Note: I wrote this in November 2005 as an email on a game-design mailing list. It has been lightly HTML-ized but is otherwise intact. The game in question, "To Hell With Johnny" was written as part of a write-a-game-in-72-hours competition, and, as I recall, some people on the list had thought it pretty compelling.
Ok, what the heck, I'll comment, although I'm not that pysched by THWJ and found Falldown pretty boring, here we go:
Any of you who've read my GDmag columns may have noticed I like to start off by providing historical context. I often do this to point out that an idea is tried-and-true but forgotten or ignored. In practice in computer science, if some idea was first published by X some years back, if people are currently using the idea, it really can almost always be traced back to X.
This doesn't seem true in game design. Most game designers don't seem to have much historical perspective about games, although the proportion may be quite different between "average" game designers and the best game designers. So I have a different reason for pointing out historical antecedants: games are still wishy-washy, touchy-feely subjects in terms of our analysis, so I think have a larger sample pool to draw upon when talking about a style of gameplay may be useful. So my point isn't "this is unoriginal", just "hmm, what can we say".
In the case of 'To Hell With Johnny' and the extremely similar [*] Falldown (and its sequel Falldown 2), I'd trace their historical antecedants (even if there's little actual casual chain) to the original "racing" games where you dodged objects coming down the screen. Since the faux-LED minigame version of this in System Shock was called (IIRC) "Race", I'll refer to those games as 'Race'.
[*] It may not seem extremely similar at all; I hope that the historical context will make clear why I say this. The non-historical discussion later will delve into the differences, anyway.
So the basic idea in Race is that you can move back and forth, and avoid being hit by the falling objects (which may have fallen at different speeds). Your only UI control was the ability to move your avatar left and right.
You can imagine inverting the gameplay: instead of dodging the falling objects, you're trying to hit (catch) them. This was a classic gameplay style of the same era; one famous example was Activision's Kaboom for the Atari 2600, but the gameplay predated it.
But there's another way to invert it: instead of having single objects descending the screen that you're dodging, you can invert that object into a wide bar with a gap in it that you must go through. There was an 2600 cartridge, one of the 60-different-games type, which had gameplay like this that was 'skiiing': you moved back and forth at the top and had to go through the one gap in each giant rising bar. Generally you could see the next bar and its gap as you went through the previous one, so you could try to optimize your movements (although IIRC it used a paddle control so it had fairly rapid direct positioning). [I don't actually recall, but I imagine some of the other game variations included the more traditional Race-style games.] I don't know what the game was actually called, but let's just call it 'Ski'.
Ski game and Falldown are extremely reminiscent of each other. Both have extremely simple, abstract art. In both you are constrained solely to left-right UI. A bar rises with a gap in it, and you have to go through the gap. But Falldown adds a second dimension to your movement, even though you don't have direct control over it. In Ski, your avatar must be aligned with the gap when the bar reaches the top of the screen. In Falldown, your avatar can align with the gap at any time while it's on screen (essentially). (Another closely related design to this basic idea of Falldown are things like the levels in SMB3 where the screen scrolls at a fixed rate while you still have to jump from platform to platform and avoid being crushed against the left wall.)
It probably wouldn't take much coding to turn Ski into Falldown or vice versa; it would take a lot of code to change Falldown into To Hell With Johnny. Nevertheless, I think in terms of pure abstract gameplay, the leap from Ski to Falldown was bigger than the leap from Falldown to Johnny. (Although, as I said before, I don't think the designers were actually making this leap. They probably never saw Falldown, much less Ski. In fact, if it derives from anything, I'd guess Motherload ( http://www.xgenstudios.com/play/motherload/ ) [a, umm, arcade-RPG-with-no-action? A cross between Dig-Dug and a 4X game? a 3X game? Ok, rat hole. It's an exploration/resource-aqcuisition game where you return to town to resupply and upgrade.], and that's more a general aesthetic sense (going down underground) rather than game mechanical.)
Because the leap here is opening up that 2nd dimension even while keeping the UI constrained to just left-and-right. (In the interests of context, I should mention that abstract flash game checker showed me with just one button of input; you were always turning up or turning down. But it's not really that interesting; you could implement FD or Johnny with just 1 bit of input producing either left or right, and to hold still you have to rapidly toggle it.)
The difference is that Falldown does almost nothing with that extra dimension, and Johnny leverages it in a lot of different ways that makes it a far more interesting game, at least until the speed overcomes my speed of planning.
Here's how Falldown uses the 2nd D:
Back when I first played Falldown, I didn't think of Ski, but I did notice that I didn't like how it sped up. As it sped up, it just moved the bars up the screen faster, but you never saw more than one bar at a time. Clearly equivalent but more interesting is to have the bars move at the same time but become more closely spaced--and this was what Ski did, IIRC. So I'd argue that Falldown is an ok concept but poorly thought through (I guess you can't call this 'tuning', but I'm not sure what you do call it. 'Revision' in writing.)
THWJ I believe just speeds up, but that's actually the _right_ thing to do when you have narrow bars and wide gaps. The equivalent mistake in THWJ would be for it to not speed up, but instead for the platforms to become more and more widely spaced vertically (which would reduce your planning/reaction time more). Keeping the platforms at a fixed density and speeding up the descent works much better.
So, my conclusion is that THWJ radically expands the gameplay possibilities from Falldown to make a much more interesting game with a surprising amount of interest and choice from only left-right controls. But I find the gameplay a little too unvaried and the constant tension monotonous as well to think it's much more than a demonstration that this mode of gameplay can be entertaining. But I think I knew that from the SMB3 levels.
I discussed the gameplay/mechanics but I forgot to comment on the obvious aesthetic differences between THWJ and Falldown.
Do we have a word for "everything that's not gameplay"? In this case I want to talk about the visual aesthetic and the setting (i.e. the textual backstory in the readme) together. I'm going to call this the "context" (that is, this is the context provided for the abstract game mechanics).
Although there may not be anything wrong with abstract aesthetics, I think THWJ is a pretty strong example for the whole "people are more interesting" argument. Even if it doesn't really matter while playing that you're pursuing your kidnapped girlfriend, I think it's a nice (albeit cliche) touch that probably gives some people (not me) a little more motivation to try to win. And clearly seeing a guy on the screen is a huge win in terms of "oww"ness and peril. So yeah, hands down THWJ leverages the superior production value successfully.
Ok, I never actually used the word "context" I defined. Nevermind.