Page 2 of 2

Posted: Sat Feb 20, 2010 12:52 pm
by Smivs
goran wrote:
Smivs wrote:
And can you imagine Zero-G water-polo?
I just did, only with female teams... 8)
Mmmmm...zero-g wet T-shirts!

Posted: Sat Feb 20, 2010 1:02 pm
by DaddyHoggy
Stop it! We're not that kind of forum! :oops: :wink:

Posted: Sat Feb 20, 2010 1:20 pm
by Commander McLane
Ahruman wrote:
Commander McLane wrote:
Oh, crappy pseudo-random generation! :evil:
Trivia: there are 23 873 086 025 possible planet descriptions (before expanding the %-escapes), with varying probabilities. It’s not surprising that some of them are unused. :-)
Well, yes, of course. But I wasn't refering to the complete different descriptions, but marvelling that there seem to be whole areas of the system_descriptions which are unused. In this case I stumbled across one of five items in array #35 and two out of five items in array #34 which are never used at all. Now I wonder how about arrays #0-33? How much "filler" do they contain? Funny descriptions devised by B&B, which due to the pseudo-random generator never actually made it anywhere into the game? For instance it seems that all of the words in #7 (beset, plagued, ravaged, cursed, scourged) sound familiar from the F7-screen. But what about "exuberant" (#19,4)? Has anyone come across it in real F7? Quick search: yes, it exists in a total of four planet descriptions. So maybe #34,0; 34,4 and 35,3 are the only victims of chance?

And of course it is also strange that #34,1 only ever gets combined with #35,4; and #34,3 only ever mates with #35,2. And #34,2 only gets mixed with ##35,0 and 35,1. Would that change in a bigger sample, with different galaxy seeds? Just wondering...

Posted: Sat Feb 20, 2010 3:45 pm
by JensAyton
Commander McLane wrote:
And of course it is also strange that #34,1 only ever gets combined with #35,4; and #34,3 only ever mates with #35,2. And #34,2 only gets mixed with ##35,0 and 35,1. Would that change in a bigger sample, with different galaxy seeds? Just wondering...
Probably. There are two factors at work here: 1. the PRNG used for this stuff isn’t very good, and 2. in any random or pseudo-random series of significant size, you’re likely to get clumping effects.