Reviews on COTA caught may attention very strongly, so that I'm considering buying it. However, I'm don't use to play online, so that I will mainly be playing against the computer. I just hate games that present no challenge, no matter if the engine is great and the simulation precise. So, before buying, I would like to know if the computer opponent presents a real challenge and the replayability value of the game.
Thanks.
computer opponent strenght
Moderator: Arjuna
RE: computer opponent strenght
damezzi,
Welcome. You know I'm going to be a tad biased here, but I don't think you be dissappointed buying COTA. I'll leave it to others to comment on the challenge but I can say that our use of generic AI planning - ie it develops its plans based on the situation at the time not according to some scenario specific script - provides a hell of a lot of replayability.
Welcome. You know I'm going to be a tad biased here, but I don't think you be dissappointed buying COTA. I'll leave it to others to comment on the challenge but I can say that our use of generic AI planning - ie it develops its plans based on the situation at the time not according to some scenario specific script - provides a hell of a lot of replayability.
RE: computer opponent strenght
ORIGINAL: Arjuna
Welcome. You know I'm going to be a tad biased here,
For an unbiased opinion you might want to check out these reviews :
http://www.armchairgeneral.com/articles ... e=1&cat=59
http://www.wargamer.com/reviews/conquestoftheaegean/
The ultimate accolade in my opinion is the fact that Conquest of the Aegean was elected Wargame of the Year in all 3 open WOTY elections on the 'Net, i.e. Usenet's War-Historical newsgroup, The Wargamer and GameSquad.
Greetz,
Eddy Sterckx
-
Count Sessine
- Posts: 204
- Joined: Thu Oct 30, 2003 4:13 pm
RE: computer opponent strenght
The way COTA is designed (AI, tacAI, orders) makes singleplayer very challenging, no worries there 
-
jhdeerslayer
- Posts: 1224
- Joined: Sat May 25, 2002 3:24 pm
- Location: Michigan
RE: computer opponent strenght
Actually the AI is one of my favorite aspects.
-
HansBolter
- Posts: 7457
- Joined: Thu Jul 06, 2006 12:30 pm
- Location: United States
RE: computer opponent strenght
As a beta tester for Panther I am a bit biased myself, but the strength of the AI in this game engine is one of the big reasons I became so enamoured with it that I became a tester in the first place.
Here is a very candid excerpt from my AAR of a beta test of the BFTB scenario Race for the Meuse:
"There are three ghost outlines stacking up on the track in the woods north of the FJ. This turns out to be one of the best deployments I made the entire scenario.
After the frontal asault downhill on the town fails, the Amis take another tact and make their second concerted push at the town with a move through the woods.
btw: I was very impressed by the AI as it kept coming back at me with different approaches and tactics.The entrenched FJ absolutely refuses to budge and gives every last unit that comes down that trail a severe bloody nose.
There will be a stack of surrender/destroyed crosses there by games end."
Note the reference to how impressed I was by the AI.
Here is a very candid excerpt from my AAR of a beta test of the BFTB scenario Race for the Meuse:
"There are three ghost outlines stacking up on the track in the woods north of the FJ. This turns out to be one of the best deployments I made the entire scenario.
After the frontal asault downhill on the town fails, the Amis take another tact and make their second concerted push at the town with a move through the woods.
btw: I was very impressed by the AI as it kept coming back at me with different approaches and tactics.The entrenched FJ absolutely refuses to budge and gives every last unit that comes down that trail a severe bloody nose.
There will be a stack of surrender/destroyed crosses there by games end."
Note the reference to how impressed I was by the AI.
Hans
RE: computer opponent strenght
Yup the AI kicks! Get this game already.[:)]
Regards,
Grell
Regards,
Grell
RE: computer opponent strenght
Well, that's the first time I don't see a testimony of the kind 'I can beat the computer opponent easily... the way to go is online play!"
I'm almost convinced of buying it (if AI is as good as you state), but among the reviews listed in the official site I found such a testimony:
And he gave the game 8/8; so, I suppose he is not trying to depreciate it. Anything about this argument?
Is there some space for scripting? Wouldn't an AI without any kind of scripting play always the same way, becoming predictable?
Just one more question: is there any plan to widen the time scope of the game in order to represent other historical periods? Is there any possibility of doing it with the game on the present state? I'm not a wargame expert, so I like games like TOAW, which offers a large variety of scenarios covering different historical periods, so that I don't have to learn a whole new rules set (don't have much time for gaming, sadly) any time I become interested on a specific battle.
Thank you for the answers
I'm almost convinced of buying it (if AI is as good as you state), but among the reviews listed in the official site I found such a testimony:
However, the strategic enemy AI, that issues the orders for the computer player, is too easy to defeat. I’ve found that, in a lot of games, the enemy AI will just sit there and not actively try to occupy their objective locations for some reason, even though they could have easily defeated me with the amount of force at their disposal. They say the AI is improved, but the computer opponent just isn’t aggressive enough to provide a worthy challenge in some of the scenarios. In short, the enemy strategic AI is not aggressive enough at taking objectives, and makes beating the game a little easier than it should be. Really, the strategic AI works better as a defender than an attacker since defending involves less movement.
And he gave the game 8/8; so, I suppose he is not trying to depreciate it. Anything about this argument?
Is there some space for scripting? Wouldn't an AI without any kind of scripting play always the same way, becoming predictable?
Just one more question: is there any plan to widen the time scope of the game in order to represent other historical periods? Is there any possibility of doing it with the game on the present state? I'm not a wargame expert, so I like games like TOAW, which offers a large variety of scenarios covering different historical periods, so that I don't have to learn a whole new rules set (don't have much time for gaming, sadly) any time I become interested on a specific battle.
Thank you for the answers
RE: computer opponent strenght
Re Strategic AI. I don't know where that review came from but I'd have to disagree. Maybe he only played it once or twice. Here's a quote from Eddy, one of our intrepid beta testers. He's just been testing our upcoming BFTB title. Yes it's a new game but it's based on the same engine and I have not tweaked the strategic AI ( yet [;)] ).
Re Scripting. Actually, scenario specific scripting is by far more predictable. Even if this included conditional branching ( eg if A then do X ) at best it can cover several, maybe a dozen options. With generic AI, you will get as many options as there are permitations in the way the units deploy, make contact, fire, react etc. In otherwords as the situation changes so too then the plans change, just as in real life. Sure if the scenario designer limits to objectives to say just one location, then that will constrain the primary options/approaches somewhat. But even then, our AI includes a proper decision making cycle, where plans are developed, executed, forces react and reassess, modifying and replanning as the situation dictates. So an original plan changes over time.
Further, we utilise the diferent commander values to affect the decsions made and the plans developed. Aggressive and timid commanders, those with good and poor judgement, those that are lazy and efficient, they all affect the nature and the speed of the plan developed and executed.
In this way it far less predictable. What might have happened this time won't necessarily happen next time.
Re Scope. Our prime focus to date has been WW2. We do have a data design team working on a modern version, but that will be sometime away yet. I would like to say this though, that developing a game like this that can realistically simulate operational warfare is a massive undetaking ( around 40 person years of effort ). While in theory the engine is very scaleable and adaptable, the effort required to make the changes necessary for other time periods ( eg the combat system, communications, weapons etc ) would be considerable. If someone wants to bankroll a Napoleonic or Ancients version then please call...we're always happy to listen to a well funded business proposal.[:)]
Surprise ! This time the Germans opted for going for Manhay over the main highway. It seems that every time I play this scenario the Germans try something new - not always the best choice in hindsight but no complaints on the replayability front [;)]
Re Scripting. Actually, scenario specific scripting is by far more predictable. Even if this included conditional branching ( eg if A then do X ) at best it can cover several, maybe a dozen options. With generic AI, you will get as many options as there are permitations in the way the units deploy, make contact, fire, react etc. In otherwords as the situation changes so too then the plans change, just as in real life. Sure if the scenario designer limits to objectives to say just one location, then that will constrain the primary options/approaches somewhat. But even then, our AI includes a proper decision making cycle, where plans are developed, executed, forces react and reassess, modifying and replanning as the situation dictates. So an original plan changes over time.
Further, we utilise the diferent commander values to affect the decsions made and the plans developed. Aggressive and timid commanders, those with good and poor judgement, those that are lazy and efficient, they all affect the nature and the speed of the plan developed and executed.
In this way it far less predictable. What might have happened this time won't necessarily happen next time.
Re Scope. Our prime focus to date has been WW2. We do have a data design team working on a modern version, but that will be sometime away yet. I would like to say this though, that developing a game like this that can realistically simulate operational warfare is a massive undetaking ( around 40 person years of effort ). While in theory the engine is very scaleable and adaptable, the effort required to make the changes necessary for other time periods ( eg the combat system, communications, weapons etc ) would be considerable. If someone wants to bankroll a Napoleonic or Ancients version then please call...we're always happy to listen to a well funded business proposal.[:)]
-
Deathtreader
- Posts: 1058
- Joined: Tue Apr 22, 2003 3:49 am
- Location: Vancouver, Canada.
RE: computer opponent strenght
Hey damezzi,
If we look hard enough we'll find negative reviews on everything under the sun..... but for every negative review on this game (from those who have actually played it more than once or twice) I'll wager there will be at least 7 or 8 that praise it. It's AI is not perfect, but it is far and away the best out there. I hate losing to machines..... yet when I lose to this one I always learn something of value in tactics etc. I've never lost and walked away feeling that some gimmicky AI aid, or stacked die rolls or some pro-AI coding caused me to lose. I have all the games in the series, the AI just keeps getting better and better, meaning I have to also.
Buy it !!!!
Rob. [:)]
If we look hard enough we'll find negative reviews on everything under the sun..... but for every negative review on this game (from those who have actually played it more than once or twice) I'll wager there will be at least 7 or 8 that praise it. It's AI is not perfect, but it is far and away the best out there. I hate losing to machines..... yet when I lose to this one I always learn something of value in tactics etc. I've never lost and walked away feeling that some gimmicky AI aid, or stacked die rolls or some pro-AI coding caused me to lose. I have all the games in the series, the AI just keeps getting better and better, meaning I have to also.
Buy it !!!!
Rob. [:)]
So we're at war with the Russkies eh?? I suppose we really ought to invade or something. (Lonnnng pause while studying the map)
Hmmmm... big place ain't it??
- Sir Harry Flashman (1854)
Hmmmm... big place ain't it??
- Sir Harry Flashman (1854)
RE: computer opponent strenght
Funny you say that, Dave. I remember when I first got TacOps in 1996 and there was one scenario that I just couldn't win. I basically tried it about 30 times until I essentially had the scripted movements of OPFOR memorized, and only then could I defeat it on a consistent basis.
ORIGINAL: Arjuna
Re Scripting. Actually, scenario specific scripting is by far more predictable. Even if this included conditional branching ( eg if A then do X ) at best it can cover several, maybe a dozen options. With generic AI, you will get as many options as there are permitations in the way the units deploy, make contact, fire, react etc. In otherwords as the situation changes so too then the plans change, just as in real life. Sure if the scenario designer limits to objectives to say just one location, then that will constrain the primary options/approaches somewhat. But even then, our AI includes a proper decision making cycle, where plans are developed, executed, forces react and reassess, modifying and replanning as the situation dictates. So an original plan changes over time.
RE: computer opponent strenght
Crimguy,
Note that I said scenario specific scripting ( SSS ) was more predictable, not that therefore it was more beatable. Now I played Tac Ops a lot back in the mid 1990s and I played every scenario released many times. The fact that you quickly knew what the enemy was going to do doesn't mean you have the resources to defeat him. That depends largely on the allocation of forces.
Typically, the way most wargames that rely on SSS counter this predicatability is to provide the AI side with overwhelming forces. So the challenge is not in working out how, when and where the enemy will strike but how I can preserve my forces long enough to attrit the enemy down. Now this may be fine as an exercise in itself but it robs the player of half the challenge and half the fun.
PS there is one other logical explanation for why you couldn't win against the sripted opponent but I'll refrain from raising it here. [;)]
Note that I said scenario specific scripting ( SSS ) was more predictable, not that therefore it was more beatable. Now I played Tac Ops a lot back in the mid 1990s and I played every scenario released many times. The fact that you quickly knew what the enemy was going to do doesn't mean you have the resources to defeat him. That depends largely on the allocation of forces.
Typically, the way most wargames that rely on SSS counter this predicatability is to provide the AI side with overwhelming forces. So the challenge is not in working out how, when and where the enemy will strike but how I can preserve my forces long enough to attrit the enemy down. Now this may be fine as an exercise in itself but it robs the player of half the challenge and half the fun.
PS there is one other logical explanation for why you couldn't win against the sripted opponent but I'll refrain from raising it here. [;)]
RE: computer opponent strenght
You mean my stupidity? I resemble that remark![:D]
IIRC, TacOps had a number of scenarios, with different variations to make them more difficult, e.g. TF Johnson I, TF Johnson III, which was more difficult, etc. Major took pains it appeared to make them all winable, and some even had certain tactical lessons to be had. My problem was indeed bad tactics and an insufficient concentration of force (amazing how you remember these things 12 years later!). But, when the ai becomes predictable, you must agree it takes most of the difficulty out of the equation.
It could be equated to a first-person shooter, where one has to beat the Boss at the end. It took me about 25 tries when I played Marathon as a kid.
Regardless, your AI, whatever it's doing, is very solid, arguably the best in the business.
IIRC, TacOps had a number of scenarios, with different variations to make them more difficult, e.g. TF Johnson I, TF Johnson III, which was more difficult, etc. Major took pains it appeared to make them all winable, and some even had certain tactical lessons to be had. My problem was indeed bad tactics and an insufficient concentration of force (amazing how you remember these things 12 years later!). But, when the ai becomes predictable, you must agree it takes most of the difficulty out of the equation.
It could be equated to a first-person shooter, where one has to beat the Boss at the end. It took me about 25 tries when I played Marathon as a kid.
Regardless, your AI, whatever it's doing, is very solid, arguably the best in the business.
ORIGINAL: Arjuna
Crimguy,
Note that I said scenario specific scripting ( SSS ) was more predictable, not that therefore it was more beatable. Now I played Tac Ops a lot back in the mid 1990s and I played every scenario released many times. The fact that you quickly knew what the enemy was going to do doesn't mean you have the resources to defeat him. That depends largely on the allocation of forces.
Typically, the way most wargames that rely on SSS counter this predicatability is to provide the AI side with overwhelming forces. So the challenge is not in working out how, when and where the enemy will strike but how I can preserve my forces long enough to attrit the enemy down. Now this may be fine as an exercise in itself but it robs the player of half the challenge and half the fun.
PS there is one other logical explanation for why you couldn't win against the sripted opponent but I'll refrain from raising it here. [;)]
RE: computer opponent strenght
Yes, as I said earlier the AI is outstanding, and the replay value is very high because the AI always does something different, or it seems like always.[:)]
Regards,
Grell
Regards,
Grell
-
06 Maestro
- Posts: 3997
- Joined: Tue Oct 11, 2005 10:50 pm
- Location: Nevada, USA
RE: computer opponent strenght
damezzi
I'll chime in here saying that CotA has the best AI around. It is one of the 1st things that will strike you about the game is how the AI will respond differently to your very similar moves.
I play against the AI about 3/4 the time, and online the rest. Playing against the AI is not as difficult as playing against a competent biological unit, but it will always give you a good fight, without the AI cheats which are inherent in most other games. The CotA AI is quite capable of defeating a human player, as I have seen on more than a few occasions. There are some particularly difficult historical battles included in the game which will take much more than one go to achieve victory.
CotA is as good as it gets for an operational level game. It is also my 1st choice for gaming overall.
I'll chime in here saying that CotA has the best AI around. It is one of the 1st things that will strike you about the game is how the AI will respond differently to your very similar moves.
I play against the AI about 3/4 the time, and online the rest. Playing against the AI is not as difficult as playing against a competent biological unit, but it will always give you a good fight, without the AI cheats which are inherent in most other games. The CotA AI is quite capable of defeating a human player, as I have seen on more than a few occasions. There are some particularly difficult historical battles included in the game which will take much more than one go to achieve victory.
CotA is as good as it gets for an operational level game. It is also my 1st choice for gaming overall.
Banking establishments are more dangerous than standing armies.
Thomas Jefferson
Thomas Jefferson
RE: computer opponent strenght
ORIGINAL: 06 Maestro
There are some particularly difficult historical battles included in the game which will take much more than one go to achieve victory.
I hope that the majority of them demands more than one go to beat.
As to the predictability, I agree with you Arjuna that when playing a scenario a lot of times, scripted AI becomes very predictable, but with scripted AI, when you change scenario wouldn't it show a greater difference in the way computer plays than a non-scripted one, since it will reflect the strategy of the designer, while non-scripted AI would use more or less the same variables and principles aways? I agree that an AI that could be 'creative' wouldn't need any scripting when it comes to strategy, but since I'm not a programmer, I don't know if that could be the case here.
I decided to buy the game, but just want to ask one more question: is there a reasonable number of user-made scenarios on the net or are the scenarios that come with the game all we get? naval and air battles aren't part of the game, or...? what about events?
Anyway, I would like to thank for your help... I was impressed with the good will and promptness of people in this forum.
-
06 Maestro
- Posts: 3997
- Joined: Tue Oct 11, 2005 10:50 pm
- Location: Nevada, USA
RE: computer opponent strenght
ORIGINAL: damezzi
I hope that the majority of them demands more than one go to beat.
That is a very safe bet.
Banking establishments are more dangerous than standing armies.
Thomas Jefferson
Thomas Jefferson
RE: computer opponent strenght
ORIGINAL: damezzi
As to the predictability, I agree with you Arjuna that when playing a scenario a lot of times, scripted AI becomes very predictable, but with scripted AI, when you change scenario wouldn't it show a greater difference in the way computer plays than a non-scripted one, since it will reflect the strategy of the designer, while non-scripted AI would use more or less the same variables and principles aways? I agree that an AI that could be 'creative' wouldn't need any scripting when it comes to strategy, but since I'm not a programmer, I don't know if that could be the case here.
Enjoy!I decided to buy the game,
but just want to ask one more question: is there a reasonable number of user-made scenarios on the net or are the scenarios that come with the game all we get? naval and air battles aren't part of the game, or...? what about events?
There are a number of user made scenarios - around a half a doxen at the moment but I know more are in the works. You can download them from the COTA website - http://cota.matrixgames.com/ - or from the Wargamer's Games Depot site - http://www.wargamer.com/gamesdepot/. I would recommend you check out the Scenario Design and Modding sub-forum - tt.asp?forumid=284.
You're welcome. Thanks to all who responded.Anyway, I would like to thank for your help... I was impressed with the good will and promptness of people in this forum.