From news-rocq.inria.fr!jussieu.fr!univ-lyon1.fr!howland.erols.net!www.nntp.primenet.com!nntp.primenet.com!news.texas.net!news.kei.com!newsfeed.internetmci.com!in1.uu.net!news.new-york.net!news.stormking.com!PK6811S@ACAD.DRAKE.EDU Sun Oct 6 14:18:04 1996 Article: 6145 of rec.games.corewar Path: news-rocq.inria.fr!jussieu.fr!univ-lyon1.fr!howland.erols.net!www.nntp.primenet.com!nntp.primenet.com!news.texas.net!news.kei.com!newsfeed.internetmci.com!in1.uu.net!news.new-york.net!news.stormking.com!PK6811S@ACAD.DRAKE.EDU From: PK6811S@ACAD.DRAKE.EDU Newsgroups: rec.games.corewar Subject: multiwarrior hills Date: 5 Oct 1996 14:05:18 -0400 Organization: Storm King Ind. Inc. Lines: 89 Sender: server@news.stormking.com Distribution: world Message-ID: <01IAA2V2UKIA006880@ACAD.DRAKE.EDU> Reply-To: PK6811S@ACAD.DRAKE.EDU NNTP-Posting-Host: valhalla.stormking.com Originator: corewar-l@stormking.com Franz wrote: >well it would be a good round for the tournament but i don't think it >would be good for the hills ... second there are already the multiwarrrior >hills where vampires find easy prey soon .. since tehre are 10 programs >they are liable to sink their teeth into some good proccesses :) Vampires _could_ find easy prey if they didn't face certain handicaps, like being outnumbered 9-to-1 by a bunch of silk replicators, and being generally weak against imps :-) On the other hand, my little 'test Sp' program is sitting pretty on top of the multiwarrior '94 hill: Your program test Sp fights 500 times: Number of wins with zero, one, two, .., 10 other winners and losses 4 6 4 4 3 6 3 2 4 12 402 50 [402 ties + 50 losses = 452 opportunities to do better!] # Name Author Score Age 1 test Sp P.Kline 5609 2 2 sisyphus Kafka 5380 20 3 IMPossible! Maurizio Vittuari 5235 1 4 Evol Cap -- John Wilkinson 5234 21 5 Die Hard P.Kline 5220 42 6 IMPossible! Maurizio Vittuari 5098 9 7 Super Evol Cap John Wilkinson 4979 19 8 Son of Imp Steven Morrell 4961 55 9 test S P.Kline 4902 4 10 TESTI Maurizio Vittuari 4695 30 11 60% Cotton Wilkinson 4626 38 Here is the code, maybe we can drive off more replicator/imps and rebalance this hill: ;redcode-94m ;name test Sp ;kill test Sp ;author P.Kline ;assert CORESIZE == 8000 ;strategy kill paper-imps on the multiwarrior hill ;strategy - spiral and forward spl/dat wipes in one loop ;strategy - two 8-process 3-point imps swGate dat swWipe ,100 swPtr dat swWipe ,swTail swDat dat #swDat-swGate ,swLoop-swGate+1 dat #swDat-swGate ,swEnd-swGate swWipe spl #swDat-swPtr ,swEnd-swGate ; spiral and forward wipers mov *swGate ,>swGate add.ab #381 ,swPtr mov *swGate ,@swPtr mov *swGate ,>swGate swLoop djn.f -4 ,>swGate dat 1000,1000 swEnd dat 1000,1000 for 18 dat 1000,1000 rof impSpc equ 776 Imps dat imp1-(Imps*2667),imp1-(Imps*2667)+impSpc for 7 dat imp1-(Imps*2667),imp1-(Imps*2667)+impSpc rof slImps mov imp1,imp1+impSpc spl 1 spl swWipe spl 1 spl 1 spl *Imps,}0 spl @Imps,}0 jmp swWipe imp1 mov.i #0,2667 for 20 dat 1000,1000 rof swTail dat 0,0 end slImps In general for non-replicators on this hill you do better to have a single compact loop rather than multiple ones as having more copies greatly increases the risk of being overwritten by paper and starting your own useless replications. I tried 1,2,3, and 4 copies of a Harmony-like scanner and though losses decreased so did wins, the overall effect was negative. Paul Kline pk6811s@acad.drake.edu