A study of Hut Outcomes

[sirsnuggles]

Yeah, I wonder what version Solo was using. If he's using FW or Gold that would explain the differences and I think that your studies would still reconcile.

Yeah, I reloaded the original civ2, and it doesn't include the extra advances.

Yet, perhaps Xin Yu was still correct, however. Perhaps the game engine's default calculation is based off of 100 (even though only 93 exist in the rules.txt). I think this plausible, because it would provide a natural explanation why Microprose (if that's who they still were at the time), added 7 Extra Advances rather than some other number.7 They knew that it was based off of 100, so they completed the @civilize to 100 (an 8th extra advance might have thrown the game engine out of kilter; thus a reason for their warnings about changing the rules.txt file!).

Anyway, I've really enjoyed your work Samson. I've got to read the rest of your posts now.

[solo]

The plot always thickens!

My memory is that it was MGE being used for my tests and that Samson is almost never wrong!

[samson]

My proposition is quite testable.

If we add 100 more extra advances to @Civilize in Rules.txt (bringing the total techs, in MGE, to 200) I predict the chances of getting Alphabet from a scroll-tip (with no starting techs) would then be 114/200 or 57%. This assumes that there isn't some limiting factor on the number of total advances one can have in @Civilize.

[sirsnuggles]

Now that would be very interesting to see! Please try it.

I think I read once, in one of the Microprose handbooks, that there was a limited number of technologies available. But I could be totally wrong (or Microprose could've been wrong too!)

[SlowThinker]

Usually Microprose handbooks are wrong.

[War4ever]

wow 100 more techs?

[samson]

Originally posted by solo
The plot always thickens!

My memory is that it was MGE being used for my tests and that Samson is almost never wrong!

Well, I was this time. I tried adding extra techs to the 2.42 Rules.text to see if it changed the chances of Alphabet and it didn't.

Apparently, the denominator is hard-wired at 93 for Civ2.42 and CIC and at 100 for FW and MGE.

[sirsnuggles]

Thanks Samson. I love your work. So, be encouraged, all of this discussion has indeed proven your previous work. Your data has been reconciled with Solo's too!

[Cobra]

To steal Advances with brain!

I didn´t read this whole thread before, so I made my own tests and came nearly to the same results as Samson and Xin.
Because it fits straight about, here still some references, when it makes sense to steal an Advance. To steal an unfavorable Advance is rather a handicap. With an embassy everyone can see all Advance which can be stolen and and decide for himself, if it really makes sense. Exactly as with Advance finding in huts, are also with Advance stealing the probability unequally distributed. Again in mentioned list all advances which potentially can be stolen are marked and the difference to the above indicates the probability in percent. So the theory. I made test rows with 100 attempts, with which only 3 Advances were to be stolen. The first numbers are the expected results (x/100), the second numbers are the received results
Ceremonial Burial......44...45
code of law..................3....4
pottery........................53... 51
The result should as proof be sufficient.

The discussion-Thread about this (in German): http://civ2-forum.gamigo.de/showthre...5&pagenumber=1

[rah]

Damn, learn something new every day. Good job.

[Julius Brenzaida]

Re: Re: A study of Hut Outcomes

Originally posted by Elephant
Some questions/comments:
You might want to test to see if having Pyramids suppresses the chance of getting a Granary. In your spare time... Just for the fun of it...

No, having Pyramids does no supress the chance of getting a granary.

I am happy to provide the most useless answer of this otherwise great great thread.

Congratulation for the superlative job

[Elok]

Statistics make my head explode, so I'm not going to comment except to say:
A. Samson did one helluva job, and
B. The terrain-dependent mercenaries rumor may be false, but it would make sense, wouldn't it? A forest tribe wouldn't have a bunch of skilled horsemen to offer in service, after all...

[Scouse Gits]

Originally posted by samson
.Apparently, the denominator is hard-wired at 93 for Civ2.42 and CIC and at 100 for FW and MGE.

Hmmm! 93 & 100 have quite distinct Hex values - this makes a hexedit of the source fairly viable -- do they need changing?

SG[1]

[duke o' york]

Has no-one done a definitive test on the mercenary thing then?
I always thought that it would run along the same lines as the appearance of barbs on terrain squares around a hut, but do notice that the best unit you can get is a rifleman, and you never get any others in the (very) late game.
Someone with the cheat menu open and lots of time on their hands could easily set up huts on different types of terrain and enter them repeatedly to see what copmes out. I assume that the units are also influenced by both your tech level and some kind of average tech level.
Having said that though, I ain't volunteering!

[La Fayette]

Originally posted by duke o' york
Has no-one done a definitive test

No.
There is no such thing as a definitive test.
That's why this game is great.
(that's also why testers should be admired )