I was playing winbolo at school andI I was looking at the features and it has autopilot :!: so I clicked on it and the tank was moving and doing cool 8) stuff and you have options like if you want the tank just to get bases tank kills pillboxes and it can lay mines and more.So I think you should make a winbolo 1.14 with autopilot :D.
yes we should... and we do: brains and bots. but I agree that it will not be a better playa. It is, after all, the computer playing, not you 8)
Seems like, oh, about a million years ago, Mac bolo had a brain called "aIndy 3.1" that could actually kick a newbie's ass pretty handily. Eventually, of course, it becomes a joke to play against, but I used to practice Mac bolo by pitting myself (all running on one computer, mind you) against about five aIndies at once, and it was a handful, to be sure.
Last I checked, we were nowhere near being able to do this on winbolo.
I can't cast aspersions, since I can't program my way out of a paper bag, but the sad fact is, winbolo is still not where Mac bolo was about ten years ago when it coumes to AI bots to practice against, or use as an ally in a pinch.
:idea: I believe that winbolo needs a specialized team that immagines, fabricates, designs, creates and releases borgs and brains. A team of designers working together would gain better and faster results than singular "codemasters" working seperately. I think we should organize bot-makers together and speed up the software technology to aIndy 3.1. Comments? 8)
I would love to see some advanced brains created. I'm sure it's a ton of work, but think of it...with the computer in control, a brain could line up for pill takes; always lining up right the first time and sending in its builder at the safe moment to rebuild. If it could be programmed with some of the good rudimentary tactics and be careful with its lgm........it would be a worthy opponent or training brain.
Dont know if this one has been mentioned on or, but here it is...
http://ryan2.servehttp.com/
Features Include...
Options for Base Taking
Pill Taking
Base Guarding
LGM Trix
Plus options for Assasin, Attack MISC, Build MISC, Trees, and other misc options.
It is the best one I have found so far, but then again, nobody lets me in on any secrets, so...
re: qmandabomb
Fighting 5 aIndies is good for honing your reflexes and your ability to keep track of multiple opponents; it is, however, not really like pitting yourself against multiple noobs. The problem is that aIndy 3.1 (while the best bolo brain I have seen, all due respect to the designer), consistently does some seriously stupid things under certain circumstances, whereas noob stupidity is usually expressed in highly individualized, and, if you're not familiar with the noobs you're playing, seemingly random ways. Throw in the occasional smacktard or two, and sometimes it's just annoying chaos.
It becomes difficult when overmatched to not deliberately exploit some of aIndy's more quirky weaknesses to beat it. e.g., you can repeatedly trick the brain, by encouraging it to plant a pill deep in your quad, to send its builder on a kamikaze mission to fix the distant pill. It's possible to kill off multiple aIndy lgms in rapid successsion using this technique. I won't spoil it all by giving away more stupid aIndy tricks, but you get the idea. I used to try to restrain myself from picking on the mindless brains like that, but when they got me in a tight spot and I felt in danger of losing, a wave of irrational killer rage would sweep over me, and I would then do what I had to do to survive. So there is a danger to practicing with aIndy 3.1 too much: If you get good at it, you sometimes forget that people aren't so predictable, and if you use your rear-guard tricks to defeat aIndy on real brains, they may surprise you.
Hence my humble request: A good AI should build on aIndy 3.1's strengths, while mitigating some of its weaknesses. I think there might be an easy way to do this: If somebody can program as good a general fighter as aIndy, but add some stochastic element to its responses under various situations, it might be possible to make it seem more wily and less of a mindless droid. There may well be three or four good ways to deal with a particular problem that the programmer could think of. One of those ways might be the best one, but the others might not be so bad. So have the brain implement its counter-strategy at random, choosing from the short list of possible responses. This would make the brian difficult to anticipate with full accuracy, and would help mix things up a little.
I gave up my attempt to create a bot that didn't suck, the brain interface doesn't even provide as much information as a normal user gets, so how can a brain ever be as good as a human player? ..
Min
lol ...i just played against ryan2 borg..and ..no offense to whoever made it...it was really stupid......cant they even be programmed to do pilltakes...i think sticks mentioned one that could...ill look at it later.......but ya cant they at least be made to do a proper not get hurt at alll two pass hardliner (like mine) (well sometimes i can do it without getting hit.....but usually i get hit once or twice :) )....but ya cant they be made to do that turn so they dont gte hit by bullets? i watched the ryan_2 and all it did was go up and suicide pills....... :o :(
Hey Min, glad to see you around...
Since I'm not a coder, I'm not sure what you mean by "brain interface". Are you saying a brain has less access to critical info. than a player does?
I seem to remember, with Mac bolo, that brains could be given an "advantage" if one so chose. I've kept that option on by default, so I don't know what difference it makes, but the "advantage" seems to imply that brains have at least, if not more access to info. than a human player does. And, as Elvis pointed out above, there's a fair amount of parity between the winbolo and bolo brain interface, does that mean that brains ought to have adequate info?
As I alluded to, I probably don't understand the problem completely. Having said that, I know from experience that aIndy 3.1 is quite a capable brain. Even standard autopilot can whip the virgin noob at first. aIndy's problem is it only seems to respond one way to a particular condition, and hence it's easy to learn how to exploit its predictabilities and stomp it. Having said that, prior to figuring aIndy out, it's a decent opponent. If winbolo provides similar tools for the brainmaker that bolo does, I can't see why producing at least another aIndy shouldn't be possible. Am I missing something?
Elvis you should make winbolo with auitopilot. it will be fun :!: :)
umm, the things I'm complaining about I wrote to you in detail in private message elvis. In regards to the interface being identical to the bolo interface. well in my opinion, both the interfaces are fundimentally flawed. you can't tell weither any enemy tanks are carrying pills, you can instantly tell when a pill has been pickedup by a enemy tank, you should also be given instantatious information in regards to when bases are taken, etc, just like human players do via the pillbox and base counters.
p.s. I've never used winbolo sounds. So I have no idea weither the interface is given that information.
Min
and they should be made so they start when someone cals go and stop whe someone says hld....but the hold feature...people might take advantage of it...so i dunno........ :?
I was play bolo at school autopiolet is cool
Some good Resources
http://www.lgm.com/bolo/guides/brains.html#Resources
You will find the source for Indy 2.02, Decayed Brain 0.4 Code, used in Bolo 0.99.7's Standard Autopilot, Ladmo, Indy 2.02 and aIndy.
I would help with porting one for winbolo, but time is money and I don't have much of either.
I tried using the Ryan2 bot and it does not even compare the the Mac aIndy bot. I also remember playing MacBolo and would make a team of aIndy's that would hold their own for a while. I made a team of Ryan2 bots and they just ended up driving in circles. Is there a aIndy bot for Winbolo?
