of going out of logic (or even clearing checks in logic, potentially, with the new logic setting). The PopTracker makes what is and isn’t possible (with and without assists) expressly clear, as well, which is incredibly handy, basically serving as a better Berry Camp. I love playing this implementation every time I get the chance to; it’s so much fun and adds a surprising amount to the game. The Archipelago mod is also very compatible with utility mods, which is quite convenient.

I think a fair complaint is that levels like 6A and 6B, as well as 8A and all of the B-sides, don’t feel as rewarding as the other levels with a much larger pool of strawberries. I imagine running roomsanity would even things out a bit, and it has been something I’ve found really fun in the Big Async slot I have, but it’s definitely unfortunate just how much of a disparity there can be here.

Logic: 6/10#

I’ve got mixed feelings on the logic side of thing: Each check sent feels rewarding, but due to the check-item imbalance that can exist at times, especially at lower strawberry counts, multiworlds can be flooded with useless filler a lot of the time, especially if you’re running some of the extra checks, but for maybe half of your pushed items to be complete trash in any run can be very upsetting. I’ve heard complaints of people getting BKed behind Celeste players; ironically, this was during a sync where I was BKed for a long time as a Celeste player and goaled pretty late into a sync, something I literally woke up, hopped into immediately, and went to bed immediately after it finished, taking up an entire day where I did spend quite a bit of time waiting. It’s an incredible game for asyncs and can be pretty good for syncs, but definitely go into it with caution.

Customization: 10/10#

The settings in the newest versions of the APWorld are genuinely incredible with absurd levels of personalization that makes the game really stand out as one of Archipelago’s best, certainly the most thought-out implementation I’ve played myself. Multiple goal areas, logic difficulties, traps, many different levels of -sanities and included/excluded levels/features, movement/useful item restrictions, and cosmetic options, as well as deathlink and even trap link: It’s not hard at all to make Celeste your own here, and I’ve done that already with three premade YAMLs that I’m ready to just start tossing into things once I’m less bogged down by async slots. It’ll be very fun to play around with some of the other challenges in logic once I’m more ready to play the game, as well.

Overall: 9/10#

I do think the logic is definitely a point of contention, but the gameplay is so fun that I think it warrants this high of a rating. This is one of the best APWorlds out there and it’s definitely the premier example of “Archipelago got me back into this game I never thought I’d play again” for me. Now I want to do a proper vanilla near-100% save one of these days.

Content Warning#

I ran the Monster Hunter goal in an async as a solo slot and had to force-release my slot after techincal issues started coming up, so I expect some pretty heavy bias in my review that could be fixed a few ways. It should be stated outright that it’s been a long time since this implementation saw any activity, so be wary when running it anyways.

Gameplay: 0/10#

Yes, I cheated, and I’m okay with that.

It’s a friendslop game, and I really went into it solo. The issue with this is that there’s no one to save you when you go down (and you do go down pretty frequently), and if you go down, you basically kill your run right there and then, making some later game checks a nightmare. By the end of it, I had to download multiple cheaty mods just to not constantly die and kill my runs, something I felt guilty about while still putting myself up to try to make footage I’d never send anywhere entertaining, throwing off game balance and making progression items feel moot, somehow even more than they already did due to just how bland the game felt with or without them.

It should also be mentioned how RNG-dependent the entire game is, from hats to sponsors to spawns in the Old World, to the point where some checks that really should be easy are impossible. Many checks are bugged to not even register, in fact; I never got my Arms checks and it never counted against anything, which inevitably forced me to release, and I never saw the Black Hole Bomb again after the first time I died with it in my camera roll, meaning I was straight up locked out of my goal. I’m sure that I’d have a better experience with a different goal, but I’m already too burned.

Did I mention that filler items are sent on connection and are often sent before you even load your save? This makes certain checks impossible in an async, especially as Meta Coins, which are required to buy hats, are just straight up not sent to an offline save.

Logic: 0/10#

Almost everything is a sphere 1 check. This game really wasn’t meant to have good randomizer logic and it shows. I’m sure this is an issue lots of friendslop games have, but this one definitely has it the worst with everything else stacked on top of it. There’s going to be a pattern in this document of games that have horrid logic having so many things in sphere 1, and it’s painful when you finally do get a sphere 1 check because it turns out someone’s progression item is behind something you should’ve been able to do if the game didn’t intentionally screw you over. Why are Meta Coins marked as filler when they’re literally required for sphere 1 checks, anyways? It’s just a mess.

Customization: 7/10#

Multiple goal conditions, inclusion and exclusion of goals and checks, inclusion and exclusion of monsters, and a surprisingly thorough set of rules for a run that would probably feel a lot better if the game was actually good, even down to filtering in multiplayer-exclusive logic and deathlink.

Overall: 0/10#

They call me 007: 0 Gameplay, 0 Logic, 7 Customization. Nothing could possibly save this dumpster fire of an implementation short of a major clean-up operation, potentially a complete overhaul.

Elementipelago#

I was never able to goal the solo I put this in and I’d love to potentially give it another try as a result, but I did play it for quite a bit and have thoughts on it regardless.

Gameplay: 1/10#

This is an alchemy type game with absolutely no logic as to what goes together with what. Gameplay of this is predominantly performed by trial and error of a million combinations of random entries in your list of literally any type until something finally happens, after which you just have to do it all over again because you forgot what works with what and what doesn’t. I’d love to see recordings of trial and error or a proper hint system for combinations because I don’t know when I’m BKed or when I’m not, and it just becomes a mess after a lot of items come through; maybe a system that shows what items have checks in logic, still?

Logic: 7/10#

Take this rating with a grain of salt, but the logic… actually isn’t that bad? The theming of what you get in-game literally being the checks other people send you the items with is a fun touch, and there’s heavy customization when it actually comes down to how complex you want the logic to be. It’s quintessential BK game, which is really going for it.

Customization: 3/10#

You can customize exactly how long or short your game will be, as well as if compounds matter beyond sending checks, which is a fun touch. You can’t disable intermediates entirely, disappointingly, and you can’t set how many checks there are directly, which is definitely fair if the implementation demands it a certain way.

Overall: 2/10#

I’ll give this game some credit for being a solid BK game, but there are a million better BK games to run to the point where I don’t know why you’d bother.

Geometry Dash#

I never goaled the slot I played of this due to the difficulty of the game that I’m still bad at.

Gameplay: 3/10#

It’s Geometry Dash. I never could really get into it, as much as I tried. It’s not even custom level oriented, so it’s just old levels and pain. Electrodynamix is a 12 star level, somehow. The game resets your configuration and login when you enter this, which is also very disappointing and I’d love to see an alternative to that to keep my cool icon and my very important settings.

Logic: 1/10#

Checks are in the low double digits total, even with everything turned on. No partial level progress, which would fit this game very well: Just level clears (and secret coins, apparently, but I never saw those be sent) and that’s it. It’s hard to tell what the goal even is. It’s incredibly punishing for people who can’t beat every level up to either Electrodynamix or Deadlocked, and definitely not as fun as I’m sure many custom levels would be. It’s just disappointing. I’d love an implementation of this that was custom platformer focused so that I could put some energy into it, but that unfortunately just doesn’t exist, and I’d expect that the logic for making something like that just isn’t there, in part. Give me more.

Customization: 5/10#

There’s some, actually, so it’s not that bad: Deathlink is present, coins can be toggled, there’s a dedicated checks shop… Some fun stuff. I just don’t know if any of it works.

Overall: 1/10#

Try to steer clear of this one until it gets completed, or play one centered around custom levels if it exists, which I hope it does.

Muse Dash#

I’ve goaled this as part of the Big Async. My normal YAML is harder, and I’ve played some of it, but have not goaled that.

Gameplay: 10/10#

Rhythm games generally make wonderful APWorlds, and with Muse Dash being the quintissential rhythm game implementation, it’s perfect for the cause. I don’t even like Muse Dash that much and I definitely have quite a bit of fun with it in multiworlds; I get to play random songs and just have a ton of fun doing it. It’s great.

Logic: 10/10#

No need for filler if your implementation is this airtight: Just songs and macguffins. Pacing just works and the checks don’t need to be high for you to have a ton of fun. It’s perfect.

Customization: 9/10#

Customize how long your run is, your song pool, your goal song, clear requirements, difficulty of songs, and enable deathlink and certain VFX/SFX traps depending on what DLC you have. It’s a pretty small YAML compared to some other Core-Verified games, but it’s still quite good, and you can’t expect too much out of it.

Overall: 10/10#

If you enjoy Muse Dash, you’ll enjoy this APWorld for sure. If you don’t enjoy Muse Dash but enjoy other rhythm games, you’ll probably have more fun with this APWorld than you usually do with Muse Dash. If you don’t enjoy rhythm games normally, maybe this would be a good way to expose yourself to a new genre of Archipelago games.

Neon White#

I’ve goaled this game a number of times in a number of slots on different settings. It is one of my two go-to Archipelago games.

Gameplay: 10/10#

This is maybe some of the most fun I’ve had playing Archipelago. The progression items being the cards themselves is very interesting, and the logical progression to the goal is clear to someone who knows the game well, as I do. It’s also pretty analagous to actually playing the game on its own, except for the fact that you’re playing levels in the order the randomizer gives them to you. It’s a ton of fun; I love figuring out ways to squeeze times both in and out of logic and also seeing all the creative ways people have already done it.

Logic: 5/10#

There are two key issues with this APWorld’s logic, being the WIP logic and the Ticketpocalypse, and I’ll explain these individually.

Neon White’s APWorld is in alpha as logic is constantly being refined to make things playable, which means that things in your skill level might not even be accessible to you at certain points in time: At the present moment, you’re almost expected to search for time out of logic just to get items in, so if you’re not a fan of going out of logic, this game really isn’t for you. Logic is constantly being updated, however, so we’ll hopefuly be seeing a complete version of this APWorld when the time is right.

The Ticketpocalypse, on the other end of the coin, is how I affectionately refer to the massive amount of trash items this game has: With a total of around 100 levels, with or without sidequests, and Ace medal max, being a far more generous version of the more commonly played Dev medal max that is still competitive for a skilled player, sending out four checks per level, around 3/4 of the item pool coming out of the game (assuming you’re playing Levels unlock mode, but it’s pretty similar for the others) is filled with the game’s trash item, the Heavenly Delight Ticket, which (currently; the devs are planning a shop to make them less useless) does absolutely nothing. That’s 300 filler items for a game with about 100 progression items; with Dev max, that bumps up to about 400 filler items. Especially considering many games are around the 100 check range, or even lower, someone will absolutely be complaining about the amount of tickets that are in their world, and you’re going to have to deal with that.

It was worse before. Those items were marked as Useful before. I had to bitch three seperate times in the thread to get that changed, and I’m so glad I did.

Customization: 8/10#

Customization is incomplete at present, but is already very robust: Two different logic difficulty settings, max medal selection, toggleable extra checks (gifts and sidequests), and three different level and goal unlock methods, with additional logic challenges, progressive checks, deathlink, and traps in the works. It’s as robust as it can get for this game, so it gets my approval.

Overall: 8/10#

Not for everyone, especially at this point, but also just in general, but if you enjoy it and understand what you’re going into, it’s a blast. I know it’s gonna be even better when it’s finished.

osu!#

This is my other go-to Archipelago game and how I got into Archipelago at large; suffice it to say, I’ve got a ton of familiarity.

Gameplay: 9/10#

This is such a unique and refreshing way to play through osu! in any mode and is ultimately what got me back into the game, now at the point where I play literally every mode. It’s a shame that the random selection is limited to featured artists, but that pool is already surprisingly big to where I’m only getting one duplicate every other run out of a pool of 50 songs in a mode that doesn’t really have the map depth that others do at my skill level.

I will dock down a point thanks to the fact that if you try to gen for multiple modes, there’ll always be a huge bias towards the oldest mode you include in a single file. I like to gen each mode individually if I can help it; I’m doubtful if this can really be helped in the APWorld much.

The setup and interaction with the game can be annoying for some due to taking place in an entirely seperate text client that connects to the osu! API, but that’s not a big enough deal to dock another point.

Logic: 10/10#

Another airtight, trashless rhythm game APWorld. It’s perfect.

Customization: 9/10#

Customizable song count, minimum grade requirement, difficulty restrictions, toggles for loved maps, converts, and explicit lyrics, mode selection with individual difficulty ranges for each modes, and options to include or exclude songs from the pool (unfortunately, songs can’t yet be forced to goal songs like they can in Muse Dash). It’s remarkably robust for what it is.

Overall: 9/10#

Almost perfect, just like Muse Dash, and with a lot more flexibility as a game, at that; in my opinion, even though Muse Dash has a better implication, this is the more fun game. If you enjoy osu!, you’ll enjoy this a ton. If you enjoy any game that’s similar to a gamemode in osu!, you’ll enjoy this, still, like I did. I made a habit of marking the highest pp plays I get during runs in my Cheese Tracker; it’s kinda awesome.

Rabbit & Steel#

I’ve goaled this before in a solo and am pretty close to goaling this again in an async. I run with kingdomsanity and classanity as must-haves, as well as set-based itemsanity.

Gameplay: 6/10#

It’s pretty damn close to the original experience and it’s great for that. Limited item pools make it possible to stack dupes of the same item, which can lead to some very fun builds. The only thing that feels off is when runs cut off early, which can feel very limp and generally disappointing. It should also be mentioned that runs do take a very long time, making it a pretty difficult sell in a shorter sync setting unless you’re running a very short 1 Shira 1 Witch YAML or something like that. It’s certainly not for everyone, but it’s not bad per se, and you know what you’re getting into.

Logic: 9/10#

At present, run length settings and progressive regions are broken, which one of my runs uses, leading to some very long and unfun stretches of BKs, but if we count that out… I’ve genuinely never seen this implementation send out trash, and it’s actually pretty flexible on how you get to your goals. It’s a challenging experience that’s light on checks and items unless you turn on a -sanity, and with a curated YAML, the logic is pretty airtight and enjoyable. It’s a very solid roguelike randomizer for what it is.

Customization: 9/10#

Fits the bill literally perfectly. Multiple different randomization vectors as -sanities, varying goal conditions, and even forced kingdom orders. You can even turn off randomization entirely, from the looks of it… but at that point, what are you even doing? It’s that robust.

Overall: 7/10#

It’s an incredible roguelike randomizer; as such, it’s not for everyone, but if you’re a fan of this game, you’re gonna have a ton of fun here.

Refunct#

I’ve goaled this in a single-game run; in fact, I literally bought this game just for Archipelago. It’s so cheap that I did it by selling Steam cards, even.

Gameplay: 10/10#

It needs to be said outright that this is more than the base Refunct experience by a mile, and it’s great. The main game itself is very nice and meshes well with Refunct as a move randomizer that harkens to any well fleshed out randomizer out there, but the selling point for this APWorld is the minigames, custom variants of the base game with their own set of checks that are just incredibly fun to play in an already fun movement set. You’ll be completing Refunct in a million different ways with so much of the character of Archipelago inside it. It feels like a love letter to both the game and the community it’s based for. It’s incredible.

Logic: 6/10#

Much of this game’s item pool is filler, and the check pool can be massive, leading to a case of Ticketpocalypse; however, this is offset by the use of macguffins on top of a decent number of progression checks, which can offset it quite a bit; however, minigames can be forced to contain trash after a certain point to make reliance on them less necessary and decrease the amount of flowers in the multiworld. Minigame selection is random, however, so this can be an unpredictable solution.

Customization: 9/10#

Macguffin count length control, many options on how a goal would work (including unknown goal conditions), unlockables and minigame count to control the amount of checks somewhat, multiple levels of traps, and some fun cosmetics, on top of a low-stakes deathlink implementation. It’s incredibly robust, and the default settings already lead to a level of replayability.

Overall: 9/10#

It’s an incredible implementation, an expansion of Refunct that practically serves as its own game, and for a game so cheap, it’s an incredible experience. Just be careful if you’re running it in a smaller lobby.

Twisty Cube#

I’ve goaled the 3x3x3 cube.

Gameplay: 6/10#

It works well as a BK game, but becomes more frustrating the more thought you have to put into it, I find. The gameplay is reasonably boring, not much more than I’d expect, and it’s kind of annoying to get new pieces early on when you’ve set something up already, especially as someone who still can’t wrap their head around commutators. It’s somewhat offset by how satisfying this game’s late game can be as the cube comes together and you can actually start to solve it like a regular cube, but it can be very annoying otherwise.

Logic: 10/10#

This is as airtight as logic gets, with checks being linear despite pieces being as random as they are. It plays no different from the non-Archipelago experience, just with added delay between checks, and that does it so many favors: No trash, all action. It’s a perfect small addition to any run.

Customization: 7/10#

It certainly doesn’t need much, but it has what it needs to: Different starting sticker counts, random color layouts, and how soon or late you want your goal if you just so happen to accidentally solve the cube without seeing everything (a statistical improbability). It’s not much, but it doesn’t really need to be.

Overall: 7/10#

It’s a fantastic BK game and will serve you perfectly there, but it falls flatter and flatter the more time you spend thinking about it and the less time you spend thinking about anything else.

ULTRAKILL#

I’ve goaled this before in an async.

Gameplay: 10/10#

It’s peak, straight up. Unlocks feel natural even when they’re out of order, and the challenge of running this game with whatever weapons you end up getting is interesting and leads to some very fun ways to play around logic. The collectibles lend themselves well to the game and the trash still actually does something, even if it’s ultimately moot. The game already has systems to make running low weapon counts near the start interesting, and including your arm in the randomization pool could also be an interesting experience. It really speaks to the challenge of ULTRAKILL in a unique way.

I’ll also mention here that you do need to be very careful of how you disconnect from the lobby to make sure data saves, but obviously, that’s just an issue for the end of a session and doesn’t cut down points.

Logic: 9/10#

The game definitely serves up a bunch of trash at times, especially with the more checks you put in, but the trash at least feels satisfying. There’s definitely no strict logic for progression items, thankfully, as long as what you’re doing leads to goal and you feel confident with what you’ve got to learn. I definitely don’t like the fact that clearing a level on its own does not send a check and only contributes to the goal, which definitely forces me to go for P-ranks to contribute to the multiworld if those are on, but things are very flexible around that, even, so I’m chill.

Customization: 10/10#

It’s so, so robust. Custom starting level, goal level, clear count requirement, special handling for certain types of unlocks, multiple types of filler and traps, levels of trap toggling, “-sanity” settings that can go pretty deep down, starting weapon pool, many levels of loadout handling, movement restrictions, and even cosmetic randomizers, as well as deathlink with amnesty. It may not be as jam-packed with features as the biggest implementations, but that’s downplaying how advanced what’s going on here is.

Overall: 10/10#

I can’t pretend I don’t have the most minor of qualms with the logic that do throw me off, but that doesn’t stop a masterpiece of an APWorld in its tracks on its own. Go frag out and have fun.

Yacht Dice#

I haven’t goaled this game yet, but I’ve got a ton of experience with the game in preset YAML asyncs and my own YAML submissions to the point where I feel qualified to speak on it. I think the distinction of Bliss and the core-verified version being different games is stupid, so I’ll just treat them as one, with Bliss being my version of choice.

Gameplay: 9/10#

Despite being luck-based, it’s a surprisingly satisfying implementation with a solid gameplay loop. The inclusion of alternative categories makes for a lot of interesting play with the most random of things popping up as important chases, and the buildup of power feels satisfying, especially with the exponential explosion in power that just an extra die, roll, or category makes for the run. Logic being as clearly defined as it is means there’s not as much pressure to overperform, but hitting things way out of logic is incredibly satisfying. The dice skins are a fun cosmetic touch, as well.

Logic: 10/10#

This APWorld designs itself around low trash counts unless you explicitly ask for them, and I don’t think I’ve seen anything useless be sent to it to date, which makes for very clean multiworlds. While it does have a pretty small check count, at least how I have it set up, it still feels like enough for a lobby much of the time.

Customization: 10/10#

Customizable logic difficulty on multiple axes, game length, check count, category filtering, and lots of fun ways to fill the last bits of remaining space.

Overall: 10/10#

A random branded BK game being core-verified should be a sign of its merit as a solid APWorld. If you need a BK game in a panic, come give this a try.

YARG#

I’ve not goaled this one before, but I do have experience in it and its forks from various solo runs.

Gameplay: 8/10#

The game itself is solid, as all rhythm games tend to be, but there’s definitely less freedom of play. I’m bummed that the randomizer can’t filter out songs that don’t have a chart for my instrument of choice, that being pro keys, without me having to turn on instrument shuffle, which isn’t viable for only one instrument. The song selection, however, is great, and the game is very flexible.

Logic: 8/10#

Three checks per song is a bit much. I guess it’s offset by higher macguffin counts, but it doesn’t feel right running anything higher than 50% on macguffin generation. YARG could almost become a load-bearing game for progression. The higher rating is mostly because I haven’t had any issues with it yet.

Customization: 7/10#

Not bad, actually. Deathlink, energylink, run length, setlists included/excluded as needed, song exclusion, and an instrument shuffle mode for those who care about more than one instrument. Well thought out.

Overall: 8/10#

Not a stellar, airtight rhythm game implementation like they tend to be, but I don’t even think it’s bias speaking when I say it’s still a solid APWorld for a game I’m sure many find enjoyable.