Matrix Games Forums

Nah you guys just gota get used to what is challenging and leave the history lesson and total realism out of it. I can play against 1000 extra divisions and still enjoy the GAME of it. These wargames aren't historical anyways they are merely simulations based on some documentation that could be considered questionable at best. I just want the game challenging and hard I don't really care how accurate it is with history. Of course I'm talking higher difficulty levels. The normal or easy mode can be anything others want or perceive it to be, but, I certainly want that whatif built into the difficulty settings. 1000 Russion division sounds like fun to me.

AI's need to be programmed to WIN. That's the issue I see with many games is the AI really isn't programmed to win only delay or keep the human from winning or reaching his/her objective in a given time or turn. That's what developers need to get away from and program these AI's to win to seek the objectives to hold them and to KNOW WHEN IT IS WINNING. Anything I hate more is an AI that is actually beating me and then because I got one little objective it goes balistic trying to get that one objective back and then it loses terribly. We can also take the Total War series what is the WINNING goal for the AI in that game?? Hrmmm come on someone tell me what the AI's goal for WINNING in that entire series is? There isn't one, it's only objective is to slow you down there is nothing or a goal for it to accomplish.

What's the AI's winning goal in Alexander, Caesar, Panzer Command:OWS? Deride? And don't say to preven the human from reaching his/her objective. Why doesn't the AI have a winning goal that beats the human without just being a delaying activity?

ORIGINAL: ravinhood

Nah you guys just gota get used to what is challenging and leave the history lesson and total realism out of it. I can play against 1000 extra divisions and still enjoy the GAME of it. These wargames aren't historical anyways they are merely simulations based on some documentation that could be considered questionable at best. I just want the game challenging and hard I don't really care how accurate it is with history. Of course I'm talking higher difficulty levels. The normal or easy mode can be anything others want or perceive it to be, but, I certainly want that whatif built into the difficulty settings. 1000 Russion division sounds like fun to me.

AI's need to be programmed to WIN. That's the issue I see with many games is the AI really isn't programmed to win only delay or keep the human from winning or reaching his/her objective in a given time or turn. That's what developers need to get away from and program these AI's to win to seek the objectives to hold them and to KNOW WHEN IT IS WINNING. Anything I hate more is an AI that is actually beating me and then because I got one little objective it goes balistic trying to get that one objective back and then it loses terribly. We can also take the Total War series what is the WINNING goal for the AI in that game?? Hrmmm come on someone tell me what the AI's goal for WINNING in that entire series is? There isn't one, it's only objective is to slow you down there is nothing or a goal for it to accomplish.

What's the AI's winning goal in Alexander, Caesar, Panzer Command:OWS? Deride? And don't say to preven the human from reaching his/her objective. Why doesn't the AI have a winning goal that beats the human without just being a delaying activity?

We seem to be complete opposites. I couldn't care less about a "challenge" from a computer game. If a challenge is what I want, I won't be sitting in front of a computer monitor, I'll be detailing and painting a tank model or a painting an oil portrait.

What I want from a computer wargame is an opportunity to immerse myself in the flavor of a time period of history and explore plausible alternatives to the historical outcome. As soon as that flavor is lost, all incentive to play is also lost. Challenge is NOT what it's about. Immersion is what it's about.

It seems designers are the one's with the real challenges. For every gamer like you that they win over by making their AI challenging, they risk losing one like me if the mechanism they utilize to achieve the goal of "challenging" is the sacrifice of historic accuracy and/or reasonably plausibly realistic alternatives to history. If you give the Russians 1000 divisions, why not just give them hover tanks with laser cannons? Get the point?

I think the term "horses for courses" applies here. Surprisingly, again AND in the same thread, I'm with ravinhood.

I want the game to be a challenge and, of absolute utmost importance, fun. If it's neither of those, then I have lost the will to live or, less drastically, can't be arsed to play it.

I like an element of history in it, and it's true to say that I would like to "explore" alternatives...but not at the expense of fun or challenging and I do not want to replay history. It's done. That's why it's history.

Someone on the C:EaW was complaining about start dates on offensives and why this and that didn't happen....I don't really care. If I'm playing Germany in that game and Russia DIDN'T attack me because I left them alone, then that is an element of history (playing Germany in WWI) but with alternatives (Russia left me alone because I left them alone)...

If you give the Russians 1000 divisions, why not just give them hover tanks with laser cannons? Get the point?

 
Yeah I could go for that also, secret weapons of the Luftwaffle. I don't have to have historical accuracy for silly immersion. I play all these games like I would a chess game or a game of checkers. Everything is just numbers anyway there is no actual realism at all except the Title of the Game depicting that the game represents some major war or battle. That's where the realism ends plain and simple. Get the point?

ORIGINAL: ravinhood

If you give the Russians 1000 divisions, why not just give them hover tanks with laser cannons? Get the point?

Yeah I could go for that also, secret weapons of the Luftwaffle. I don't have to have historical accuracy for silly immersion. I play all these games like I would a chess game or a game of checkers. Everything is just numbers anyway there is no actual realism at all except the Title of the Game depicting that the game represents some major war or battle. That's where the realism ends plain and simple. Get the point?

Nope, not in the least.

How can anyone possibly see any value in the silly crunching of numbers.....that's what we created computers to do for us.
For me, gaming will always be about the history and the plausible alternatives.

I see no point whatsoever in what you do.

You probably see no point whatsoever in what I do.

I think we can agreee to disagree.

ORIGINAL: ravinhood
What's the AI's winning goal in Alexander, Caesar, Panzer Command:OWS? Deride? And don't say to preven the human from reaching his/her objective. Why doesn't the AI have a winning goal that beats the human without just being a delaying activity?

A fair question where you probably won't like the answer.

The goal in TS:Alex and TS:JC was to gain tactical advtange at all points in the game. Usually, this involved keeping objectives, defending against enemy strong points and keeping some units in reserve for counter attacks. The AI was really only as strong as we made our 'expert advisors' in the game -- the ones that decided which moves made sense on a unit by unit basis and then the strategy advisors that looked at the move overall and decided if it made sense or not. Unfortunately, it was still rather short term focused.

But, as you suggest, in the end, the goal of the AI was rarely to outright WIN. It's main goal is to put up a challenge and fun fight. And, in the end if you fail to beat it, it has won. Again, putting yourself in our shoes, that's what we hear from the vast majority of game players -- they want an AI that is beatable but challenging. This is a very fine line -- and it takes testing to make it effective.

Deride

P.S. I didn't work on the enemy AI for Panzer Command.

ORIGINAL: HansBolter
How can anyone possibly see any value in the silly crunching of numbers.....that's what we created computers to do for us.
For me, gaming will always be about the history and the plausible alternatives.

I see no point whatsoever in what you do.

You probably see no point whatsoever in what I do.

I think we can agreee to disagree.

Ravinhood's sense of game fun seems to be much the same as a card player's. Yours seems to be sandboxing toy solders, with only historically proper match-ups allowed. Both are viable in their extreme states, but the real challenge for game designers and developers is to build their customer base, by appealing to the widest possible range of players while remaining consistent and true to their own vision of what their game should be.

Not an easy job, by any measure.

If wargames had real good AI's, it would likly beat most people 90% of the time. Who would really want to play a game where they lose 9 times out of ten games?

MEEEEEEEEEEEEEEEEEE! But, it ain't gonna happen. lol I could ask the same question who wants to beat the computer AI 9 out of 10 times or everytime as is most times in my case? I'd personally like just options of difficulty settings to make the game more challenging to an unlimited level. Just some sliders or dice roll modifiers or bonus to odds tables to -1 per level of difficulty. Like say the die roll is 6 and I set the difficulty to -1 then that roll would become a 5, if I set the difficulty to -2 then that die roll would be 4, etc. etc. Just one idea. I just can't see why developers can't develope the game to be more difficult by PLAYER INPUT. Just makes no sense that THEY setup the final difficulties instead of letting the players do it. I think CM and SPWAW do a good job of this, but, I'd like to see this in more games than just these two. The AI in either isn't that great, but, with the modifiers and advantages and handicaps I can give it at least makes playing them challenging and I can and do lose occassionally. I'm just asking for more of the same in a broader selection of games.

ORIGINAL: JAMiAM



Ravinhood's sense of game fun seems to be much the same as a card player's. Yours seems to be sandboxing toy solders, with only historically proper match-ups allowed.

An astute observation.

Kind of reminds me the older Civilization series AI, in which u can see phalanxes ( or swordsmen, or whatever) beat up on modern armor. Try as I might, I could not justify my tanks being taken aout ny medieval dudes. ( unless the survivors managed to reach them and stick their swords in the vehicles' parts, whatever) It use to po me, but then, it's just a game in the end. Exit and reload..

ORIGINAL: Deride
Trust me, I can always build an AI that can beat you at just about any game (Of course, my algorithm must be allowed to run as long as it wants whenver making any decision.)

I doubt this is really true.
Assuming you give the human player unlimited time too, most games should converge towards draw overall (equal amount of wins/losses) or never end at all.
If a perfect strategy can be found, all players will find it given unlimited time.
Unless you can prove that a computer can do something that a human can't?

Some will not agree, but this is my opinion only

Space Empire IV The Mods especially TDM Mod and some other, especially if low bonus is used, Good Luck.. Ai will give tough fite

Dominions 2 & 3

Medevil Total War I & II also the napoleonic mods

Remember please, this is only my opinion from playing numerous games not everyone will agree, but keep in mind we are talking SP here not mp

as for the one I would vote best in MP mode with fog of war would be war plan orange... The fog of war makes this game truly nasty and sneaky

ORIGINAL: Shades of Grey
Unless you can prove that a computer can do something that a human can't?

1) Have a short term memory greater than 5 +/- 2
2) Have unlimited expandable short and long term memory
3) Live longer than 100 years

Though, for the argument of running an algorithm for 1,237,128,371,298,129,381 years, I suspect that just #3 is sufficient.

Deride

ORIGINAL: Deride

ORIGINAL: Shades of Grey
Unless you can prove that a computer can do something that a human can't?

1) Have a short term memory greater than 5 +/- 2
2) Have unlimited expandable short and long term memory
3) Live longer than 100 years

Though, for the argument of running an algorithm for 1,237,128,371,298,129,381 years, I suspect that just #3 is sufficient.

Deride

What is your premise for the assumption regarding #3?

One can site many examples of humans living longer than 100 years.

So far, no computer has ever acheived that goal.

Furthermore, please site an example of a computer having "unlimited" memory expansion.........as an astronomer I often wonder if people actually have a grasp on just how huge "infinity" is when they bandy it about so cavalierly!

B

ORIGINAL: Deride
1) Have a short term memory greater than 5 +/- 2

Pen and paper does wonders for your short term memeory.

2) Have unlimited expandable short and long term memory

In the Turing computing model memory is represented as an unlimited
strip of paper. Give the human player a unlimited strip and they are on
equal foting. Both players should have access to the same unlimited resources.

3) Live longer than 100 years
Though, for the argument of running an algorithm for 1,237,128,371,298,129,381 years, I suspect that just #3 is sufficient.

As HansBolter state below, not a sure assumption. Warranties on computer parts are not that long and not many old computers around. And assuming unlimited time for plotting moves somewhat assumes that the players will be around when it is their turn.
Otherwise every turn based game would reduce to who lives longer, ie refuse to move
and wait for victory.

You guys aren't allowed to use logic... that's not fair [:D]

ORIGINAL: Shades of Grey
As HansBolter state below, not a sure assumption. Warranties on computer parts are not that long and not many old computers around. And assuming unlimited time for plotting moves somewhat assumes that the players will be around when it is their turn.
Otherwise every turn based game would reduce to who lives longer, ie refuse to move
and wait for victory.

I guess my point is simply that a brute force algorithm that can explore all possible moves, then all possible counter moves, then all possible counter counter moves, etc. etc. will have the ability to pick the 'best' move on each turn of a game.

For a simple game, a human would be able to tie against this computer. Take, Tic-Tac-Toe, for example. I can write an unbeatable AI for it -- though it will tie most games against a competent human opponent.

If you allow me to do the same for chess, I can create a program that is unbeatable -- worst case a tie (let's just ignore the black vs. white thing for now.) The problem with chess, however, is that the combinations are huge compared with tic-tac-toe, so a brute force algorithm won't work unless you are willing to wait a very long time (given today's computing power.)

Tin Soliders is actually (from a combinational perspective) more complex than Chess because I move all of my pieces each turn instead of just 1. Again, if you allow me to write a brute force algorithm (randomness of dice aside), I can write an AI that is guaranteed to at least tie (or at least perform to the highest possible level given the setup, etc.)

If my PC today can process enough moves to make brute force possible for Tic-Tac-Toe, perhaps in 100 years, I can write a chess algorithm that can do the same in an acceptable amount of time. Maybe 100 years after that, I can do the same for Tin Soldiers.

Computing power has continued to double every 18 months. If that trend continues, eventually, that will all be possible.

And yes, a human could *potentially* have resources that match a computer. But, realistically, the day will come when a simple PC will have more computing resources available than a human brain. (According to Kurzeil, that should occur around 2029, I believe.) And, only 18 months later, there will be a new model with 2x that capacity. And, in just 50 years later, a computer with 2^50 (1,125,899,906,842,624) that capacity -- a combined computing power greater than all the humans that have ever lived.

Deride

P.S. I think human 'computing' capacity doesn't double every 18 months. Perhaps that should have been on my list instead. [8D]

P.P.S. There are several computers that still run after hundreds of years including Babbage's difference engine, several Chinese abacuses, etc. Though, those would all probably find WitP a bit difficult to deal with [;)]

P.P.P.S. If infinity is so huge, how come it isn't even in the class of numbers that are bigger? [X(]

ORIGINAL: Deride

P.P.P.S. If infinity is so huge, how come it isn't even in the class of numbers that are bigger? [X(]

I'm not a mathematician but my guess would have to be because infinity is not a number at all and therefore cannot be categorized along with numbers?

IIRC a discussion regarding computer modeling of the universe pointed out that in order to model the universe the computer would have to be able to process a number of digits at least equal to, if not greater than, the number of interactive particles in the universe. Think we'll be there by 2029, or will it take a bit longer?

p.s. Isn't the notion that anything can be greater/larger than infinity an oxymoron? Attempting to explain reality with pure theoretical mathematics is what has led an entire generation of cosmologists down the rabbit hole.

Computer with the kind of power you are talking about Deride will have to be played in the Artic or Antartica (providing those areas haven't melted out by then) I believe the power is already here, it's the heat of that power that has to be kept in check. I can only imagine the air conditioners required to keep something of that kind of power at a nominal temperature.
 
I'm not so sure the more core method is going to pan out. The more things you have to go wrong the more things that will go wrong with future systems. I'm already reading too many technical reports with issues with just dual core systems. Can you imagine the technical issues with quad core or dual quad core systems in the future?

Quad core's here, and unlike other "simpler" machines that had to contend with Ram use, storage space, expansion slots, the First concern for a Quad core pc is whether ur unit is sufficiently cool or not. We're talking about cooling the CPU, cooling the GPU, internal water cooled system ( for other hot zones) AND fans galore. Too far, too fast....[:-]

ORIGINAL: ravinhood

Computer with the kind of power you are talking about Deride will have to be played in the Artic or Antartica (providing those areas haven't melted out by then) I believe the power is already here, it's the heat of that power that has to be kept in check. I can only imagine the air conditioners required to keep something of that kind of power at a nominal temperature.

I'm not so sure the more core method is going to pan out. The more things you have to go wrong the more things that will go wrong with future systems. I'm already reading too many technical reports with issues with just dual core systems. Can you imagine the technical issues with quad core or dual quad core systems in the future?