Re: call for duelmasters help

From: johnjohn@cco.caltech.edu (John White)

Date: Sun, 25 Oct 1992 15:01:06 +0000

In article <00962831.716BD800@abby.chem.ucla.edu> ben@abby.chem.ucla.edu
writes:

I wrote:
>>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
>learnes 2 skills/fight...

I'm not, but then again, you're running guys with 13-15 wit while I'm
running teams of warriors that average above a 17 wit.  Let me put it this
way.  In order to do well in the Champions tourney (i.e. TV), you probably
need to have 75-80 skills.  At two skills a fight, thats around 40 fights.
How many 40 fight warriors do you see in the Champions TV bracket?

>In Khalhums, an ally (Chris Winslow) had an original
>convention warrior, Pint, a PL.  With a 14-16 wit he learned <2
skills/fight
>for 48 out of 50 fights!

Well, yeah.  He had a 14-16 wit.  That's a tremendous argument
against making warriors with < 15 wit.

>>>I don't
>>>belive that each warrior has a 'luck stat' that influences his learning.
>>
>>I do.

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

>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...

So?  What does that prove, really?  Just because it didn't appear on
a scrap of paper that you happened to pick up, it doesn't exist?  Come
on...

>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.

Whoa... You didn't say anything about a "luck stat," you said a stat that
affected learning.  I definitely don't really think that there's a luck
stat, though I do think there are a lot of "hidden" stats (whether or
not they appeared on that screen dump).

>>My understanding of the effect of wit on learns is this:  The higher a
>>warriors wit, the more skills it's possible to learn in a fight.  I.E. it
is
>>impossible for a 3 wit warrior to ever learn 6 skills in one fight, though
a
>>21 wit warrior can.  One thing I have noticed:  warriors with 17 wits
>>AVERAGE around 2 skills a fight.   That's all I can say.

>Interesting, I hadn't thought of that.  Though I did see a 6 WT AB (wierd
>manager) learn 3 @ tourney...  Do you think It sets a minimum as well?  Do
21
>WT warriors EVER learn 0 skills?  What's the most skills anyone ever
learned in
>a fight?  What WT did they have (prob. 21...)? Obviously, the overall
minimum
>skill learn is 0.  Is there a maximum, of say, 10?  Help, I need more Data!

The max is 6 in one fight.

>>>  Do high levels of Attack in the

>>>teacher give you a greater chance of learning one or more Attack skills?
>>
>>At the face to face, the one of the RSI employees made some comment about
>>how she would modify the "house" teams for odd-man-out matchups in AD.
She
>>said that all those warriors were maxed out in all skills in order to
                                    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>>maximize their teaching potential.  However, this might be just how SHE
  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>>understood teaching to work.
>>
>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...

Of course it does.  There isn't a really conceivable upper limit in rela-
tive skill level, and even if there was, do you really think that house
warriors would have them?  I mean, come on, Oski Noble with Arch-Master
everything?!  He'd romp over ANYONE in AD.

>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.

Isn't that what I just said (check out the ^'s)?

>>Well, you have mine.  Do any of you play in 37?  The Dealer claimed that
the
>>Dark Circle had program information there a couple turns ago.  Hmmmm.

>Dark Circle, huh? Well I have a contact, (Moriarty, a.k.a Prof) I'll see
what I
>can do.  In the mean time, I think I'm going to write a SimDuelmasters code
in
>C++.  This program would try to emulate DM , rather than being a
stand-alone
>game... This way I (or anyone with the source) could try out new theories
with
>out spending a bazillion dollars running 45 teams.  You could also DIRECTLY
>test different aspects of the code...  The other reason for doing this is
to
>learn how to program in C++...

How are you going to make this true to DM if you don't have the DM
source code?  I mean, things will have your biases (i.e. longer fights
will always net more skills, etc.)  .....

>Eventally, I going to try to collect huge amounts of data from other
managers,
>but this will primarily be for character generation, i.e., initial skills.
I
>have the Bloodgames chart, and have seen the RatPak, but these, IMHO, are
far
>from complete.  I also need better info on Hit Points, and especially
>Endurance... Comments from any Mega-managers out there?

Hmmmmm... I know I'm not a mega-manager.... Garth is the only one I know
of on the net...

1) What's the Bloodgames chart?
2) How do you plan on emulating the DM starting skill levels without
   knowing how it's done in the DM source code?

Hmmmm... just looked back over this, and I didn't mean to sound as abrasive
as it came out.  Sorry.  Just trying to get my thoughts across clearly.

:-)

References

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
Fri, 23 Oct 1992 11:58:52 +0000