Hi, this is Mr. Gimmick!

As you can see he’s pretty small. That’s because he was made in 1992 for a NES game with 240p resolution.

If we wanna play his game on screens with higher resolution, we have two options. Scale him up, or, make our screen emulate a lower-resolution screen.

Scale him up

The fun thing about scaling him up is that there’s all this information in between the pixels that is unknown to us:

There’s lots of options of how to fill in those gaps. Build him out of triangles, or blurry circles, or try to interpolate and recreate an underlying drawing the way a DAC turns audio samples into sound, like ScaleFX does. You might have seen this option: building him out of big squares:

That’s not necessarily the default or only approach. That is one specific algorithm called “nearest neighbor”. Each empty pixel is given the color of the nearest “real” pixel.

Importantly, there’s no single one answer how the artists who made each game would want their work scaled up. Super Mario Bros. 1 was made on graph paper, which Super Mario All-Stars was made to reflect with its assumptions that pixels are square. While Chrono Trigger was made for wide pixels, by artists that knew how the TVs of the time would display that game. But that’s just two games. The creators of those games left clues in the game (by clues I mean the coin logo in All-Stars, and the moon background in Chrono Trigger) on how they were making the art for their games. For most other games we just have no idea.

It’s okay that you, the player, set the game to look in a way that you yourself feel is comfortable to play and lets you enjoy the game.

ScaleFX which is less blocky and more vectory is to me a very relaxing and inviting look for 8bit games, especially new-to-me ones where there’s less need for nostalgia. It just makes me wanna play the games for hours. (I haven’t found a scaler for SNES and Mega Drive I like as much. I do use ScaleFX after an mdapt pass on them first, but it doesn’t as feel as 💯 as it does on say SG-1000 games or Master System.) ScaleFX doesn’t move around or melt or distort any of the real pixel data. It just makes different assumptions on how to fill the gaps between them, by a more detailed interpolation algo than just big blocks.

Emulate a screen

The other approach is to turn your screen into a lower-resolution screen instead, by emulating exactly how that older screen would look through a magnifying glass. NES and SNES famously used a “double-strike” rendering method which made it look like there’s black lines between the scanlines. And pixels weren’t square, they looked like sideways glowing ellipses almost. While on older Game Boy systems you can see a small cell divider between each square in the matrix.

I’m really grateful for this approach because sometimes nothing beats that magic nostalgic feeling of looking at the video game the way it looked back in the day, or if you really do prefer a CRT screen but can’t have one for space reasons this can be a way to make what you have look like what you want.

But I’m really sick of the approach online that this is the only way and that all other approaches is “destroying” the games; it’s 100% fine if you personally treasure the screen emulation approach but the level of vitriol and attacks against the scaling approaches is what’s not as welcome.

In the CRT era and the original DMG-001 era, I loved the games but they were so tiring. I’m grateful can emulate that look for the occasional nostalgia hit but I don’t want to make it my bread and butter. I was so grateful when TFT LCDs came around! I personally liked the GBC and original GBA so much more than gaming on a CRT.

Screen emulation also kind of assumes that the screen that is doing the emulating is some sort of platonic neutral that doesn’t have any traits of its own. I’ve seen expensive high-end FPGA Game Boy clones advertising a 615 PPI resolution vibrant IPS screen… and then it defaults to emulating the blurry old dot matrix green yellow retro screen that so many people hated back in the day! I personally kinda liked the original Game Boy screen unlike CRTs. Not sure I’d be able to go back to it though. It’s funny how many hours I logged on the original DMG Game Boy, then after Game Boy Pocket I couldn’t go back, and then after GBC I couldn’t even go back to the GBP (even though I tried a couple of times because my GBC had a DC hum audio problem, as did my GBA, a problem that my DMG and GBP didn’t have).

Newly made screens can also have all kinds of interesting traits, like they might be RLCD or even memory displays, or someone might be gaming on an actual CRT. Scaling up the game in a way that suits your screen might be the best or only option then.

Preservation vs nostalgia vs play

When we try to preserve classic movies we don’t always try to add extra film grain or lower the framerate to the silent era’s 16 FPS or jam extra chromatic aberration and VHS tracking errors onto it. We just want the movie itself.

Sure, movie players can add that kind of stuff or we see those kinds of effects be valuable for flashback scenes or a retro vibe (just as how in indie games pixel art is often great for low-memory, smaller artist teams etc along with the aforementioned nostalgia hit), but when you just want the movie and just wanna watch the movie for the movie’s sake, no-one is gonna bite your head off saying “you’re destroying the movie”. (Outside of that one HBomb video when he tried to convince everyone to go back to VHS tapes for movies.)

There’s also some value in hybrid approaches. The Sega Mega Drive famously had an issue where transparency effects wouldn’t work on an RGB TV connection, only on composite, at the expense of color fidelity, but with mdapt+RGB you can get some of the best of both worlds. Or, people from the scale-up side of things can have fun with some glowies or other effects originally created for screen emulation.

Point I’m trying to make here is as per ushe that we’re condemned to choose; that we can case-by-case it. You can use one setup on your handheld and your daughter can use another on her TV. Or even different setups for different games or even different moods.

Too late to fix now but I wish the button labeling on the SNES was rotated 90˚ clockwise.

On NES, A was for acrobatics and B for brawn or battling or bullet speed. Usually jumping with A.

On SNES they kept the classic NES grip except the labeling was garbled so Y was “the B” and B was “the A”. And the A was now specials like the weird twisty jump in Mario. Instead of being the most basic button it was relegated to being a weirdo button.

Made infamous in Super Mario All Stars where X or Y was “the B” and A or B was “the A”.

So far no good but then on Virtual Console they made it even more confusing because now the A was the A again and B was B. And it was unremappable. But they did a smart thing: they made X also be an alternate “B”, so you could have the literal A=A and B=B (like Ditko intended) or you could have the NES grip with A=A and X=”B”. A perfect compromise solution to a problem that never shoulda happen in the first place if they had been three percent more sober when making SNES.

But then on Super Mario Bros. Wonder they brought back the problem and returned to the All-Stars layout. And you can’t even set controls indivudually on multiplayer—everyone needs the same layout.

Two years ago a couple of friends (who don’t know each other) both convinced me to get an iPad, since I needed to join huge Jitsi meetings while sick in bed, but there’s been a lot of drawbacks being on a proprietary OS.

One fun thing is that they recently added support for Nintendo JoyCons so I started playing Starfield. I know, I know, proprietary video games are immoral—I managed to go like five years with only ScummVM, Gargoyle, and retro emulators, but here we are. I immediately started kvetching about not being able to invert the X axis, only the Y axis, until I realized it was possible to invert either at the OS level, which solved my problem completely.

But the curiosity of why someone would only wanna invert the Y but not the X sent me down a rabbit hole.

I’m calling “inverted” what I want, and “regular” what I don’t, just like those people who (unlike me) peel bananas upside down acknowledgingly call it “peeling it from the bottom first” or similar. In other words, I am using the now-common nomenclature where tilting the stick forward to look down is called “inverted” even though that was the default & only option back in the day.

Reg: “grabbing the frame or reticule & dragging it” or “stick is controlling a dot in front of the viewer”
Inverted both: “stick is in back of viewer’s head” or “grabbing the game world itself and moving it”
Inverted Y, reg X: “stick is on top of viewer’s head”. I guess.

I was playing a lot of video games in the N64 era.

Super Mario 64 for example has both X and Y inverted (it wasn’t called inverted back then obv) and so did The Ocarina of Time, with no option to change it (and then the re-released version had it reg instead, again with no option to change it).

Then I noped out during the GameCube, I was only playing retro games then. NES and handhelds. And then I did get a Wii but there I used the motion controls to point where I wanted (like in Twilight Princess).

So the first modern “3d-with-thumbsticks” game in the 2000s I ever played was Breath of the Wild. I know, I know, sounds wild to miss seventeen years of games. I went into the settings before even starting the game, and I saw options to invert X and Y. So I thought: “OK, here’s what I’m gonna do. I’m gonna go into the game, and as soon as I get to control the character, I’m gonna try to look up, just let my thumb’s natural instinct do whatever it wants to do when I think “look up”. If that leaves me staring into the floor, I’ll invert Y. And then, I’ll do the same for “look left”.” And doing that experiment led to it being super clear that I needed invert both.

I’ve sinced tried to read up a bit and found that that almost no other games have an invert-X option and many games don’t even have an invert-Y.

And I’ve read up some of how it works. If you wanna move a reticule on the screen or the frame of the screen, reg X and reg Y makes sense. For example, The Flame and the Flood doesn’t let you invert but since it looks more like a 2d game with panning, I don’t have a problem. But me as an invert, I think of 3d games more as having my thumb on the back of the camera. I move my thumb higher to look down, further left to look right, I am “looking” with the thumb.

Using a touch screen has a similar dilemma. On those ugly & dinky “scroll wheels” some mouse used to have, you’d scroll the wheel down to scroll the viewport down. But on touch screens you instead drag your thumb or stylus up in order to scroll the viewport down (because you’re not actually moving the viewport, you’re moving the underlying browser canvas).

I 100% respect people who use reg X & reg Y, and people who use only one of them inverted, and people who like me invert both. And I’m sorta familiar with the history of both inverted (the N64 generation basically) and the history of neither inverted (sorta. I dunno, Xbox?). Reading up it seems like “only Y inverted” came from mouselook on PC and I vaguely remember that making some amount of sense with a mouse. That clears it up.

Still a li’l strange to me that we went from one thing being the unchangeable default to the complete opposite being the unchangeable default 🤷🏻‍♀️

So I had played like three seconds of Popeye on a home computer (guessing the C64) at a distant relative’s house and had two Game & Watches (Life Boat and a non-Nintendo space shooter, both of which I loved), but the first real video game I really played was Duck Tales.

Loved it, amazing game, kind of sad to start with one of the best games of all time since it was all downhill from there, but point of the story is that I play a couple of levels, get to Transylvania, start talking to Webby, and… I have no idea how to move forward. I was stuck on the same screen for an hour.

Because I didn’t know to press A to continue the dialogue.

I must’ve accidentally hit A without realizing it because I was stuck again the next time, a li’l bit shorter since that time I did figure it out.

That was winter of 90/91. Today, 32 years later I was playing OlliOlli World and first it took me a while to realize that once I’m done with making my character, I need to back out all the way using the B button (the actual making of the character was a li’l bit tricky, too; it says that A “selects” which I at first thought meant confirms that choice. “I don’t wanna confirm this dorky beanie, aren’t there any hijabs?”—and then, some select screens with sub-areas (like selecting your start runs for example) needed one more A to start selecting than the highlighting seemed to indicate, and throughout, “B” for backing out also means confirm).

Then after the first level the visual cues seem to indicate that A would restart the same boring push-five-times tutorial but nope. This time it’s A that’s moving forward. I remember that the original OlliOlli game, which I completed, also had misclick-prone UI with some unintuitive menu flows.

Conclusion: UX does matter in game dev. We want the game part to be challenging but we also want the systems around the game to lead us into that fun part of game.

Link’s Awakening DX was released on Switch (the version with depressing color scheme and the can’t-skip-dialog–bug for Pieces of Power and Guardian Acorns, and some other de-charm tweaks like how dialog changes and stone tablets are now that annoying owl) but comparing DX to the remake on the same device, it’s remarkable how much better all of the DX audio is (instrumentation is crisp and clear, compared to the more muddled sound on the Switch Remake, with “the dropping a toyshop down six flights of stairs” instrumentation of Animal Village a nadir of the latter) and how much faster and more fluid the DX controls are. And also shorter—albeit non-zero—loading times on DX.

The remake only lets you use stick to move, the DX version you can play with either stick or D-pad.

Ironically using the DX version with stick feels much nicer than the remake with a stick. The remake fixes the sqrt(2) issue where in the original, moving diagonally let’s you move 41% fasterthan normal. There’s also no diagonal animation. So small movements and adjustment feel fast and nice on DX whereas on the remake you’re a freight train that needs to get oriented and then get going.

That is primarily a visual illusion since, looking at your core box, it’s not that slow on the remake, it’s just that the visual jitter makes it feel like an ordeal.

Weirdly enough, it’s all the parts that aren’t traditional Super Metroid that I like the best about Dread. Dashing, sliding, mêlée counters… This is Mercury Steam so of course there are gonna be QTEs.♥

QTEs can be really awful for two reasons (or I guess kinda the same reason explained two ways). One, they make gameplay feel samey. They reduce all gameplay to Dance Dance Guitar Hero style “time button X right Y plus to this beat please”. Two, they make gameplay feel very symbolic and extradiegetic: they lessen the feel of “OK, I’ll shoot, I’ll bash, I’ll run” and turn it into “I’ll hit A, then X, then down” cold and boring reality.

In Dread, they’ve fixed it (some of this started in Samus Returns). All timed events are X button, all timed events are bashing with Samus’ arm, all bashing is the X button. It’s a 1:1:1 correspondence which turns it from feeling playing Parappa da Rappa (which their earlier game, Mirror of Fate, did) to “Aw, I’ll bash it when it gets close enough”. I.e. awesomely tense and immersive.

I also found the sneaking and dodging and running from E.M.M.I.s to be really fun.

I loved all the sliding and bashing and dashing. The Aeon flash ability was also really fun. I like Igavania games, I like how the control feels as you’re dodging and weaving and flying around.

What caused me to nope out was the Shinespark and Grapple beams. Sorry, Super fans, but I just never could get used to them and this implementation doesn’t feel better, it has all the same problems.

Then all the problems of traditional Metroidvania gated exploration started rearing their heads. The old “You need a Frobnicator Ray to pass these Frobnicated blocks. Is this awesome? y/n” problem.

It’s no secret that I have a problem with traditional Metroidvania. Many of my favorite games are Metroidvanias, but that’s in spite of these features, not because of them.

So then I dug out my old copy of Bloodstained: Ritual of the Night, which I still haven’t finished, and played it some more, making it from just after one boss to the next. I was having such a blast with the way Miriam moves and the richness and solid implementations of all the items. And, while Dread’s controller felt like it was getting more and more cluttered and control-meta-cokebottle as I was unlocking more and more beams and rays and aeons, RotN’s controls are elegant and flexible with their shard setup, swapping shortcuts, and weapon techniques. And then I beat the next boss on the first try just as I had the last, and then died in a death spiral from some corridor weenie. Ugh. I guess RotN has some balance issues. I’m really bad at video games for having played them for so long, but, I might’ve stumbled onto a spell or something that’s really good against bosses.

Oh well. Back to visual novels and card games.♥

It helps me understand how to play it if I think of it more as some sorta hyper-charged Cave Noire than in the Wario’s Woods / Magical Drop family. Since it’s not on the beat, it doesn’t feel similar to Cadence of Hyrule at all either. I was still struggling but I happened to get Shield Knight as my first unlock and I’ve been playing as her. She has a Bofuri style build which is easy to play and made the game flow a little better (and riskier). I then unlocked Specter Knight but I’ve been staying with Shield Knight.

I like the game overall but it does the thing some rogues do (or as the game eyerollingly refers to them, “roguelites” with a “t”), where you unlock stuff permanently for later runs. That doesn’t disqualify it from the genre, but it’s not my preference. It takes away some of what I think is fun about rogues, that every run is a fresh start. The game is absolutely designed around this and it wouldn’t be an easy change to make (unlike Cadence of Hyrule which does have a setting to make it that way). For example, unlocking knights.

I’ve made it to the Chromatic Caverns. I play with 1 stock, rogue style, but again with Shield Knight which to me is an easier character. While I’ve only seen three knights so far, it seems like they did a good job with balance. In Mr Driller Drill Spirits (which this game feels less like than I would’ve guessed from videos–again, the closest relative is probably Cave Noire, maybe with a little Witch and Hero mixed in, both in terms of overworld and Defender-like flow of enemy herds. If I have to pick a puzzle game to compare it too, Chain Shot! comes closest, but not that close. Again, as I was watching gameplay footage, I was like “this looks like Mr Driller with a shovel” but it doesn’t feel that way at all), the unlockable drillers are so much weaker than Mr Driller. More fun, so I’d often use them, but at a huge hit to my success rate. I like that that isn’t the case here and that Shield Knight led to the game feeling comfier and (for me and my play style) easier.

A problem with the design is that in the “camp site” (a between run area sort of like the backpack in The Flame and the Flood) you can unlock relics (which work similar to the relics in Slay the Spire, i.e. they’re a huge and vital part of your strat) you later can buy in the main game when you find Chester (a character from the Shovel Knight series). However, he gives you a menu of three relics to shop from. So, as any deckbuilder player knows, there’s an incentive to not unlock bad relics because they will dilute your choice of good relics. This is a permanent mistake, i.e. persists between runs. I’ve steered clear of the ones I don’t want, for the most part, because I realized this right away, but then I misclickingly unlocked one I was just checking out.

Overall, the game is addictive. Normally, I get the most addicted to variety games like Wario Ware or Rhythm Heaven, and here there are some nice bonus games on offer like Mona’s reflex game (sometimes they can take over, like Kirby’s Brawl Ball (a pinball bonus game) in Kirby’s Mass Attack. I had no idea that game was gonna be in there but as soon as I unlocked it, I gave up on the main game and only use the cart for Brawl Ball, still one of my most played DS games on my 2DS. I busted the R-key on my DSi marathoning it) but here the main game is on its own super addictive. Some levels flow well and others are frustrating. It’s sometimes unclear what’s going on, like stepping on water (which to me seem indistinguishable from floor tiles).

Overall I like the game.

Sometimes a sequel makes the original worse just by comparison; The Addams Family Values is just such a better movie than the one it followed, and New Super Mario Bros. is comparatively the weakest in the NSMB series which all outshine it. But when they first came out, The Addams Family movie and the original NSMB were awesome! They were just overshadowed, almost deprecated, by even more awesome sequels.

That’s not what I’m talking about here.

Instead, take something like The Matrix Revolutions. The original was a potent metaphor for how culture shapes us, for how “fish can’t see the water”, for how a dominant ideology—a hegemony, even—is something that affects us in ways we’re not even aware of. (Frustratingly, the movie’s imagery is used not only by my own side of the Overton window, but also the opposite, to the point of it’s basically ruined for all time by all the “redpilled” jerks.)

The third movie comes across as if all the philosophy in all three movies is only relevant for the specific diegetic conceits of the trilogy. It makes it all about the proverbial midichlorians.

The first movie still works on its own, but the second movie becomes completely pointless. A lot of us liked the second movie, The Matrix Reloaded, but a lot of that was because it seemed to be building up to something interesting, which was ultimately a letdown. It’s not that Revolutions is a bad movie. It’s that it’s just a science fiction adventure movie.

Another example is Analyze That, which made me realize that “Oh, I guess that in this and in the original Analyze This, they were making fun of mental health issues all this time.” Uh, I get that more perceptive people probably realized that as soon they saw the first movie. That’s fine, that only shows that “the sequel makes the original worse” is such a personal journey.♥︎

Sometimes, a sequel or prequel can make the original more compelling. The Perry Mason prequel that came out the other year made me more interested in the original books because of how it really humanized the characters in a way that was missing. This sort of overlaps the concept of shows that take a while to become good, like Adventure Time or Star Trek: The Next Generation.

Of course, ideally we want both parts to be good right away. I love an original that works, that’s fantastic on its own, and then a sequel that has really changed with its audience, subverting and destroying the original in a poignant and meaningful and ultimately respectful way. Archie and Afterlife with Archie comes to mind; zombie apocalypses are a dime a dozen but when the cast is a set of characters that I love, it hits that much harder and more meaningfully. Another example is The Secret of Monkey Island and its sequel Monkey Island 2: LeChuck’s Revenge. The tonal changes were shocking but, for me at least, they worked perfectly.

In BSD robots (which I’ve played since my early teens—we had a Windows port of “Daleks”—and never really figured out), I’ve always wondered if you can teleport into trash heaps (and accidentally telefrag yourself that way). Checking the source code just now, it turns out you not only can’t, you also can’t teleport into a robot! However, the robots move after you teleport so when you die after porting, that’s because you landed right next to a robot.

So the myth is that the japanese Super Mario Bros 2 was considered too difficult for the western market and instead they reskinned another game and that’s what we got. Then in the SNES “All-Star” collection just a few years later we finally got the Japanese original, named “The Lost Levels”.

OK. But in reality…

A lot of the levels from “The Lost Levels” had been released here. Just not on the home console, it was only for the coin-up arcade (“VS Super Mario Bros”). So obviously not “too difficult”. That was one factor, but “too similar” was a much bigger issue, including the six levels that were reused from “VS”. Overly similar-looking sequels were frowned upon in the post-crash eighties video game landscape. That’s also why Castlevania II and Zelda II were so different. But yeah, there was a brief trend of making western versions easier which, as the rental game blossomed, quickly changed to make western games harder to screw over renters. Like Ninja Gaiden III where the western version is almost boringly difficult with how far back you need to start over, compared to the Japanese orginal.

Yes, Super Mario USA, a.k.a. “Mario Madness”, was built on an advertising game that was given a narrow release in conjunction with a television corporation’s tech demo festival. “Dream Factory: Heart-Pounding Panic” (or “夢工場ドキドキパニック”), starring the mascots for that particular television manufacturer’s expo event. But that “Heart-Pounding Panic” game itself started life as a Super Mario Bros sequel!

They started working on a sequel to Super Mario Bros, couldn’t get it done in time, rushed out a glorified level pack (on disk only) as “Super Mario Bros. 2” (adds wind, poison, and a separate Luigi), and then resumed work on their sequel, and then they were hired to quickly put out the ad game and decided to use their original SMB2 prototype to do that, gave it a disk release that received a 31 Famitsu score. So they reskinned a Mario game into “Heart-Pounding Panic” and then, when Nintendo of America asked for a different take on SMB2, Nintendo of Japan insisted on reskinning “Heart-Pounding Panic” back into a Super Mario theme which was released in 1988 in America. That Mario version of the game was later released in Japan as well, on a cart called Super Mario USA in 1992, only one year before Super Mario All-Stars and two years after Super Mario World and the Super Famicom. But still.

And characters from it such as the “Pokey” cactus showed up right away in SMB3 while others, like the Shy-Guy, didn’t appear again until Super Mario World 2 for the SNES.

“Mario Madness” a.k.a. “Super Mario USA” is a real Mario game and arguably a sequel to SMB1. That’s not to say that Lost Levels isn’t fun too or that it didn’t deserve it’s FDS “Super Mario Bros 2” moniker. It’s also great. I’m glad there were three true sequels on the NES (LL, SMUSA, and SMB3). I’m sad the NES couldn’t have a longer life. I don’t feel ready to upgrade to SNES yet… It’s S-SMP sound subsystem sounds like screaming into a pillow under water compared to crisp square waves and triangles of the NES.

I dearly love many Metroidvania games. But now that there is a fandom and a very strict definition – stricter than I would’ve liked – I want to speak out against those two design principles: ability-gating and guided non-linearity.

Some games are great because they forfeit those principles, some games are great despite that they are hampered by those principles, and only a rare few gems are well designed with those principles and use them to their advantage.

My fondest memories of Metroid are the platforming, the fighting, the exploration. My fave Metroid game is the 1991 Return of Samus and I have fond memories of trying to find hidden nooks and crannies with the space jump and the spider ball and traversing the weird terrain, finding metroids hidden in deep sand etc.

My fave memories of the Igavanias (which I do love – Dracula’s Curse is my fave Castlevania but I like both the classic and the iga style of Castlevania) isn’t “Oh, I guess I can’t go here until I get some sorta super jump or water traversal power.” It’s the tight boss patterns, the equipment/spell optimization, the feeling of discovery as I enter new areas.

Link’s Awakening was my favorite Zelda up til the release of Breath of the Wild. (Yeah, yeah, I know that it’s not a metroidvania because it’s not wholly side-scrolling.) It is very ability-gated, more so than the 1986 Legend of Zelda. But I don’t love it because of the ability-gating, I love it in spite of the ability-gating. None of my fondest memories of it is “oh, now I have this item, I can finally go here”. The feather and the lift-bracelet are fun to use, not just fun because they are “keys”. It’s fun to jump around and to throw stuff around.

I’ll concede that ability-gating was part of the reason why many of those old games are so good. Ability-gating is a tool to curate progression which can lead to a gradual reveal of story. Some of my fave moments of Link’s Awakening are due to the fact that the owl conversations, the stone tablets and the trade chain are revealed in a certain order.

But some of my other fave games, like Breath of the Wild (again, not side-scrolling) and VVVVVV, don’t have ability-gating. You can just go anywhere, do anything. There are micro-stories throughout the otherwise open world, pockets of linearity. In VVVVVV you’ll stumble upon a series of linear rooms. In Breath of the Wild you’ll find places where there are locked doors that you need to unlock. These are used to create narrative, both story narrative as well as (more importantly) ludonarrative gameplay arcs.

There are also games I love that aren’t even non-linear; they can be level-based like Celeste or Dracula’s Curse, or single-screen like Clu-Clu Land, the 1980 Fire, or Tetris.

Ability-gating is a way for games to try to have the cake and eat it too. Sometimes it works, but often it just feels forced. The hammerable poles in A Link to the Past felt especially egregious. The hammer felt like such a superfluous item (especially coming from Link’s Awakening [which I played before A Link to the Past] where you can flip some with your shield) only there to act as a “key”. If you want to put in linearity in your game, just do it. Like maybe the defeat of a boss (I’ll save the “I wish more games were non-violent for another post, because I’ve been really grooving on non-violent games like Fire, Snake Pass and VVVVVV lately) opens up a new area, or you need to find literal keys, like Cave Story. Or design it to work open world, maybe with linear segments, like VVVVVV.

What’s wrong with ability-gating

Two things:

When you don’t know for sure that an area is gated

There are some spots in Hollow Knight and plenty of spots in Alwa’s Awakening that are really hard to tell whether or not you’re supposed to be able to make it up there. I wasted hours in Alwa’s Awakening trying to jump up a ledge or over a lake (combining a few of the powerups I had at that point) only to give up. And then later find another powerup that made that spot trivial. “Oh, had I known this power was coming I wouldn’t have wasted my time trying to reach that unreachable ledge…”

Especially coming from games that made the gate really clear (like the unscalable wall in The Messenger) where there are some very difficult sections to traverse but the only way to do it is to try over and over again (or be 上手 (jōzu) af, which I’m not). I don’t have to waste my time trying to scale the wall, I just know I need some item. And I am rewarded putting 20 minutes into making a difficult section of cloudstepping because that’s the only way to get to that shard.

So that’s one way to do ability-gating “right”, to make it really clear it’s gated. This can sometimes look super forced, like the beforementioned hammer in A Link to the Past or the upgrades in Duck Tales 2. “Hmm, this suspicious hook, and this suspicious ‘iron block’, seem to be here only for the purposes of some item I don’t have. Gee…”

The other, even better way is to make it completely hidden, to grant an ability that’s a complete WHOAH! I didn’t know I was going to be able to do that!

“What, I can walk on water? What, I can now turn into a painting and walk along the wall? What, I can now fly?!” to open up areas that you never even considered to be reachable.

When the powerup needs to function both as key and as gameplay

The best acquired abilities are when the item is in and of itself

1. fun to use, but
2. not essential enough to the gameplay to be available from the start.

The Messenger satisfies the first criteria. Once I had the abilities I really loved stringing them together to complete difficult challenge rooms and bosses. You upgrade your li’l ninja to become a core part of a very well designed game experience.

Buuut it kinda whiffs on the second criteria since new game+ is much more fun since you have all your powers right away.

Having so many restrictions on what constitutes a good acquired ability along with so many restrictions on what constitutes a good key leaves an intersection of very limited design space.

Separation of key and ability can lead to a good design. Again, Return of Samus is an example of where yes, there are some ability-gated areas, but there are also areas that are gated by other things.

It sucks in an otherwise ability-game to get an item that is just a key. “Oh, man… just some boring Book of Mudora, I was hoping to be able to fly around or skate or something else cool”.

Summary of practices

Consider not gating the progression at all

Wholly linear and wholly non-linear games can both work just as well as “guided non-linear”, depending on other factors. If you have a good platforming game mechanic, maybe it doesn’t need to be yet another Metroidvania.

Snake Pass is an amazing game but it’s a clean, level-based design. There are gated areas (opened by levers) but you have all your abilities from the start. The levels are tiny open pockets, micro-sandboxes, that you can complete in separate sittings. But if you have to gate the progression, then…

Consider not separating the keys from the gates

In other words, have gates that you open up by completing a challenge right by the “gate”, without needing a key.

You can have a gate behind a boss room that opens up once you defeat the boss that’s in there. You can have linear sequences (like VVVVVV) reachable from a larger, open hub area. But if you have to have ways to expand your hub area in many non-adjacent directions…

Consider separating “keys” from “power-ups”

The 1990 Super Mario World is an example of this done right. Switches reveal blocks and open doors in a way that is is 100% wholly separated from your character’s abilities. You never get disappointed finding a key switch and you never get disappointed finding a cape or fireflower.

I’d argue that Cave Story also does this right. It’s a tight linear game with exploration segments between finding the right keys.

The acid lake in Return of Samus is another example.

But if you have to have abilities that unlock new areas…

Beware the “you can almost make it there”

Never tease. Never put a ledge “juuuust out of reach”.

The good way is to make it immediately apparent that they can’t get there yet.

The awesome way is to make it such a part of the scenery that the player doesn’t even consider that they can go there.

The first good way is a bit boring, it’s immediately apparent that there’s no way I can make it past there, I just need to wait for the right power…

the second is the “Wait, what… does this mean that I even can go here? I never even considered that!” out-of-the-box style ability. But these kind of mechanics are rare finds and maybe your game would be more fun if you had that ability earlier. Super Mario Bros Usa (SMB 2 Mario Madness) gives you the power to lift things up right away. It doesn’t hold on to this ability in order to use it as a gate.

Closing words

Uh… maybe I put that li’l summary in a bad order. It leads you down a path to do do ability-gating after all, just as long as you follow this or that best practice. But in the end I want to reiterate the first point there. Consider to not use this kind of gates in order to guide progression. Often, either just opening it up wholly or just making it linear is better.

PS. As I was editing this essay after posting it I found that I already had another post on this topic. But I re-read it and that post and this post don’t have a lot of overlap. Phew, what a relief! I’ll forget my own head next!

So core to the style of video game called Metroidvania, Igavania or Metroid-like games is the idea of gated areas and then items to unlock those gates.

But would’t it be pretty awesome to have a game in that style but without those gates?

A 2d sidescroller (sprites or polygons, doesn’t matter) similar to something like Metroid or Symphony of the Night but where you could go to Mother Dracula right away if you wanted.

Combine the openness of Breath of the Wild (and to a lesser extent A Link Between Worlds) with sidescrolling 2d!

You’d still have an incentive to explore the Castle Planet because you’d want items, powers, become stronger or just because it’s fun.

Hearts vs XP

Metroid and Zelda makes you stronger through hearts (“tanks” in Metroid) – you’re more likely to survive more dangerous fights. But you can still have a chance against the most dangerous enemies in the game.

That’s much better for an open world experience than XP that some of the Igavania games use. We don’t want areas that you can’t meaningfully interact with because your attacks are too weak. We do want areas that you might think are risky or dangerous when you are low on hearts.

This is the one innovation that Dungeons & Dragons added in their fifth edition, going to a Zelda-like “you just gain more HP, and more attacks, as you level up” over their previous system where you needed to level up in order for your attacks to even hit at all vs stronger foes. (Castlevania and many other video game series were inspired by D&D, with the Famicom edition of Simon’s Quest even using the cover art from a D&D book, I6 Ravenloft.)

Theme and gameplay

I’ve had an unrelated idea for a video game for a long time, and maybe it’s a good idea to mash it up with this one, or maybe it’s better as a separate project, but here goes:

Inspired visually by NES era ninja games like Ninja Gaiden, Shadow of the Ninja, and Wrath of the Black Manta, especially the cut scenes and varied visuals of the Ninja Gaiden series (the old ones) , but inspired gameplaywise by… WarioWare! Your goal as a ninja is to infiltrate a castle and you’d do that by playing well-designed, discrete minigames that you’d branch in and out of as you explore the castle. Going into the air ducts and you get the air duct controls, trying to disguise yourself as a cook doing kitchen work and you get the cooking controls etc etc. With platforming controls as the glue between those sections.

Ideas for minigames:

• Dodging arrows
• Kicking and fighting enemies that approach you from all sides, Game&Watch style
• Disguising yourself – matching the appearance or mannerism
• Following someone through a maze without being spotted
• Trying to avoid being spotted
• Solving puzzles to unlock doors
• Doing the work associated with various disguises – cleaning, cooking, soldier drills
• Poisoning food
• Trying to charm your way through a dialogue similar to the alien translator minigame in Rhythm Heaven

Again, that might be a fun separate game and not necessarily be this game, but… that’s at least a separate idea.

A new old take on Rogue

The original Rogue-style games (of which I’m most familiar with Nethack) generated a large world for you to explore. It didn’t have the emphasis on short runs that later roguelite games like Cave Noire added. Instead, it was a full, multilevel affair with many nooks and crannies.

So one idea is a randomly generated Castlevania castle, but the entire castle (and if this is too amitious, maybe for the sequel).

And, (unlike Nethack, where you get one try at each dungeon) you can type in custom seed numbers! You can keep on exploring the same castle over and over again, continue when you die etc. Let’s say you roll up seed 141251. You’d then play at that seed, try over and over again, continue, respawn (perhaps quick respawns a la VVVVVV, one of my fave Metroid-like games). You’d still be in the 141251 Castle Planet for as long as you wanted. And when you wanted a fresh challenge you’d roll up a new number. You could type in the same number as a friend, sharing ideas and progress without either of you being able to resort to external walkthroughs and guides.