Monday, April 17, 2017

The racing line, F-Zero X, and Super Meat Boy

at 7:00 PM
I was thinking - if you designed a platformer around a racing line, would the result be Super Meat Boy? This was a spinoff of trying to apply the concept to platformers and quickly coming to the conclusion that I like it even less as an analytical tool. Dustforce is another obvious one, but anything from that hardcorer genre or whatever they call it. Crungeformers?

Basically, when you're looking solely for route optimization you're asking the player to minimize dynamism, and Super Meat Boy et al. just come straight out and say anything that isn't exactly right is death. I mean, if the player is always trying to avoid walls and ice and pits, why even assign them different dynamics, right? It's like Gamasutra's approach treating all hazards as "off course". A traditional racing game compensates by ramping up the mechanical complexity (or, to look at it the other way around, the mechanical complexity of traditional racing games requires restraining dynamism in order to present a coherent game space), which drives greater suspension, which multiplies the effect of any dynamics, even at a very simple level. Boosting makes F-Zero X seem incredibly dynamic, but increasing acceleration for a fixed time isn't really that complex.

I guess that accounts for the unique design space the meatformer occupies. High speed, high reflex route pruning with immediate payoff. It's stripped of all the suspension and context that nonlinear car mechanics create and makes it such that you actually can define, and quickly test, a meaningful racing line.

Also, I realize I'm not exactly sure what "dynamics" means in the Critical Design model (since it's any word at all in the English language, I'm assuming it's in there). I'm just thinking of it as second order effects. As in, accelerating/braking/steering are mechanics because they are static input/output mappings; machines can always accelerate/brake/steer. Boosting, drifting, ice, etc. are dynamics because they change how the mechanics work (acceleration higher under boost, steering less responsive under drift, etc.). I haven't put a lot of thought into applying this definition outside F-Zero.

Wednesday, April 12, 2017

Vectorman revisited

at 7:00 PM
Hm beating Vectorman really just came down to getting a 5x1Up. That was enough to ride the wave to a few 2Ups and 3Ups and ultimately the end credits, even though I'd never seen the final four stages. I guess it was neat how the penultimate stage was set in the first stage of Pulseman, in the context that I was debating which to beat first. Now Pulseman will feel like a fun spinoff! Which incidentally has the same problem of a too-close camera.

Question I'm surprised I don't hear more often: how come V-man is yellow during cutscenes and green during gameplay? Paint shortages? Where's Super Mario Color Splash when you need him?

Monday, April 10, 2017

Super Turrican vs. Mega Turrican

at 7:00 PM
I picked up Mega Turrican after doing the deed on Super C. Poking at them in emulation, Mega and Super always seemed close in quality, but everyone assured me Super was The Good One. Not so much the case. Mega Turrican is every bit a traditional Big Showy Genesis Run-and-Gun, and a legitimately great one. The levels are strongly conceptually driven (the sewer starts out dry and gradually fills up with water! the junkyard has steep mountains with enemies falling from above! the Aliens level has tight angled corridors that neutralize your weapon while shit jumps out at you!), the excess mechanics occupy unique design space (morph ball is the fastest way to move and allows dodging otherwise unavoidable attacks! grapple is the only way to reach certain areas! there isn't a laser beam that freezes every enemy in the game!) while still providing strategic variety (grapple can be used to short-circuit platforming! weapon choice influences aiming!), and the whole game looks and sounds awesome.

Super Turrican is fun, but it never gets over its weird identity of Mega Man meets SNES-style maze platforming (I take this to be the canonical Turrican identity, but it doesn't mean I like it). Enemies feel meaningless since clearing the path is only part of the battle, and the freeze beam doesn't add dimensionality to the shooting the same way grapple does for Mega's platforming. In fact, it flattens the design, since it turns every enemy into a hunk of stone. Individual levels still have a unified sense of purpose, but I'm not sure they go deep enough to describe as theme - they're more like well-done gimmicks. The conveyor belt level gradually transitions through three types of conveyor, upping the hazards and narrowing the pathing as it goes, but it never comes together as more than "the conveyor belt level". Note that shooting doesn't in any way inform that. Mega Turrican gets more mileage out of organic structures and combinations - emphasizing a grueling climb up a hill of trash with tiny enemies bouncing down, then transitioning to a bunch of huge wrecking balls you need to duck through enemy-filled pits to get past (but of course you bounce off enemies, so getting in the pit is complicated!). It's probably easier to deliver these high-interplay scenarios when the player is railroaded on a single path... but that's just an argument for Mega Turrican having a better approach. Even Super's great ice level fails to coalesce into a conceptual space beyond the slipping gimmick (although isn't it neat that each game takes its best level from a Most Reviled Platforming Trope?)

Mega feels more complete than Super, from the dynamic setpieces (the crazy intro to the junkyard where you fly through the atmosphere jumping between spaceships!) to the platforming robustness provided by the grapple (jumping around in Super feels really boring afterwards), but also in content. Understandably, Super's final level was cut, but even ignoring that it feels like an alpha version or a game's worth of filler - doubly so when the big moments like Master Hand and Alien Train turn out to be recycled from Mega (or possibly earlier series entries?). Its level themes are ruins, volcano, factory, ice, Alien - a fairly generic set - compared to Mega's lab, sewer, junkyard, Alien, factory, which doesn't sound a lot more interesting, except that they all draw from the same techno-mechanical aesthetic and give the game an overarching personality.

There are some minor quibbles with Mega, like the difficulty balancing being off to the degree that I smart-bombed the final boss to death (both phases!), or that the later levels are long as fuck and meting out vital morph ball energy over 15 minutes is a lot to ask even from a seasoned player, but all told I'm kinda surprised it's never been reclaimed as one of the hidden treasures of the Genesis, being so characteristic of the library. I get why Turrican fans frown at it, but it complements RKA, Gunstar, Ranger X, etc. quite nicely. And it's a hell of a lot better than that goddamned Revenge of Shinobi. It's also got the best 16-bit cinematic this side of Phantasy Star IV.

I don't know why I wrote this really long review of Mega/Super Turrican. Maybe because I was surprised how different they are and that I ever assumed Mega was just more of the same. Maybe because HG101's coverage is entirely about how the Amiga port has different backgrounds.

Monday, April 3, 2017

Level knowledge in Soldier Blade

at 7:00 PM
I was practicing a bit of Soldier Blade over the weekend, trying to focus on just one level at a time and learn the layouts. It's hard to compare shmups over a long span of time, as even minor skill improvements can reframe an entire game. But looking at just the few I've played recently (Axelay, Cho Aniki, R-Type, S&P if you want to count that), two things stand out about Soldier Blade.

First is that it's really fast - not an original or particularly interesting revelation. But compounded with the 45-minute length (longish, compared to the 30-min norm), it means tons of content spewing from just a single loop. Not that I've taken the time to count waves, but it at least feels like three of R-Type's levels could fit in one of Soldier Blade's.

That said, the excess is packed with a hefty supply of popcorn*. Admittedly, Axelay and R-Type are self-consciously stripped down games, and Cho Aniki is... something else in terms of level design (not good), so Soldier Blade's ratio feels closer to average. It's not even bad popcorn - most of it comes at weird angles that make your weapon level relevant (always a bad time to have Blue 1) - it's just that at the rate the game moves, it adds up quickly. It isn't afforded the same level of variation as the primary waves (nor should it be if it needs to remain low-engagement), so you see the same couple waves in every level.

A third note that I knew already, but contributes to the same conclusion: the x-axis orientation of some enemy patterns is determined reactively by the player ship's position. That is, if Soldier Ship is on the left half of the screen, a string of popcorn may stream in from the top right corner, but if positioned on the right the same wave would come in from the top left. Since this behavior is specific only to certain waves, the stage doesn't just flip-flop - relative enemy positions will actually change.

What this adds up to is a game resistant to idle memorization. I don't consciously commit shooters to memory (I don't know that I could), but after a degree of practice it becomes instinctual. With high speeds requiring high response time, enemy spawn points derived from player position, and a relatively limited weapon swap, Soldier Blade places value beyond the obvious on knowing what's next. Yet it readily defies the patterns that make this knowledge a subconscious product. The layout density creates an excess of information to store. The uniform popcorn puts just enough space between distinct encounters that the concatenation of "Wave A, then Wave B, then Wave A, then Wave AB" requires an investment of active memorization (while fighting popcorn: "Wave C came last, Wave C came last, Wave C came last"). And the reactive spawning does more than mix things up - it adds pre-wave activity to the necessary data set for layout prediction.

It's an effect well-coordinated with Soldier Blade's presentation and balancing to make the game feel especially lively and engaging through repeated plays, moreso than just being "fast". Particularly satisfying is seeing this kind of spontaneity evinced through techniques not reliant on randomly generated elements or structural flattening.

*"Popcorn" is a term used by shooting game enthusiasts to denote waves of one-hit-point enemies that typically don't fire projectiles. For example, the first few waves of any given stage of Gradius. Popcorn isn't meant to defeat the player, instead it keeps them lightly engaged while creating some breathing room. Think of it like the foam packing peanuts in a shipping container - not what you're after, but necessary to fill out a stable, symmetrical structure. I came up with that metaphor because I thought the packing material was called popcorn, then as I wrote it out I was like, wait.