My thoughts on 1.03 (w/ a few screenhots)

Empires in Arms is the computer version of Australian Design Group classic board game. Empires in Arms is a seven player game of grand strategy set during the Napoleonic period of 1805-1815. The unit scale is corps level with full diplomatic options

Moderator: MOD_EIA

User avatar
delatbabel
Posts: 1252
Joined: Sun Jul 30, 2006 1:37 am
Location: Sydney, Australia
Contact:

RE: My thoughts on 1.03 (w/ a few screenhots)

Post by delatbabel »


Most of these are already reported as issues in Mantis.
ORIGINAL: testcase4321

1. I find that after conquering bits of Turkey my corps remaining in conquered Turkish provinces sometimes randomly teleport back to Russia. In fact, this has happened 3 out of 4 times so far, and I have no idea why. In fact, it just happened in the game I am currently playing (troops transported to Odessa). It tends to happen in Winter (February) as this is usually right before my spring campaign

Issues 165 and 108.
2. I can no longer invade England across the channel in France. This occurs even when there are no blocking British fleets.

Are you sure you have the crossing arrow turned on? This is now an option.
4. Placing fleets from Free States. This happened in previous versions as well, and Im not sure if it's a bug or just something goofy. When I conquer Sweden or Denmark and turn them into free states I seem to be able to place their fleets whereever I wish. This includes in the Black Sea, which I have done often. It would seem that a restriction to placing fleets in the home country makes much more sense.

Issue 40. It can also happen by accident.
2. I have noticed that the AI chooses the 'right' options for alliances. i.e. Austria/Britain/Prussia always seem to ally. However, even though they ally they are not able to work together. In fact, while declaring war on every major power in the game I find that only Turkey has invaded my Russia, and even then the Turks seem to be content to attack only provincial corps and not push the attack.

Issue 87, et al.
4. My declarations of war against Austria/Prussia/Britain have led to exactly zero corps invading Russia, coming towards Russia, or threatening me in any way. The game tends to pile up troops in the capital cities (particularly Austria) leaving the rest of the country open.

Issues 35, 147 and friends.
5. Fleet actions are still a bit suspect. I find the Brits dont concentrate their forces against the Russia Navy, leading to significant British losses.

That's not been logged before I suspect.
6. I don't see the French conquering any of the big power territories. They do a better job with the minor states at least, but are not aggressive enough. They also tend in my games to leave large forces in the Swiss area, and do not pull back until Paris is threatened.

Issue 35.
7. I have no idea what the Spanish are doing all game, but with close to no combat and at the Hard AI level, they seem to be my closest competition for game winners.

Issue 167.
1. Instituting the "Alternative great power" rules is really a must (mostly for PBEM play, sa the AI isnt up to challenging humans yet). It shakes up the game quite a bit and throws a wrench into the standard Turkey/Spain/France alliance vs British/Austria/Prussia/Russia alliance.

Issue 92.
2. This game really needs something like a 'badboy' rating. Another term for this is a 'Hate' rating (depending on what other games you have played). These terms refer to the computer opponents reacting to human players on the basis of the human player actions. Essentially, the more aggressive and powerful a human player is, the greater the likelihood of the computer players all ganging up and attacking the human player. I cant tell if EIA has this built into it, and from what I can tell it does not. But my expectation is that at some point, if I as the Russian player have conquered 1/2 the globe, then I should be a target for the AI.

Amusingly enough, I note the following quote from the AI log on Marshall's latest dev release:

Prussia is carbon-based. All humans must die!
3. The small map in the right bottom corner of the screen just bothers me. Its almost, but not quite useful. In addition to marking colors for conquered nations, it really should have a similar scheme for Free States. It is sometimes difficult to tell how powerful nations are, or how much they have spread, when a significant portion of their holdings dont show up on that map. I dont see why we cant have all the conquered and free state nations the same color.

Right click on the map and you will see a menu showing what can be highlighted. You can choose any combination of conquered, free states, and home territory.
--
Del
Thresh
Posts: 393
Joined: Mon Dec 25, 2006 4:19 am
Location: KCMO

RE: My thoughts on 1.03 (w/ a few screenhots)

Post by Thresh »

This is actually part of issue # 35 in Mantis so if the save game could be posted there that'd be great.

There is a general issue with France being too timid from the beginning of the game onwards.

I don't think thats it at all.  I think with the PP bonus's the AI MP gets, they don't have to risk a DoW in oder to try and win their way up to the dominant zone.  An AI france that has 20% of the PP'sneeded by the December of 1805 phase doesn;t have to do a thing forthe rest ofthe game, and will win by a comfortable margin...

Thats what itlooks liketome, I was going to look at it somemore this weekend.

Todd
User avatar
Marshall Ellis
Posts: 5630
Joined: Tue Oct 02, 2001 3:00 pm
Location: Dallas

RE: My thoughts on 1.03 (w/ a few screenhots)

Post by Marshall Ellis »

Todd:
 
No, it's not that she is already winning and in fact I cannot see why she would do this because the AI's craving for PPs is never-ending??? I am collecting games where France is just staling out so send some if you can so that I can further analyze this.
 
 
 
 
 
Thank you

Marshall Ellis
Outflank Strategy War Games


Thresh
Posts: 393
Joined: Mon Dec 25, 2006 4:19 am
Location: KCMO

RE: My thoughts on 1.03 (w/ a few screenhots)

Post by Thresh »

Marshall,

I started a game where I was Spain, with a bid of 13.
Every other MP was on Hard AI.

The bids I used were
France 38
GB 36
SP 13
Pr 15
Tu 10
Ru 23
Au 21

In January of 05, after setup and diplomacy,I checked the VP status.

France was at 13%
GB at14%
Sp at 0%
Pr at 14%
Ru at 15%
Au at 21%

France needs 400 VP to win the game.  At 4 PP phases per year, 44 PP phases Total, France needs 9.09 PP per phase (at aq bid of 0) to win.  The Hard AI  bonus is giving France 52 points before the game even begins.

That means France needs 7.9 points, call it 8 PP per economic phase to win the game.  Starting at the D2 position, France can sit and manipulate to stay in the dominant zone, and never have to do another thing PP wise to win the game.

Which is what I think is happening.  I think the French AI is looking around, sees it doesn't have to doa thing, and does nothing but build.  Any DoW's against France are the result of either a Call, or a random die roll that I haven't figured out.

The only way around such bonus's I see at the moment are to bid ridiculously high numbers for the AI...

I may very well be wrong on all of this though, as programming is not my forte.  :-)

Todd
testcase4321
Posts: 11
Joined: Mon Feb 04, 2008 1:16 am

RE: My thoughts on 1.03 (w/ a few screenhots)

Post by testcase4321 »

Thank you Del for your very helpful responses. I especially appreciate the small map help, it makes the game much clearer.
 
Also, where is this 'Mantis' list or whatever? Id love to take a look at it.
 
Finally, just wanted to share that I was thinking the most recent update was to be a AI enhancement update. I was thinking of this as I pulled my Russian troops off the Turkish border in 1805 after Turkey declared war. For the remainder of the game I watched as the Turkish troops made the daring move one province in to my territory and occupy Podolia/kamenetz. For the next 7 game years, we saw no activity on the Turkish border. I guess this is an improvement over previous versions of the game, where the turks would declare war and then sit on their side of the border, but overall I would say the game dynamics havent changed much. And this is a problem.
 
 
User avatar
Marshall Ellis
Posts: 5630
Joined: Tue Oct 02, 2001 3:00 pm
Location: Dallas

RE: My thoughts on 1.03 (w/ a few screenhots)

Post by Marshall Ellis »

ORIGINAL: testcase4321

Thank you Del for your very helpful responses. I especially appreciate the small map help, it makes the game much clearer.

Also, where is this 'Mantis' list or whatever? Id love to take a look at it.

Finally, just wanted to share that I was thinking the most recent update was to be a AI enhancement update. I was thinking of this as I pulled my Russian troops off the Turkish border in 1805 after Turkey declared war. For the remainder of the game I watched as the Turkish troops made the daring move one province in to my territory and occupy Podolia/kamenetz. For the next 7 game years, we saw no activity on the Turkish border. I guess this is an improvement over previous versions of the game, where the turks would declare war and then sit on their side of the border, but overall I would say the game dynamics havent changed much. And this is a problem.


You are correct that we didn't get all of the AI fixes into this release that we wanted and this was done because we did find some crash bugs. I wanted to get these out ASAP. I apologize for this but keep in mind that 1.04 will have more AI improvements.




Thank you

Marshall Ellis
Outflank Strategy War Games


User avatar
Marshall Ellis
Posts: 5630
Joined: Tue Oct 02, 2001 3:00 pm
Location: Dallas

RE: My thoughts on 1.03 (w/ a few screenhots)

Post by Marshall Ellis »

ORIGINAL: Thresh

Marshall,

I started a game where I was Spain, with a bid of 13.
Every other MP was on Hard AI.

The bids I used were
France 38
GB 36
SP 13
Pr 15
Tu 10
Ru 23
Au 21

In January of 05, after setup and diplomacy,I checked the VP status.

France was at 13%
GB at14%
Sp at 0%
Pr at 14%
Ru at 15%
Au at 21%

France needs 400 VP to win the game.  At 4 PP phases per year, 44 PP phases Total, France needs 9.09 PP per phase (at aq bid of 0) to win.  The Hard AI  bonus is giving France 52 points before the game even begins.

That means France needs 7.9 points, call it 8 PP per economic phase to win the game.  Starting at the D2 position, France can sit and manipulate to stay in the dominant zone, and never have to do another thing PP wise to win the game.

Which is what I think is happening.  I think the French AI is looking around, sees it doesn't have to doa thing, and does nothing but build.  Any DoW's against France are the result of either a Call, or a random die roll that I haven't figured out.

The only way around such bonus's I see at the moment are to bid ridiculously high numbers for the AI...

I may very well be wrong on all of this though, as programming is not my forte.  :-)

Todd

The VP bonus should not influence the number of DOWs. She will always want to win by as much as possible. The latency on the AI's end is odd. It looks like the diplomacy side is better and even the tactical side is better BUT she doesn't close the deal very well (Forming large stacks that moving little) and these types of games are what I would like to see emailed to me so that I can experiment with various scenarios.

Your observations are helpful.



Thank you

Marshall Ellis
Outflank Strategy War Games


Tater
Posts: 60
Joined: Tue Dec 25, 2007 7:06 pm

RE: My thoughts on 1.03 (w/ a few screenhots)

Post by Tater »

ORIGINAL: RayKinStL

I am not sure how you got France to surrender as GB. If I make a run at Paris, the one thing they will do is move that monster stack to protect their capital. Plus the reason you aren't winning is the way Matrix did difficulty levels. Knowing the AI was garbage, they made up for it by giving the computer more VP (by percentage) each turn depending on the difficulty. Difficuly has nothing to do with the way the computer moves, just the way VP is calculated.

That is another thing...if the GB & French start at war option is used the Conditional surrender option should not be available. All I did was win a battle in the Paris area and breach Paris. Next turn France surrendered conditionaly.

OK, now the VP makes sense...it's just a cheat. Hmmm...but then wouldn't this encourage the AI to not be aggressive? If the AI is sitting fat, dumb and happy VP wise what drives the AI to take aggressive/offensive actions?
Later-

Tater
Tater
Posts: 60
Joined: Tue Dec 25, 2007 7:06 pm

RE: My thoughts on 1.03 (w/ a few screenhots)

Post by Tater »

ORIGINAL: Reiryc

Darth tater?

Is that you? [:)]

Yep...it's me in all my electronic glory. [;)]
Later-

Tater
User avatar
Marshall Ellis
Posts: 5630
Joined: Tue Oct 02, 2001 3:00 pm
Location: Dallas

RE: My thoughts on 1.03 (w/ a few screenhots)

Post by Marshall Ellis »

ORIGINAL: Tater

ORIGINAL: RayKinStL

I am not sure how you got France to surrender as GB. If I make a run at Paris, the one thing they will do is move that monster stack to protect their capital. Plus the reason you aren't winning is the way Matrix did difficulty levels. Knowing the AI was garbage, they made up for it by giving the computer more VP (by percentage) each turn depending on the difficulty. Difficuly has nothing to do with the way the computer moves, just the way VP is calculated.

That is another thing...if the GB & French start at war option is used the Conditional surrender option should not be available. All I did was win a battle in the Paris area and breach Paris. Next turn France surrendered conditionaly.

OK, now the VP makes sense...it's just a cheat. Hmmm...but then wouldn't this encourage the AI to not be aggressive? If the AI is sitting fat, dumb and happy VP wise what drives the AI to take aggressive/offensive actions?

There is a stalling effect going on but it is not due to the fact that it is winning. Try the MED PC setting (No bonus VPs) and you will see the same behavior.


Thank you

Marshall Ellis
Outflank Strategy War Games


iamspamus
Posts: 433
Joined: Wed Nov 15, 2006 10:23 pm
Location: Cambridge, UK

RE: My thoughts on 1.03 (w/ a few screenhots)

Post by iamspamus »

ORIGINAL: ndrose

Update on the game played as Austria:
[snip]
What should the AI have done? I can see the difficulty of writing a general rule that would cover this case. And of course I don't know how much money it had. (Though one would think, having done nothing, it should be rich.)

Ideally, it would already have been set up on a depot to the east of Milan when the war started. But barring that, it could either make the calculation to spend the money for one turn to move its whole army forward (and it could have placed a depot in Milan to cut the cost, rather than supplying from Turin). Or (even better in this case, since Spain was about to win the game on points), declare against Spain and eat Parma and Genoa as a snack before advancing on Modena. The Spanish army was in Portugal and posed no threat.

Hmmm. I'd say move in two or three mutually supporting, adjacent stacks, could be an option with the Frenchies and their great leaders. The other option is to go whole hog after Vienna taking Venice on the way. Or finally stomp Venetian corps (since they're spread out), but once again in mutually supporting, adjacent stacks.

How's that?
Post Reply

Return to “Empires in Arms the Napoleonic Wars of 1805 - 1815”