Path: ibmpcug!penny.ibmpcug.co.uk!pipex!howland.reston.ans.net!news.sprintlink.net!news.clark.net!rahul.net!a2i!olivea!nntp.msstate.edu!nntp.memphis.edu!nntp.rhodes.edu!harlie.mathcs.rhodes.edu!graham
Newsgroups: rec.games.corewar
Subject: Help with Hunter
Message-ID: <1995Feb22.122843.1965@rhodes>
From: graham@harlie.mathcs.rhodes.edu (Randy Graham)
Date: 22 Feb 95 12:28:42 -0500
Summary: Tips to improve my favorite warrior
Nntp-Posting-Host: harlie.mathcs.rhodes.edu
X-Newsreader: TIN [version 1.2 PL2]
Lines: 52

OK, here's where I have gotten my warrior so far.  This is exactly the
latest warrior I have on the beginner's 94 hill.  So, anyone wanting
to beat it can just target it and try to knock it off.  The real
reason I am posting it here though is to ask for help improving it.  I
realize going to a cmp-scanner instead should be better, but I have
yet to get one correct.  Mine all seem to end up with ~50/0/~50, so I
know I am doing something wrong, but can't figure out what.  I will
work on this and ask for more specific help later.  Right now, what do
you think of hunter?  Can anyone give some good tips for improving
him?


;redcode-b
;Name Hunter v2.6
;Author Randy W. Graham
;assert CORESIZE>1
;strategy a-scanner based on XTC and rave.  Don't bomb self, lay
;strategy carpet of spl bombs.  After avoiding hitting ourself
;strategy several times (twice through core), sweep and sit.
;strategy As is true of all my programs, I still can't beat imps.
;strategy v2.5 Hey, '94 rules -- .f scan!!!  Now catches imps some.
;strategy v2.6  Was using wrong 3/9 point imp gate.

OPT     equ     98              ;mod 2
BOMBS   equ     12
LEN     equ     (fini-hunt)
TIMES   equ     16      ;how many times to hit ourself

hunt    add     #OPT,   ptr
ptr     jmz.f   -1,     ptr+OPT
        slt     #LEN+1, ptr
zero    djn     hunt,   #TIMES
        jmz     spltrap,        zero
        mov     #BOMBS, traps
        mov     spltrap,        >ptr
traps   djn     -1,     #0
        sub     #BOMBS, ptr
        jmp     hunt,   <2667
spltrap spl     0,      #0
dosweep mov     1,      <-1
        dat     <2667,  <2667*2
fini    end     ptr

Just some additional notes.  I never put v0.1-v1.3 on the hill, and
v2.3 was buggy.  I have tried larger and smaller jumper sizes, and
subtracting instead of adding.  None seemed to make much difference.
I am trying to add a qscan, but so far it hasn't made a difference.
It seems to me I should be able to eliminate one line of code
somewhere, but I can't figure out how.

Thx in advance,
Randy