Current difficulty? (Full Version)

All Forums >> [New Releases from Matrix Games] >> Empires in Arms the Napoleonic Wars of 1805 - 1815



Message


von altair -> Current difficulty? (2/26/2010 5:45:20 PM)

Hello! I have been visiting this forum to see current situation about Empire in Arms. I have always been intrested
about this game. Have been at edge to buy this game many many times. But always when I come to this forum, I am
seeing a lot of unhappy people.

My most important requirement for the game is working decent AI. Could someone please tell me how it works now
and what about incoming 1.08 patch which seems to be under work. I would like to learn the game and play
against decent AI before moving forward in MP games. Not going to spend the money on the game which AI can't
even offer baby sitter difficulty level. I have seen too many of them. Thanks.




StCyr -> RE: Current difficulty? (2/27/2010 2:35:04 AM)

Marshall still did not get the rules and needs to ask for assistance to understand how the game works- how can you expect a decent AI from such a person ?
There is no chance for a working decent AI.

But I think you donīt need it to understand (and "learn") EiA itself.




pzgndr -> RE: Current difficulty? (2/27/2010 3:31:39 AM)

quote:

I am seeing a lot of unhappy people.


Well don't confuse unhappy people with those who are pessimistic versus optimistic. To some degree everyone is unhappy because there are still issues with the game, pbem issues, and AI issues. I'm not exactly happy. Most here are optimistic that these issues will eventually get resolved. A few like StCry have a decidedly negative attitude and continue to bash the game regardless.

quote:

I would like to learn the game and play against decent AI before moving forward in MP games.


Fair enough. If you are starting from scratch, be prepared for a steep learning curve. It takes time to learn the many nuances of the game, and time to play through each of the 7 major powers and learn their strengths and weaknesses. The current AI, despite its weaknesses, is still useful as a learning tool. Why not get started and learn the game? Each patch is making the game better and the AI is continuing to improve bit by bit.




Dancing Bear -> RE: Current difficulty? (2/27/2010 3:37:29 AM)

StCyr, have you ever played a game were players have not argued over the rules? The original EIa rule book has holes you can drive a truck through, so I think it is reasonable for Marshall to ask for assistance.

Von Altair, you can not expect any AI to play like a human. As Aresmars said on a post a long time ago, humans make much bigger mistakes than an AI. For this game, you want to play the PBEM. The AI is good for the mechanics, but no AI will ever give you the unpredictable aspects of a human game. Buy the game in a couple of weeks and practice with the AI. Join a PBEM game with version 1.08.

quote:

ORIGINAL: StCyr

Marshall still did not get the rules and needs to ask for assistance to understand how the game works- how can you expect a decent AI from such a person ?
There is no chance for a working decent AI.

But I think you donīt need it to understand (and "learn") EiA itself.





NeverMan -> RE: Current difficulty? (2/27/2010 3:49:02 PM)

The AI is pretty lame at the moment. It's not "terrible", but comes across more like it doesn't really have any interesting in playing the game, IMO.

I would not expect good AI play from this game... anytime soon, JMO.




von altair -> RE: Current difficulty? (2/28/2010 1:18:14 AM)

Thanks for replies!

I am not worried about learning curves. I've played strategy games back from ZX-81 to this day. If some can play the games like
War in the Pacifig: Admiral Edition, I think this wont cause any problems either.

quote:

ORIGINAL: Dancing Bear
Von Altair, you can not expect any AI to play like a human. As Aresmars said on a post a long time ago, humans make much bigger mistakes than an AI.


Well, there are very good AI in many strategy games. They even made AI which can beat a human in a chess! (IBM Deep Blue). Its possible, its just how much each case have resources/will to do it (IBM had a lot of both :)

As I am positive guy, I belive Marshall can make decent AI in time, while he may have fewer resources. But with good will and community support, it will come better and better in time. Thats why I have been here, checking out, if it might be this month, or mayby next one :)




Skanvak -> RE: Current difficulty? (2/28/2010 9:55:41 AM)

AI for chess cannot be compare as there is one winning serie of moves that garantee victory (this is a mathematical theorem useless to event try to argument).

Here you have to take into account your opponent diplomacy profile and randomness. Though I think that an AI for one player is feasible, and ultimatly an AI for diplomacy too. I just don't think that Marshall Alone can do anything better than an AI for a battle scenario.




NeverMan -> RE: Current difficulty? (2/28/2010 11:19:06 PM)

Not to be an ass, but yeah, the chess comparison is not an accurate one.

And Deep Blue is not a program, it's a machine.




Grognot -> RE: Current difficulty? (3/19/2010 8:02:17 AM)

A friend and I are playing against five 'hard' AIs, with us as Austria and Prussia.  Observations iin this 1.07.01 game include:

- Russia DOW's Sweden and didn't even enter Finland, let alone Sweden.  Instant lapse.

- Russia is nominally at war with Turkey, with no forces anywhere near, and with its army mostly in St. Pete.

- Turkey DOW'd Egypt, sent only a single feudal corps into the Sinai, ran into my in-your-face defense w/ one Egyptian corps, and let the war lapse by running back across the border.  Hello, Austrian Egypt... and even during the present Austrian-Turkish war, the Turkish fleet has made no attempt to land anybody... anywhere at all.  The Turkish transport fleet *did* make an unescorted trip to deliver a massive force of zero corps to a port in Palestine, where there were also exactly zero Turkish factors anywhere near that could be picked up.  This got them blockaded and annihilated when the Egyptian corps dropped a one-factor garrison there.

- Turkey has frequently had long depot chains to nowhere, and even with the extra money from 'hard' is having trouble managing its finances resulting in starvation.

- Turkey is nominally at war with Austria, but has made no attempt to invade other than a tentative solo one-area advance with the Syrian corps... which ran away when confronted.   The Austrian forces near them have mostly been patiently waiting, not rampaging, and the majority of Austrian forces are much closer to the French. 

- Likewise, Turkey is at war with Russia, but Turkey appears to make no attempt to actually wage war on them, nor to declare a peace.  Russia and Turkey are acting like a non-existent white peace is in effect.

- Great Britain's fleet has spent the majority of its time hanging around off Stockholm.  Ditto for the British troops.  Never mind that Great Britain is at war with nobody but France and Turkey, neither of whom poses an obvious threat to Sweden under most circumstances.  Must be the blondes.

- France's fleet is content to let itself be blockaded even with massive odds in its favor and the near-certainty of capturing Admiral Nelson, who tends to be on tiny stacks.

- France's Napoleon is often found managing the home front, leading a small rear guard at Paris, while Ney was given the I and IV Corps.  When attacked by an Austrian stack of  a few corps under Charles (not all, the rest were glaring at the Turks), Ney -- not exactly a leader with high strategic rating, and with the advantage of defending against a river crossing -- well, Ney chose "withdraw" and failed, resulting in the near-total destruction of the previously full-strength I Corps with precisely zero losses to Austria.  There *is* a dangerous-looking French Stack of Doom, but Napoleon is nowhere near.

- France has managed to lose all its minors in the first few months by landing in instability. 

- France has pointlessly sacrificed PP by attempting to illegally declare war on a minor when there were British corps in French national provinces.

- I have seen exactly zero attempts by the French AI to really exploit the double-move potential, e.g. by attacking two stacks in a row with the same Stack of Doom or the like.

- Spain DOW'd Algeria, and let the war immediately lapse.


Quite a few of these strike me as significant regressions. The Russians used to march a single corps into Finland (but not the capital) to prevent that war from lapsing, the British fleet tended to be blockading the French fleet in appropriate strength rather than hanging out with the Swedes, the Turks used to move in force on Egypt, and so forth.




NeverMan -> RE: Current difficulty? (3/19/2010 9:17:46 PM)

yep, the AI for this game SUCKS, it's HORRIBLE...

... will it get better? Maybe. How much better? Probably not much.




von altair -> RE: Current difficulty? (3/21/2010 11:19:25 PM)


quote:

ORIGINAL: Grognot

A friend and I are playing against five 'hard' AIs, with us as Austria and Prussia.  Observations iin this 1.07.01 game include:



gulps... okay seems that you have quite good knowledge about AI... I must admit that those things doesn't courage me to
buy the game...This is strange... when this game released again?




Dancing Bear -> RE: Current difficulty? (3/22/2010 3:05:34 AM)

This is a political game first and a military one second. You won't get too much political from an AI. Practice with the AI, but you really want to play this game PBEM (version 1.08).




Tater -> RE: Current difficulty? (3/22/2010 3:42:31 PM)


quote:

ORIGINAL: NeverMan

yep, the AI for this game SUCKS, it's HORRIBLE...


Very true...

quote:

... will it get better? Maybe. How much better? Probably not much.


Disagree...the AI is so bad that almost anything would be an improvement. I think the AI could be improved tremendously (because it is so very poor even incremental improvements would be tremendous). The AI may never reach the point of winning any given playing...but , I think it could be improved to the point that it acted as though it knew the rules. Right now the AI acts as though it has absolutely no idea what to do in any given situation...it can't seem to even manage very basic maintenance requirements much less manage tactical considerations.

It has been that way from the start...it hasn't changed much...and it won't change until a decision is made by Matrix to invest the resources required to make the AI better.




Tater -> RE: Current difficulty? (3/22/2010 3:44:36 PM)


quote:

ORIGINAL: Dancing Bear

This is a political game first and a military one second. You won't get too much political from an AI. Practice with the AI, but you really want to play this game PBEM (version 1.08).


Partly true...however, shouldn't the AI at least be able to recognize when to feed it's troops?




pzgndr -> RE: Current difficulty? (3/22/2010 5:52:52 PM)

quote:

I think the AI could be improved tremendously ... It has been that way from the start...it hasn't changed much...


Of course it can be improved! Unfortunately the AI hasn't received much attention since release due to higher priority efforts to fix bugs and improve PBEM. When Marshall gets around to AI improvements and enhancements, we should see better computer opponent(s). Even a few modest improvements will go a long way. Real improvements resulting in challenging gameplay (not brilliant, as some seem to unreasonably expect/demand?) will take some time and effort. Marshall says he's in it for the long haul, so I'm patiently optimistic.




Grognot -> RE: Current difficulty? (3/22/2010 6:59:37 PM)

High-level strategy is difficult.   However, there are lower-hanging fruit, such as --

- avoiding stacking leaders where a much worse superior overrides an inferior (e.g. Bernadotte overrides Massena or Soult, because it's B-vs-C); more obvious for France where they almost always have a choice of good leaders (Napoleon, Massena, Davout, Soult...) than the Bernadottes of the world.   Russia likewise should be reluctant to use Alexander; there are normally better alternatives.

- not spending money maintaining depots that are the ends of supply chains and which have nobody using them

- not choosing 'withdraw' when it has superior numbers and superior quality on its side (or, alternately, when led by low-strat leaders like Ney) -- Ney's choice turned what should have been a nasty and far-from-certain fight into a one-sided slaughter.  It's more useful for the very outnumbered.  It's also possibly useful to deny battle in other times, but the AI probably shouldn't unless the withdrawing leader has a high odds to withdraw successfully, because otherwise the risk of catastrophe is pretty high.

- Russia and Turkey really, really shouldn't let their obvious first-minor-target wars lapse unless they are embroiled in a major war, because they're that strategically important.





Tater -> RE: Current difficulty? (3/23/2010 1:46:23 AM)


quote:

ORIGINAL: pzgndr

quote:

I think the AI could be improved tremendously ... It has been that way from the start...it hasn't changed much...


Of course it can be improved! Unfortunately the AI hasn't received much attention since release due to higher priority efforts to fix bugs and improve PBEM. When Marshall gets around to AI improvements and enhancements, we should see better computer opponent(s). Even a few modest improvements will go a long way. Real improvements resulting in challenging gameplay (not brilliant, as some seem to unreasonably expect/demand?) will take some time and effort. Marshall says he's in it for the long haul, so I'm patiently optimistic.


Why does it have to be a sequential effort? Why is Matrix not putting the resources into a game they took money for to fix problems more rapidly in a broad based format rather than a piecemeal fashion? Couldn't they have some one working bugs while someone else works the AI? Seems to me Matrix took the money and ran while throwing a few crumbs to those who bought the game.

I could be a lot more optomistic if I really saw some serious effort from Matrix to fix the game. Sure, Marshall does what he can but this game needed a team of folks working the issues from the start...it has never recieved serious effort from Matrix...and it still isn't.




NeverMan -> RE: Current difficulty? (3/24/2010 3:54:33 AM)

Tater,

Here's the problem and the truth of it:

You paid $70 for this game. You think that's a lot of money to pay for a game (and I'd agree). Sadly, not that many people have probably bought the game (I'll leave it up to you to do the math with this one, but even if they sold 3000 copies, that's not much money for the YEARS this game has been in development, not even enough to pay for one developer). They would have to sell like 10000 copies before they even start thinking about breaking even probably. This game has been in development/released for like 6 years. An experienced developer getting paid ****ty money makes 70k (****ty money for an experienced developer), again, you do the math.

So, now how many copies would Matrix have to sell in order to be able to have two full time developers on this game. This does not include distribution, storage space for game downloads, marketing, release engineer, etc, etc, etc.....

The VERY SAD fact is that these games simply don't have the market to throw a lot of resources at them.

In fact, even Marshall has admitted that he's doing this thing part-time, so they can't even afford a full time developer.

Personally, I honestly believe that is all the more reason to make this thing open source or at least have some closed source API.




Tater -> RE: Current difficulty? (3/30/2010 1:47:39 AM)


quote:

ORIGINAL: NeverMan

Tater,

Here's the problem and the truth of it:

You paid $70 for this game. You think that's a lot of money to pay for a game (and I'd agree). Sadly, not that many people have probably bought the game (I'll leave it up to you to do the math with this one, but even if they sold 3000 copies, that's not much money for the YEARS this game has been in development, not even enough to pay for one developer). They would have to sell like 10000 copies before they even start thinking about breaking even probably. This game has been in development/released for like 6 years. An experienced developer getting paid ****ty money makes 70k (****ty money for an experienced developer), again, you do the math.

So, now how many copies would Matrix have to sell in order to be able to have two full time developers on this game. This does not include distribution, storage space for game downloads, marketing, release engineer, etc, etc, etc.....

The VERY SAD fact is that these games simply don't have the market to throw a lot of resources at them.

In fact, even Marshall has admitted that he's doing this thing part-time, so they can't even afford a full time developer.

Personally, I honestly believe that is all the more reason to make this thing open source or at least have some closed source API.


If that is the math then Matrix is guilty of "fraud in the inducement".




Skanvak -> RE: Current difficulty? (4/6/2010 7:26:21 AM)

quote:

I honestly believe that is all the more reason to make this thing open source or at least have some closed source API


I think so too. Especially for AI. We need more people on it. Beside it is fun to design an AI.




pzgndr -> RE: Current difficulty? (4/6/2010 1:43:53 PM)

quote:

make this thing open source


Sure. Let's make all proprietary information open source while we're at it, and then see what happens to free enterprise.

There's nothing stopping Neverman from starting an open source EiA project on SourceForge.net, as long as he's not making any money or profit in violation of copyright laws. So, where's that at?? If it's all so simple... [8|]




Grognot -> RE: Current difficulty? (4/6/2010 8:54:15 PM)

"Trademark violation" is certainly not "nothing", and profit is not necessary at all to justify litigation.




NeverMan -> RE: Current difficulty? (4/6/2010 9:31:50 PM)


quote:

ORIGINAL: Grognot

"Trademark violation" is certainly not "nothing", and profit is not necessary at all to justify litigation.


Yep! Pzn failed law school 101.




borner -> RE: Current difficulty? (4/7/2010 3:32:13 AM)

Oh, if only everyone had as positive an outlook as Ole PZ the world would be a better place... heck, a hurricane could roar through and he would probably be happy his yard was getting watered!!! [sm=00000436.gif] I would not say Matrix is gulity of fraud, but i think 95% of the people that bought this the first month after it was released agree it was far from ready, and Marshall has been forced to scramble to address one bug after another as they came up. As such, while at first glance you would think the AI could/should have been looked at by now, there have been more important things to get fixed. Plus, is there any AI out there for any game that really deals well with the diplomatic side of any game?




Grognot -> RE: Current difficulty? (4/7/2010 5:55:15 AM)

Personally, I'm more bothered by the apparent regressions than the the rest.  I distinctly remember fewer significant lapses-of-war in previous versions, for instance -- e.g. the Russian AI would normally stick an infantry corps just into Finland (but not the Finnish capital!) for this very reason.

I also don't recall the AI having been quite as self-mutilating when choosing whom to sue for peace -- twice now in the same game, an AI has chosen to unconditionally surrender to the opponents... who *aren't* the ones occupying their capital in force, leading to an additional surrender (and thus another round of nasty terms and PP loss) to the one who is.  There are cases where this can make sense (like one of those other enemies having captured a large number of factors and a good leader), but those cases are rare enough that they're the exception to the rule.   Strange.





Skanvak -> RE: Current difficulty? (4/7/2010 7:00:54 AM)

Opening the AI source code, without the main code, won't impact profit margin. Second you can make something open without making it right free. The idea would be to let anyone code the AI. There are people out there that just like to code AI just for the challenge.

quote:

Plus, is there any AI out there for any game that really deals well with the diplomatic side of any game?


No, I made research about the game "Diplomacy" they have been developping AI for years now and still don't give good result as I read on their website. There is a club of guy that do that.

But we could have a sparing partner like military AI. Thyat is in the realm of the possible. I just think that we need someone else to work on it.

As I see, the problem would be to make the link between the diplomacy and the strategic thinking for the computer.




Grognot -> RE: Current difficulty? (4/7/2010 8:24:10 AM)

quote:

ORIGINAL: Skanvak
Opening the AI source code, without the main code, won't impact profit margin. Second you can make something open without making it right free. The idea would be to let anyone code the AI. There are people out there that just like to code AI just for the challenge.


That's not likely to be feasible unless the code base was designed to be that modular from the very beginning. If the AI code is woven throughout the rest of the code base or requires direct knowledge of the implementation of internal data structures rather than of just some abstract API, then it's useless to open it up without also opening up those details. And *that* may have other problems, such as revealing too much IP about the engine itself or facilitating misuse of the saved game files.




pzgndr -> RE: Current difficulty? (4/7/2010 12:38:39 PM)

quote:

Opening the AI source code, without the main code, won't impact profit margin. Second you can make something open without making it right free. The idea would be to let anyone code the AI. There are people out there that just like to code AI just for the challenge.


Again, this is totally beside the point. It would be a sensitive business decision for Matrix Games and the individual developers like Marshall and Outflank Games to release their proprietary game code, in whole or in part. Not likely. Some folks need to go back and review Business 101, or try starting their own business and put up or shut up. [;)]




NeverMan -> RE: Current difficulty? (4/7/2010 4:42:59 PM)

While I agree with Skanvak.... I'm inclined to believe what Grognot believes, and that is that this game was poorly designed (ie. is not coded robustly or modularly), so it's a moot point.

If it was then I do think it would be ok to release that part of the code, if done in an API fashion.

Skanavak, if you are interested in designed the AI, you can certianly do this without knowledge of the code, since you have knowledge of the game. If you come up with a design and it's good, maybe Marshall will implement it.




borner -> RE: Current difficulty? (4/7/2010 5:26:12 PM)

So we should all agree with Pz or shut up? [sm=00000506.gif]




Page: [1] 2   next >   >>

Valid CSS!




Forum Software © ASPPlayground.NET Advanced Edition 2.4.5 ANSI
1.25