[Coco] Tandy Color Computer 3 RND Function

[James Jones]

What you're doing is called "Monte Carlo simulation". It's been around
quite a while, and can indeed be done on a modern PC. BASIC comes from the
days of linear congruential pseudorandom number generators (some aren't
bad, but then there's the infamous RANDU from the 360/370 Scientific
Subroutine Library). A lot of work's been done since then. Nowadays they
accumulate "entropy", measuring things (often having to do with I/O) and
using those values as a source of randomness.

There are a number of tests for random number generators--there has to be,
because any individual test can be faked out. A bucket test can be faked
out by a "random number generator" that just returns the value of a
counter. A range test like the original poster's can be faked out by a
"random number generator" that alternates returning a number close to 0 and
a number close to 1.

On Thu, Mar 26, 2020 at 1:51 PM Steve Ostrom <smostrom7 at comcast.net> wrote:

> And this all angles back to my original problem with the marbles in the
> bag.  I wasn’t looking for a true probability answer, but just an estimate
> of the true probability.  And the Coco gives us this tool with RND.  If you
> run enough iterations, and your program is set up correctly, you can get
> this good estimate of the various probabilities of these problems.  You
> really can’t do this on a modern PC, running Linux or Win10, like you can
> with the Coco.  The Coco is simple to operate but quite powerful.  Even if
> RND is not all that good, it’s plenty good for these problems.
>
> --- Steve ---
>
>
> Sent from Mail for Windows 10
>
>
> This is exactly right. Randomness is lumpy.
>
> --
> Coco mailing list
> Coco at maltedmedia.com
> https://pairlist5.pair.net/mailman/listinfo/coco
>
>
> --
> Coco mailing list
> Coco at maltedmedia.com
> https://pairlist5.pair.net/mailman/listinfo/coco
>