(no subject)

[mattlistener]

This is an interesting toy from a scientific perspective.

The process is akin to trying to get a self-sustaining nuclear reaction going. Every moving particle has N% chance of triggering another particle, sometimes two.

The boards it leaves behind are akin to a ferrous metal cooled in a magnetic field -- regions of identical orientation.

The longer this random process goes on, the more ordered the board appears at the end. That challenges a common-sense understanding of entropy -- how does the starting board have more energy than the ending board? Analogous to crystal formation (warm, disordered solid dissolved in liquid loses energy and forms cool, ordered crystals).

The corners and edges are dampeners due to their exposure to non-reactive locations. So a board-spanning reaction will always take the form of a substance cooling from the outside in.

Game-comments... the best board to start from is one that has few snakes and doughnuts, and few white bands where pieces point away from each other on each side. It's good to start in an area with many layers of surrounding pieces tending to point inward. Nirvana is a reaction in which several areas of activity are turning pieces toward each other.

Sometimes pieces turn more than 90 degrees. I believe this happens when on arrival at 90 at least one of the reactive surfaces is pointing at a piece that isn't done turning, in which case an opportunity to react on that face would be missed. It would greatly dampen reactive areas otherwise.

It's impossible to create a looping condition. Any set of tiles that might be proposed to be involved in a loop have corners among them which would shut off. If the board edges wrapped (ie it's a doughnut), this form of proof wouldn't work. Yet my sense is it'd still be impossible to get a looping condition, since the ordered regions always get larger and fewer over time. Would like to play with that!

My best score: 2605

Comments

[pinkfish.livejournal.com]

My observation of the more-than-90 turns is that the current policy dampens the reaction considerably (I have seen entire rows turn right past a row of waiting edges, which were not triggered). In fact, all of the plans I have made to get high scores were based on rows triggering each other. It was quite frustrating to watch the edges roll right past the point where they were to trigger, with nothing happening. My high score was only 21xx.

[mattlistener.livejournal.com]

Interesting -- my approach has been to hit reset until I get a board I like, and then trigger one tile. That's how I got the 2605. I can get a 1500+ every few minutes.

[chanaleh.livejournal.com]

Thank you for these notes! I've been mulling this structure over as I've played the game the past couple days, but without enough scientific background to make this kind of sense out of it.

[mattlistener.livejournal.com]

Glad they were appreciated. :-)