Target Selection

[Loser]

Personnaly I applaud any attempts to remove useless micro out of any RTS. All these discussions about fairness with AI, isnt it realy a bit too soon. I mean if one person gets a realy good AI and if he would share it then it could be impleemented as the default AI when the game is finnished benefiting the game as a whole.

Funny that people thinking that having an diffrent AI is would make a game unfun. In university i made a game where the whole point was to make an AI and let them duke it out on a medival battlefield.
(then again the game never caught on perhaps that is why)

Well all in all i would simply like to say at this stage of the game Group AI would be a GREAT addition. It could always be limited later.

[Kixxe]

I think group ai should be allowed in all it's forms. Why? since we have no problem yet!

you dont get a anti cheting device untill someones cheating right?

And if the problems arises, then SY bans Group ais temporarly, and we find a solution. My sugestion is to have an open list of evry group ai there is on the server for each player, and make it hard to change the name unless you fiddle whit code deep down in the programing.

That way pepole will notice and probaly kick the one using it(or allow it, up to the host.) and we can still devolp and use them.

[aGorm]

Why is it no-one gets what were fighting for.

Look, don't you all play RTS's? Have't you all noticed they have things in commen? They are REAL TIME STRATAGIE games. Not REAL THICK SIM.

If you just add in lots of different groupe AI's then it will boil down to the game being decided in the first few clicks...

I've stated what would be an edition, an advancement to the game play, but no, you all keep wanting to have it compleatly open. You do realise the only reason is because in the back of your heads theres a voice going..

"and i'll make an AI that will do this, And this, and this, and it will be the best and i'll beatt hem all and everyone will go woow, your so cool..."

That is whats making you want it all open.

you dont get a anti cheting device untill someones cheating right?

and to that I say... no, you get it when you can see the cheaters coming, even when there still 100 miles away.

I refuse to get angry... but I will get upset.

aGorm

[Kixxe]

If you just add in lots of different groupe AI's then it will boil down to the game being decided in the first few clicks...

And if the problems arises, then SY bans Group ais temporarly, and we find a solution.

The problem being that someones using the group ai to cheat/andvandge/whatever.

Quote:
you dont get a anti cheting device untill someones cheating right?


and to that I say... no, you get it when you can see the cheaters coming, even when there still 100 miles away.

Cs cheaters and ''codes'' (cheat codes) maybe is esay to find and spot, but andvacend coding into a group ai to make the unit better? what programmer/virusmaker will do that? And as i said:

And if the problems arises, then SY bans Group ais temporarly, and we find a solution.

For now, this problem dont exsist. Anit even likely that it will ever. and if it comes, we will deal whit that later. For now lets just keep all the coding as it is and if someone want to change group AIs/radar targeting/somethingsomething then do so. And tells us and we will (probaly) Lisen. ( i know ''lisen'' is not a word, but i wont drag upp word just to check that 1 word that i know i mispelt (<--- i mispelt mispelt ) And it's 1:40... me tired anyway. why the hell em i upp at night when i can do this in the morning instead?))

[Warlord Zsinj]

My post:

I think my stance has been misinterpreted.

I am not against the advent of AI's that manage mundane tasks in Spring.
For example, I think that AI's that control my metal makers to turn off when I'm about to run out of energy is fantastic.

I do think that some AI's make redundant bits of TA that were not chores, but rather made the game good.
For example, if I can tell my units to attack, I rather like being able to tell them to split up, converge, retreat, etc. I feel like I am actually in command, as opposed to a link between my AI that is controlling my base, and my AI that is controlling my units in battle.

I feel that if I was to turn off any AI's and control things on my own, I should never be at a disadvantage, but rather an advantage. This is the crux of my argument. An AI should never be a replacement for developing skill at the game.

I think many of you are having difficulty drawing the line between "this is a mundane task" and "this is a necessary battle skill".
Metal makers on/off - mundane task.
Units targetting radar targets - perhaps a mundane task (if it didn't make not having radars such a deathblow it'd be alright)
Unit pathfinding - mundane task
Telling your units which targets to attack during a battle - necessary battle skill
Controlling the way your units react in battle situations - necessary battle skill
(Note that simply responding to a threat is something which OTA did using the roam/maneuvre/hold position stances. This is a far better solution then having people make their own unit stances)

Some things which you are taking away are the things that divide good players from bad players. Tactical movements are irreversibly linked to grand strategic gameplay, and you can't honestly think that by removing tactical gameplay you will leave grand strategy untainted.
Making the learning curve shallower for newbies is one thing, but the effect of it is that it lowers the heights that those curves will ever attain.

--------------------
Now to answer various points:
--------------------

Ugh kixxe, you are one of the people that makes everyone throw up their arms in frustration.

I think group ai should be allowed in all it's forms. Why? since we have no problem yet!

Of course we have no problem, because it doesn't exist yet! Why should we expend time and energy on something that is highly controversial, and will be abandoned if it fails? Surely there are better things to spend your time on?

The problem being that someones using the group ai to cheat/andvandge/whatever.

Of course that is the problem. It is also the point. The point is that an AI will give you an advantage over your enemy. The problem is that where this advantage becomes unfair, and whether it makes the game boring by absolving player skill

-------------

aGorm: I feel that the radar targetting is a rather pertinent example of the divide between "this is a useful and helpful feature" and "this removes large amounts of intelligence in the game, as well as largely unbalancing it". But I do feel that it is detracting attention from the argument at hand, so I think I'll drop it.

------------

SJ:

Well in that case we should fix the problem with the engine and not shot the messenger. This is especially true since most things that can abused by an AI could also be to some degree abused by a player that just learns to click fast enough. We might not be able to get all bugs in a single release but spring is supposed to be an ongoing process and over several versions the bugs will be found and eliminated.

It is not a manager of shooting the messenger, but rather of constantly putting more and more bandaids to hold together someone who has a disease (no intended loaded dramatics), when it is the cure to the disease that is needed.

1: You massivly underestimate how hard it would be to create a good AI which means there will probably be only a few major ones created by groups of people which will more or less automatically mean they will be available for download.

Perhaps. But if you look at the way TA's unit making has progressed from simply giving units new stats to zwzsg's freak units, I think the AI progression will be quite quicker than you think, especially considering that someone with a better AI has the potential to beat a better player with a worse AI.
What is more, I don't think that you will get AI's that will play the game for you and beat a human player. There are no delusions there. I do think it would be possible, for example, to give the current aircraft suggestion, and give aircraft a completely automatic target list, so you click over the enemy base, and that is all you need to do.
From my perspective, this removes vital parts of the game. There is no mundanity in picking targets for your units as you attack. It isn't a matter of being able to click faster.
If I want to launch a succesful air strike, I will send over scout planes, mark out the buildings I want to attack, and give the order for my bombers. This is a perfectly logical way of performing a bombing run, and it requires a level of skill in order to decide the best targets, and perform it.
It is the crux of my argument that the AI that decides what targets the bomber should get on its own cuts out the player from the loop, and removes essential parts of the game, and essential parts of gameplay, which is what makes the game fun.
Now, I know I could just turn that AI off and do it myself. But the point is that someone who is using that AI immediately has an advantage over me. While I am busy scouting out the enemy base and selecting the best targets, my enemy can just click in my base and achieve the same results - minus the thought and effort.

2: Even if one assumes that a single person can create a superior AI what will give him the biggest ego boost, winning a few games or seeing everyone else use his AI?

You are missing the point.
It isn't a matter of that person getting an "Ego boost", it is a matter of that player winning, without defeating his enemy through greater skill or intelligence.
And if everyone gets the AI, all the worse for Spring. Because as I argued earlier, they are no longer playing the game, the AI is. They cannot do it as well as the AI can, so they have to use the AI to be competitive.

3: Assuming that someone keep his superior AI secret, why is it worse to lose to someone because he can build a superior AI than because he can click faster than me ?
4: To make everything fair should I also outlaw computers above 2.6 GHz and mice with more than 1600 dpi since those might give the opponent an advantage over me ?

As I said earlier, any AI that removes the mundanity of the game without affecting the necessary skill required to play it is a good addition to the game.
I haven't advocated options whereby the person who can click faster will win. I have advocated options where the person who makes the best strategic and tactical decisions will win. This often results in fast clicking, but it isn't essential. OTA runs on similar principles, and you don't hear people screaming "no fair! you could click faster than me!". Again, I realise Spring isn't OTA; but there are certain elements that OTA got right the first time that it would be silly to ignore.
Consider the example of the aircraft picking their targets automatically via AI. This removes strategic decision, and tactical intervention of the player. It removes gameplay from the game, and yet if it were implemented, everyone would be forced to run it, or risk becoming uncompetitive versus a player who is running it. This is an interactive game we are talking about. Telling your units what to do is the whole point. I'm not saying that units should be absolute morons and sit there being pummelled without response, but at the same time they shouldn't go off and fight the whole war for me. You have to strike a balance; one which I think you are teetering over the edge with.

Then the answear would be to weaken defenses, not to force the player to click like mad on the minimap. Although as a counterpoint i have also seen several ppl saying that defenses are now too weak and must be strengtened ....

Another bandaid fix.
"making defences weaker" doesn't fix the fact that radar effectively extends units LOS to radar range, so that they will always fire at their maximum range. This means that any unit which has a radar supporting it becomes instantly twice as powerful. And units only really have radar support, until the late game, when they are defending. What it effectively has done has turned radar from something which lets you keep an eye on your enemy, into something which is effectively a level 1 targetting facility.
Surely there is a reason why Cavedog decided not to make autofiring a default, but rather include a unit to allow it?

Assume that player 1 know about metal extractors but player 2 dont. Should we then outlaw metal extractors since these will clearly give player 1 an advantage against an otherwise equally skilled player 2

This is a silly reductionist argument that completely misses the point.
What happens if player 1 and player 2 are of equal skill, but player 1 has a better AI. Surely the game should be very close, and player 2 should have an equal amount of winning as player 1, based on their skill levels. unbalanced AI's prevent this.
What happens if player 2 is a slightly better than player 1, but player 1 has a better AI then player 2? Shouldn't by all logical rights, player 2 defeat player 1? Can you say that with AI's, this will always be the case?

---------------
Felix:

Yes, because we all know that elitism is the answer to all problems.

Perhaps with Storm there is a level of elitism, but he isn't entirely wrong with the level of intelligence pervading these forums. Although I am happy that this conversation has turned much more rational. For the most part.

--------------
Yeha:

You want your units to fire on radar dots on the map, your units should know that and fire on them by themself without you having to constanly clicking on them.

While this addresses micro management issues, it does not address the numerous gameplay issues which arise from this change. Frankly, I would rather have patrolling spotter units to increase the range of my units, occasionally instructing my units to fire on radar signatures, then make attacking in the first part of the game impossible, porcing insanely easier (which always makes for a boring game) and aircraft useless.

The unit AI should help you execute your orders to your wishes, not stand and do nothing because you don't have enough time to click on every singel one of them. You should be able to lay out a good strategy and have you unit execute it and knowing your intention without having to individualy babysit them for maximum performance.

I partially agree with you, and partially disagree with you.
Yes, units shouldn't sit there and do nothing when fighting. But this didn't happen in OTA (usually), and there were no complex unit AI's. I could even control how I wanted unit responses with holdfire/returnfire/fireatwill and roam/maneuvre/holdposition. These were very limited unit AI's, which stopped my units from doing stupid things, but never made my units so clever as to think they are smarter than what I have told them to do.
I think I should have to babysit my units for maximum importance. This is the whole idea of controlling an attack. Its the same as closing your eyes in the middle of a game of sport. Evidently you aren't going to do as well, because you can't respond to threats, and alter your attack route, etc.
"babysitting" is a derogatory word for "tactically controlling" your units. Obviously, I want my units to shoot back at threats. This is helpful AI. But I don't want my units to think they can make tactical decisions for me.
For example, a common tactical strategy is to divide your forces in the middle of combat. The effect of this is that it divides your enemy's firepower between two targets, while you focus on one. Of course, if I am in the middle of a battle, and I see that there are enemy units to the side, or that dividing my forces will unnacceptibly weaken my main force, I won't do it. I don't want my units making their own decisions for themselves, because I am able to take in many different factors, while they are only able to run what they algorithm tells them to run.
For all I know, my units will try a flanking maneuvre, meaning that half my forces runs into an HLT and gets flattened, and the other half is now too weak to respond.

... or do you want them to cowardly escape home when damaged a bit?

This is exactly what I don't want. If I want my units to retreat I will give them that order. What if there is a nuke that is about to fire, with a sliver of health left, and my units decide they can't win the battle and retreat? Of course they can't win the battle. I don't need them to survive, I need them to kill that nuke!

-----------

Alantai:

because I'm tired of leading the only project that seems to have gotten anywhere while possible other projects could be hiding waiting for me to complete first and be harassed and flamed, remember there are high expectations of me and I haven't got much help.

Your project has gotten anywhere? Really? Care to show us how far your AI has gotten compared to something like, Buggis map program, perhaps?
And the only reason there are high expectations of you is become you have trumped up your AI so much.

----------

Sorry for the long post. Thank you for reading it all, if you've got this far.

[mongus]

at first i was reluctant.. but kixie has convinced me...
why not
if those ais are there for everyone, that is fair.
all we can use it, all we can learn of it, all we can heate them.
Its there, its open, its free.
transparecy in the scripts is a nice thing to have.. some ala share map system?

we must get our own experience.

and, of course, there could be the option to let them off (mod?), like a good old battle, pure clicking power! not really, just making fun of it, we know its not only clicking that separates good and not-so-good players.

These only adds depth to the gaming experience.

An standard set could help settle fears. (mine too).

[aGorm]

Don't worry Warlord Zsinj I'm with you. Your right on all accounts. The Groupe AI is Good for a few thigs. But some things just don't need an AI, but people will invarably write one for it, and a lot of idiots will go "yay i dont have to do that anymore" and it will become the norm. And then no one can say hay your an idiot useing that, becasue all tyhe lame ass people that can't polay will go, "what, its a realy good AI. Were no playing you you Twat"

aGorm

[smokingwreckage]

For me the threat of group AI blowout is more to do with the fact that at my peak I have more finesse and skill than the AI will in pretty much any RTS (not for a few years now). While an option to assign certain basic tasks to AI would be nice (making a builder in a group repair patrol the nearby mil units of same group), having an AI mess with my plans would suck hard.

Go here, blow this up, you others go here on suicide diversion, you last go THERE! "Nah boss, we're all going to go over to this other thing and get shot to hell by the mines you *guessed* MUST be in such a lightly defended chokepoint."

[Kixxe]

Quote:
I think group ai should be allowed in all it's forms. Why? since we have no problem yet!

Of course we have no problem, because it doesn't exist yet! Why should we expend time and energy on something that is highly controversial, and will be abandoned if it fails? Surely there are better things to spend your time on?

Quote:
The problem being that someones using the group ai to cheat/andvandge/whatever.

Of course that is the problem. It is also the point. The point is that an AI will give you an advantage over your enemy. The problem is that where this advantage becomes unfair, and whether it makes the game boring by absolving player skill

Your still missing the point.

We have no problem yet yes? Then why are we even dicusing something that dosent exsist?

Ehum!

And if the problems arises, then SY bans Group ais temporarly, and we find a solution.

Not just aboandend it. Problems are meant to be solved.

Okay, if your relly worried that somethings gonna happen, SY can maybe ADD A CHECK BOX THAT READ: DISABLE 3rd PARTY AI's!

problem solved? yea probaly. and if we get a SUPER ai that nothing can beat in combat, we will count it as a cheat and ban it alltogeter.

For now, this problem dont exsist. Anit even likely that it will ever.

lets deal whit it then, and not dicuss anymore about it. Lets be just play and wait and do stuff more creative then argue forward and backward whats wrong for spring and whats right.

At least i got stuff to do-.-*

Bye!

[AF]

Now your about to go on about the fact player 2 cant write his own not mattering. But it will. To counter the arguments that have already been mentioned (these are the ones Alantai gave.. if theres more tell me, i cant remember all of them.)

I never said it doesnt matter, you have misunderstood.

For now, this problem dont exsist. Anit even likely that it will ever. and if it comes, we will deal whit that later. For now lets just keep all the coding as it is and if someone want to change group AIs/radar targeting/somethingsomething then do so. And tells us and we will (probaly) Lisen. ( i know ''lisen'' is not a word, but i wont drag upp word just to check that 1 word that i know i mispelt (<--- i mispelt mispelt ) And it's 1:40... me tired anyway. why the hell em i upp at night when i can do this in the morning instead?))

doesnt exist!?!?!?!?!?! The only method fo progress in skirmish AI that I am aware of that doesnt invovle globalAI leads directly to this. *breaks promise shortly, though I am sorry for that* TAI is designed around GroupAI, and there is no reason whyt hat coudlnt eb abused and used by a human player to automate thigns. Though I think by itself it would be a crappy helper at that.

and, of course, there could be the option to let them off (mod?), like a good old battle, pure clicking power! not really, just making fun of it, we know its not only clicking that separates good and not-so-good players.

These only adds depth to the gaming experience.

Whatever the outcome,ability to turn off and disable GroupAI is a must.

I feel that if I was to turn off any AI's and control things on my own, I should never be at a disadvantage, but rather an advantage. This is the crux of my argument. An AI should never be a replacement for developing skill at the game.

This is negated by the previous post by smoking wreckage. AI will not replace skill, they cannot perform sufficient strategic assesment to win a game. Moreover any attempt to would interfere with the player and become an annoyance. Also, do you realise that those who do use AI will coem to rely on their AI too much and become weak, and thus strong players will fidn patterns in the AI and use abstract strategies to defeat it and cripple the clueless player. An AI that adapts is extremely hard to code, and it is something we humans are adept at, be unpredictable and abstract and you will defeat any AI.

This is a debate betwene the micromanagers who want strategic freedom, and the tacticians who fear redundancy at the hands of their machines.

Firstly, I think we've all put our points across and we all think the others havent understood ebcause they ahvent tried to quash sufficiently.

Those of you against GroupAI such as aGorm and Warlord zsinj, but think that it shouldnt eb eliminated as it still ahs its uses, you'r worried about it's ufnair use and the effect it could ahve on the community of players and the negative effects to gameplay.

Those of you who take the stance of SJ, Yeha, me, and kixxe etc, you want them to stay, but you want them to get rid of mundane tasks as warlord puts it, you do nto focus so much on the bad side because whatever it meant it cant possibly be as good as the good side that you see in it. You want a fun game, and the groupAI does soem fot he tasks that you find a nuisance.

The first group are scared of the implications, they look at the bad side, and the latter see the good side and dont want to loose it.

Warlord, you have good points, the reason you have not had a sufficient arguement yet is because you are not arguing an opinion you are arguing a possible fact. IT IS POSSIBLE that this could happen. And nobody here doubts that therefore you wont get any arguements without them seeming silly. We should be focusing on how do we prevent unfair use of the interface, not wether we get rid or what the unfair use is, as we've already established that. Due to presure from others at the AI forums, anyone who releases a skirmish AI will likely do so open source, and any GroupAI or GlobalAI in those projects can easily eb adapted for use as one of these AI that will be misused, although I point again to smoking wreckages post.

[zwzsg]

Apparently there's strong disagreement on these issues. Why can't you just add a tick box in the lobby to enable and disable them? Specifically:
- radar targetting: autotarget (like in spring now), only manual target (like in TA, save that dots are shown on the main screen), no targetting possible (if you try to target them, you fire at the ground instead)
- radar blips on main screen: on/off
- ghost building after being seen once: on/off
- group AI: allowed/disallowed
etc...

Yes, these are changes that can change the balance alot. But did you know that in Total Annihilation it was possible to play without permanent full LOS all over the map? It changes balances and gameplay a lot more, and yet the cavedoggies left the choice to the players. This mean that people like Storm can have games with old TA settings, why others can have games with all the new help features. All playing with the same Spring on the same lobby. No splitting of community, no people leaving spring angry, just lots of different gametype and people choosing whatever they want. After a few years of playing maybe some standard will appear, but it won't be something driven by theory and by the coders, but by popular consensus. For instance the people at Cavedog thought the more logical way to play was with map unexplored, but nowadays all TA games are played with map explored.

Make everything you can optional and easily selectable in the battleroom. I remember it was even written in the guideline of code editing: any change must be optional.

Yes I do want a battleroom filled with buttons and controls!!

[Storm]

Yes, because we all know that elitism is the answer to all problems.

Elitism? I'm just fucking aggravated that no one has managed to adress the points I made I don't know... fifty times... Not a single person.

[mongus]

storm, maybe if you post your concerns in its single thread, with the correct arguments, but here this is a mix of things, so ppl who cares about the same will post there.

[10053r]

Look. If it is possible to build an AI to play better than the best players, then we will have a problem. Frankly, however, I haven't seen anyone suggest that that is possible, and anyone who is hasn't studied computer science and AI, and doesn't know what they are talking about.

It is IMPOSSIBLE to create an AI that will take hints on what aircraft to build, build them, send them at the enemy base, and effectively bomb better than a good human can. It just can't be done with today's tech. Anyone who can write this code can make millions programming for the Pentagon.

Now, is it possible to make an AI that will harass your opponant with aircraft? Absolutely. Should it be done? Absolutely.

We should view your relationship with the AI when you are playing the way a CEO interacts with a company. If the CEO takes a week off, the company doesn't stop. But if the CEO is any good, the company either postpones decision making until the CEO returns, or makes decisions that are not as smart.

Will this allow nubes to play better than they do now? Of course, but calculators allow math nubes to do better multiplication than they did in 1923. It is natural to incorporate computer processes into our own thinking processes where they give us an advantage.

"But what happens if we automate so much of the game that there isn't anything to do any more, because the computer is better than people at EVERYTHING!?!?" you ask in despair. I reply, "We'll make more bits of game that only people are good at."

Anything that can be coded by any of the people on this forum and will run in realtime on our home computers won't be better at strategic decision making than an even vaguely good human. Yes, as automation increases, the game will change from deciding what to do to deciding when to override the computer and what overall policies to set, but this is a good thing. That is strategic decision making.

[aGorm]

NO no... 10053r... That is the point that we that are against are trying to make. Of course you wont map an AI so good it will out perform a human. We have never ever said that. But... Imagin 1 man vrs another man of equal skill plus the help of an AI player.
Your maths example is good, but you haven't seen the actull flaw in your own argument for your cause. A genius back in 1923 could maybe work somthing out as fast as an idiot with a calculatore now a days... but if he was trying to beat another genius that had a calculator he would be left standing.

And Alantai ... i know you did'nt say that the fact ply 2 cant code does not matter, and infact i never siad you did, i mearly said that i expect someone on your side of teh argument would be likly to say that.

aGorm

[10053r]

Your maths example is good, but you haven't seen the actull flaw in your own argument for your cause. A genius back in 1923 could maybe work somthing out as fast as an idiot with a calculatore now a days... but if he was trying to beat another genius that had a calculator he would be left standing.

Who said that people would play without the AI? Everyone would play with AI assistance, just like we do now. Do you tell your units what paths to take? No. That's AI helping you. Would some people play with custom AI? Perhaps, but if someone can rewrite Spring, good on them (and they can do that now...) The majority of AI assistance written will be released back to the public, who will use the best of it. There will always be a level playing field, unless someone writes special AI and doesn't release it. And if they write special AI and don't release it and it is better than me, then they are pretty badass anyway and deserve to win. It's NOT a problem.

[aGorm]

Yes, but as I'm a heavly defencive player, I would not play with an AI that would keep wasting my resorces on sending out small raiding parties. It would be no help. Now whiles't I could probablie find a use for the AI to help me defend (other than one of the uses of AI already in the game, note we dont want them gone) others wont be able to find anything to help and will have to play manuly and lose becasue they cant do everything as fast as everyone else is.

aGorm

[HellToupee]

no ones talking about ais u have no control over, this is more about ais that handle tasks you set for them, if you dont want ais to waste resources on raids then dont assign an ai to it. To send or not send the raid is a strategic level decision which is not what people want to automate, even if they could do it better than even a bad player which i doubt.

[AF]

aGorm:

If you assign a metal maker to the metal maker AI did you knwo you can still issue ti commands and turn it off? Even thought he AI ahs turned it on? And it will stay off as long as the resource situation is the same at which point the AI intervenes?

Builders udner central build AI can still be controlled and given commands, you can still micromanage.

It is IMPOSSIBLE to create an AI that will take hints on what aircraft to build, build them, send them at the enemy base, and effectively bomb better than a good human can. It just can't be done with today's tech. Anyone who can write this code can make millions programming for the Pentagon.

That sort fo AI = extensive trial and improovement or prediction = infinate game trees or huge system resource and time wastage.

aGorm, as for your point that a lesser player may win against a better player because they have an AI assisting, please look at zwzsg post, I'm sure any pro player will have groupAI turned off or only play if they knwo the enemy can be trusted not to use anything more than metal maker type AI. Sufficive to say I've been saying what zwzsg said all along, I wish I'd ellaborated now like he has.

[Warlord Zsinj]

Originally posted by Alantai
AI will not replace skill, they cannot perform sufficient strategic assesment to win a game.

Please refer to the answer I gave SJ here:

What is more, I don't think that you will get AI's that will play the game for you and beat a human player. There are no delusions there. I do think it would be possible, for example, to give the current aircraft suggestion, and give aircraft a completely automatic target list, so you click over the enemy base, and that is all you need to do.
From my perspective, this removes vital parts of the game. There is no mundanity in picking targets for your units as you attack. It isn't a matter of being able to click faster.

Clearly I do not believe that we will see I, Robot. I do think that the AI will be able to do certain tasks better than humans which a human should be left to do as part of the game that we are playing.

And Alantai, please don't nominate yourself mediator here. You are in no position to do so. Please don't put words into my mouth. You phrase my arguments in a slanted manner, and I don't like it. If people want to read my arguments, they can go and read them. You portray the people who are "against GroupAI" (an incorrect statement which you have shoved into our mouth) as a bunch of old TA salts who are afraid of change.

Warlord, you have good points, the reason you have not had a sufficient arguement yet is because you are not arguing an opinion you are arguing a possible fact.

This line here rather infuriates me; I've put the effort into explaining my views, and they are dismissed as "possible facts" that could happen. You go on:

We should be focusing on how do we prevent unfair use of the interface, not wether we get rid or what the unfair use is, as we've already established that.

Yes, I made it quite clear in my post that I do not think that all AI's are bad. Assuming they are only alleviating nuisance, and not cutting out essential parts of the game, I have no issue with them. By allowing 3rd party's to make AI's to their will, I think that there will be no way to mediate this. Therefore, 3rd party AI's should cause synching errors unless both players have them. What's more, I think that AI's in use should come pre-packed with Spring. This would act as a mediating process, and would make sure that every AI that is included is given apt scrutiny.

---------------

As for those who say "why not have a tick function to disable it?". I think this is a bad idea for two reasons:
1) Firstly, as I said, I think that some AI's are very useful. I love the AI that turns metal makers on and off. I think an AI which I could give to a construction unit telling it to cap the closest mexxes available would be useful. Why should I be forced to lose out on the advantage because a number of people with no mind for balance have made the equivalent of the OMGKROGOTH in an AI?
2) Secondly, this will mean that those who are pro-AI develop AI-dependant skills, while those that don't may suffer because of AI-assumed game design. It basically creates a divide which is not good for either party, and will mean that each will practically never be able to play with the other.

Originally posted by 10053r
Now, is it possible to make an AI that will harass your opponant with aircraft? Absolutely. Should it be done? Absolutely.

Why? It isn't like it is difficult to tell your aircraft to harass your opponent. Simply make them patrol in an area over a region, or give them a target while on another patrol route. Not only is this not a nuisance to control, but it removes a level of skill from the game. A player should be able to balance ordering support strikes with controlling his ground units. If he can't manage it; he should get better, not have something do it for him.

Originally posted by 10053r
"But what happens if we automate so much of the game that there isn't anything to do any more, because the computer is better than people at EVERYTHING!?!?" you ask in despair. I reply, "We'll make more bits of game that only people are good at."

Can anyone else see the logical fallicy of this argument?
... And I didn't ask that in despair. I am not worried about the human becoming redundant. If that happened, people simply wouldn't play the game. I'm worried about this dumbing down the game by making important gameplay aspects nonexistant.

Originally posted by 10053r
That is strategic decision making.

Where is the strategic decision making in an aircraft picking its own targets for me? (This is what this topic was originally about)

Originally posted by 10053r
Do you tell your units what paths to take? No. That's AI helping you.

Ugh, please read what I wrote. I noted several times that AI is not a bad thing - even pointing out the move orders and fire orders.
Please read my post and note where my criticism lies. I do not want to repeat myself.

Originally posted by HellToupee
no ones talking about ais u have no control over, this is more about ais that handle tasks you set for them, if you dont want ais to waste resources on raids then dont assign an ai to it.

See, I think automating a raid is a death blow to the game. Controlling your units is an essential skill which players should not be allowed to ignore.
Of course, if I select my units, and order them to attack somewhere, I don't expect them to walk into the enemy defences and be absolute morons. I want my AI to stop my units from doing stupid things - but that's it. I don't want them to go making tactical decisions for me. I'm the commander, not them, and definitely not someone sitting infront of 1000 lines of text.

-----------

I am finding that I am repeating myself over and over. I am assuming that everyone who is posting has read all my posts. I can only hope that you have done so carefully enough to warrant criticism.