superficial question

General Stuff about Armagetron, That doesn't belong anywhere else...
Post Reply
|||
On Lightcycle Grid
Posts: 23
Joined: Sun Aug 15, 2004 6:47 pm

superficial question

Post by ||| »

I'll be short:

Are armagetron bots doublebinders ? :)

That's all, i was playing today with bots and i am not sure if they do.

Thanks
User avatar
Tank Program
Forum & Project Admin, PhD
Posts: 6712
Joined: Thu Dec 18, 2003 7:03 pm

Post by Tank Program »

Wow... That's a good question... With some of the stuff they can do, it sure does make you wonder...
Image
User avatar
Your_mom
Match Winner
Posts: 653
Joined: Sun Jun 06, 2004 1:45 am

Post by Your_mom »

bots dont double bind...,at least i dont think so, they can turn tighter then any double binder or non-doublebinder ever could.
User avatar
iceman
Reverse Adjust Outside Corner Grinder
Posts: 2448
Joined: Fri Jan 09, 2004 9:54 am
Location: Yorkshire, England. Quote: Its the fumes, they make one want to play
Contact:

Post by iceman »

im sure the bots are octuple binders ;)

actually your_mom is correct, the bot program can react much faster than a human (if programmed correctly)
Image He who laughs last, probably has a back-up
Image
Image
sorry about the large animated gif
User avatar
root down
Round Winner
Posts: 344
Joined: Fri Jun 11, 2004 8:46 am
Location: closer
Contact:

Post by root down »

i would imagine how quickly bots can turn is only affected by the cycle delay, which would define their minimum turn time
stakes is high
User avatar
Lucifer
Project Developer
Posts: 8751
Joined: Sun Aug 15, 2004 3:32 pm
Location: Republic of Texas

Post by Lucifer »

aiplayers.cfg wrote: # 1st column: Program name
# 2nd column: Program reaction speed (meta: program execution speed)
# 3rd column: Reaction in an emergency
# 4th column: Program viewing range
# 5th column: wall tracing ability (meta: how good can the program handle texts?)
# 6th column: close combat ability (meta: how good can you control the prog with keyboard alone?)
# 7th column: pathfinding ability (meta: mouse control quality)
# 8th column: ability to detect closed areas and to close the enemy in (meta: is the program secure?)
# 9th column: ability to evade or cross enemy cycles (meta: is there a feud going on with users of another similar program?)
#10th column: ability to avoid dangerous tunnels ( meta: is it stable?)
#11th column: ability to react on enemy cycles tracing the own wall (meta: meaningful error messages)
#12th column: state the AI starts in (0: survive, 1: trace, 2: closecombat, 3: pathfind)
#13th column: time the AI drives straight on at the beginning
#14th column: time before the first state change
Most of these are numbers from 0 to 10, with 10 being the best. Here is everybody's favorite bot in Hell:

Rage 6 4 8 4 10 8 8 8 6 8 2 2 2

You'll have to count the columns and stuff, sorry.
ishAdmin
Match Winner
Posts: 625
Joined: Sun Jul 04, 2004 12:11 am
Contact:

Post by ishAdmin »

I'm going to have to fiddle with the bots. Make them all different in their abilities. Make at least one of them nice and dumb!
Image
User avatar
iceman
Reverse Adjust Outside Corner Grinder
Posts: 2448
Joined: Fri Jan 09, 2004 9:54 am
Location: Yorkshire, England. Quote: Its the fumes, they make one want to play
Contact:

Post by iceman »

ive made some crazy bots in the past that wiggle all over the place and then spiral in and die

but i didnt save the settings sorry
Image He who laughs last, probably has a back-up
Image
Image
sorry about the large animated gif
User avatar
/dev/null
Shutout Match Winner
Posts: 819
Joined: Sat Sep 04, 2004 6:28 pm
Location: Chicago-ish

Post by /dev/null »

Heh, ive noticed a few things,

1. The bots seem to ignore cycle_delay ( They might listen to it, but it sure seems the bend the rules.
2. Has anyone else noticed that the faster your cpu, the smarter the bots?

On my old duron 700 with the default bots they were ok, but still easy to kill. On my new Athlon64 3000+ with the same default bots they are retardedly hardcore with the ai at 50. I recently turned the ai up to 100, hopefully people learn to fear my bastard bots, hehe.
User avatar
Tank Program
Forum & Project Admin, PhD
Posts: 6712
Joined: Thu Dec 18, 2003 7:03 pm

Post by Tank Program »

lol... i played 100 on my 2.4xeon/linux... wasn't imposssible. won about 70% of the time...
Image
|||
On Lightcycle Grid
Posts: 23
Joined: Sun Aug 15, 2004 6:47 pm

oh

Post by ||| »

Oh sorry but I was really lost, I was changing from one distribution to other and i needed all time of wordl to compile system in my slow pc.

Thank's for all answers and now is time to practise with all questions and of course, let's try how "Rage 6 4 8 4 10 8 8 8 6 8 2 2 2" bot it works.
User avatar
Lucifer
Project Developer
Posts: 8751
Joined: Sun Aug 15, 2004 3:32 pm
Location: Republic of Texas

Re: oh

Post by Lucifer »

||| wrote:Oh sorry but I was really lost, I was changing from one distribution to other and i needed all time of wordl to compile system in my slow pc.

Thank's for all answers and now is time to practise with all questions and of course, let's try how "Rage 6 4 8 4 10 8 8 8 6 8 2 2 2" bot it works.
Heh, Rage is what Yaza would be if he were slightly less aggressive and a bit more "perfect". In fact, I not only modeled Rage after Yaza, I tested him against Yaza. Rage is a bot that can consistently kill Yaza. I place his success rate about 50%, but before I disabled the bots Yaza was starting to figure out Rage's tricks. I suspect that given another month of Rage, Yaza would have finally mastered beating him.
Post Reply