AI Opponent

[chaos king77]

Well I was thinking alittle yesterday and I thought up this idea. Why not have an AI (Artificial Intel.) Opponent to practice your setups on?

Now here's how it would work. You would make your set-up AND the computer's set-up and then you would fight the computer in a 1v1 Match. Now of coarse no stats would be given out in fear of people trying to find ways to get free stats but this would be a great way to test your set ups against bombs, turtles, Etc.

Now I know this could be a large set up but I think it's worth a chance to mention it. It would be jus like a chess AI player.

Comments?

EDIT: If a Mod sees this please lock this topic. I think everyone made their points.

[Megabyte]

Feel free to program an AI capable of handling all the variable setups and moves, not to mention strategy/counter-strategies. Could be large? Think massive.


FYI, you've made this thread...2 years ago...So you should know how often its been suggested.
http://www.tacticsarena.com/forum/showthread.php?t=10805

[uniquinous]

yeah there are pretty much 2 options for that
1) a highly predictable setup and attack pattern which wouldn't help anyone after the first or second game, or
2) as mega mentioned - an incredibly complex coding endeavor. Think of how many people try to work on the dynamics and tactics of chess. In chess, at the least, units are removed upon attack. Here you'd need to program for HP anticipation and blocking percentages. Impossible.

A better suggestion would be an AI learning too, where the computer makes certain pre-programmed moves and offers text on what it's doing at the time.

In conclusion - dont remake the exact same thread at a later point in time. It wasn't good then, it's not good now, and it won't be a good idea 2 years from now.

[Volcom]

whats wrong with a clan member ?

[Matt 34.5]

It would be nice, but as mentioned its just too much.

[Hellblazer]

This has been suggested and it's an advantage for boosters. No.

[Jeffery]

This has been suggested and it's an advantage for boosters. No.
PSST...

Read what he wrote before commenting at least.
"Now of coarse no stats would be given out in fear of people trying to find ways to get free stats but this would be a great way to test your set ups against bombs, turtles, Etc."


Hrm..

I think it is time to update the "don;t suggest or we ban you" list.

[Megabyte]

This has been suggested and it's an advantage for boosters. No.
whats wrong with a clan member ?


both of these suggest the intelligence of people above machine...
I'm pretty sure we could half ars out an AI that was smarter than most players.

Not that I'm suggesting or volunteering, I just want to mock people as a whole.

[deleryn]

Those chess playing simulators sure outsmart me. For a few moments I think I'm being all smart, and then... WHAM! There goes my queen.

It would be funny if there were a TAO version of Deep Blue. We could even have one of those message bots for the chat window.

[The Coder]

the problem is, there aren't a lot of people that are tao masterminds and programming masterminds... Im the only one I know of...

Anyway, the way those bots work is that the units all have a value... king is like infinate, queen is second highest, rook is third, etc... Tao is a little more complex.

[FullMetalFreak]

goos idea but not even mods would spend that much time and effort on creating something a human clan member can do

[The Coder]

goos idea but not even mods would spend that much time and effort on creating something a human clan member can do

moron... mods don't create things. none of the forum staff create things. If you noticed, mod stands for moderater. they control gameplay and the flow of the game. They don't create the game and its edits.
Its called programmers.
We create these things.
Not the mods
Id kill to work for seed

[Megabyte]

the problem is, there aren't a lot of people that are tao masterminds and programming masterminds... Im the only one I know of...

Anyway, the way those bots work is that the units all have a value... king is like infinate, queen is second highest, rook is third, etc... Tao is a little more complex.

Just so we're clear, I'm gonna really start calling your bullshit. Do it if you please, just at least feign in a manner not so blatant and easily seen.

1. I actually laughed on the "i'm the only one I know of" line...nice work.

2. "these bots" don't necessarily function on a value based system.

[chaos king77]

whats wrong with a clan member ?

So you don't have to bother a clan member everytime you plan to test out a new set up. Though alot of people don't like to show it, I'm sure they get alittle annoyed be asked time and time again to fight someone to just test their new setup while they can fight someone to win stats.


Anyway, the way those bots work is that the units all have a value... king is like infinate, queen is second highest, rook is third, etc... Tao is a little more complex.
And I can guarantee most people would rank a dragon over a knight, or some pieces over others. The Pieces can be ranked based on the formation. I know a good coder would probably be needed to do this but it is possible to do. A way this can work is maybe the lower the hitpoints the higher the rating on the piece, like the computer would save a knight at 10 hp and move it to safety instead of moving a tyrant dragon a full hp.

This has been suggested and it's an advantage for boos
Didn't know it's been suggested more than once and by the way. Read the full post before hitting the reply button. I stated that you would NOT get any rankings from losing/winning against the AI Opp.


yeah there are pretty much 2 options for that
1) a highly predictable setup and attack pattern which wouldn't help anyone after the first or second game, or
2) as mega mentioned - an incredibly complex coding endeavor. Think of how many people try to work on the dynamics and tactics of chess. In chess, at the least, units are removed upon attack. Here you'd need to program for HP anticipation and blocking percentages. Impossible.

A better suggestion would be an AI learning too, where the computer makes certain pre-programmed moves and offers text on what it's doing at the time.

In conclusion - dont remake the exact same thread at a later point in time. It wasn't good then, it's not good now, and it won't be a good idea 2 years from now.
Nothing is always pridictable.. I know that the pieces in chess and these are completely different (How many times do you see a King shoot a fireball at a pawn?) but still. It IS possible to make it so the computer may have a new move for every different move you make. And even if it is predictable, you got nothing to gain/lose from these fights. I keep on hearing that this will take time. Well I know it will take time, coding is a very long and hard process. But sitting and looking at the screen won't get the work done, even if it takes a few years, you have to start somewhere. (How long do you think it takes to make games like this? not to mention the coding for this?) And if you would have rather me bump a topic 2+ years old I'll be sure to do that rather than making a new thread? I'll remember that.

Anything else?

[Soda]

Go away.

[uniquinous]

Nothing is always pridictable.. I know that the pieces in chess and these are completely different (How many times do you see a King shoot a fireball at a pawn?) but still. It IS possible to make it so the computer may have a new move for every different move you make. And even if it is predictable, you got nothing to gain/lose from these fights. I keep on hearing that this will take time. Well I know it will take time, coding is a very long and hard process. But sitting and looking at the screen won't get the work done, even if it takes a few years, you have to start somewhere. (How long do you think it takes to make games like this? not to mention the coding for this?) And if you would have rather me bump a topic 2+ years old I'll be sure to do that rather than making a new thread? I'll remember that.
If you're going to attack my ideas, at least have some clue as to what you're talking about. No. It's not gonna happen. It could take decades to create a full AI version for this game. Heck they are still improving chess versions, and a TON of people have been working on that for a very very long time. So no, it's not gonna happen - and any half-assed AI would very quickly boil down to predictability. That's how computer opponents work. It's why someone can beat Soul Calibur on the hardest level yet sweat a bit when playing an unpredictable button mashing noob.

the problem is, there aren't a lot of people that are tao masterminds and programming masterminds... Im the only one I know of...Easy on the ego and arrogance. Actually, quite a few people who play TAO are coders, and create a number of websites, applications, and other useful add-ons to this game. Didn't you wonder who created the damage calculator or moderator toolbar? Did you think they came from nowhere? Sorry, you're not special.

[chaos king77]

If you're going to attack my ideas, at least have some clue as to what you're talking about. No. It's not gonna happen. It could take decades to create a full AI version for this game. Heck they are still improving chess versions, and a TON of people have been working on that for a very very long time. So no, it's not gonna happen - and any half-assed AI would very quickly boil down to predictability. That's how computer opponents work. It's why someone can beat Soul Calibur on the hardest level yet sweat a bit when playing an unpredictable button mashing noob.


I apologize for attacking you're ideas, but there are always possibilities that it MAY happen. go back 50 or 60 years, would you even dream that we may have half the things we do these days? I admit, I do like your ideas but unless you fight the AI with the same set up over and over wit hthe same moves then yes, you more than likely will recieve the same response. But why would you bother fighting the with the same formation against the same formation over and over in the first place if you're going to get the same results? What i'm saying is that every different move you make may and probably will result in the computer making a new move. Nothing is inpoissble when it comes to computers and I'm not saying that they can do this in a week, I know it takes time.

Edit: And yes making a perfected version of an AI Opp would take decades, but nothing is ever finished. Everything can be updated in one way or another. Making a sample AI Opp may be better than nothing though and like TAO itself, people can report bugs or give ways to make it better. I am fully aware how long it takes to make codes for games and programs. But even if it does take a decade, which it probably will, an AI Opp would be a great addition to this game. That is the reason I made this thread. I'm not demanding for it to be made for next week, there are alot of people here that have been playing this game for 2+ years, I'm sure even they would like an update like this. It would also be a fantastic update for new players of TAO.



EDIT to all the people against this thread: I did not come here and make this post to flame or be flamed and to just leave this thread, so regardless of how many "Hate" messages I get in my pm or on this post, I'll be here untill a mod or admin says otherwise. Other than that I have just as much rights to post my oppinions as you do.

[Jeffery]

You have the right to your opinion, but you should at least make a minimal effort to know about TAO, and how the game works, and how developement of TAO works, before making suggestions that 1) have been made before and 2) will never happen BECAUSE of how developement works on TAO.

[uniquinous]

You're still missing my point tho. Even IF TAO had a real development team and a good group of people working on this very problem, the time it would take to create a computer version that isn't highly predictable would indeed be decades. In fact, that goal of such a project would be set years after the game itself was made obsolete. Yeah, 50 years ago you wouldn't imagine some of the things we have now. But 50 years FROM now, TAO won't be around. As such, having a TAO AI would be rather pointless.

You have shown a complete lack of understanding regarding coding if you think people reporting bugs or making suggestions will make an AI better. You have also shown a lack of respect for this community by not taking the time to search for this topic before just making a thread. Yes, you are entitled to an opinion, but your current one isn't that great - so I'd ask (personally) that you please drop this already.

It's not gonna happen.

If you are so intent on making it happen - start coding for yourself.

[chaos king77]

You have the right to your opinion, but you should at least make a minimal effort to know about TAO, and how the game works, and how developement of TAO works, before making suggestions that 1) have been made before and 2) will never happen BECAUSE of how developement works on TAO.
You're still missing my point tho. Even IF TAO had a real development team and a good group of people working on this very problem, the time it would take to create a computer version that isn't highly predictable would indeed be decades. In fact, that goal of such a project would be set years after the game itself was made obsolete. Yeah, 50 years ago you wouldn't imagine some of the things we have now. But 50 years FROM now, TAO won't be around. As such, having a TAO AI would be rather pointless.

You have shown a complete lack of understanding regarding coding if you think people reporting bugs or making suggestions will make an AI better. You have also shown a lack of respect for this community by not taking the time to search for this topic before just making a thread. Yes, you are entitled to an opinion, but your current one isn't that great - so I'd ask (personally) that you please drop this already.

It's not gonna happen.

If you are so intent on making it happen - start coding for yourself.

Yes I will admit, I don't have a clue on how the coding for TAO is done nor do I know how to personaly code myself. But I do know that coding takes time and effort. You say making this happen will take decades, and I'm not saying I don't believe you. All I'm trying to do is make a suggestion, (and yes I think I have realized by now that this have been suggested a numerous ammount of times.) However I did not see it on the list which I agree with Jeffrey, the list really needs to be updated. :P

I think I have said enough on what I think may help TAO and Uniqui, you said that 50 years from now TAO may not be around? Well I hope that isn't the case and that this game does improve. I have stated my oppinion and seeing how I am vastly out numbered in this I will stop fighting it. I do however still and will continue to think that this is a great idea and I do hope that in the future (no matter how long it may be) this will be brought into the TAO. Thanks for not flaming me... Too much :P

[Riathmus]

Want to know how the coding for the forums is done? Right-click, then click "view source" I've looked at it, and it's hell. I'm learning the basics of coding now, but I still don't get this stuff.

[Jeffery]

PSST- That gives you the html output, which has very little to do with the coding running the site.

[Yazoo]

Well I was thinking alittle yesterday and I thought up this idea. Why not have an AI (Artificial Intel.) Opponent to practice your setups on?

Now here's how it would work. You would make your set-up AND the computer's set-up and then you would fight the computer in a 1v1 Match. Now of coarse no stats would be given out in fear of people trying to find ways to get free stats but this would be a great way to test your set ups against bombs, turtles, Etc.

Now I know this could be a large set up but I think it's worth a chance to mention it. It would be jus like a chess AI player.

Comments?

I remember some guy posting this exact same idea years ago.

I believe, if my memory serves me, the reason why this couldn't work is because the amount of memory that this would require, setting up offensive moves, defensivemoves, and the counter-offensive or -defensive ones would require an gigantic amount of server space. On top of that, even if we did do that, I think you can kiss Free Accounts goodbye. Or kiss such cheap Gold Accounts goodbye.

[monkus]

I could make an AI that plays better than Wayfaerer with 12 events.

[uniquinous]

Want to know how the coding for the forums is done? Right-click, then click "view source" I've looked at it, and it's hell. I'm learning the basics of coding now, but I still don't get this stuff.
*smacks forhead*
*shakes head back and forth slowly*
nooooooooooo nooooooooooob
Accessing the code of TAO isn't that easy, and includes the use a few illegal procedures. Not that... I've tried... or anything... >.>

Yes I will admit, I don't have a clue on how the coding for TAO is done nor do I know how to personaly code myself. But I do know that coding takes time and effort.
There should be no "but" after the first sentence. You make the job sound as if it just takes a bit of hard work and dedication. That's not the case. It would take the development of computer thinking this world has not yet seen. Chess is still developing it's AI strategy, and that only takes into account the movement abilities of the pieces - that's it - movement ability. This one little variable has people stumped, and has our best computer scientists working their hardest to master. TAO AI would need movement ability, wait time, HP, power, armor, blocking % assessment, healing, and a number of other variables. If CHESS, with it's ONE variable is a ridiculous endeavor, it makes TAO an impossibility.


However I did not see it on the list which I agree with Jeffrey, the list really needs to be updated. :P Did you try checking the other threads in this very subforum which you yourself created? That might have helped.

[chaos king77]

Did you try checking the other threads in this very subforum which you yourself created? That might have helped.

Yes and I havn't seen any, but I guess I should have looked harder. And not to mention the last time I was probably actually on the forums before this was when I made my last thread 2 years ago when this was rarely mentioned as a topic :P

At that time it seemed like a good idea and I guess as I said above I should have looked harder. If any Mods are actually reading this thread, would you mind closing it? I think everyone made their point and I don't see this thread going anywhere.

Thanks for clearing some of the problems with this though.

[Riathmus]

PSST- That gives you the html output, which has very little to do with the coding running the site.

I knew that, but he didn't have to know that.

[Megabyte]

Yes and I havn't seen any, but I guess I should have looked harder.


You didn't look, just say that instead of bs'ing and saying "I should look harder." I just typed "computer opponent" (since it won't accept "AI" as a search term) and found your thread in about 10 secs.


Its best to simply admit "oh, I didn't bother" than to make yourself look foolish or as a liar. I mean, lots of us don't search a lot of the time. Hell, usually I just ignore the other threads and mock people if they try to mention it.

[max2k106]

Um, no?

Eat a peach.