last night i was thinking about making decisions under pressure in games, and how i really like exploring that space as a player/designer. i've neglected it in my designs lately, preferring perfect simple information, so i wanted to make a more concerted effort to get back into realtime messiness.

i roughly thought about three points and scribbled them down in my notebook:

quick thinkingcomplex, hidden informationsoft fail states

>>> quick thinking

there's a timer that gets shorter each wave, and you need to accomplish a number of goals in that time

>>> complex, hidden information

most real-time games have a lot of unnoticed complexity. e.g. the bullets that are shot by the green poop enemy up top have complex positions through time. predicting that is not totally trivial!

also, your goals (the pink box things) are sort of a very tiny microcosm of the NP-hard travelling salesperson problem. in most cases you can usually come up with at least two decently reasonable paths and don't have the time to decide which one is the absolute fastest. so you compromise and pick one under pressure.

>>> soft fail states

i wanted to do "resources" but i think the important part of resources is soft fail states. rather than a crisp, clear indication that you made the wrong choice, i'd prefer costs that you're always balancing.

in the short-term, i think the bullets in this game sorta accomplish this: bumping into one doesn't kill you, it just has a kind of unpleasant disorienting and often time-wasting effect.

"soft fail states" doesn't mean "no hard fail states", it's just a renewed focus on these "bad" interactions that increase complexity rather than reducing it.


a note on my process for this game: i sat on my couch and pictured the game in my head for about thirty to sixty seconds and figured out what i needed to do to make it feel right. (the previous night i had built the jumping player and the level "generator".) i wasn't actively thinking about these bullet points, and a lot of what i'm saying here is only retroactively applying the pattern of my "pressure" bullet points to this one particular game.

Made withUnity


Log in with to leave a comment.

I managed 15 :) The decision to pick up the last resource at the left or right edge is usually a good one. Also sometimes had better luck waiting out a second of the timer before taking the last resource. A quick restart option would be nice (currently have to f5).

An ability that would make the salesman problem more complex and enable more planning could be teleports. Press space to place a marker, then press space again to teleport to the marker. This could use up N resources, where N starts at 1 and increases by 1 each use.

Another thing I'd love to see would be a gameover that is not instant. Maybe make the last second bullet-time and after time runs out momentum is preserved like the soft failstate when hit by a bullet? That would allow for more 'close calls' and 'last breath saves'.


Misc information:

>> My high score is 18 waves (I can usually get 16 -- getting to 18 was basically a fluke)

>> I'm working on this project on github, if you like source code? It includes the latest iteration of mynavdi sub-engine :P