Re: call for duelmasters help

From: itlm013@fella.ucdavis.edu (Captain Caveman)

Date: Mon, 26 Oct 1992 18:57:28 +0000

In article <00962831.716BD800@abby.chem.ucla.edu> ben@abby.chem.ucla.edu writes:
>>In article <1c21loINNe59@gap.caltech.edu>, johnjohn@cco.caltech.edu (John White) writes:
>>In article <009625EC.C3F63300@abby.chem.ucla.edu> ben@abby.chem.ucla.edu
>>writes:
>>
>>Oh, you want numbers?  I have a 15-14 (and like 20 tourney fights) PL with
>>_69_ skills to his name.  17 wit.
>>
>>I had a 17 wit PR who died at 20 fights; 40 skills.
>
>These are both average, yah? Even for 17 WT.  I'm HAPPY with a warrior who

I did an ad hoc study on this a while back, looking at over 1000 fights, and
found that 13 WT warriors average about 1.3 skills/fight, 15 WT warriors
1.5, 17 WT warriors 2.0, 19 WT warriors 1.9 (go figure), and 21 WT warriros
2.3.

The most I've ever seen a warrior learn is 5 skills in 1 fight, although I
have heard of warriors getting 6.  I don't believe that 3 WT warriors have
any inherent maximum number of skills they can learn (back when I first
started, I had a 9 WT warrior who learned better than my 17 WT AB.  I
thought all that extra WT was a waste...maybe then it was), just that
it gets prohibitively difficult to learn each additional skill, because
their percentage chance of learning is so low.

For Ben, I've had 21 WT warriors get 0 skills, and 19 WL warriors miss thier
first train of a stat.  Very discouraging.

>IMHO, there are two main pieces of evidence AGAINST the luck stat:

Sorry Ben, but here you're dead wrong.  RSI has admitted the existence of
a luck factor, and programming-wise it makes sense.  If you've got a
completely deterministic program, then number crunchers get a big advantage.
It allows them to clone good warriors to get identical warriors, and to
also figure out the charts and formulas within the source code.  Throwing
in a random factor makes number-crunching less effective.

>1) You would have no reason to know this, so maybe this will have some
>infulence:
>AT OrcCon this yar I found a piece of scrap printout with partial info. on a
>particular warrior.  This, I am pretty sure, was an ACTUAL screen dump of the
>program that RSI uses to 'view' warriors.  FYI, the Warrior was named 'Chaos',
>of arena 33, a LU that I think was killed at the last mail-in by a  Khalhums
>AB, of all things.  The scrap gave Starting Stats, Current stats (which,
>incidentally, were unreadable), Favorite Weapon, (SH), Favorite Activity (5),
>Favorite Off. EFfort (also 5), and Endurance number (I think 22), Hit Points
>(in the 20's) and a carrying capacity.  I did not see skill ratings, nor a
>'luck', 'learning', or 'teaching' rating.  Of course, we know skills exist,
> and I beleive in them...

I think the fact that it was a partial dump may tell you something.  Also,
I'm sure that data inputters wouldn't have access to every single piece of
sensitive information in the code.  Why should skill/luck/learning come up
on a common screen dump?

>2) There's absolutly no REASON, programming-wise to include a luck stat.  In
>real life, people are 'lucky' or 'unlucky', why not just let a string of
>random numbers to the job?  For example, Pint, mentioned above, once broke 2
>Scimitars each in two fights in a row at a tourney!  But given stochastics
>statistics, I don't believe that this is impossible without a 'luck' stat.
>Certainly, some warriors, like some people are luckier than others, but I think
>this is just an artifact of the large number of random factors used by RSI in
>differentt parts of the DM code.

Certainly I believe that a lot of things attributed to the "luck" factor
are really just dice-rlls that can go either way.  However, as I stated abvoe,
there is a very good reason to include luck in the program.

>>>  Do high levels of Attack in the
>>>teacher give you a greater chance of learning one or more Attack skills?
>>
>Well, presumably she's either a programmer or was told by one, so assuming
>she's not LYING (no reason to...) Dosn't answere whether it's total NUMBER of
>skills, or relative skills in each area...   I just had a thought!  Maybe it's
>how MAXXED out the warrior is in skills, rather than just skill levels.  That
>is to say a TP with 20 attack skills and an AdEx, would be a better teacher
>than an AB with 5 skills and a Master.

Having had some experience in the martial arts...

If it follows reality (and they keep saying that it does) then it's the
difference between the two skill levels, and not how maxed out the opposing
character is.  So...an Expert attacker fighting a Master attacker should have
a pretty good shto at learning attack skills.

Just my $0.02.

Trump
dfcleary@ucdavis.edu

References

Re: call for duelmasters help
ben@abby.chem.ucla.edu
Tue, 20 Oct 1992 14:42:13 +0000
Re: call for duelmasters help
johnjohn@cco.caltech.edu (John White)
Tue, 20 Oct 1992 22:36:40 +0000
Re: call for duelmasters help
ben@abby.chem.ucla.edu
Fri, 23 Oct 1992 11:58:52 +0000

Referenced By

Re: call for duelmasters help
ben@abby.chem.ucla.edu
Tue, 20 Oct 1992 14:42:13 +0000
Re: call for duelmasters help
johnjohn@cco.caltech.edu (John White)
Tue, 20 Oct 1992 22:36:40 +0000
Re: call for duelmasters help
ben@abby.chem.ucla.edu
Fri, 23 Oct 1992 11:58:52 +0000