Convince me WITP will be worth playing

[LargeSlowTarget]

ORIGINAL: smallie1

And my dad is bigger than your dad. geez, what a lot of childish posts by people who can hide behind an anonymous name.

You make me laugh - that's like the pot calling the kettle black. 'Smallie1' is your real name for sure and you don't even post your DOB and location in the profile either - but you accuse others of hiding in anonymity. That's really cool, dude. [8|]

[ZOOMIE1980]

ORIGINAL: PeckingFury

If Matrix were to put "PBEM only with Single Player training scenerios to learn game mechanics" on the WiTp retail box, would that be ok with you? Personally I dont believe I will ever see an AI in a game of this type that will outwit a human in my life time. (unless ofcourse your Messiah Code hits the shelves in 5 years)

Once again, for the thick.....no one expects an AI that can outwit and expert at the game. But we damned sure as HELL have the right to expect one that provides something more than a 100% predictable, simpleton as an opponent. Which is what we have in UV.

[ZOOMIE1980]

ORIGINAL: pasternakski

ORIGINAL: ZOOMIE1980
Every poster for the past two years on this forum that brings this up gets the same treatment.

I raised the issue of AI quality a long time before you ever showed up (more than two years ago, in fact), and I was treated with respect in an exchange of ideas and opinions that was worthy of the adults who engaged in it.

Your problem is that all you talk about is you and how your opinions ought to carry the argument just because they are yours.

Get over yourself all ready.

In spite of the efforts of the developers here to shut the AI issue up, it is quite CLEAR and OBVIOUS to anyone that it remains a major issue with the MAJORITY of people who will buy the game. If there is one genre of computer gaming that should be allocating a significant percentage of the development budget to the AI subsystem, it is the turn based wargame genre. It the last bastion of human vs computer gaming really left; where the vast majority of customers play by themselves against the computer. Problem is the developers simply don't listen and don't care. The day I find a decent war game on the shelf where 80% of the budget and code is devoted to the AI no matter how primative the interface, is the day I'll shut up.

[pasternakski]

ORIGINAL: ZOOMIE1980
Get over yourself all ready.

In spite of the efforts of the developers here to shut the AI issue up, it is quite CLEAR and OBVIOUS to anyone that it remains a major issue with the MAJORITY of people who will buy the game. If there is one genre of computer gaming that should be allocating a significant percentage of the development budget to the AI subsystem, it is the turn based wargame genre. It the last bastion of human vs computer gaming really left; where the vast majority of customers play by themselves against the computer. Problem is the developers simply don't listen and don't care. The day I find a decent war game on the shelf where 80% of the budget and code is devoted to the AI no matter how primative the interface, is the day I'll shut up.

Haven't you been listening? Fer chrissake, I AGREE with you. I don't WANT you to shut up.

I just wish you could find a way to be more civil.

[ZOOMIE1980]

ORIGINAL: pasternakski

ORIGINAL: ZOOMIE1980
Get over yourself all ready.

In spite of the efforts of the developers here to shut the AI issue up, it is quite CLEAR and OBVIOUS to anyone that it remains a major issue with the MAJORITY of people who will buy the game. If there is one genre of computer gaming that should be allocating a significant percentage of the development budget to the AI subsystem, it is the turn based wargame genre. It the last bastion of human vs computer gaming really left; where the vast majority of customers play by themselves against the computer. Problem is the developers simply don't listen and don't care. The day I find a decent war game on the shelf where 80% of the budget and code is devoted to the AI no matter how primative the interface, is the day I'll shut up.

Haven't you been listening? Fer chrissake, I AGREE with you. I don't WANT you to shut up.

I just wish you could find a way to be more civil.

I always have started out the AI discussion threads with civility. However, after hearing the same exuses as to why the team cannot even provide something as simple some more creative strategic scripting or multiple transport paths to a given base so players cannot just load up in the pre-ordained hex at the pre-ordained time, again and again, over and over, I run completely out of patience with whole thing.

It is quite clear they have NO INTENTION of ever improving the AI, even with some primative improvements that have nothing at all to do with "intelligence". That alone is somewhat disappointing. But to counter it with the "players don't have Supercomputers" schtick is downright irritating.

[Mr.Frag]

However, after hearing the same exuses as to why the team cannot even provide something as simple some more creative strategic scripting

Has it not dawned on you yet that there are no developers telling you anything?

You have been insulting people who do nothing more then donate free time to help find bugs so the shipping product will be stable and as bug free as possible. People like you really make me wonder why I waste 4+ hours a day helping out so YOU benefit. [8|]

If you bothered to get off you arse and read what is posted in the forums, you would find out a lot more information from people happy to provide it, but no, you want to prove to the world how smart you are ... sorry if we are not buying the line based on your style of delivery.

[Mike Scholl]

The whole point about "better computers" comes up simply because what is availabel
to the public right now (software/hardware combinations) simply cannot "learn by
doing" the way a person can. It's reactions must all be pre-programmed choices from
which it attempts to pick an appropriate response. And the multi-environmental and
multi-variable nature of possible responses in WITP (not to mention the changing
force availablity and ratios during a 4 year period) would make for a truely HUGE
"decision tree". Until software/hardware reaches the point of being able to "learn",
it's a daunting, almost impossible (given a reasonable time-frame) task.

My point about "offering some tips" to the people at 2by3 was made in this sense. If
you have any ideas how to construct a simpler and more capable "logic tree" that can
handle such a load, would you please share it with the people who need it! AI's SUCK.
The main reason I almost never play an actual WAR Game in this fashion is because
the AI's in them "suck" much more than most. Anybody who has any worthwhile ideas
on how to make them suck less owes it to the whole community to share the thoughts.

[ZOOMIE1980]

ORIGINAL: Mike Scholl

The whole point about "better computers" comes up simply because what is availabel
to the public right now (software/hardware combinations) simply cannot "learn by
doing" the way a person can. It's reactions must all be pre-programmed choices from
which it attempts to pick an appropriate response. And the multi-environmental and
multi-variable nature of possible responses in WITP (not to mention the changing
force availablity and ratios during a 4 year period) would make for a truely HUGE
"decision tree". Until software/hardware reaches the point of being able to "learn",
it's a daunting, almost impossible (given a reasonable time-frame) task.

My point about "offering some tips" to the people at 2by3 was made in this sense. If
you have any ideas how to construct a simpler and more capable "logic tree" that can
handle such a load, would you please share it with the people who need it! AI's SUCK.
The main reason I almost never play an actual WAR Game in this fashion is because
the AI's in them "suck" much more than most. Anybody who has any worthwhile ideas
on how to make them suck less owes it to the whole community to share the thoughts.

Once again, we have the computer chess player analogy, which of course has absolutely NOTHING to do with ANYTHING I have been saying. Once you people finally get that, maybe we can have an actual discussion.

[ZOOMIE1980]

ORIGINAL: Mr.Frag

However, after hearing the same exuses as to why the team cannot even provide something as simple some more creative strategic scripting

Has it not dawned on you yet that there are no developers telling you anything?

You have been insulting people who do nothing more then donate free time to help find bugs so the shipping product will be stable and as bug free as possible. People like you really make me wonder why I waste 4+ hours a day helping out so YOU benefit. [8|]

If you bothered to get off you arse and read what is posted in the forums, you would find out a lot more information from people happy to provide it, but no, you want to prove to the world how smart you are ... sorry if we are not buying the line based on your style of delivery.

If anything, the ridiculous responses most AI complaints get are insulting by their very nature. And I am not benifiting at all by anything you are doing, unless I buy the game. And yes I have gotten responses in the past from the developers themselves and those responses were very disappointing, basically stating that the AI was never and likely never will be, a central focus of the development effort. And that is too bad, because the AI will, without a shadow of a doubt, be the biggest detraction of this game. Whether you like that or not, it is so.

And when it comes to "getting off my arse" well, I do that to go do something that actually involves having a life outside a computer screen, like play golf, watch a concert, take in a ballgame, etc....

And I have to wade through mountains of completely anal crap to ever find anything really interesting. Every time I look on this forum I see arguments about torpedo attacks in ports from level bombers, whether or not a ship moored inboard could ever be hit by a torpedo, why the G variant of some plane isn't modelled in the game, whether some ambiguous British destroyer was called out of theator on what day, and on and on and on..... What percentage of the final sales do you think will go to people who actually know or could care about such drivel?

Hint: the people without a life on this forum with 1000+ posts that seem to get worked up over this sort of minutia, will amount to less than 1%!

[Mr.Frag]

Troll troll troll ... Buy the game, play the AI.

Prove you can win and I'll personally send you a check to cover the purchase price. It's called putting your money where your mouth is. As you are such an *expert*, you can play on the Very Hard setting. I'm not going away ... got a pair or are you all talk?

[ZOOMIE1980]

ORIGINAL: Mr.Frag

Troll troll troll ... Buy the game, play the AI.

Prove you can win and I'll personally send you a check to cover the purchase price. It's called putting your money where your mouth is. As you are such an *expert*, you can play on the Very Hard setting. I'm not going away ... got a pair or are you all talk?

You mean that Very Hard setting where the computer player actually plays more aggresively, and "smarter" or just gets more stuff with better accuracy? LOL!!!!

And yes, I will probably buy the game, braindead AI or not. I'll at least get one run where I won't know exactly what to expect every turn. At 1500+ turns, I should enough interest out of it to cover the initial cost. Then it will collect dust, like UV does now....

[smallie1]

steve little, Il. Got a chuckle from LST.

got AH's Midway for my 10th B-day. Man, I loved that game back then, the FOW being the cardboard shield in the middle of the board. played board games and then SSIs games on a Commodore-64 (can't believe I ever thought that was a good purchase!). hardly an expert in the games, but they center on a historical period I always found interesting.

Laid off the games until stumbled on UV, and then these forums.

so much for the boring bio. back to the topic at hand.

some of the threads are interesting, some not as much. the overall knowledge on the forums is pretty impressive (although my wife hardly thinks so).

but from time-to-time they degenerate into silly back and forth.

I mean, come on, go back and read this one from the beginning. this morning was the first time I ever saw this thread, and it wasn't exactly the most uplifting ever posted.

[Mike Scholl]

ORIGINAL: ZOOMIE1980

ORIGINAL: Mike Scholl

The whole point about "better computers" comes up simply because what is availabel
to the public right now (software/hardware combinations) simply cannot "learn by
doing" the way a person can. It's reactions must all be pre-programmed choices from
which it attempts to pick an appropriate response. And the multi-environmental and
multi-variable nature of possible responses in WITP (not to mention the changing
force availablity and ratios during a 4 year period) would make for a truely HUGE
"decision tree". Until software/hardware reaches the point of being able to "learn",
it's a daunting, almost impossible (given a reasonable time-frame) task.

My point about "offering some tips" to the people at 2by3 was made in this sense. If
you have any ideas how to construct a simpler and more capable "logic tree" that can
handle such a load, would you please share it with the people who need it! AI's SUCK.
The main reason I almost never play an actual WAR Game in this fashion is because
the AI's in them "suck" much more than most. Anybody who has any worthwhile ideas
on how to make them suck less owes it to the whole community to share the thoughts.

Once again, we have the computer chess player analogy, which of course has absolutely NOTHING to do with ANYTHING I have been saying. Once you people finally get that, maybe we can have an actual discussion.

Now you've decended to "Straw Dogs" and "name-calling" (in the post after your re-
sponse to me). I didn't mention Chess anywhere. I only stated that there are limits
in current computer technology and invited you to "share" any ideas you had for cir-
cumventing them. In spite of your claims to the contrary, I'm beginning to suspect
you are as much "in the dark" about how to improve AI's as everybody else. Nobody
is arguing with you that an IMPROVED AND MORE CAPABLE AI wouldn't be a wonderful
thing to have. But every time you're challanged to "put up or shut up", you try to change
the subject or pretend you don't understand the post. So apparently you're only willing
to sling mud from the stands. We have enough of that already, so I for one am simply
going to ignore you from now on.

[mogami]

Hi, Don't know how I got dragged into this debate. All I did was try to point out that it was harder then thought to program the AI to invade India or knock out China or invade Midway. Because the programmer would not have access to the war situation that would exist when the time to execute these operations arrived. The AI would need to begin planning such operations months in advance. If the reason for such AI action was to challange the human it would fail because preparing such massive operations would telegraph to the human what was coming and then you would have the same old problem of AI doing what the human knew it was going to do. And just as important such preplanned operations would have no relation to the current war picture.

The Allies are massing in Noumea so the AI commits all assets to invading India where the Allies have sent extra LBA formations (in response to build up of AI)

Now I question whether this can be programmed into an AI for one very simple reason.
I can't plan the entire war because I don't know what the enemy is going to do on day two. How can a programmer make a "better" AI by programming operations. Won't such operations just result in disasters for the AI and make it look just as stupid as UV sending AP into air range without CAP?
I think the AI in WITP has learned to provide CAP and it knows to send damaged ships to a size 3 port ASAP. It sends decent force to fight for hexes. (it's learned to use the "follow" command) (versus 1 at a time units that are defeated in turn)
Each scenario the AI has been instructed what bases are important. If it begins with them it defends it not it attacks. It tries to prepare these operations and use proper force ratio.
It still flies too much (higher op loss) But then you don't catch the AI with everything grounded as often as PBEM opponents catch me.

No matter how you slice it a smarter AI will have to be able to calculate. By that I mean assess the current on map situation and render a judgment call as to what the future course of operations should be. This is something I find many human players cannot do.
(I know I often have a hard time because the situation is not clear)

So I am interested in knowing how long it would take for AI to be programed. I would be excited to see it in action and offer my services as play tester. I would not expect it to be able to win. But if it could be made good enough to draw often it would be the best AI in any wargame.

Note: AI is not a preprogramed set of instructions. It is a process of choosing options that produces rational use of forces directed at objectives of value.
Telling the AI to transfer forces from Manchuria to make a major offensive in China is not an improvement. (The forces in Manchuria were there for a reason. They did provide reinforcements to other areas but always as a trade. (A veteran unit was transfered and replaced by a less trained unit)
Japan did not lose the war because of a lack of ground formations. A better AI will understand why Japan is losing and make changes to fix the problems. A better AI will have to recognize where and when the enemy is plotting operations. It will then need to make movements of assets to oppose.

I still can't see how this can be programmed and not require the AI to spend time on each turn. I still think the more time you allow the better the AI will function. However any AI that does not assess the war situation each and every turn will produce the same old AI problems.

[pasternakski]

ORIGINAL: Mogami
I still can't see how this can be programmed and not require the AI to spend time on each turn. I still think the more time you allow the better the AI will function. However any AI that does not assess the war situation each and every turn will produce the same old AI problems.

Thanks for the comments, Mog. Two things:

There isn't anything wrong with the AI spending time on each turn. The human player does.

I do wish someone would call Brad Wardell. His AI in GalCiv works behind the game while the human is taking his turn. I don't know how revolutionary this is or how it might help in AI design for other games, but the AI in GalCiv is first rate (and it doesn't cheat very much).

[mogami]

Hi, There is nothing wrong at all with the AI spending time. You trade a stupid opponent for a slower one. If the AI could do a turn per hour it would be worth while. I'd just let the program run in the background while I did something else. But then I am kept quite busy with my human games.

[pasternakski]

ORIGINAL: Mogami
I am kept quite busy with my human games.

You aliens. Always acting superior. "Let's have the stupid humans fight each other for our amusement, Nyaxxhchro."

"That's one step for Gnarrhh ... one giant leap for Gnarrhhkind."

[mogami]

Don't worry. I have placed you on the list of protected humans.

[Rendova]

ORIGINAL: pasternakski

ORIGINAL: Mogami
I am kept quite busy with my human games.

You aliens. Always acting superior. "Let's have the stupid humans fight each other for our amusement, Nyaxxhchro."

"That's one step for Gnarrhh ... one giant leap for Gnarrhhkind."

By the way Pasternakski which galexy is Tranquility Base located in [;)]

[pasternakski]

ORIGINAL: Rendova
By the way Pasternakski which galexy is Tranquility Base located in [;)]

Should we make this a trivia question, my friend? Free copy of "Pacific War" to the first responder with the correct answer.