Castlevania: Aria of Sorrow

The Castlevania retrospective continues! Well, this time around, it’s the year 2035, and you’re Soma Cruz, the cool teen reincarnation of Dracula. A welcome excuse to turn into a bat again! There’s actually a story this time, and I like it, too, even though every other character — including the legendary Alucard — spends the whole game doing basically nothing.

We’re still without X and Y buttons here, but Aria of Sorrow already feels like Castlevania’s Nintendo DS era. A lot of mechanics come together, sticking around for at least a few games: a reasonable backdash, a variety of primary weapons, and a ditching of the subweapons system, finally daring to mess with the heart-counter that was there since the NES, even though its inclusion hadn’t always made sense. I may be a little biased on that last front, as I always tend to hoard consumables for some ever-distant rainy day, even when I know it’s foolish. But Aria gives you back all your regular resources at the save point, and I like it: I think it’s better suited for meandering exploration.


This has little to do with anything, but after Aria of Sorrow came two games with shitty generic anime art, so I want to appreciate Ayami Kojima while I’m in the moment.

Aria also manages to improve the visuals without sacrificing sound quality as its predecessor did, which might lead me to believe that it was just an excuse, but word is that the team redeveloped their audio tools for Aria and dedicated “more cartridge space and processor cycles” to sound. This might mean that it never draws quite as many big sprites and effects on screen at a given time — I never really noticed — but it’s impressive that it looks as good as it does, and falls more in line with the Metroid games on the system.

I love the new soul-gathering system: you basically go around obtaining every monster’s ability, as if catching Pokemon, and you equip these as magic spells and passive abilties. It’s hard to describe the joy of killing a “Kicker Skeleton” and stealing its flying kick move. But the execution is unnecessarily grindy. When I wrote about CotM, I mentioned that I was kind of okay with that grind — and I welcomed the opportunity to catch up on some podcasts — but rare items are enough for that, and there was a missed opportunity to do something more creative here than simply killing the same monster until its soul randomly shoots out. If it had taken the form of a thief skill, or a “Blue Magic” type of ability where Soma needs to be hit by the skill in question, it could have lead to some interesting puzzles where specific conditions had to be set up, like getting a monster to low health, or placing some kind of trap down that the monster has to attack, or anything else, because they really could have done whatever they wanted with this.

There’s a ring that boosts drop rates for souls, and because I didn’t have the money to buy it, I chose to wait until the end of the game to fill the gaps in my soul collection. This meant I missed my best chances to play with some fun-looking abilities. And some of the abilities I did have were too similar to each other, like Devil, Manticore, and Khali, where the only difference seemed to be cosmetic. Even so, as it stands, it’s a vast improvement over CotM’s card combination system, and I probably found actual use for 60-75% of them, which was far better than letting one spell carry me through the whole game, which is what happened during Harmony.

you get a cat

“YOU get a cat! And YOU get a cat!”

I used a variety of weapons, too. You’re not just looking for the biggest stat stick here. I didn’t use the heavier weapons like hammers, but I would alternate between typical horizontal swords and vertically swung broadswords as the situation demanded, or even switching to a sword of another element if the situation called for it. Though we’re not quite in the DS era in this regard — you still only have one attack button, and the single-screen interface makes it tedious to pause and check what an enemy is weak against — if I were fighting something that seemed to be taking many hits to die, I found that for the first time in the series, I was willing to make that change. Sometimes.

It’s the little things, too. I can pause the game without stopping the background music. I can time an attack to land just before I hit the ground to preserve my momentum, and cancel the end of a heavy attack’s animation with a backdash, which elevates the combat to a more interesting level. Suspend saves made their first appearance here, too.

It’s not a hard game, particularly, though you do see the return of extremely rude rooms with conveyor belts that lead into spikes and medusa heads that petrify you so you fall onto the conveyor belts, so all the ingredients are there. You also never get your constitution so high that you won’t take meaningful damage from spikes, which is a little mean, but just good design. On the whole, it’s probably a little easier than CotM, in part because you’re likely to overlevel while killing everything for its soul, but I suspect the diverse weaponry and spells also mean it just isn’t going to be as tight. (It also depends on whether a CotM player is willing to turn on the stationary heal spell and put the GBA down for a few minutes.)

Though it can annoying to switch between the power that makes you sink in water and the power that lets you stand on water, the use of water is far more creative than the barriers found in Dracula’s last two castles. Though you don’t get to wall-kick this time around, there’s generally more of a focus on movement than on finding the right keys for a series of bland doors, such as the ability that makes Soma fall very slowly, allowing him to extend the horizontal distance of a jump. I’m certain Aria has the best castle design of the GBA games.

Aria’s boss battles are inconsistent. Some were decent takes on classics, like Death. Many were immediately forgettable. Legion was standard, but the lead-up, featuring creepy clay dolls in the nearby chambers all collectively shuffling over to the boss room, was an amazing touch. Seeing as you’re Dracula, the final battle was probably doomed to be anticlimactic — a boring affair against a big skeletal mass with some orbs you need to shatter — but the penultimate boss is Julius Belmont, wielding Vampire Killer and all the usual subweapons, and it’s absolutely a cut above any fight in CotM or Harmony.

This time around, postgame content includes Hard mode and an actual New Game Plus. Who doesn’t love New Game Plus? It’s possible to kill at least one boss without getting its soul, in which case NG+ is a good way to get to 100%, but I managed to do it the first time around and already claimed the ultimate reward — a piece of equipment which gives Soma infinite MP — so even on Hard, NG+ would just be a way to bully all the monsters in the castle while constantly stopping time or making yourself literally invincible. If I had played this in 2003, I might have done it to get the exclusive NG+ weapons (such as Death’s scythe), but as of now I’d prefer not to linger.

There’s also a mode where you play as Julius Belmont. It’s a great idea, and he’s fun with his MP-based subweapons — he carries the holy water, axe, and cross at all times, instead of needing to find one to switch to it — but as with Maxim mode in Harmony, it’s been shoehorned into what’s supposed to be a progression-based game. I remember actually leveling up in the DS bonus modes, but Julius still doesn’t seem to here, and the first boss dies in about two hits. I think I’m good.

This game was thoroughly enjoyed by the reviewer. It is an excellent game that may be too simple or not ambitious enough to be a 5, or there are design flaws meaningful enough to prevent it from enduring as something truly beloved. Highly recommended.



Castlevania: Harmony of Dissonance

Moving forward a year from Circle of the Moon, we have Harmony of Dissonance, which tries harder to be Symphony of the Night. That’s an admirable ambition, even though it fails.

There are many mechanics that I was happy to have back, like currency drops and merchants, maps of unexplored castle segments, an enemy encyclopedia listing undiscovered rare drops, and shoulder-button dashes. The execution of these features aren’t perfect. The shoulder buttons let you dash without any kind of cooldown, so I pretty much did all my movement by rapidly clicking them instead of using the regular D-pad moves, which seems like some kind of speedrunner bullshit, except that it’s right there in the open. Merchants are annoying to find and rarely sell anything actually needed; while there are many more items to collect this time around, most of them are just noise; you see a down arrow on a piece of gear and you never look at it again. A few pieces allow you to jump infinitely or regenerate health or some such thing, almost more like certain magic spells in the last game, but I pretty much stuck to high-defense gear and steamrolled everything.

It’s too easy, honestly. I don’t necessarily care about pushing myself with a challenge, and I don’t have the patience for most truly hard games unless they’re absolutely exceptional, but if a game is easy enough that you don’t really need to learn its tricks or make full use of its mechanics (such as elemental affinity) to adapt to different situations, it’s sort of a waste of time to even include those mechanics. A couple things to specify: first, every open-exploration Castlevania is pretty easy when you consider that you can freely pause at any time and eat an entire chicken from your inventory, but most players will at least push themselves to do so sparingly. Also, it’s not as if I never died. But given the “quick save” feature — totally unlike the suspend saves of other portable Castlevanias, as it allows you to get back to a save point with all your progress retained — I was never at any risk, and I never had to reevaluate my strategy.

It’s a shame, because I think the bosses are better than those in CotM, if you can ignore how easy they are to beat, because a few of them took long enough to die that I got to see some interesting attack patterns and tells. There are also twice as many bosses in Harmony, though only a few are memorable.

The magic system kind of sucks, too. When I last wrote about CotM, I mentioned that it had a lot of spells I obtained too late or never found a use for, but Harmony opting to tie them to subweapons — which were already hard enough to give up, not knowing when you’ll next find the cross or whatever if you do — tends to discourage experimentation. I’m sure other people had their favorites, but for me I was set from basically the start of the game when I found the ice-axe combo, which hits the area ahead of Juste, and tends to do about double the damage of a whip swing, at a reasonably affordable MP cost, from a distance, while also giving enough invincibility frames to completely avoid many heavier boss attacks. Even in the unlikely situation that I actually spammed it enough to run out of MP, I could have just turned it off to start actually spending hearts on throwing axes, but that was unnecessary too: I spent just about the whole game at full heart supply.


Animation is more fluid and interesting, especially with enemies that lunge around, drawing out the full potential of the new dash mechanics.

Let’s talk about the visuals for a bit. They’re very different from CotM. Going back to what I said about the priorities of the former game, it’s hard to make a clear stand on whether Harmony comes out ahead with the trades it makes. The animation still falls far short of Symphony of the Night, but not for lack of trying on the GBA. (I considered gathering up some comparison images for this post, but I found that someone else had already done it, so I’ll just link to that.) Sprites are bigger and use more saturated colors now, which gives Harmony a more cartoonish feeling that some might consider garish compared to CotM’s great earthy palettes. But the enemies are far more expressive now, and creative, throwing barrels around, mimicking the player’s movements, twirling weapons around and so on. Many of these were probably lifted from Symphony of the Night: it’s been a while, and I don’t recall what originated where, but there was a lot of direct asset recycling among the DS Castlevanias in particular. But if these were rehashes, they were at least redrawn with new resolutions and palettes for the GBA, and they certainly add a touch of charm that CotM lacks.

Also, the promotional art for this game from Ayami Kojima is absolutely beautiful, but I guess that’s neither here nor there.

The main trade-off is that the graphics were apparently pushed so hard that the GBA had to fall back on the old Game Boy Color hardware to fit the music in there. Voice samples and other sound effects are unaffected, but the soundtrack is only slightly more sophisticated than what an NES can manage, which is jarring, especially on the heels of CotM. It’s a shame, because I’d otherwise like to be positive about its use of original compositions, given that the CotM OST was mostly a greatest hits compilation. This one is pretty good.


I’m starting to miss Symphony of the Night. It can be interesting to work with constraints, but it’s always nice when you can be this beautiful and don’t have to mess up your audio for it.

A lot of Harmony’s creativity also shines through in the environments. There’s more fun stuff in the maps, like elevators, trap rooms, walls that get smashed apart by enemies, a big room where you race to collect an item before a giant ball rolls into place and blocks it, and various other gimmicks. It makes CotM a little straightforward and boring, but at the same time, gimmicks aren’t a substitute for good level design, and here the maps are confusing and repetitive. The same lack of information in CotM’s castle map hasn’t been addressed at all, but the obstacles to progression are far more obtuse now. It’s one thing to need some otherwise-useless block-pushing power, but sometimes Harmony will just break a wall down on its own somewhere after you kill a boss somewhere else. I’m sure I wasted an hour or two revisiting every formerly inaccessible pathway, trying to figure out where something might have opened up for me. It’s irritating to play without a guide.

Just running around one castle can be tedious enough — you can go awfully far down a fork before discovering that you don’t yet have the item you need to actually accomplish anything there — but Harmony’s main gimmick is the existence of two castles; the same place in alternate dimensions. Often you might already be where you need to go, but in the wrong castle, so you have to traverse a fair bit of distance to a warp point, switch from Castle A to Castle B, and then get back to where you were. What really sucks about this is that these are literally the same sets of rooms, not even flipped from left to right. The enemies change, and the wallpaper may differ, but that’s it, and there’s no mistaking that this is a way of padding out the game’s length relative to the level budget. At least Symphony of the Night had the self-respect to make its second castle more of an optional postgame affair, and put it upside-down. It’s just not as insulting when you’re walking on the ceilings.

It’s a shame: it actually would’ve been pretty cool if you had some Link to the Past-esque means of switching from world to world on the spot, say, to get around obstacles that weren’t exactly the same in both worlds, and you still only had one map to explore between the two of them. They could have given certain monsters the ability to switch from these Light to Dark worlds, too.

There are a few extra game modes. One is hard mode, which might have helped the game the first time around, but with the two castles padding things out, I can’t be bothered. There’s also Maxim mode, which features a different playable character. That’s my favorite kind of postgame option, but he doesn’t level up, despite the game being balanced with character progression in mind (though they do boost the health of some of the early-area enemies). Also, unfortunately, Maxim doesn’t even have the ability to get full 200% map completition without making level changes that the developers couldn’t be bothered to make for him, which kind of bugs me. (Then again, since you can apparently beat his mode in under 30 seconds, perhaps anything is possible.)

Harmony of Dissonance has some cool ideas that improve on CotM, but I think it’s the lesser of the two games, mostly due to padding and an often-trivial experience. Still, they’re not so far apart in quality, and it takes some healthy steps in the right direction, bringing it closer to the better games of the DS era.

The reviewer finds this game hard to get excited about, but still has a positive opinion of it. It may be somewhat fun, having good features or ideas counterbalanced by a few boring parts, bad design or other fundamentally irritating qualities that can’t easily be overlooked. Alternatively, it could be pleasant, but with nothing new to offer. Worth a little money if you’ve got the time for it.


Castlevania: Circle of the Moon

This was the first Castlevania title I ever played for more than five minutes, and it kind of holds up. As a GBA game from 2001, it lacks some of the refinement from the DS generation, like a backdash ability and a monster compendium. Its age doesn’t fully excuse it, though. Super Metroid, an obvious influence, put little dots where you didn’t pick up a collectible, but here, you have to write that stuff down. Being difficult and being unhelpful are two completely different things.

What I mainly remember about Circle of the Moon is that it’s probably the first game where I was ever happy to run in and out of a room killing the same enemy, hoping for a rare drop. I don’t know exactly what it is about it, but while I hate grinding for experience, and I hate the general ethos of impossibly rare drops in MMORPGs, something about the Castlevania experience works for me despite still being kind of tedious. Perhaps the drops are just cool enough, or the gameplay is fun enough even when I’m repeating one specific kill, or the drops aren’t so countless and troublesome to get that I feel like I’m in some kind of endless Skinner box trying to get some meaningless gold star. I feel the same way when trying to get something like a ghost blade in Dark Souls: at the end of the day, it’s still Dark Souls.

The level design isn’t amazing; occasionally there’s a little puzzle, or an interesting arrangement of enemies, or a well-timed HP-boosting collectible to take you back to full health, but most of it is nothing to write home about. There are a few too many long halls with repeating enemies, and a number of points where I have to gripe: would it have killed the designer to put a little connecting shortcut here or there, not to make a zone less challenging, but just for getting around later? Maybe a little lift you can turn on between the right-hand side of the Audience Room and Eternal Corridor? Or to Chapel Tower? If things had linked up more often, sprinkling around various means of terraforming Dracula’s castle, and picking the right moments to give you abilities to get around weaker enemies — like getting the Pluto/Unicorn combo before postgame — it would have been more satisfying and no less challenging, and I would have been happy to do without warp points at all. As things are, warp points are a must; they’re also awkward to use, and not one of them can ever conveniently come out next to a save point. Some of this is a direct carry-over from Symphony of the Night, this game’s predecessor, which also got messy, and not to its credit.

As is common for the genre, the player explores freely, but everywhere they go progression is blocked by a variety of barriers, some of which are surmounted by changes to the player’s inherent capabilities — like being able to wall-kick over a ledge that can’t be crossed by a double-jump — and some with direct changes to the map, for example, hitting a switch that blows up all of a certain type of gate. I think these external types of barriers are lazy, and here, the internal ones often aren’t much better. You need two separate upgrades just for getting around giant blocks that are just randomly put in your way, which is dull. Getting the power to push a block around does nothing for my general fighting capability, either.

The game has a spell system where you combine two types of cards. It’s an interesting experiment, but the outcome of this is that picking up one card has a multiplying effect as you go, and one card may suddenly give you eight to ten new abilities you may never use. Most are either severely gimmicky or entirely useless by the time you get them, and there’s no balance to speak of. There’s also a completely inscrutable elemental system with three distinct elements for plant, earth, and stone type enemies. What are they weak to? I have no idea; I always found it more useful to give myself something to increase my range or protect me from behind, rather than trying to imbue my whip with a specific element. I think it might have been better to hand out powers one by one, in situations where they’re useful, as some of the later DS Castlevanias try to do.


Nathan’s face is only three pixels, and the animations use barely any frames, but the restrictive color palette is used perfectly.

From Symphony of the Night in ’97 through to Order of Ecclesia in ’08, for a while there the 2D Castlevanias were some of the only titles keeping the art of hand-animated pixel sprites alive, and there’s no way I’d avoid talking about it. Circle of the Moon has some cool sprites, but looking specifically at its animations, it falls far short of Alucard’s beautifully-flowing moves on the PlayStation, or even what the GBA is capable of: Nathan’s run is a pathetic 4 frames, and they’re flipped when he turns, so he ambidextrously whips with whichever arm is in the foreground. Contrast with Metroid: Zero Mission, where Samus has a 10 frame run cycle, doubled to 20 if you include her running in the opposite direction. The difference couldn’t be more apparent.

You’d be forgiven for thinking that fluid animation would be a greater priority for a character swinging a whip than for one firing energy blasts from a gun arm, but priority really does appear to be the issue here, given that Circle of the Moon apparently wasn’t holding back on the GBA hardware. As it is, when enemies and spell sprites start to fill the screen, the game slows down (this might have more to do with the resolution of these sprites than how many frames they need to cycle through). Metroid is a little more zoomed-in and doesn’t put as many big creatures on screen, making it kind of an apples-to-oranges comparison. But I do think Metroid finds a better balance.

On the bright side, the soundtrack is absolutely one of the best on the GBA, with stuff like The Sinking Old Sanctuary, Clockwork, Clockwork Mansion, and Aquarius never leaving my head for long in the fifteen-plus years since I last actually played this game. I had no idea until writing this up that these were all remixes of older Castlevania songs — the SNES version of Clockwork Mansion was especially cool to hear — but the bass-driven GBA versions are absolutely fantastic. (It honestly boggles my mind that Vampire Killer still gets so much attention when these other tracks are so good.)

There are several extra postgame challenge modes, and I think I beat at least a few of them when I was a kid, but I didn’t bother getting into them this time. Magician mode offers an opportunity to use some cards that were picked up too late to be useful in the main game, but none of these reinvent the wheel the way wholly new playable characters can, and do in many other Castlevanias. Still, as the game thankfully doesn’t drag out playtime too much to begin with, only taking some ten hours to beat, it’s a reasonable way to mix things up if you still want more.

The reviewer finds this game hard to get excited about, but still has a positive opinion of it. It may be somewhat fun, having good features or ideas counterbalanced by a few boring parts, bad design or other fundamentally irritating qualities that can’t easily be overlooked. Alternatively, it could be pleasant, but with nothing new to offer. Worth a little money if you’ve got the time for it.


Metroid: Zero Mission

This holds up pretty well. The cutscene art and music seem a little crude on the GBA, though I’m not sure if the hardware’s any excuse. I always thought from games like Circle of the Moon, Final Fantasy Tactics Advance, and Mother 3 that the GBA really hits a sweet spot in terms of the instrument fidelity and enforced constraint. Super Metroid‘s soundtrack on the SNES is a minimalist masterpiece, but I didn’t really feel like the music here was anything special.

There’s still quite a bit of content after the point where the original NES game ends, and the additions are interesting. They take away your suit and make you do stealth segments, which is kind of a hard sell, but it kind of works. Getting chased between rooms by space pirates who can kick your ass, and still being able to do wall-kicks and find your own way around, doesn’t make for a bad approach to stealth. Nor is it incompatible with what the series is all about, as far as I’m concerned. I had already been thinking about what else could be done in the Metroid series without just repeating the same formula–the last three Metroid games I’ve played now have been direct remakes (of Metroid 2, 2 again, and 1, respectively), and at times it felt formulaic to the point that I didn’t know what the point was, especially with Samus Returns, which stretched this out over a much larger world map. What is Metroid, really? Am I playing all these because I think it’s important to use 99-unit energy tanks for player health? Of course not. As long as a Metroid game allows me to find my way around for myself, at my own pace, and is open enough to allow for some pretty deep advanced techniques to get some upgrades and abilities early on (while not ultimately skipping whole areas), I think I’m good. I overlooked this in talking about Samus Returns: you tend to have to do things when it wants you to do them, and it was a little joyless by comparison.

Zero Mission‘s design can be a little obtuse, though. Usually you have some indication of what area you’re supposed to be in, but even on the primary path you might need to bomb some subtly-different ground tile to progress onward. I got stuck for a while because of this. Mind you, because it was early in the game when I didn’t have too many other places to be, I was bound to find that tile sooner or later. And on the bright side, it got me to adjust to how Zero Mission likes to hide things. But it was far from the only time I got stuck. In Ridley’s area I couldn’t figure out what I was supposed to do until I gave up and turned around, only to find out that backtracking was the way to progress: a defeated boss moved away from its arena after I’d left its room, creating a new path for me. It’s as if some designer said, “Let’s have it so in this area, players have to butt their heads against a wall and then give up in order to find the way forward.” It’s a minor thing, I suppose — every player would eventually turn around, even if they could think of no other new place to visit with their current gear — but I can’t imagine a world in which that was the ideal way to have the player’s path through the level flow.

But that’s only regarding the main path. When it comes to optional ammo pickups, the missile caches, finding those gets far less intuitive. There are clever sequence-breaking tricks that the game never teaches you, but which the faithful Metroid players would already know about going in, like wall-kicking up a single wall or doing infinite bomb-jumps. I found it thrilling to use these to get up to hard-to-reach areas early on, assuming I’d otherwise be forced to wait until I had the space-jump to do them. But you don’t get to use the space-jump at all during the original NES stretch of the game. At some point these advanced tricks seemed to stop being the quick-and-dirty way to do get these pickups, but the only way. Many secrets I couldn’t find at all: I was convinced I’d be able to keep exploring Crateria after blowing up Mother Brain, despite anticipating the self-destruct timer, because there were power-bomb-yellow doors on the map, and I hadn’t found power bombs yet. Oh, I hadn’t entirely been wrong: you aren’t normally introduced to power bombs until after Mother Brain, in the new Zero Mission content. But you don’t get to keep exploring Zebes after it blows up, either — at least, not without loading up Super Metroid. So what gives? Well, if you see something to power-bomb on Zebes, the only way to do it is by using some esoteric diagonal ballspark move I never would’ve thought of, to get a different power-bomb pickup early. You know how in Super Metroid, you could use the mockball trick to get super missiles early? Imagine needing to approach that same level of esoteric nonsense for 100% completion: that was roughly how this felt. It also explains none of it. Super Metroid actually had animals that showed you the wall-kick, assuming you were willing to stop and watch. But there’s nothing like that here. I find that a little strange, given it’s a remake of the original game. Where else should a player expect a fresh start, with no foreknowledge of how things work?

To some extent I think it’s kind of cool and old-school to have secrets I’m not going to find, but I would’ve only really gone in for that if I could reload a save after the credits or something and get to return to before Mother Brain died, letting me do it at my own pace. I also resented the use of the shinespark for these. There were multiple occasions where the levels felt too cramped for it, and I had some trouble with the controls, at least in using the 3DS D-pad. (The GBA, unlike the SNES or the sort of controller you’d use playing AM2R on PC, also lacks X and Y buttons, and has to cram more functionality into what’s left.) Shinesparking is also used in more complicated ways than before, often requiring you to chain them by using a new mechanic where your launch gets interrupted on slopes, while retaining momentum. You tend to have to do things in a very specific and calculated way, hitting precise but ambiguously-textured blocks.

Even when you found the secrets, collecting powerups gets pretty challenging sometimes, but this was the kind of challenge I happened to appreciate. One room I just barely noticed near Ridley, with two missile packs in it, required basically all the skill I felt I had, even after I figured out how I was supposed to do it. I had to manage rolling into a shaft, clearing away blocks with my beam, and shooting missiles upward, before the blocks at my feet crumbled away. It was getting frustrating, but I felt extremely gratified when I pulled it off. This was ten times harder than any of the Zero Mission bosses, and ten times harder than the skill level needed for obtaining any pickups in Samus Returns: there’s no time-slowing power here to make a joke out of the crumbling floor tiles. And yet at the same time, Samus Returns had bosses that required extreme precision, being hard as hell and the highest-quality aspect of the game. Strange.

Having complained that Samus Returns’ map-revealing ability was too much, it might seem a little silly to be making a strong complaint in the other direction here, but Zero Mission has classic map stations, and they just don’t reveal enough to really matter. The desired middle-ground should be obvious, though: add map stations that take some effort to reach, but which actually reveal gaps on the map tiles where there are entrances to rooms you’ve never been in. It’s odd to me that Samus Returns did away with the map stations and just let you see everything around you in a giant radius when this other option is here. Starting blind in an area, and having the map room as an objective in itself, has always been a great way to design the levels.

I liked this one a bit more than Samus Returns. It had its own strengths, but where it didn’t offer enough to justify its longer playtime, Zero Mission tried new things that worked, and has stronger fundamentals. The openness with which you could use tricks to get over ledges early or do a sequence of rooms in reverse — without doing some directionless amorphous design where you have no sense of where you’re going at all — is top-tier among the other games in the series. But I’d have loved it if it were ported into AM2R with some changes here and there, like adjusting the shinesparking a little and benefiting from a more informative map screen. Mainly, I have some regrets that the game’s post-Mother Brain sequence got in the way of me finding everything for myself. It’s not a long game, but I’d rather not start over anytime soon. I’ve never been a Metroid speedrunner, and I think doing things slowly, in one go, at my own pace, should remain a core part of the Metroid series, too.

This game was thoroughly enjoyed by the reviewer. It is an excellent game that may be too simple or not ambitious enough to be a 5, or there are design flaws meaningful enough to prevent it from enduring as something truly beloved. Highly recommended.


Metroid: Samus Returns

There was an unofficial Metroid 2: Return of Samus remake last year; a fangame. It was actually pretty incredible. Nintendo sent the guy a DMCA request and, surprisingly, announced their own Metroid 2 remake shortly after: Samus Returns, for the 3DS. Funny how that works; Metroid 2 came out in 1991, and then few people cared about it for like 25 years. Now we have two new versions. It gives us a pretty interesting way to compare and contrast.

Being on the 3DS hasn’t worked in the favor of the official game: the 3D graphics perform sluggishly at times (at least on my old XL), and the controls are brutal on my hands: Samus Returns has to be quite forgiving for anyone aiming with a circle pad, but AM2R felt extremely tight without bothering with the new free-aim stuff (which actually would’ve worked better with the real control stick I used to play it, or, say, in an official title on the Switch). And it’s no shock which of the two games has more profit-seeking junk getting in the way: I don’t want to buy plastic toys, not to mention the actual plug-in amiibo reader, to unlock all the modes and get all the energy tanks. And I won’t. (Besides, the old Castlevania games always had awesome postgame modes where you played as new characters with new moves, but here? Fusion Suit mode is just regular mode, but you take four times the damage. With these controls? Nah.)

I also noticed some slowdown in a few areas. No doubt this is less of a problem on a new 3DS or 2DS. On my old XL, it’s normally fine, except in cutscenes. In some areas deeper into the game, it gets worse. There’s one boss with two big grinders for arms — it’s actually kind of a neat fight, except for a badly communicated weak point and the 3D effects that drag the framerate down to a basically unplayable state. I don’t think the 3D graphics look very good anyway: if they had just done another sprite game with slightly retouched Zero Mission assets — as I assume AM2R did — I think it would’ve looked great and performed better. It’s a sad state of affairs when Nintendo’s releasing titles for their flagship series which barely run on the systems they are, officially, still coming out on. It’s not like I’m trying to play Xenoblade Chronicles here: this is not a New 3DS exclusive. As I already alluded to, it’s not like they can’t do Metroid sidescrollers on the Switch too, you know? Save the 3D for there, and give us some more pixel art.

I actually saved myself some thumb pain by using a homebrew app that sends control stick input from an xbox controller over wifi. This was really cool, and mostly worked well, but sometimes the wifi link would get spotty for a while — it seemed to get laggy whenever I needed it most. And once I got super missiles and found out I had to use the touch screen to switch to them, that made using an xbox controller a little more of a hassle. Still, whenever the going got tough (like with the aforementioned Big Grinder Arms guy) it helped a lot that I could change to a real controller for ten or fifteen minutes.

Samus Returns definitely isn’t afraid to break from convention, which is nice, because Metroid 2 was kind of crude; an almost blank slate to build upon. If you first look at the gameboy version’s map, and AM2R’s, they’re very close to the same size, but with three entirely new areas added in AM2R. These are probably the coolest parts of the game, but for the most part, as one would expect from a fangame, your path through SR-388 stays pretty faithful to the source material. Though I hardly remember much of my original Metroid 2 playthrough, I think AM2R just made some occasional adjustments to the maps to let you use Super Metroid power-ups that weren’t originally there: the general flow and shape of passageways remained unchanged. The map in Samus Returns, though, seems to completely do its own thing. It’s several times bigger. The designers did whatever they wanted.

Some of these changes are nice. Some are problematic, though, like the fast-travel stations. Teleportation points aren’t worth crying about, but it’s better to have an interconnecting map; when you can just warp to earlier areas, it feels like cutting corners. AM2R’s means of getting back to old areas later — by getting shot through cannons into directly vertically or horizontally aligned rooms far away — was much more clever. I feel the same way about the scan pulse: while it sucks when a game hides things in random tiles and you have little chance of finding everything unless you play with a guide open, I think most Metroid games have already been smart enough about showing the connections to unvisited rooms, and marking which ones still hold power-ups you haven’t collected. As long as the game is straightforward enough about which tiles can be blown up, something like a scan pulse should be unnecessary, and I think putting it in there takes some of the responsibility off the level designer to keep puzzles sensible.

Even with this “corner-cutting” designer handicap, though, I had to look up how to get past one type of obstacle: the ones you propel yourself past with power bombs. The idea that you’ll only get launched if you’re using the spider ball to secure yourself just never occurred to me after realizing it did nothing in non-spider form; it’s the kind of thing that they should’ve tutorialized by locking you into a room right after you get the power bomb, making you use it (both horizontally and vertically) to get out. AM2R had some head-scratchers, but everything was communicated clearly. I don’t like looking things up in a Metroid game.

It’s a tough game, even with changes that can make it feel casualized. Power-ups even get sucked in when you’re not holding a charge beam now, which is a little sad. More to the point, though, you can continue from outside of a boss door when you die now. In exchange, though, those bosses tend to be more than you can possibly take without several practice runs, and even then will test your limits, mainly because getting hit by the wrong attack can empty three energy tanks when you have maybe seven total. Blame the framerate and thumb pad if you want, I certainly will, but either way some of those encounters (like Big Grinder Arms guy) took me half a dozen tries. These fights can last several minutes, too.

But to speak some more on departures from the traditional mechanics, let’s get to the obvious one: they gave Samus a melee bash that parries enemies, which is about as out-there as it gets. It’s neat, but you have to play a little too reactively, and I’m not sure that’s the right fit for Metroid. When you’re up close already and need some breathing room, it makes sense to have it, but enemies take far too many hits to die when you’re not killing them with a parry and counter-blast. If you’re not already in a position the enemy will attack from, you have to go out of your way to line one up, making yourself unnecessarily vulnerable, so I found myself resenting the addition at times. It may have been smarter to simply avoid those enemies, but I find you’re usually hungry for at least one kind of ammo or health pickup, or you’re still checking out the room and don’t want to leave the threat there, so it’s best to kill everything outside of a speedrun.

It’s admirable that the game doesn’t try to sell itself purely by repackaging the old, though as I say this I actually feel contradicted by the use of the Lower Norfair music used in the lava areas, which wasn’t actually part of the series until the third game. It seems to have been thrown into Metroid 2 retroactively, for a kind of backwards application of nostalgia. Of course this is a nerdy thing to get mad about, but what better time to actually expand the repertoire of Metroid themes than when you’re going back to a time when there was so little to build from? Like Super Metroid itself did, with such incredible results? (I’ll make an exception for Ridley’s fight music, though a new version of his NES lair theme could’ve been cool too.)

It seems that the only place where Samus Returns and AM2R both really stick to the Metroid 2 gameplay is in throwing in dozens of repeat fights against the same metroid minibosses; first the alpha, then the gamma, zeta, and omega life stages. Both remakes added several of their own completely new bosses, and both got pretty creative there, though Samus Returns definitely goes the extra mile in boss complexity — it probably says something about my preferences as a whole that during a few boss fights in the 3DS game, I found myself thinking about how cool it would be if someone added them to AM2R in a patch. But the miniboss repetition doesn’t seem like something that really needed to be preserved, and we could’ve tried fighting just one metroid at each stage of life — maybe then throwing in a gauntlet of each in a row near the end? — and just getting to kill some more metroids scattered around in their iconic larval state elsewhere — especially in Samus Returns, with its deeper interest in change.

Super Metroid is my favorite Metroid; I think the series only had the barest inkling of what it was supposed to be before then, and it’s lost that inkling numerous times since. I still haven’t played Zero Mission, but it’s safe to say that for players with no existing associations or attachments, AM2R was up there with the best of them, doing what the series has always been praised for, and for bonus points it did it on the best possible hardware, which is to say PC. But this isn’t a review of AM2R. Samus Returns fluctuates between bold and formulaic, but in both the old and the new, it’s a mixed bag. On better hardware, I might have called it a very good game.

The reviewer finds this game hard to get excited about, but still has a positive opinion of it. It may be somewhat fun, having good features or ideas counterbalanced by a few boring parts, bad design or other fundamentally irritating qualities that can’t easily be overlooked. Alternatively, it could be pleasant, but with nothing new to offer. Worth a little money if you’ve got the time for it.


This was a pretty funny game, even though the hero was an insufferable horndog nerd. I liked Zera, as I’m a sucker for magical companions only the protagonist can see, and the voice acting wasn’t bad. Apparently that was a later addition to the game, and it was a little buggy for me: I had to quit and reload sometimes when the audio would stop playing.

Coming from La-Mulana, which didn’t let players choose a difficulty setting, I chose to play on Unepic’s Hard Mode, which disabled autosaving checkpoints. I wished I hadn’t needed to make that choice, since one game’s Very Hard is another’s Normal, and I’m usually in no shape to make an informed decision at the very start. Though I had my share of deaths and troubles, the game was typically on the easier side; every boss had you puzzle out some trick to how you were supposed to attack it, and skill wasn’t usually involved. The trickiest challenges were independent of difficulty, like avoiding falling rocks for 60 seconds. You also have to do a fair bit of pixel-precise platform jumping, which I wasn’t very good at–but that’s mostly a matter of adjusting to the later-than-expected timing and getting over the oddly-vertical jumps. The “game feel”–and I mean stuff like the jumping distances, speed, and hitboxes–seemed a bit awkward and under-polished, but I actually kinda liked it: it reminded me of old ‘90s PC games.

My usual deaths were a result of my greed: wanting to push just a little further when I was running on fumes and could’ve teleported back to the save point, but then dying to a trap or ambush. You can’t pause the game while in a panic and teleport from the menu like you can in La-Mulana. But I appreciate the added difficulty of that, and anyway, it gives value to the endgame skill that lets you establish a recall point that remains even if you save and reload (which respawns enemies). That was my favorite spell. Being able to place a customizable warp point, Morrowind style, even from within a boss room, was a sign of more freedom than most RPGs would allow, although it’s a little trivial by the standards of today’s building games like Terraria.

What was hardest about the game, though, was having to commit skill points to a particular build. It was very easy to screw up my character, and the skills didn’t have equal utility. If I’d known earlier what I know now, I’d have tried dropping Fire and Frost entirely and used Bows. I’d also have dropped Robes and Staves and all melee skills except for Axes, which I never tried and, according to forums, were unequivocally better than the weapons I used. Even though I wanted to try making a magic-oriented character from the beginning, robes and staves had extremely narrow utility. For example, there wasn’t an item that reduced the cost of Alteration spells, or one that increased the casting speed of all magic types across the board. Even after a one-time respec, I made some bad choices, but I was at least able to drop daggers, which could possibly be useful in lower difficulty settings where enemies turned their backs more often, but seemed useless to me in Hard. But in order to deal with all the late-game threats with less frustration, I’d have needed all the points I could get in the cooler magic schools like Alteration, Protection, Light, and Mental (and Arcane just for fun). They were also comparatively overpowered: I got through the last part of the game by polymorphing every enemy into a harmless chicken. They couldn’t resist or dodge the spell, and of course there was no cooldown on it either.

Having to get deep into the castle and unlock better types of magic wasn’t a terrible idea, but it would’ve been far better if you didn’t have to beat the bosses in linear sequence, and if daring players could potion through the tougher areas to nab the best skills at the beginning of the game. I’d have loved exploring randomly and having to turn to one of a dozen other loose ends on my map because some strong enemy totally owned me there. Maybe they could’ve done away with skill points altogether and linked skill progression more to unique pieces of equipment. Some could still be permanent investments, such as from a sidequest where the player chooses one of two rings as his or her reward.

Realism seems to have been a sticking point in the game design: you end up with an empty flask after drinking a potion, swords suck at breaking barrels open, leeches stick to you until you peel them off, you can cast Frost magic on yourself if you’re on fire, and you could even wear up to 8 rings at a time (though none of the rings were especially cool). These realistic touches would’ve been more impressive if the central mechanics and skill balances were in better shape, but I liked some of them, and luckily Unepic stopped short of a hunger system or an inventory limit of two or three weapons. In the case of the potion bottles, it’s probably the best way of doing consumable healing items, as potions can be brewed as many times as you like yet the bottles limit how many you can bring with you, and it’s hard to find the time to drink one in the middle of a fight. I was satisfied to find that I was actually making use of them in my preparations for dealing with certain enemies and challenges, Witcher style.

I played Unepic with the keyboard, sometimes using the mouse in menus, and I couldn’t imagine playing with a gamepad. Hotkeys were troublesome and confusing enough even with a keyboard. I’d have suggested a single horizontal action bar for the number keys, which would swap out to show the Ctrl-# hotkeys when Ctrl was held down and so on. As things stood, I’d have been better off if I rebound movement to WASD (Terraria controls?) and used the mouse to activate hotbar spells. But trying to change keybindings was a nightmare: it wouldn’t detect half my keys, and I’d have to move things off to a blank key it would detect before putting something else on it. Certainly confusing.

I didn’t get into the multiplayer, but it looked interesting. From what I saw, it wasn’t just a tacked-on game mode. I’m giving Unepic a rest now, but if it suddenly shows up in a friend’s steam collection, multiplayer is something I’d be psyched to take a look at.

The reviewer finds this game hard to get excited about, but still has a positive opinion of it. It may be somewhat fun, having good features or ideas counterbalanced by a few boring parts, bad design or other fundamentally irritating qualities that can’t easily be overlooked. Alternatively, it could be pleasant, but with nothing new to offer. Worth a little money if you’ve got the time for it.


La-Mulana is a good one, somewhere in the realm of Castlevania, but with obvious Zelda influences, too. It’s a challenge-oriented game that avoids handholding, but it’s not really so rough, at least as long as you’re not going for some extreme no subweapons, hard mode, speedrunning achievements. You can tank through most bosses’ attacks and kill them without really learning their patterns, and the most physically challenging element in the rest of gameplay is in mastering how to time your jumps when you can’t remaneuver yourself in midair, in knowing how to avoid losing your momentum when you step off platforms, and so on. Jumping works as intended; it was designed to take a bit of practice to pull off a jump of the exact distance needed to land on a platform, but the controls are reliable. Not being able to quickly release from ladders or grab them mid-jump can be annoying if you’ve played other sidescrollers in which you can easily do those things, but is also a deliberate and defensible design decision.

I’m not so sold on swimming, though, which can be incredibly frustrating. If you’ve been hit just once, you can really get Mario Karted by plumes of volcanic rock and your own backwards momentum.

The movement controls are a great fit for the way the stages are actually laid out, but I did at times feel it would have been in La-Mulana’s favor if it had prioritized the “joy” of movement, with faster, more dynamic jumps and Super Meat Boy wall-kicks. A few of the items seemed like they would have been a great opportunity to enhance general character mobility, but aside from the double-jumping feather, most are on the extremely situational end of the spectrum: the place where boring Zelda-esque lock-and-key inventory puzzles live. Also, having an item halve the damage I take is tremendously useful, but HP and defense are just numbers, and those stats are far less interesting than, say, reduced knockback, or an item that allows aerial momentum recovery after a second of flinch, or climbing spikes that prevent sliding on ice.

I also want to call some attention to the unconventional health pickup system, where players have to gather an amount equivalent to their total HP and then they get a full heal all at once, instead of getting it back piecemeal like they would with hearts in a Zelda game. It’s not too hard to suppose what the design intent might have been: this way, you can’t trade hits with random enemies and get the health back at the end of each little skirmish. But the heals are entirely unreliable, unless you calculate and grind out to just short of a full heal so you can get one off a skeleton just outside of a boss door, which is a silly thing to incentivize. An Estus flask or Zelda-style consumable fairy in a bottle might’ve been better choices, but also would automatic full-heals in front of a save point and no healing at all otherwise, since a player can warp around the world all they want at no cost. That stands out as an oddly friendly choice for an allegedly brutally hard game–free teleporting to save points is worth more than infinite free weights and subitem replenishments combined would have been.

Which brings me to a bigger topic: initially, when I started La-Mulana, I was determined to beat the whole game blind; no advance knowledge outside of knowing the basic tenets of gameplay, no guide lookups when I got stuck. I cleared out the first few bosses and then saved my game in a permanently unwinnable state, trapped in the Temple of Moonlight, without the warp item–which is actually a hidden item that becomes available not five minutes into the game and makes things a hundred times easier. I thought I was playing the way the game was meant to be played, but in googling around for the problem, I realized that my getting stuck was, to other people, a quaint hypothetical. After the warp item, the game didn’t seem too tough at all.

Fez was one of my best gaming experiences in the past few years, and I was hopeful that I’d see something like that in La-Mulana: fair, rewarding puzzles that are best challenged with patience and reams of graph paper. After losing a bit of progress in having to reload to that second save slot back in town–when you can’t warp or heal whenever you want, it’s natural to expect risk in keeping only one save slot, deep in a labyrinth with a finite supply of HP and the weights used to prop doors open–I was a little less determined to never check a guide again. I made it to the Gate of Illusion and then had to check a hint guide twice. By I’d finished the Chamber of Birth, I must have looked up a few dozen things that had only kept me stuck for no more than a couple minutes. I couldn’t tell that a texture was supposed to be interpreted as a climbable ladder, so I checked the hint guide. I couldn’t immediately tell which of my hundred notes from faraway levels applied to a puzzle I was dealing with, so I checked the hint guide. I didn’t want to run back to every area of the game in sequence to see what door opened up after killing a boss, so I checked the hint guide.

Ultimately, I think playing La-Mulana blind is a fool’s errand. I’m sad to say it, because I really enjoyed parts like drawing up my own maps (even though they didn’t always turn out to be spatially coherent), and figuring out how to read the La-Mulanese cipher when I had yet to find all of the in-game paleographies. It’s not a bad idea for new players to push themselves to try and get as far as they can on their own first, but ultimately, even the absolutely necessary, main-questline puzzles were prohibitively arcane. And no matter how strong your dislike of handholding in games is, that warp item should’ve been forced into the player’s inventory from the beginning–or should not have been included at all, if you want to be that much of a purist old-schooler.

Some riddling tablets were too far away from the puzzles they referred to; others were given too early; some were too vague to be useful, and only served to pour salt on my wounds when I finally figured the puzzles out by chance or perseverance. It’s not babying or patronizing to show players where a door has opened after they kill a boss or they weigh down a hard-to-reach platform. While I don’t think it’s ideal for a game to focus-test all its interesting edges off, as typically happens to big studio games, La-Mulana is on the level where even a few extremely dedicated friends of the developer wouldn’t have been able to figure certain things out if they took all week trying. My reaction to learning the solutions was too often “Well, that’s some bullshit,” and not often enough “Ohh, I should have thought of that!”

My gut says that some of this stuff was meant to be figured out by crowdsourcing the problem after the game’s release, and not by individuals–certainly, figuring out how to get inside the postgame “Hell Temple” was like that, and I didn’t even get around to doing the temple itself. But the problem with this approach is that an “online community” can only figure something out once–anyone who ever looks online hoping to do such a thing afterwards is really just “checking the guide”, without participation.

Esoterism aside, I can’t complain too much, and most people probably wouldn’t have been as stubborn as me. (My stubbornness seems to vary depending on how prepared I am to like a game, and its genre, in the first place.) The difficulty wasn’t brutal, but wasn’t entirely trivial either, and I had the option to enter hard mode if I ever thought otherwise. There were also some very ambitiously designed (not to say difficult) boss fights, good lore, funny jokes, catchy tunes.

Ultimately there were only a few design issues holding back something fundamentally great. It’s getting to be a bit cliche to say that a game is “not Dark Souls”, but you can hardly look elsewhere for an exemplar of more nuanced approaches to challenging gameplay, and of doing less to incentivize constant looks at the wiki. La-Mulana’s got a kickstartered sequel in production and I’d love to see it take a step in that direction. But what’s here now is certainly worth playing.

This game was thoroughly enjoyed by the reviewer. It is an excellent game that may be too simple or not ambitious enough to be a 5, or there are design flaws meaningful enough to prevent it from enduring as something truly beloved. Highly recommended.