GDC 08: Nuances of Design



Yesterday, Jonathan Blow organized a talk called "Experimental Gameplay Sessions," where he pointed out some interesting, experimental indie games. I posted links to those games, but intentionally neglected to explain them for the purposes of forcing you to play them. This game may be made by some Seattle Game Design School students, it was playable enough glitches aside.

That's not what I'm doing with the games mentioned in the "Nuances of Design" lecture. Before the talk started, Blow walked around with a USB drive of all the games he'd be talking about, so we could play them as he described what made them so special or unusual. Similarly, I'm gonna post you links to the games he talked about (the available ones, anyway) before describing their importance.

So, hit the jump for that. 


A game whiich is almost unfairly difficult at every single turn, flywrench by messhof has some really interesting gameplay mechanics, but also seems determined to piss the player off as much as possible. The discordant soundtrack is anything but encouraging to the player, the flight mechanics are absurdly punishing (one flap does almost nothing, two flaps makes you hover, three flaps shoot you up ridiculously high), and the level design just gets harder and harder.

Yet, the game ties the mechanics of movement to where you can go; you can't pass through white bars unless you aren't moving, red bars unless you're holding the flap button down, or green bars unless you're spinning. This combination of movement controls with goal achievement give the game its kick, despite the fact that it's really damn intense and frustrating. With that in mind, Jonathan Blow created:



While not yet available at the time of post, Blow promised to upload nicewrench sometime within the near future. nicewrench basically takes all the mechanics of flywrench, but removes the frequent death and intensity. Rather than focusing on simply getting to the end of the level, the player is meant to explore the playspace and grab stars. Rather than dying upon touching a bar of the wrong color, the player simply bounces off. It's flywrench, but calmer.

Which is why it isn't as good. However frustrating or unfair flywrench is, it's still pretty damn fun, while nicewrench is kinda dull; in creating nicewrench, Blow showed how different two games can be while still sharing the same central mechanic.


Stars Over Half Moon Bay

It's not out until next week, but Rod Humble's Stars Over Half Moon Bay is a very meditative, symbolic experience. Humble was present to explain his game, which surprised me given how intentionally minimalistic the actual game is.

Essentially, the player leads a black and white ball around a starry night sky with the cursor. As a wall of darkness slowly rises from the bottom of the screen, the player can make the black and white balls run over stars, creating a long (but temporary) trail of stardust. The player must then take the white and black balls to the darkness. At the point the balls hit the darkness, a box appears whose size is relative to the length of the stardust trail you had before moving into the darkness.

The wall of darkness rises and the player continues to drag stardust into the darkness, until the entire screen becomes black save for the little gray stars the player has amassed. The blackness then moves away, and the entire night sky becomes blank save for the little gray stars the player created, with which he can then create and share constellations.

Basically, the whole thing is a huge metaphor for creativity; you take aspects from the outside world and internalize them, then implement them, then release them back into the world. It'll make more sense when you play it.


Mr. Heart Loves You Very Much

Zaphos created this game for the Gamme256 contest, and he wanted to play with the idea of non-contiguous space in a puzzle game. Once he came up with the idea of this non-contiguous space (in case you haven't noticed by now, you can push the individual rooms around), he had to deal with how to create clever, interesting puzzles around such a mechanic.

Ultimately, Zaphos decided that each puzzle should only have one solution, leading him to abandon several previously-implemented mechanics. You used to be able to jump really high, rotate in midair, and even create clones of yourself who could then be used as bridges (a la P.B. Winterbottom). Once he found these aspects made the game too easy and nonlinear, he axed them. 

EDIT: Everything I said was pretty much wrong. Here's what Zaphos was trying to communicate in the seminar, from the man himself:

"Apparently I fail at communicating ... I guess it wasn't clear in the talk, but I didn't decide that puzzles should have only one solution. I just decided that _one_ strategy to build puzzles that surprise the player is to (1) make something 'loose' with likely many (unknown) solutions, (2) find the solutions and (3) make 'tighter' puzzles around specific solutions you find. I do think restraining to one solution makes the game easier to think about as a designer, but I don't think it's a general "should" of puzzle design. It certainly wasn't meant to be the general take-away from the talk!

Also I never implemented the 'a is for asexual reproduction' thing; I ditched it because it was hard to think about and hard to code, and Mr. Heart just a two-weekend project. I was just mentioning it as an aside in the talk because I thought it was a fun thing to think about.

So, yeah, apologies for apparently saying stuff that I didn't intend to convey at all.

(Odd that the stuff that I just stuck in the talk as random asides, labeled as interludes, is apparently the only stuff people got out of the talk? I guess the lesson is not to include that stuff?)"


Crayon Physics Deluxe

Interestingly, creator Petri Purho had a completely opposite philosophy to puzzle design: he loved the idea of there being infinite solutions to any one of the levels in Crayon Physics, which may be why the game won the iGF grand prize. Anything else you need to know about the game can be more or less extrapolated from here.



You're cooperating...with yourself. This sort of gameplay forces the player to think in a completely nonlinear way, forever planning ahead and using their time wisely.



Timebot takes a similar time clone mechanic from cursor*10, but tweaks it a little bit and makes the game eight hundred times harder. Rather than having characters lives based on time, the player can make clones at any moment they deem necessary; however, you can only make a limited number of clones, and they will always repeat exactly what you did while you stay in the same position you were in when you hit the "clone" button. The game gets ridiculously hard as it progresses, but, like flywrench and nicewrench, shows how different two games can be with very subtle changes despite containing essentially similar central mechanics.


You are logged out. Login | Sign up



Anthony Burch
Anthony BurchContributor   gamer profile

Lead writer of Borderlands 2, curator of  more + disclosures



Filed under... #GDC



You're not expected to always agree, but do please keep cool and never make it personal. Report harassment, spam, and hate speech to our community team. Also, on the right side of a comment you can flag nasty comments anonymously (we ban users dishing bad karma). For everything else, contact us!