I haven't written much about Super Meat Boy, but it's not because my fingers and hands are too ruined to type (although that's nearly the cause). It's a fantastic experience, basically the Platonic platformer, I just didn't have anything to add that was more substantive than "Yeah, it's really good."
Earlier this week though, I was listening to Jorge and Scott's conversation about grinding and it reminded me of something Kirk wrote about "skill grinding." Jason linked that piece and referred to it in his review of SMB for Paste (and thus the circle is complete). What Kirk calls "skill grinding" is basically referring to progress in a game that emerges not from a character earning experience points and gaining levels, but from the player themselves actually getting better at the game's challenges. Character improvement vs. player improvement. Kirk was writing about this in the context of Demon's Souls, but it applies even more so to SMB.
What I think SMB illustrates is games about player improvement operate best when the loop of the player attempting a challenge, failing and being able to try again is as tight as possible. In SMB, you barely have time to angrily sputter a new concoction of profanity before you're back in the fray. This allows the player to dash themselves again the rocks as much as possible, and while that sounds frustrating to some, it's actually the best way to improve. The frustrating part isn't failing, it's waiting to try again. Or worse, it's being forced to overcome a previously surpassed challenge just to reach the one in question simply due to arbitrary checkpointing. The satisfaction and improvement comes from analyzing the situation, attempting a solution, seeing what didn't work, refining it and trying again. Anything that sits in the way of that will make player improvement slower and less satisfying.
Of course, putting up barriers to player improvement with intent and purpose can actually be used to good effect. As the game that inspired Kirk's discussion of this topic in the first place, Demon's Souls certainly doesn't have the tight loop that SMB does. And failure in Demon's Souls is far harsher than SMB, where death is swift but relatively without consequence. As such, Demon's Souls feels far tenser than SMB, with much higher stakes. The only time I've found SMB to be particularly tense was a few of the longer levels near the end and when I finally neared unlocking The Kid. Demon's Souls plays the player, asking them to constantly weigh banking the souls they've got versus pressing on for more. Demon's Souls isn't superior to SMB or vice versa, but I'd wager more people could pick up and improve at SMB. Demon's Souls demands a certain ... dedication.
In the opposite direction, games that communicate progression of a character do so through lengthening the loop between moments of progress. Character progression is all about contrast. Progress that comes too quickly muddies the waters, making it so the moments of progress seem less impactful and distinct. Again, this style of progress isn't inferior to player improvement, merely different. Mass Effect 2 uses it quite well, going so far as to decrease the number of skill points you gain at each level. It certainly feels like fewer than Mass Effect 1, and the parade of "SpaceCorp Gun I-VII" is largely toned down as well. Whenever your Shepard progresses in Mass Effect 2 it feels substantive.
The trick bit with character-based progression is that it's easier to abuse than player progression. Turning the reward progression into an ever-increasing time commitment that offers little in return otherwise is dangerously nearing Skinner Box territory. And many free-to-play games make showing up really the only condition for success, but if that success still isn't coming fast enough, you can pay to increase the rate of reward.
I'm not going to say one is better (some of my favourite games are RPGs that are primarily about character progression), but there's something that feels very "tactile" about player progression in games. It's a manufactured skill you're improving, but honestly, you have learned to do something better. It's a made up something, but is it any more arbitrary than, say, juggling? I'd like to see more player progression games with a feedback loop as tight as Super Meat Boy. We've seen plenty of strong character progression games, but the trick bit with them is they almost always need something else to lean on to stay engaging (Mass Effect 2 as a pure shooter would be ... lackluster). Creating a great story or expert pacing is hardly a given for many studios, even very competent ones. I'd love to see more folks experimenting with player progression games instead. As SMB demonstrates, this kind of satisfaction is hardly played out.
And of course, the player-character progression dichotomy is only one axis a game can be satisfying on and it doesn't need to be particularly prominent in all games. And it's certainly a spectrum (multiplayer games with level advancement blend these things interestingly, and I might have more to say about that in the context of Assassin's Creed: Brotherhood). But I'll drop a stake to say player progression works best with a tight loop and character progression works best with a lengthy. Deviating from this can be useful, but it should be intentional. And it should certainly never be a punishment. I don't think many favourably remember Too Human for exactly that reason.