Latest update: Mu is the current version.
MiniBot is gone and it's all about RR@H now, so here are his current ranking details:
As of last check he went straight to first in RR@H and hadn't yet dropped. Not enough matches to call it stable (need 500).
Kappa continues to perform solidly and ranks around #2 in everything except the minibot grand finale. I believe I may have had a spot of bad luck in match selection there, or NanoSatan isn't comfortable in a top bots only format - or both.
Kappa has won at least one MiniBotChallenge.
Iota wins 2nd in MiniBotChallenge (20030525)! Still can't scrap it's way past Kakuru 1.20, though :(
Iota has also won the Robocode Outpost league in the Semi-Pro division, season 4.
Here are some results of the development versus FunkyChicken: [700 rounds vs FunkyChicken]
Here is a simulated MiniBot, just to see how I can potentially do. Might run this alot more rounds: [Minibot Pretest]
Tentatively released [Kappa].
Alpha and Beta use a spiral movement which is essentially just half of an old movement pattern I made a long time ago that doesn't really work against the big boys anymore. This modification has been doing nicely in nano.
Gamma uses a variant of it I call nano pattern smasher movement.
Delta uses a regular oscillator with alternating movement shapes on a different period from the oscillator. Seems to somewhat confuse pattern matchers while not opening up too much to simpler targeters. This could possibly yield some results in the future, given someone put some development time and thought into it.
Iota, uses some very simple rules for reacting as a movement. They can in theory cause it alot of pain, but so far they work scarily well.
All credit in this department goes to Albert and Michael Dorgan (miked0801) with their nano PatternMatching. It initially used my variation of Albert's work in alpha, beta and gamma have moved on to an almost verbatim version of mld's work.
Bots after gamma discard some of the changes that made it unique and go back to Michael's style of shooting. :(
Still can't think of much that'll let me hit Kakuru more. Only a few hooks in his pattern that a nano-sized gun can pick off.
It basically relies on the fact that linear/circular/random/direct/pattern are the only really viable platforms for a nano. It looks circular but isn't, is unhittable with linear, and as of gamma is hardened against pattern matching. Direct and random fire are sadly the most reliable options, resulting on DuelistNano being one hell of a ProblemBot.
Delta relies on switching between a linear and circular pattern while roughly strafing, this seems to confuse every kind of nano-sized targeting somewhat, as they all have some parts of what they look for in it.
Iota uses a rule-based system that doesn't dodge bullets, but reacts in such a way as to (hopefully) make advanced nano targeters sufficiently broken that it doesn't matter.
It chooses the losing option alot more in melee. The only bright point is that it's movement has apparently made it the highest ranking user of pattern matching in the nano melee on MiniBotChallenge this past season.
Up to Delta this trend of scoring unexpectedly high has continued, largely because of the little attention they pay to enemy location. I doubt Iota will continue with this trend, however.
Probably whoever it sees first. Might accidentally work.
It stores nothing between matches, but between rounds it keeps a detailed array of movement history on the opponent.
It was originally a nano-size platform for the movement code found in JesuSatan?, thus the name. Since I'm never going to write that one up, I should probably explain that name. It comes from a joke on my real life self, where people say my look and actual personality contrast like the former and latter.
Knock yourself out.
Hit Kakuru 1.20.
Obviously Moebius and NanoLauLectrik. Only vaguely JesuSatan? at this point.
Cool! Five bytes are what I need to improve NanoLauLectrik. I'w wait for it. -- Albert
Actually, NanoLauLectrik will probably get six :) Moebius needed an extra byte of movement code to make the change work. -- Kuuran
Tobe, I saw your comment on the MiniBotChallenge page. Yes, he was most definitely meant to be entered, thanks :) -- Kuuran
Cool! I found the difference and now I have some more bytes! Thanks a lot. Wait for the new NanoLauLectrik! -- Albert
It's a silly edit, isn't it? ;) Hope the new NanoLauLectrik can beat the new Kakuru, he's a monster. -- Kuuran
Just to annoy you, I just released FunkyChicken's movement in a HaikuBot. NanoSatan Kappa doesn't always lose to it, but it doesn't beat it all the time, either (neither does Kakuru 1.20). But congrats of course on Kappa - it's the only nano to beat FunkyChicken 1.0 in 27 battles over 3 seasons on the MiniBot Challenge.
Meanwhile, are you up in arms for the final MiniBot Challenge Showdown? Just a week and a half to pump up NanoSatan to beat the next version of FunkyChicken (kind of hard to do when you don't have the next version of FunkyChicken, eh?) which features a movement loosely based on Fhqwhgads 1.0's movement, trying to out-do SmallDevil in the field of flat nano movements. -- Kawigi
Heh, fun. :) I guess as you've since discovered I was on break at the time, I kind of regret missing it, but NanoSatan still placed reasonably well against all these new and monstrously upgraded bots. Congrats to everyone who made top 10. -- Kuuran
It seems my development versions here have turned up a bot which can utterly trash FunkyChicken over short (say around 10 rounds) matches, loses by about 55% over medium (75 or so) round matches, and ties over long (200 or so round) matches. Unfortunately this may not be suited to the 35 round environment. It may be time to work on my gun. -- Kuuran
would you mind taking out one of the nanosatans? two is a little too much to deal with... --andrew
![[Home]](/images/RoboWiki.png)