Archived talk:Thorn 20091005

From Robowiki
Revision as of 05:52, 8 October 2009 by Robobot (talk | contribs) (moved Thorn/Archived Talk 20091005 to Archived talk:Thorn 20091005: move to talkarchive)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search
       Archive        This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.     

Hey, you did it! =) Top spot with a first version is way cool, great job. -- Voidious

Thanks! It's amazing how easy it was when I had two top-level bots to work off. -- Kev

Not sure how much of a fluke this was, but...

Fighting battle 6 ... kc.micro.Thorn 1.2,pe.SandboxDT 3.02
RESULT = pe.SandboxDT 3.02 wins 2769 to 2763

That is just incredible for a MicroBot! Keep it up, man...

-- Voidious

Unfortunately, it still lost quite a few points in the rumble. I'm suprised because 1.2's scores were way better than before in the movement challenge tests I was doing. I'll submit an expanded version of it's FlatMovement to the TargetingChallenge2K7 since I think it is one of the hardest non-surfing bots to hit. -- Kev

Nevertheless, I'd say DT sacrifices some rating for PL prowess itself, and even the top MegaBots have a hard enough time with it. If that MC score vs Dookious is consistent (from TargetingChallenge2K7/Results), I'd definitely call this the hardest to hit non-surfing movement out there right now. -- Voidious

ThornHT is not using the exact movement code from Thorn; the distancing and wall avoidance are a tweaked a bit and the "random factor" is calculated a little more accurately. Nevertheless, Thorn 1.2TC was getting movement scores several points higher than 1.1. Too bad it didn't translate to the rumble. I also have a hunch that Raiko might have better random movement than Thorn, but I'm not sure about it. -- Kev

Oh you didn't. Don't make me come back there. -- Simonton

Will this be the first in The2000Club/Micro ?? -- GrubbmGait

That would be awesome, but it looks like it might fall about 10 points short. Still, given the rating distribution and other factors in the MicroRumble, I think its current rating is more impressive than 2000 in General, anyway... Great job, Kev. -- Voidious

Truly, this is a VERY nice improvement. -- Simonton

Yay! Thorn's only ~10 points off of 2000 in the micro rumble, and has probably snagged the number 1 PL spot too! But I think those last points might be hard to get. And congratulations to you GrubbmGait by the way, for your rapid ascent in the micro rumble. -- Kev

Hey Kev, you don't think that the reduced rankings could be due to the fact that the entire rumble has changed since 1.142 was released? Maybe you should re-release 1.142 as 1.145, just to check where you actually stand today. - Skilgannon <who is proud of his current number 1>

I'm not sure what you mean when you say "the entire rumble has changed since 1.142 was released," but I doubt its rating would change significantly from a re-release. And congrats on your number one spot by the way... it looks like I've been dethroned from #1 in micro as well as nano :P. -- Kev

Well, Simonton dropped by about 10 points when he re-released WeeklongObsession. I'm just going on what he said. And it just goes to show, in Micro/Nano, PatternMatching is the best option =). -- Skilgannon