Matrix Games Forums

Forums  Register  Login  Photo Gallery  Member List  Search  Calendars  FAQ 

My Profile  Inbox  Address Book  My Subscription  My Forums  Log Out

The CSA Turtling Problem

 
View related threads: (in this forum | in all forums)

Logged in as: Guest
Users viewing this topic: none
  Printable Version
All Forums >> [Current Games From Matrix.] >> [American Civil War] >> Forge of Freedom: The American Civil War 1861-1865 >> The CSA Turtling Problem Page: [1]
Login
Message << Older Topic   Newer Topic >>
The CSA Turtling Problem - 6/16/2010 9:04:19 PM   
jscott991


Posts: 530
Joined: 4/23/2009
Status: offline
Was this game ever patched to avoid the CSA turtling in Richmond problem? I'm guessing no.

I don't want to restart the debate over this problem, but if the game hasn't changed, I'm probably going to (very sadly) retire it from my hard drive and shelf.
Post #: 1
RE: The CSA Turtling Problem - 6/19/2010 8:06:58 PM   
Gil R.


Posts: 10821
Joined: 4/1/2005
Status: offline
We've been too busy working on the new tactical game ("Brother against Brother") to work on a FOF patch.

(in reply to jscott991)
Post #: 2
RE: The CSA Turtling Problem - 6/20/2010 4:51:26 AM   
jscott991


Posts: 530
Joined: 4/23/2009
Status: offline
That's sad.

If that one problem was fixed, this would be in my top 5 of strategy games ever.

(in reply to Gil R.)
Post #: 3
RE: The CSA Turtling Problem - 6/20/2010 5:19:58 PM   
GIJim

 

Posts: 76
Joined: 1/28/2007
Status: offline
What do you mean with the term "turtling"?

(in reply to jscott991)
Post #: 4
RE: The CSA Turtling Problem - 6/21/2010 2:34:59 PM   
jscott991


Posts: 530
Joined: 4/23/2009
Status: offline
The AI, over time, will pack its capital province (Richmond in this case) with troops, abandoning the western theater and making the game a complete cakewalk.

In my playing experience, on difficulties up to captain, I could not avoid this issue in a single game, but others do not seem to experience it.

There are a few threads where I post tons of screenshots on this, but it's a moot point.

(in reply to GIJim)
Post #: 5
RE: The CSA Turtling Problem - 6/30/2010 2:09:36 PM   
janh

 

Posts: 1216
Joined: 6/12/2007
Status: offline
Unfortunately same for the Union.  Also the "superaggressivity" of AI at the strategic level, often depleting itself in short time, is a bit of a bugger.  It also is slow in feeding reserve brigades into the depleted containers in the western theaters, while always completely trying to fill out the containers in the east. It would be better if AI had a decreasing probability with number of brigades in container (1: 100%, 2: 100%, 3: 85%, 4:50, 5: 25%, 1.st artillery on top: 100%, 2nd art: 0%).

Tweaking down AI importance factors in ACWConstants.txt helps a bit, but not much.   I redid the provinces.txt files, making each state its own province, and Washington and Richmond (James) again a separate province.  That helped both problems substantially, but a real hardcode fix is still necessary. 

Another thing I would like to see moddable would be separate minimum unit strengths for Artillery, and the routing limits in detailed combat (to <250). I would like it to me more realistic in strength, i.e. battalion sized (max 640 men).  I did change this with a hex editor, it required a few edits of the routing strength threshold, the "upgrade threshold" (AI won't upgrade art with less than 2500 men it seems) etc.  It would be nice if all this could be put in the moddable txt files.  Correspondingly I scaled up art efficiency to make up for the smaller crews, but now I can give every of my divisions an artillery battalion without having a 2 million man army (yes, I allowed more units per divisions and div per corps and army as well to allow for a real formation "Army of Potomac"; I tuned down max infantry strength to 2500 men, which is way closer to average brigades in the CW; I also disabled all those unrealistic unit attributes and special  leader abilities, which usually work badly against AI and make the game feel like science fiction; and I tuned weapon, unit and building costs, times, etc to adapt them to the other changes and result in historic equipment and building rates).

Finally I also allowed all art to use indirect fire without the upgrade (more realistic), and tuned down rifle and carabine ranges to typically 2 fields, and max 3 fields for the expensive sharpshooter rifles.  This makes battlefields seem bigger, i.e. more maneuvering space and fun.  Also AI artillery doesn't get slaughtered so easily if the rifles are weaker and have shorter ranges.  I would say AI needs to be tuned to use artillery from longer ranges, and keep it safer in the back.

One more crucial fix I would hope to see would be the promotion system.  I have modified the Commanders.txt to get all the famous names already right at the beginning and kind of assign them as brigade commanders.  It would be nice to be able to correctly assign them, but that is not what bothers me.  AI usually get also way more commanders, and then begins filling up all the containers East with dozens of leaders, while the Western armies and divisions barely get a head.  Leaders should be limited ideally to the number of brigades in the container, plus one for the container leader itself (that goes onto the supply wagon in detailed combat).   And AI should really assign them proper same way. 

Finally, since AI only uses a 20% random in assigning leaders, it doesn't go through a "learning curve" like Lincoln in the CW:  McDowell, McClellan, Pope, McClellan, Burnside, Hooker, Meade, Meade&Grant, to give one example.  It would be nice if (A) the "uncover general ability" parameter were moddable, and I would tune it down a lot.  Then, I would like AI to take into account only known attributes, duty age and present rank, and not right at the start of 1861 put Grant at the head of AoP because he has the best factors.  He should slowly climb up the ladder one rank by one, and only once AI knows his true strengths, put him into promotion.  Maybe one way would be to also allow generals to learn "leadership experience" in battle like units do:  say max of +0.25 gain/loss for the "leadership" category per won/lost battle.  And have all generals start out with similar values except the ones that in 1861 were really of rank: 
0-3: can only be brigade generals
4-6: division capable
7-8: corps
9-10: can be promoted to Army or higher.
Then only allow AI to use this leadership category (plus Cavalry) to assign leaders.  That kind of could mimic the duty age and promotions succession in a simple way and probably is a small enough chance to be modded!

Anything else, like increase the possible number of provinces and subdivision into smaller areas (especially Shenandoah, Fredericksburg, James, Rappahanock or the much contest areas of Chattanoogha) would be awesome, but maybe not as easy to do.

I would really love another patch for this great game, that could be with little effort even much greater!  Or an FOF2!






< Message edited by janh -- 6/30/2010 2:10:16 PM >

(in reply to jscott991)
Post #: 6
RE: The CSA Turtling Problem - 7/1/2010 12:32:14 AM   
Gil R.


Posts: 10821
Joined: 4/1/2005
Status: offline
janh,
Please put any specific ideas in the Wish List thread, since that is the "thread of record." We would certainly like to do more with FOF -- at least a patch, but ideally FOF2 -- but because of other projects are unable to start on either just yet. But we still definitely want any and all suggestions. (And yes, jscott991, tweaking the AI so that the behavior you and some other players have seen becomes less likely would be among the things we'd work on.)

I'm not sure why you think the AI gets more commanders. I just asked our programmer if this is the case and he said it isn't (but does think it would make a splendid AI cheat, so kudos for the idea). And I'm not sure what you mean with the 20% business.

If we do a FOF2 one of the things we've already decided we would want to include would be a more complex system for promotions/demotions, so please do add your thoughts to the Wish List thread, since we'd definitely want to consider any and all ideas.

_____________________________

Michael Jordan plays ball. Charles Manson kills people. I torment eager potential customers by not sharing screenshots of "Brother Against Brother." Everyone has a talent.

(in reply to janh)
Post #: 7
RE: The CSA Turtling Problem - 7/1/2010 11:02:09 PM   
janh

 

Posts: 1216
Joined: 6/12/2007
Status: offline
Sure Gil, I will put them in the wishlist.   And thanks for the reply.  As for commanders, I set the appearance probabilities of all famous names to 200, and most of the other to 50-100, so there are pretty much all the ones that served as bridgade rank or above in the eastern theater, western armies or any other famous campaign.

I read it either in one of the patch readmes, or the manual, that AI uses the leader abilities plus a 20% random of that for determining promotions.  I can check.

(in reply to Gil R.)
Post #: 8
Page:   [1]
All Forums >> [Current Games From Matrix.] >> [American Civil War] >> Forge of Freedom: The American Civil War 1861-1865 >> The CSA Turtling Problem Page: [1]
Jump to:





New Messages No New Messages
Hot Topic w/ New Messages Hot Topic w/o New Messages
Locked w/ New Messages Locked w/o New Messages
 Post New Thread
 Reply to Message
 Post New Poll
 Submit Vote
 Delete My Own Post
 Delete My Own Thread
 Rate Posts


Forum Software © ASPPlayground.NET Advanced Edition 2.4.5 ANSI

0.703