gametechmods

Robot Arena => Discussion => Topic started by: Clickbeetle on April 21, 2011, 09:24:36 PM

Title: Ironbot test run
Post by: Clickbeetle on April 21, 2011, 09:24:36 PM
So, I'm running a (hopefully) quick test run of Ironbot using 8 Starcore AI bots.  The purpose of this test tournament is:


1. To find out if the Ironbot idea is possibly feasible on a large scale, or if it would be too much work/just doesn't work.


2. To build up hype.


3. To...


*puts on sunglasses*


...iron out any kinks in the rules. (YYYEEEEEAAAAA)


Right away I've made two small tweaks to the rules.  They are as follows:


- At the start of each round, all the bots that won their fight in the previous round (according to standard RA2 criteria) are put in one group and all the bots that lost in the previous round are put in another group.  The bots in each group are randomly paired off and fight.  This ensures that the same two bots will never fight two rounds in a row.


- If there is an odd number of bots, the bot in the loser's group with the most damage gets the bye.


Otherwise the rules are the same as I posted here https://gametechmods.com/forums/index.php/topic,6280.0.html (https://gametechmods.com/forums/index.php/topic,6280.0.html)




THE COMPETITORS:
2 Sinister (ACAMS)
Alien Queen (Starcore)
Battering Ram (Starcore Luigica)
Hunter-Killer Tank (infiniteinertia?)
Maelstrom (Starcore)
Neglected Waterbug (Rejected)
Papa Roach (foolishevil)
Rabid Pit Bull (Rejected)
















Title: Re: Ironbot test run
Post by: Scrap Daddy on April 21, 2011, 09:36:51 PM
coooool

btw Lugica built Battering Ram
Title: Re: Ironbot test run
Post by: Pwnator on April 21, 2011, 09:38:33 PM
To...


*puts on sunglasses*


...iron out any kinks in the rules. (YYYEEEEEAAAAA)

Totally saw that coming. :P

But I have no idea what's gonna happen when a 36HS meets a 36HS. It's practically impossible to keep count. O_O
Title: Re: Ironbot test run
Post by: GroudonRobotWars on April 21, 2011, 10:25:36 PM
cool
Title: Re: Ironbot test run
Post by: Clickbeetle on April 21, 2011, 10:56:29 PM
But I have no idea what's gonna happen when a 36HS meets a 36HS. It's practically impossible to keep count. O_O


I have a special AI that prints the ID's of broken components to a txt file, so I can check it after the battle, compare it to apanx's exporter, and make sure I remove the right components.  It will be a hassle with two 36HS's but it will be doable.


If I didn't have that AI, I would have given up on this long ago.
Title: Re: Ironbot test run
Post by: Sage on April 22, 2011, 12:08:09 PM
For some reason I think Inf built the new hunterkiller. Not sure, but that's what I'm remembering.
Title: Re: Ironbot test run
Post by: Vertigo on April 22, 2011, 02:29:24 PM
Didn't quite understand first time I read it

Really nice idea though, I'd like to see this go through as a tourney :D
Title: Re: Ironbot test run
Post by: freeziez on April 22, 2011, 02:33:36 PM
This looks cool, I should do my own parody.
Title: Re: Ironbot test run
Post by: Clickbeetle on April 24, 2011, 09:23:07 PM
Round 3 is up!


I'm really starting to like this format.  It's really unpredictable who's going to win once the bots start taking damage.  It doesn't seem to take super long either as I think we're pretty close now to getting down to 4 bots to participate in the Fight To The Death.


Also, I was talking to Firebeetle yesterday and he gave me an idea for a huge timesaver.  Right now I'm going into the bot lab and manually removing all the components that break, then copy/pasting the .bot files into the AI folder again.  But he wisely pointed out that with my list of broken component ID's that gets printed to a txt file, I'm already halfway to simply having the computer automatically destroy those components at the start of the match.


At first I thought that wouldn't work because it would create a pile of loose parts in front of the bots that could interfere with driving, but then I realized I could do the battle in an empty arena with no floor, and as soon as the match starts, create the real arena under the bots.  So all the broken parts would fall under the floor and disappear.


So what does this mean?  It means there will really be no limit on the number of entries I could have for this.  Previously I was thinking I might have to cap it at 32 due to the amount of work involved in manually removing all the broken components... but this new method would eliminate all that.  Just gotta figure out a more economical way of doing the splashes and I'm set.
Title: Re: Ironbot test run
Post by: Enigm@ on April 24, 2011, 09:25:09 PM
Round 3 is up!


I'm really starting to like this format.  It's really unpredictable who's going to win once the bots start taking damage.  It doesn't seem to take super long either as I think we're pretty close now to getting down to 4 bots to participate in the Fight To The Death.


Also, I was talking to Firebeetle yesterday and he gave me an idea for a huge timesaver.  Right now I'm going into the bot lab and manually removing all the components that break, then copy/pasting the .bot files into the AI folder again.  But he wisely pointed out that with my list of broken component ID's that gets printed to a txt file, I'm already halfway to simply having the computer automatically destroy those components at the start of the match.


At first I thought that wouldn't work because it would create a pile of loose parts in front of the bots that could interfere with driving, but then I realized I could do the battle in an empty arena with no floor, and as soon as the match starts, create the real arena under the bots.  So all the broken parts would fall under the floor and disappear.


So what does this mean?  It means there will really be no limit on the number of entries I could have for this.  Previously I was thinking I might have to cap it at 32 due to the amount of work involved in manually removing all the broken components... but this new method would eliminate all that.  Just gotta figure out a more economical way of doing the splashes and I'm set.
I like the Battlebots style of brackets though. This is pretty coo though.
Title: Re: Ironbot test run
Post by: Somebody on April 25, 2011, 12:56:37 PM
FB Strikes again!
Title: Re: Ironbot test run
Post by: R0B0SH4RK on April 25, 2011, 04:42:30 PM
Just wondering, how are you calculating the damage %? Is it just # of components lost/# of total components?
Title: Re: Ironbot test run
Post by: ACAMS on April 25, 2011, 05:09:38 PM
Too bad you did not use Ft Knocks instead of 2Sinister.
Title: Re: Ironbot test run
Post by: Clickbeetle on April 29, 2011, 07:37:12 PM
Just wondering, how are you calculating the damage %? Is it just # of components lost/# of total components?


It's the current weight of the bot divided by the original weight.  I'm not sure if that's a more or less accurate way of assessing damage than % components remaining, but it's how I'm doing it.


Anyway, round 4 is up.  And we're a little closer to the Fight to the Death.


Also, a few rules clarifications I'm posting here to make sure I remember them:


- When pairing off bots for the next round fights, if there are an odd number of bots in the winner's group and loser's group, the remainders in each group will fight if they didn't already fight in the previous round.  If they did fight in the previous round, the bots will be randomized again.


- If a bot loses some wheels so that it has an odd number of wheels left, but more than one, and it doesn't have any spares, one of the remaining wheel motors will be unwired (unpowered) so the bot can drive straight.


- Havok Explosions: Matches are only restarted for Havok explosions if one of the bots gets stuck in the floor, flung into a hazard, etc., such that it causes components to be broken.  It doesn't matter if a Havok explosion causes a bot to lose the match as long as no components were broken that wouldn't be broken otherwise.  There is no penalty for causing Havok explosions.


- Crawlers are banned.  (Lol, just realized how awesome Bumble Bee would be in this.)
Title: Re: Ironbot test run
Post by: freeziez on April 29, 2011, 07:42:31 PM
Title: Re: Ironbot test run
Post by: Somebody on April 29, 2011, 08:10:37 PM
I'm really liking this test tourney. Its giving me a lot of ideas for what to enter.
Title: Re: Ironbot test run
Post by: Clickbeetle on April 29, 2011, 10:18:20 PM
Its giving me a lot of ideas for what to enter.


They wouldn't happen to be rammers, would they? ;)


Judging by Battering Ram and Papa Roach's performance, it seems this is one tournament where rammers could actually do well.
Title: Re: Ironbot test run
Post by: Clickbeetle on May 01, 2011, 03:56:02 PM
OK let's get this thing done and move on to the real thing.


Round 5, 6, 7, 8, and the final are all up!


And with the close of the test tournament, a few more rules revisions:


- Bots without active weapons are required to use the "Charge" or "Ram" tactic (in other words, rammer AI).  This includes bots that had active weapons but lost them.  This is to speed up the fights once bots get heavily damaged, so we don't get tug-of-war fights where nothing happens.


- The maximum number of casters allowed is reduced to 8.


- All bots must have damaging weapons.


- Immobility warnings are only disabled in the Fight to the Death for the first 2 minutes; after that bots can be immobilized.


Finally, for the real Ironbot tournament, I will use a custom arena I'm planning on making that will be full of hazards.  This should increase the amount of damage being dealt even if bots lose all their weapons like RPB.


Once the arena is done I'll post the official rules thread. :mrgreen:
Title: Re: Ironbot test run
Post by: freeziez on May 01, 2011, 04:04:56 PM

Great job Click, and I can't wait to see how this will work.
Title: Re: Ironbot test run
Post by: Somebody on May 01, 2011, 06:39:57 PM
Well my idea extends a lot farther than, ENTER A RAMMER.