Altar Prototype 1

If you publish Flash games in some capacity and would like to fund further development based on this prototype, don't hesistate to contact me at


LEFT and RIGHT: run
UP: jump
DOWN: defensive spell
Mouse: Aim. TAP button: attack. HOLD AND RELEASE button: big attack.
SPACE: select attacks/spells, see a map.
'f': fullscreen, 'escape': leave fullscreen.
'm': mute/unmute sound (not in fullscreen)

GOAL: Explore the level and get to the exit without dying.

Please wait for this to load - there's no progress bar. Click the flash app a few times to give it input focus.
There's no way to restart, so press F5 to refresh the page to play again.

How to Play Well

   Explore the level carefully - it has many nooks and crannies. Remember to hold down the mouse button, then release it to make a huge attack. Tap space for the inventory, and you can try out different attacks, big attacks, and defenses. This demo is rough, so don't expect much depth here. Make sure to hit space for your map if you get lost.

Design Notes

Systems Thoughts

   I transitioned to this design after Sparks and Dust in 2009. S&D had, to my aesthetics, a fascinating bundle of intersecting rules between player control, monsters, powerups, and levels (my design tastes run strongly towards games from Treasure, like Ikaruga - so simple but deep intersecting rules that yield richly emergent game play grammers). Here, I tried moving in a more straightforward design direction. This was the incomplete result. So, levels, experience points, skills leveling up, an inventory, weapon ammo. Very discrete, and more obviously complex at the outset but without complexity building during play. It doesn't sit well with me, though. Should I return to this game's assets, I'll return more to the direction I went with Sparks and Dust.


   I put up this demo because it demonstrates this rendering technique well in the context of a fleshed out game. Nevertheless, putting it up pains me. Compared to Sparks and Dust, the gameplay is shallow. Much content is roughed in (in-game scripting and dialog, enemies of a variety of types, lots of player attack types, multiple characters fleshed in, lots of traps and in-world interactives, powerups, and so on), but almost all is in an incomplete state. Player-world collision is rough and distracting. Player and monster art is very unfinished. Monster behavior likewise. I don't love the game rule systems. Take it all with a grain of salt.

Graphics Technology

   I wrote more about this rendering approach here and here.

A Mini-Essay: Level Design, and the Relationship Betwee Knowledge and Goals

1. Platformer Levels and Uncertain Goals

   I love that Mario and Doom maps have what I'll call a fractal structure. Main paths are easy to reach and see, wide and well-marked. But side paths and ledges, and areas behind walls (in Mario) or areas behind windows (in Doom), will float at the edge of your perception. You won't be sure if those ledges or side paths are even reachable. But with some cleverness and problem-solving, many will be. Powerups will lazily bob, and enemies will skulk, and traps will gleam, to draw the eye to those corners as well. Some corners will succumb to your exploration more quickly than others. The spaces have a rich, deep, varied density.
   This arrangement keeps players on their toes, active and engaged. Space gets suffused with a kind of vague potential. And heightening that potential is that this is all implicit and non-verbalized, a function of physics and physical space, not disembodied arbitrary game designer proclamations. Where many modern games turn to story, dialogue, and non-spatialized interface widgets so player goals (unparsable in their noisy, art-rich worlds) remain explicit and well-marked, this older model leaves "figuring out what goals could be" as an ur-goal, ever-present. It's entirely a function of how space is shaped and the nature of interactive objects in the world.

Doom and Super Mario Brothers 3

   Play the first area of Super Mario Brothers 3, and with the first coins leading off into the sky, where you can't quite jump and reach, your curiousity is immediately grabbed. Bypass them and you'll finish the level, but there's an nagging incompleteness as you do. Only once you collect the feather powerup and fly are you finally granted resolution.
   Or take the first room of Doom 1, with its giant windows overlooking the open air courtyard. Enticing blue armor sits in a pool of toxic sludge in that courtyard, but no obvious path leads the player to that item. It's a secret; it's off the well-marked path. It lingers in the player's conciousness as unfinished business until the player determines how to reach it.
   Mario could have relied on explicit, disembodied interface information to prod the player to fly and collect those coins. Mario could have injected more scaffolding. But, by relying on explicit communication rather than implict, Mario would have denied players the chance to wonder about every other intersection of objects in the level. It would have trained players to expect that the game explicitly marked, in an attention-grabbing way, everything worth caring about. In such a scheme, if the interface didn't draw attention to an object, that object wouldn't matter. Players would gather that heightening their attention to level details would be a waste of time. All the "stuff" in the world would just art; or, really, just noise. Not signal. Games communicate this assumption to players all the time. But not Mario and Doom.
   Mario and Doom use space in an incredibly powerful way. It's a rich idea, and one I attemped to emulate here (and elsewhere - I spec'ed out many, many such levels while working on this game).