Re: [stella] entropy & randomness

Subject: Re: [stella] entropy & randomness
From: Kevin Horton <khorton@xxxxxxxxxxx>
Date: Wed, 10 Oct 2001 18:18:58 -0500
At 10:26 10/9/01 -0400, you wrote:
>The _TTL Cookbook_ only goes up to 31 stages, and they suggest bits 28 and
>You should call the random # routine at least once a frame, even if it is
>not being used.  This harnesses the randomness of the player to your
>advantage, since it will skip many states when the program does not require
>a random number.  Say, you start at the same random # seed (see below for
>setting one on startup) every time, depending on which frame the player
>pressed the start button, he will start in a different place unless he has
>*good* timing and can hit on the correct 1/60th of a second to get in the
>same place on the sequence.

FWIW, I did exactly this with INV many moons ago - heck, it sounds like Kevin's
reading from one of my posts from back then :) Although I just used a static

Actually, this is how it was done on Kevtris (for Colecovision) 6 years ago.

seed, since more detail really isn't necessary -- the random-bit-every-frame-before-
start deal was sufficient.

I did that on Kevtris as well. Seemed to work good enough because the randomness also depended on how long it took the player to "land" each block (due to the generator being called once a frame).

(Speaking of tetris... Nintendo's version on the NES (dunno about the Tengen ver) uses the same seed for the LFSR, and the blocks always come out in the same order. I've always wondered if this was good or not for the player; I would think that after more than 100 pieces, it would be hard for the player to remember the exact sequence.)

- Archives (includes files) at Unsub & more at

Current Thread