Europe at War - great game, if it had AI (Full Version)

All Forums >> [Current Games From Matrix.] >> [World War II] >> Commander - Europe at War Gold



Message

Vale -> Europe at War - great game, if it had AI (6/24/2007 3:07:37 AM)

Hello all.

I have a deep interest in WWII history, and as a child spent many hours playing WWII board games, such as Tobruk. When I heard about a WWII game that looked - from the screen shots - like a strategic WWII game, I just had to buy it.

Initially, I was pleased with the game. The game seemed to have some depth, which I appreciated, and it was stable. The glow faded after an hour. The Achilles' heel of the game - the lack of AI - became apparent very quickly. Playing axis on the default setting, and without even reading the manual, I quickly rolled over Poland, France, and Great Britain. Granted, Poland and France should have been easy, but Great Britain should have been a tough nut the crack. British troops never counterattacked, part of the reason the isle was so easy to conquer. My guess is that the British kept building battleships instead of tanks to reinforce the home defenses. At no point did the British navy challenge me as I crossed the Channel.

Currently, I have two transports in the Irish sea off Dublin. I've had them sitting there, virtually unprotected, for two turns. A huge Allied navy is sitting just a few squares away off the coast of N. Ireland, but has yet to attack the transports.

On the Eastern Front, the Russians throw a seemingly endless number of garrison units against my defenses, but the lack of a coordinated assault and the minimal use of combined arms makes the attacks seem more like suicide runs.

In conclusion, I believe Matrix did a fine with most aspects of the game, but the passive AI is a crippling flaw. Perhaps AI was not a priority because Matrix was hoping people would jump into a player-vs-player match. While it's true player-vs-player will always be more challenging than player-vs-computer, some of us have neither the time nor the inclination to engage in online play. As a result, those of us hoping for a reliable computer opponent will find our wallets a bit thinner after buying the game, and end up disappointed.








Banquet -> RE: Europe at War - great game, if it had AI (6/24/2007 3:22:23 AM)

Well this probably won't make you feel any better but Hearts of Iron - which in various incarnations has been out for YEARS now - in my latest game (as UK) I am fighting the Japanese in India in early 1944 while the USA has over 100 divisions sitting in Guam of all places - with their feet up, on the beach from what I can see!

I usually play as UK on HoI and have yet to see the USA AI do anything meaningful on their own. This after all the years that the game has been released, had expansions, sequels, add ons, etc!

Don't get me wrong, I love HoI but I can't play a game without some aspect annoying me. The simple truth is, though, that AI is hard to program and I think it's all the harder at a Strategic level.

You can't do right.. make an AI too agressive and they're too easy to invade/counter/employ gamey tactic against. Make them too passive and they don't do anything of much use!

So - CEaW isn't the only strategic WWII wargame with problems - it's a hard nut to crack.. and bearing in mind it's a first release, as yet unpatched.. and the devs seem a dedicated bunch - there's reason to be hopeful [:)]

Edit: PS, 1st post I see.. welcome to the Matrix forums :)



ijontichy -> RE: Europe at War - great game, if it had AI (6/24/2007 5:02:08 AM)

Fully agreed, Vale. Firepower (the developer) has to be made aware that the AI is simply not up to scratch, in an otherwise good game, and this has to be their focus in future patches and releases (well, I would like to see support for higher resolutions, too, but AI is more important). This is the last time I'm going to talk about AI; I hope the message gets through.



David Heath -> RE: Europe at War - great game, if it had AI (6/24/2007 6:18:28 AM)

Hi Guys

The developer is listerning and I expect we see some results in the near future.

David




Rocko911 -> RE: Europe at War - great game, if it had AI (6/24/2007 6:40:11 AM)

I hope in the future that Matrix puts in a little more review of AI before they sign up to distribute for someone. I would love to see a rating in the future for all games releases under a AI heading. Perhaps a 1-10 scale or A General Ranking from Patton to ? . I mean I landed troops in England and yet I have not been attacked in force by either England or the Americans after 6 turns. The Brits have their entire navy on the opposite coast of England. I just keep sending troops over the channel without a challenge?



Hard Sarge -> RE: Europe at War - great game, if it had AI (6/24/2007 8:53:25 AM)

What levels are you guys playing on ?

I just got the game and started my first game, the Allies landed in France in the middle of 43 (drove it off) and then made a major landing a few months later, and got a soild little toehold (with all the Navy and Air support, any Axis troops close to the coast is ground into dust)

learning the system, I took too long to take out France, so was only building up my forces on the Eastern Front, when the Russians attacked me, I am smacking them around pretty good, but they still delaying me, but again, they are bleeding me dry

NA needs some work, it put up a good fight for a bit, but then never supported it

which, I not going to say the AI is or has been great, but it will fight back, and with learning the system some more, may be able to exploit it more

but from what I see, it does not just sit back and let you do what you want, when you want




Dave Ferguson -> RE: Europe at War - great game, if it had AI (6/24/2007 10:46:21 AM)

It seems developers of games at strategic level are in a catch22 situation. Developing a AI that can handle WW2 is IMO orders of magnitude harder than a tactical or single battle AI, yet gamers still demand more than they can realistically get. Developing such a AI isnt going to happen anytime soon as no one has the resources to do it. Compare this to operational games where a game recently had a AI so good that it put players off playing? That AI was built using techniques developed as part of a government contract!! a taxpayer subsidy? I bet all developers would like to have that. This is not being critical of that game developer though, its just business.

Dave



Zakhal -> RE: Europe at War - great game, if it had AI (6/24/2007 10:49:58 AM)

Galciv2 is the only game that comes to mind that has decent AI. The whole game is (more or less) built around it to offer superb challenge.



Fred98 -> RE: Europe at War - great game, if it had AI (6/24/2007 11:35:16 AM)

I don't play any AIs.

Hitler never played an AI and he lost.
-



Dave Ferguson -> RE: Europe at War - great game, if it had AI (6/24/2007 11:52:11 AM)

I don't play AI's either but the problem for the game developers is the vast majority of gamers only play the AI, and may have unrealistic expectations of what an AI can do.

Here is a paradox. If a developer had a AI that was as good as a human player would most players want to play the game? Maybe not!

Do gamers prefer a AI versus human play because they are NOT playing a real person and so don't get the embarressment of being beaten, 'its only a AI' can be a nice excuse after having your a*se kicked!!

So do we want the AI to be a sort of untermensch, brave but not so smart that they can outsmart us?

So CEAW developers now have their own paradox, do they concentrate all their effort on improving AI play or listen to the head to head players who want more tools for making CEAW a better game for them.

dave



IainMcNeil -> RE: Europe at War - great game, if it had AI (6/24/2007 12:03:46 PM)

I think there is a problem with the labelling of difficulty more than anything. We only intended the normal setting for beginners new to wargaming to make it easy to pick up and play. For expereienced wargamers you really need to start with a medium AI advantage to get a challenge.



Dave Ferguson -> RE: Europe at War - great game, if it had AI (6/24/2007 12:16:46 PM)

So using normal setting is why I am struggling  as the Axis v human[:)]

A human opponent is so much better than a AI that we should have been playing with an axis advantage?

Is it listed anywhere exactly what the differing levels of hardness do the unit setups, production etc as with a game lasting hundreds of turns you want to make the right choice before you start!

Dave



IainMcNeil -> RE: Europe at War - great game, if it had AI (6/24/2007 3:11:34 PM)

The aim was to make it equal at the basic setting but for experienced wargamers like we have here its not enough of a challenge. Real time games are much easier to make teh AI tough because you can just make a lot of stuff happen at once making it impossible for a human to react to it all. Turn based is much harder to have challenging AI on an equal setting (read impossible - even in a simple game like chess they cant make an AI that will beat a human :) and this has infinitely more possibilities.



JudgeDredd -> RE: Europe at War - great game, if it had AI (6/24/2007 4:20:08 PM)


quote:

ORIGINAL: Iain McNeil
...even in a simple game like chess they cant make an AI that will beat a human...

I always get beat at computer chess games...I guess you can see my level of wargaming!



Warfare1 -> RE: Europe at War - great game, if it had AI (6/24/2007 4:23:35 PM)

It is tough for most large scale strategic games that are released today to balance difficulty for the human player for all sides.

Provided the AI is competent, and it's not asleep, then all that is needed are good scenario and unit editors to give the human player a tough challenge.

It usually takes a scenario designer or user-designed scenario to give a player a tough challenge. The user or modder, using the scenario and unit editors, can then mod the default scenario from one side's point of view. For example, in this game, a scenario can be designed to be played ONLY by the Axis or Allied side. Then ALL game parameters can then be tweaked to give the player a tough challenge.

In addition, previous DOS games such as Steel Panthers, Civ2, etc, had a scenario editor that allowed the following: If a city was captured by the human player, then this caused units to appear for the AI side.

For example, if the Axis player invaded Russia and captured Minsk, then the scenario designer, using the editor, could designate the number and quality of troops and tanks that would appear as Russian reinforcements on predetermined hexes anywhere in Russia. Thus, the further into Russia the Axis player went, and the more cities he captured, then the more Russian troops the player would encounter. Note, this involves NO AI adjustment. Provided the AI is competent (not great), then the Axis player will receive a very tough challenge.

The above is very doable for any game provided the game comes with good scenario and unit editors. There is a reason why Steel Panthers is still being played: It has great editors!

No game designed today should ever ship without user-friendly editors.



willgamer -> RE: Europe at War - great game, if it had AI (6/24/2007 4:51:18 PM)

quote:


In addition, previous DOS games such as Steel Panthers, Civ2, etc, had a scenario editor that allowed the following: If a city was captured by the Axis, then this caused units to appear for the Allied side.




Didn't know that; but I like it.

Here's another idea. Gamer's generally are allergic to the idea of the AI cheating and seeing the whole map (FOW On). What if, however, the AI was allowed to see the whole map just occasionally, either at random, or when a significant event occurs like a city capture. Call it AI Insight.

The idea is that a certain amout of fully disclosed "structured cheating" could go a long way to give the AI an appearance of I.



Warfare1 -> RE: Europe at War - great game, if it had AI (6/24/2007 5:03:30 PM)

quote:

ORIGINAL: willgamer

quote:


In addition, previous DOS games such as Steel Panthers, Civ2, etc, had a scenario editor that allowed the following: If a city was captured by the Axis, then this caused units to appear for the Allied side.




Didn't know that; but I like it.

Here's another idea. Gamer's generally are allergic to the idea of the AI cheating and seeing the whole map (FOW On). What if, however, the AI was allowed to see the whole map just occasionally, either at random, or when a significant event occurs like a city capture. Call it AI Insight.

The idea is that a certain amout of fully disclosed "structured cheating" could go a long way to give the AI an appearance of I.



This is another great idea that has been used for years by game designers and modders.

Again, this involves NO AI ADJUSTMENT.

I don't have a problem with the AI having no fog of war. This makes the AI "smarter" without needing a huge change in programming. It also forces the player to stay on his toes: he can't leave provinces, hexes, and cities unoccupied.

One could also give the AI side "recon" units. Special units only for the AI that helps it to explore the map both on land, sea and air.

This is not rocket science. This sort of thing has been done for years. It's game design 101.

And I have modded older games for years and have given myself some very tough games to play, where the AI has kicked my butt time and time again. And all of this without any new AI programming.

I note that Carriers at War comes with an editor that is so obtuse and that only those with a technical degree have any idea how to use it.

In 2007, there is simply NO excuse for a game to be released without multiple user-friendly game and unit editors.



firepowerjohan -> RE: Europe at War - great game, if it had AI (6/24/2007 5:54:12 PM)

We understand the basic problem here. Diffculty levels sometimes seem like inflation to me. We could have a "normal difficulty" where AI got big production bonuses but we could instead have a normal difficulty where human and AI everyone is on equal terms. It is not easy to know what is best.

Making a touch normal difficulty will mean some beginners that do not have much experience in these kinds of games or if it is their first strategy game experience might feel it is too difficult. On the other hand, making it too easy (=equal terms) will mean some good players will be able to easily beat the AI first time they play the game. "One size fits all" is not so easy to make here and that is why we have difficulty levels ;)

It is also difficult to know, those who post here are a vocal minority more active players who play games more in general. Those who never post might not think the game is as easy.

Also, the thing about deciding how the AI style should be is almost impossible.

Make Allied AI too much navy happy and ppl will complain Axis Subs are useless. Make the Allies neglect ship building too much and some good players will find a uber German sub strategy and complain "why are the 5 remaining allied ships and UK doing nothing to stop my 15 German subs that sink every goods ever put into the Atlantic ocean?"

Similarily, if AI is too aggressive then ppl will complain that it send too many units across the map and that they could make a Sea-Lion too easy while if AI is too cautious ppl will wonder why the USA and UK are staying put on their home Island in 1942 despite Axis having zero defence in France. The answer is that a aggressive AI can be countered also using a "lure and destroy" strategy where you tease it to attack and then throw everything in and crush it.

When it comes to difficulty levels, if we make a totally neutral normal diffculty then some ppl will find it a fun hobby to do these rush strategies for instance to stall Germany a while in Poland and throw everything "all-in" in western Germany using France and UK to either destroy or seriously cripple Germany for the rest of the game. If we instead make a normal difficulty where AI player get a bonus (which in proper meaning instead should be in the other difficulty levels) we risk newbies being scared off or some war buffs complain they cannot play the real setup since there is no "normal AI on equal terms".

It is very popular to invent a strategy and then go out and scream that the AI is crap or broken or anything but we still do not know what the vast majority of ppl that do not even post or maybe not even read forums much think of the difficulty levels but we will try to fix as many problems as possible.






Warfare1 -> RE: Europe at War - great game, if it had AI (6/24/2007 7:03:06 PM)


quote:

ORIGINAL: firepowerjohan

We understand the basic problem here. Diffculty levels sometimes seem like inflation to me. We could have a "normal difficulty" where AI got big production bonuses but we could instead have a normal difficulty where human and AI everyone is on equal terms. It is not easy to know what is best.

Making a touch normal difficulty will mean some beginners that do not have much experience in these kinds of games or if it is their first strategy game experience might feel it is too difficult. On the other hand, making it too easy (=equal terms) will mean some good players will be able to easily beat the AI first time they play the game. "One size fits all" is not so easy to make here and that is why we have difficulty levels ;)

It is also difficult to know, those who post here are a vocal minority more active players who play games more in general. Those who never post might not think the game is as easy.

Also, the thing about deciding how the AI style should be is almost impossible.

Make Allied AI too much navy happy and ppl will complain Axis Subs are useless. Make the Allies neglect ship building too much and some good players will find a uber German sub strategy and complain "why are the 5 remaining allied ships and UK doing nothing to stop my 15 German subs that sink every goods ever put into the Atlantic ocean?"

Similarily, if AI is too aggressive then ppl will complain that it send too many units across the map and that they could make a Sea-Lion too easy while if AI is too cautious ppl will wonder why the USA and UK are staying put on their home Island in 1942 despite Axis having zero defence in France. The answer is that a aggressive AI can be countered also using a "lure and destroy" strategy where you tease it to attack and then throw everything in and crush it.

When it comes to difficulty levels, if we make a totally neutral normal diffculty then some ppl will find it a fun hobby to do these rush strategies for instance to stall Germany a while in Poland and throw everything "all-in" in western Germany using France and UK to either destroy or seriously cripple Germany for the rest of the game. If we instead make a normal difficulty where AI player get a bonus (which in proper meaning instead should be in the other difficulty levels) we risk newbies being scared of or some war buffs complain they cannot play the real setup since there is no "normal AI on equal terms".

It is very popular to invent a strategy and then go out and scream that the AI is crap or broken or anything but we still do not know what the vast majority of ppl that do not even post or maybe not even read forums much think of the difficulty levels but we will try to fix as many problems as possible.


I don't think anyone has any problems with a game having multiple difficulty levels. This is usually standard in most games I have played.

It is up to the game designer to clearly state what each difficulty level does, and which setting should be used for the newbie. I don't think anyone has any objections to this format.

I usually tend to think that a "normal" setting should provide the player with a good challenge. Higher levels should be very challenging. The highest level should defeat the player if he is not fully experienced.

I don't think anyone has any objections to being defeated in a game provided he is given an opportunity to build units, to plan strategies, and to work out a way to win.

The game designer should design the game in such a way so that the gamer is challenged on all levels. The AI should attack, when attacking is indicated. It should surprise the player. It should pop up where it is least expected. There should be some randomness in decision-making. And lastly, there should be game editors to make all of this possible. This all makes for a highly re-playable gaming experience.



firepowerjohan -> RE: Europe at War - great game, if it had AI (6/24/2007 7:25:10 PM)

Game starts on normal difficulty by default so if we want to make it challenging then we still have to decide if that should be for the newbie player who never played wargames before, regular casual player or expert who played all games in this genre for years and years?



Warfare1 -> RE: Europe at War - great game, if it had AI (6/24/2007 7:31:33 PM)

Let's remember that CEaW has just been released. It's also a huge strategy game and patches are going to be needed to tweak the game.

I think it has HUGE potential.

Some ideas:

1) Allow for multiple invasion times, and not just in 1944. Have the AI invade and/or probe, with different size forces, French North Africa/Middle East/Sicily in 1942, 1943. Keep the player on his toes.

2) Give the AI "recon" units: Aircraft with long flying ranges to explore the map; a special ship that can "see" 4-5 hexes away; and land recon units which can see 5 hexes away. The player can destroy them, but the AI can replace them AT NO COST. The AI could even receive special "radar" installations that gives it long range "eyes".

3) Unit, tech and scenario editors. Let the player modify everything in the game.



Dave Ferguson -> RE: Europe at War - great game, if it had AI (6/24/2007 7:34:30 PM)

idea 3) is what will give me long term interest, without them in a few months I will probably move on to another game.



Warfare1 -> RE: Europe at War - great game, if it had AI (6/24/2007 7:38:55 PM)

quote:

ORIGINAL: firepowerjohan

Game starts on normal difficulty by default so if we want to make it challenging then we still have to decide if that should be for the newbie player who never played wargames before, regular casual player or expert who played all games in this genre for years and years?


From reading all the posts on this forum, it looks as though the normal setting is good as it stands for new players. The only problem I am seeing is that the AI does nothing in the Mediterranean area. So i would tweak the AI to be more aggressive.

The higher settings should then ramp-up the difficulty settings. The higher levels should: add more AI units; add higher tech levels; add larger invasion forces; add more experienced units; more places for the AI to invade; AI invasion forces should invade sooner and in different locations (randomness), etc....

If possible, and if the coding allows it, add in an editor or a function whereby new units will appear in the game for the AI defender once certain cities have been captured by the player. This is part of scenario design and is something a user can do at the start of a scenario: ie use the editor to indicate how many units, what units (their experience level and locations) will appear for each city that is captured.



benpark -> RE: Europe at War - great game, if it had AI (6/24/2007 7:44:45 PM)

Simply giving the AI "more" is the default method for most games in dealing with a computer opponent. I'm not a fan of this method, and won't play it. I realize that building a CO is no easy task whatsoever, so I hope this post isn't taken as complaining too much. I am finding the games AI so far decent, but it could use some tuning up without just giving it more of everything.

Making an AI that is truly challenging takes imagination on the developers part. I often wonder if more variability based upon historic possibilities can't be utilized through scripts that are treated as a kind of series of the old Basic lingo "If:Then" rules.

ex: "If Germany holds France in May-Sept 1944: Invade with no less than X number of units around x-z chouice of beach hexes". This would be a sort of choice "A" of around 10 variables for the Allied AI. Another could be invasion through Italy, etc.

This would require that the AI is split inti two sections, one that takes in all of the strategic situation (which utilizes the variable scripts for invasions and DoW etc), and one that is operational (and much like we have now, choosing what units to attacke, etc)

These semi-randomized choices would make a more challenging AI that would appear "smart", and provide that no two games turn out the same.



Warfare1 -> RE: Europe at War - great game, if it had AI (6/24/2007 8:03:02 PM)


quote:

ORIGINAL: benpark

Simply giving the AI "more" is the default method for most games in dealing with a computer opponent. I'm not a fan of this method, and won't play it. I realize that building a CO is no easy task whatsoever, so I hope this post isn't taken as complaining too much. I am finding the games AI so far decent, but it could use some tuning up without just giving it more of everything.

Making an AI that is truly challenging takes imagination on the developers part. I often wonder if more variability based upon historic possibilities can't be utilized through scripts that are treated as a kind of series of the old Basic lingo "If:Then" rules.

ex: "If Germany holds France in May-Sept 1944: Invade with no less than X number of units around x-z chouice of beach hexes". This would be a sort of choice "A" of around 10 variables for the Allied AI. Another could be invasion through Italy, etc.

This would require that the AI is split inti two sections, one that takes in all of the strategic situation (which utilizes the variable scripts for invasions and DoW etc), and one that is operational (and much like we have now, choosing what units to attacke, etc)

These semi-randomized choices would make a more challenging AI that would appear "smart", and provide that no two games turn out the same.


I certainly agree with you that just flinging more forces is not ideal.

Fortunately there are many ways around this.

I certainly agree with the "IF, THEN" type of programming. CEaW does use scripts.

Interestingly so does Civ2. Anyone using notepad can edit the events script in Civ2. It is so easy to use, yet one can include literally hundreds of pre-written commands for the AI.

What would be great to see would be an editor (or part of a larger scenario editor) that would allow the player to designate an invasion force, its composition, location, and time of entry into the game (month/year/turn), etc....



Primasprit -> RE: Europe at War - great game, if it had AI (6/24/2007 8:04:17 PM)


quote:

ORIGINAL: Iain McNeil

[...]even in a simple game like chess they cant make an AI that will beat a human :) [...]

Sorry for being OT but just couldn't resist: Since several years chess AI's beat every human. (And labeling chess as a 'simple game' seems a bit strange...) [8D]



Warfare1 -> RE: Europe at War - great game, if it had AI (6/24/2007 8:06:55 PM)


quote:

ORIGINAL: Primasprit


quote:

ORIGINAL: Iain McNeil

[...]even in a simple game like chess they cant make an AI that will beat a human :) [...]

Sorry for being OT but just couldn't resist: Since several years chess AI's beat every human. (And labeling chess as a 'simple game' seems a bit strange...) [8D]


Have to agree - chess AI - toughest AI I have ever played against. I am beaten even on normal settings.



targul -> RE: Europe at War - great game, if it had AI (6/25/2007 1:21:18 AM)

Personnally I no longer play this game with AI at all.  I can actually hear the AI snoring even with sound effects off.  I love the game verses Humans though.  But the TCP/IP is risky it seems to have a serious flaw other like games do not have.  The lack of replay in PBEM makes that difficult. 

I can live without the AI it isnt there now so I dont need it in the future but I sincerely hope they quickly address the connection problems many of us seem to be experiencing in TCP/IP and get us a PBEM feedback.

Jim



firepowerjohan -> RE: Europe at War - great game, if it had AI (6/25/2007 1:31:48 AM)


quote:

ORIGINAL: targul

Personnally I no longer play this game with AI at all.  I can actually hear the AI snoring even with sound effects off.  I love the game verses Humans though.  But the TCP/IP is risky it seems to have a serious flaw other like games do not have.  The lack of replay in PBEM makes that difficult. 

I can live without the AI it isnt there now so I dont need it in the future but I sincerely hope they quickly address the connection problems many of us seem to be experiencing in TCP/IP and get us a PBEM feedback.

Jim


I have played several TcpIP games with beta version against one of the testers. Only problem was if someone had a restrictive firewall because then that person could not host. But, once we connected and played only occasionally did it crash but this was because physically my opponent lost Internet connection and that is nothing a game can prevent. It is on the physical side, not the gaming side. Game autosaves every end of turn so even if opponent loses Innternet connection you will both have one save from each ones last end turn to continue.

Targul, once you did connect how often did it crash and was it because anyone of you or opponent had a bad Internet connection?



targul -> RE: Europe at War - great game, if it had AI (6/25/2007 1:36:01 AM)

Once it did connect it would crash after each players turn.  Since I played prior and after that with IP connections on different games I MUST believe it is this game not my connection.

There is a possibilty of a router between the two locations but I see so many people complaining about the same problem here I have trouble believe many of us dont have descent internet connections.



Page: [1] 2 3 4   next >   >>

Valid CSS!




Forum Software © ASPPlayground.NET Advanced Edition 2.4.5 ANSI
0.84375