It's close to midnight, so that means it's time for another Mento's May Mastery. This "do all the homework minutes before it's due in" system has worked well for me in the past, but I wonder if I shouldn't shift the schedule around a little so I'm not writing all these while half-conscious with fatigue. Still, I guess that means I have an excuse for this blog being illegible. Where does that leave me in explaining all the others?

I wasn't kidding about Life of Pixel being a twofer, turns out. While I'm not quite done with it from a completion standpoint, I am pretty close to being done with it from a patience standpoint. It, alas, suffers the same pitfalls (pun possibly included) as many other masocore platformers of its kind, in that by escalating the difficulty (which is an entirely reasonable thing for a designer to do in most circumstances) it unfortunately disrupts the careful equilibrium of its chosen format. I feel like I've gone into this before with other games, but I'll elaborate in more detail after I put a header in the way.

Life of Pixel

There we go. So the thing about Life of Pixel is that the progression mechanic is predicated on collecting the full amount of shiny objects across a stage and then reaching its exit. For its early single-screen stages, this works perfectly. As the consoles evolve and become more elaborate in their stage design, this format becomes increasingly unwieldy and, eventually, frustrating and completely unworkable. There's no checkpoints, because there's no way a system like that would function when the level design is left open - an aspect of the "collect everything" format is that it has you hunt for objects in every direction, and there's the possibility that you would get far into a stage, hit a checkpoint, and realize you left an object behind at the start that you'd now have to integrate into your route on every subsequent attempt. And, of course, with this being a masocore platformer and all, a single touch of a spike or laser or bottomless pit will instantly kill you and force you to restart the whole stage (you can hit twice by enemies before you die, which is a little more lenient, but enemies aren't usually the problem).

You can see where this is going. Let's say instead of ten objects, there's over fifty. It will take around three to four minutes to find them all. You reach about forty objects (or three minutes, time-wise) into this scavenger hunt before getting popped by some rapidly protruding/receding spikes you had no idea were there. Let's say this happens about ten times in a row. That's half an hour, roughly speaking, of zero progress. It's not even like Super Meat Boy, in which you could easily spend that long on a single thirty-second stage if it required pinpoint precision (and you were particularly obstinate), because with Life of Pixel the stages become so long that it doesn't even matter if you've memorized the entire layout and can follow the most economical route in your sleep: the chances of making an error due to carelessness or tiredness or being on edge after dying too many times increases exponentially the longer the stages get in a game with no checkpointing. It's why the latter stages of Super Meat Boy aren't anywhere near as fun as those in the sweet spot (around Hell, I'd judge) because they're simply too long and too easy to trip up on accidentally, rather than because you were facing a challenging series of jumps that truly tested your skills. Put it this way: if you were to average one stupid death by unforced error for every twenty expertly-timed jumps, which is still exceptional if you aren't a speedrunner or cybernetic to some extent, a level that requires forty such jumps (or fifty, or a hundred) will still be close to insurmountable regardless of your excellent batting average.

I ran into this issue with Mutant Mudds as well, hitting situations where I'd spend way too long on one section only to lose all my progress with a simple mistimed hop, though in that game's defense at least grabbing all its collectibles wasn't essential to completing each stage. Life of Pixel's gone from a competent little platformer with a neat historical framing gimmick to a logistical trainwreck of irksome fake difficulty. It's a shame the developers didn't take a step back and consider if they handled the difficulty curve the correct way.

Protip: Same small number of objects per stage, but made slightly harder to reach each time. Done. Not scouring a square mile for an entire jewelry store's worth on a single life. Because that would be stupid and no-one would want to play that.

That said, I'm still enjoying the presentation of these later worlds. We've moved to the 16-bit era (the game's technically over after the initial 8-bit systems, and anything after is considered a bonus world) and are now gallivanting around the Super Nintendo and Amiga, two systems I'm very familiar with (or at least I am with the Atari ST rather than the Amiga, though they were more or less identical). The game's chiptune music is pretty good, especially recapturing the SID and Laura soundchips of the two respective Commodore systems, and I'm also enjoying the way the visuals change with each new system. The Amiga levels, especially, seem to be making a lot of visual references to games of which I'm all too knowledgeable. I include one below:

Overall, while I still think it's a neat little package of nostalgia, I'm a little disappointed in Life of Pixel. Despite being an incredibly simple game from a mechanical perspective (my thanks to user @bdhurkett for correcting me about the double hit points thing. which are there from the offset; this also means that the game's core mechanics are entirely consistent throughout), it still manages to loose a few brown pixels onto its pixel bed in a few ways. Besides the aforementioned mishandled difficulty, certain weird bugs and glitches will make themselves known from time to time, usually involving sprite collision - e.g. if you push a box into an enemy, the enemy might just warp around it somehow and kill you. Many enemies are set to randomized movement patterns, especially the bats and robots, which must bring no end of joy to anyone attempting to speedrun the game. Moving platforms actually lend their momentum to the player character, so if it's going around in a circular motion and is heading down when the player jumps off, they will be pulled downwards to match the momentum of the platform they just left. Ditto for when it's moving quickly from side to side. I'm not sufficiently cognizant of thermodynamic physics to say if real objects actually work that way, but they're certainly not meant to in video games. Like many control issues, it's hard to effectively explain or demonstrate why this feels super wrong, but it will make itself known very quickly to someone actually playing it themselves. If it wasn't for the game's production values and attention to detail in other areas, I'd think these were all issues a pack of game design college students making their first platformer for a second year project would encounter. It's a little perplexing.

Still, LoP has a lot of heart at least. How many games feature a love letter to the BBC Micro or Amstrad CPC? Not that most people have any reason to of course (hey, whatever, the Micro was the first system with Elite), but it's not like you hear those names too often. What next, shout-outs to the Sharp X68000 and FM Towns Marty?