Axis GC normal vs AI - very historical

[ETF]

ORIGINAL: Rasputitsa

If, however, you want realism, then the AI can provide just as a rewarding experience. How do you replicate the behaviour of the Red Army in 1941, the 'stand fast' orders that helped destroy the German Army later in the war. No sensible human opponent will do these things, but the AI can.

I see what your getting at, but having a crappy AI and then congratulating it for acting historically is a little strange. What happens in 42-45 with that poor "historical" AI. That is the point AI's are deficient period. Who wants to play the Soviets vs. the German AI (non cheating) when the German AI refuses to do almost anything re. Attacking. But hey each to there own.


I play some turns on one title then move to another, sometimes coming back months later to finish a game. Only the AI can wait patiently that long and won't give up, or get bored.

That would be obvious. Can be fun at times I do the same sometimes.

The developers are not going to trash one, or other, aspect of the game, just because of comments on the Forum. The polls on this and other sites show significant votes for both SP and MP and, therefore, we all pay our cash to the cause and should each have a chance to enjoy the game in our own way.

I agree but there is a shift (finally in the wargaming community [:D]) and I wouldn't expect to see any significant changes in the AI capabilities anytime soon. Don't see to many on the forums really happy with any AI period. I do think though for the beginner and casual gamer it is certainly needed.

[JWW]

ORIGINAL: Mynok


Superior how? If you mean in how one enjoys the game, certainly you are correct. But if you are insinuating that the AI will challenge you just as well as a human player, you are grossly wrong.

Good grief. Enjoying the game. I figured that was obvious.

[EisenHammer]

ORIGINAL: ETF
I personally look forward to the day when the AI is reduced to tutorial status [X(].

Then I guess you look foward to the day when real wargames are no longer made.
Remember there not that many wargamers out there right now, as is. In time there will be fewer. The Masters of wargaming are growing thin lately, and in time MP will dominate and both sides will be even, and no one will be able to tell the difference between the two sides. It will be like a chess board instead of history. Multiplayer is killing the history part of the game by making it even for both sides.

[Rasputitsa]

ORIGINAL: ETF
I agree but there is a shift (finally in the wargaming community [:D]) and I wouldn't expect to see any significant changes in the AI capabilities anytime soon. Don't see to many on the forums really happy with any AI period. I do think though for the beginner and casual gamer it is certainly needed.

To survive, these games must appeal to the widest possible customer base and you can be a casual player of WiTE, but still be a committed gamer, just with wider interests.

Those people who are happy, or at least satisfied whilst waiting for the next level of improvement, probably don't feel the need to complain, but either way, nobody has to lose, we can all get (within reason) what we want. There are preferences, options and potential alternative scenarios to keep us all going. Seeing how far the industry has come, it's not either/or, but more and more. [:)]

[ETF]

ORIGINAL: Rasputitsa

ORIGINAL: ETF
I agree but there is a shift (finally in the wargaming community [:D]) and I wouldn't expect to see any significant changes in the AI capabilities anytime soon. Don't see to many on the forums really happy with any AI period. I do think though for the beginner and casual gamer it is certainly needed.

To survive, these games must appeal to the widest possible customer base and you can be a casual player of WiTE, but still be a committed gamer, just with wider interests.

Those people who are happy, or at least satisfied whilst waiting for the next level of improvement, probably don't feel the need to complain, but either way, nobody has to lose, we can all get (within reason) what we want. There are preferences, options and potential alternative scenarios to keep us all going. Seeing how far the industry has come, it's not either/or, but more and more. [:)]

True good points. Wargaming though I would suggest has grown from the PC side of things significantly in the last 10 years. Mind you it is a small genre to begin with [:D] That being said its growth curve dwarfs the board wargaming/miniatures as these are both in decline. The ability to publish games and D/L and remove the retailer has been a welcome boast to the industry.

[Rasputitsa]

ORIGINAL: ETF

ORIGINAL: Rasputitsa

ORIGINAL: ETF
I agree but there is a shift (finally in the wargaming community [:D]) and I wouldn't expect to see any significant changes in the AI capabilities anytime soon. Don't see to many on the forums really happy with any AI period. I do think though for the beginner and casual gamer it is certainly needed.

To survive, these games must appeal to the widest possible customer base and you can be a casual player of WiTE, but still be a committed gamer, just with wider interests.

Those people who are happy, or at least satisfied whilst waiting for the next level of improvement, probably don't feel the need to complain, but either way, nobody has to lose, we can all get (within reason) what we want. There are preferences, options and potential alternative scenarios to keep us all going. Seeing how far the industry has come, it's not either/or, but more and more. [:)]

True good points. Wargaming though I would suggest has grown from the PC side of things significantly in the last 10 years. Mind you it is a small genre to begin with [:D] That being said its growth curve dwarfs the board wargaming/miniatures as these are both in decline. The ability to publish games and D/L and remove the retailer has been a welcome boast to the industry.

The way this end of the gaming industry has opened up is fantastic, with games being produced with modding in mind and embracing the input from the users. The quality of the mods that have been made available from various individuals and groups, who have provided them freely, is exceptional. That doesn't mean that everything is perfect, but it can only get better. [:)]

[ETF]

True very true, except don't hold your breath on the AI [:D]

[Rasputitsa]

ORIGINAL: ETF
True very true, except don't hold your breath on the AI [:D]

It goes back to what you expect from the AI, if that is to act like a human gaming opponent, then there is little hope in the short term. However, in replicating historical leaders, taking into account the blunders and bad strategic decisions that fill the historical record, the AI can do that quite well. The game can still be challenging, whilst preferences and options can provide fine tuning to each user's needs.

Some pre-set options packages would be useful here, e.g. a set-up to replicate revised Soviet Armoured Doctrine, no Red Army purge, full Soviet mobilisation, Hitler deposed - full release of reserves, victory in the West-increase in German OOB, etc.. However, Joel has already said that this workload would be beyond the development team, at this time, but there is a very capable modding community out there and some of their work is already available.

I know that this is not enough for a major portion of the gaming community, for whom PBEM is the only real solution and I would expect that this aspect of the game should receive its full share of development, even though I personally will probably never make use of it. [:)]

[Jeffrey H.]

ORIGINAL: ETF

Face it the AI in any wargame in the las 20 years is a best an informed beginner. No game company could invest in a human like AI not possible. Or should I say cost effective. MP games are the only way if you want any sort of challenge. Anyone can rape the AI. I did in my first a game.....and I suck at MP games
This will never change in my life time.......ps I'm 41 haha.

That's defeatist IMHO. We should not just give up.

[pzgndr]

However the AI has improved since then but not overwhelming in the last 3 decades. AI development from what I have seen beta testing games can take horrendous amounts of $ and man-hours away for marginal improvements in SP performance.

This is certainly a valid business decision issue for developers to consider. Decision Games made the upfront decision to develop a computer War in Europe without AI. Heck, any game could be computerized for pbem using VASSAL or similar programs. If you don't want AI, you can have it.

And yet, the promise and allure of computer wargames over these past several decades has been for players to have a convenient computer opponent to challenge and entertain them. Ready and willing whenever we are, if we have an hour to spare ar a whole day. Nothing at all against real gaming versus another person, just a matter of preference. I've played several games versus other people over the years via pbem and/or tcp/ip and it's fun but the reality is that my schedule is hectic and I'd rather play solo and not commit to any long-term game.

Something I've noticed is that more $ and man-hours of game development seem to go into ever-increasing graphics and complexity and not into computer opponent development. Why can't more simple boardgames of yesteryear be converted to PC with emphasis on decent and challenging AI play? Games like AH Afrika Korps and Russian Campaign and such. Nothing terribly fancy or complicated, but reasonably realistic and historically accurate (for those scales) and enjoyable to play.

The good news is that some developers do care. MWiF appears to be on track with a very interesting AI, hopefully one that will prove to be both decent and challenging for a game as complex as WiF. The Strategic Command series has developed a compelling computer opponent, driven not only a decent fuzzy-logic generic AI but also with extensive use of editable events and AI planning scripting (see Using Strategic Command 2 AI to Play Advanced Third Reich). 2by3Games and others are also leaning forward in the foxhole to provide what PC wargamers want to buy and play. For that I am thankful. [&o]

[Captain]

The issue is not lack of will on the part of developpers, but sheer impossibility to produce a thinking AI. The AI is just a mathematical program which is told to do action B if A occurs, action D if C occurs or action E if both A and C occur.

In a FPS or a flight simulator, it is easier to fake a "smart" AI because the range of responses is very limited.

You cannot program an AI to think or reason or plan ahead, certainly not in a very complex game like Wite. Chess is an extremely simple game compared to Wite: 64 hexes, 16 units, limited moves, but even there, programs rely on brute force to calculate all possible moves and pick the move they were programmed to choose.

If someone could design a thinking AI, they would not waste it in a computer game. [;)]

[pzgndr]

The issue is not lack of will on the part of developpers, but sheer impossibility to produce a thinking AI. The AI is just a mathematical program which is told to do action B if A occurs, action D if C occurs or action E if both A and C occur.

No this is not correct, the AI is not necessarily deterministic and predictable. It is true that AI is programmed and doesn't "think" but enough variability can be introduced to make the AI as unpredictable as yourself. I could play you as an unknown pbem opponent and not know anything about your intentions and you will do one thing or another and I react to whatever new situation is presented. Ditto for an AI. Whether it's "thinking" or not is basically irrelevant to me, the result on my end when I start my turn is the same.

The difficult issue for AI programming is whether it can "plan" several turns in advance and then execute those plans, rather than react immediately to whatever the current situation is with rigid AI algorithms. The Strategic Command AI for example can be scripted to plan in advance and then surprise you. Is that much different than you scheming your moves in advance and then doing something I might not expect? I say no, again the result on my end is the same. In some respects, some human players are predictable in their own ways and don't "think" very well either, which can easily ruin a game. [;)]

[ETF]

Using Strategic Command 2 AI to Play Advanced Third Reich - hmmm will have to give that puppy a trial run. I find it highly unlikly SC2 can come even close to a PBEM partner without overt cheating. Hey I'm gamethough.
In the mean time I will be in the "sheer impossibility to produce a thinking AI" camp. [:D]

[jomni]

ORIGINAL: barkorn45

sorta ot but for a very good ai try "Hannibal Rome and carthage in the 2nd punic wars"it is a single player game and the ai can be wicked!

That's because it's a single player game... and you can only play one side. The game mechanics all point to the development of a good AI. No game balancing issues as well.

Those Panther Games Operational titles have good AI. They act like real commanders and you can leave the game to play itself.