Rebs March Deep Into The Union Rear

[Black Cat]

I like this Game a great deal, and try to play the land warfare part of it in a historical manner using the opperational doctrines of the times, the actual historical timeline driving my gameplay. Like most here, I just play the AI.

Mutually supporting Corps, formed of Divisions, with mixed arms included, establish or capture a supply depot. They re-organize, and then repeat the procedure. Along the road to Richmond or Washington they will win or lose battles. Because of that plans will have to be altered to take advantage of Victories or fall back from Defeats.

The developers of this amazing Game have given us all the tools to recreate those aspects of the ACW.

Playing with patch 1.06 as the Union, I started to notice that the AI _ now _ seems to exhibit the suicidial behavior with large formed units, up to, and including small Corps that use to characterize those deep kamakazie Cav. raids in the 1.05 versions.

In Sept. of 1861 a lone Reb division got itself cut off south east of Cincinnati, and after weather attrition losses through Nov. was wiped out by Grants Large Division.

Now, in July 1862 Grant is about to assault Bowling Green.

However Bragg breaks through McClernad`s strong Division at Grayson Ky., on Grant`s right ( with an overwater attack that costs him little losses BTW )

Bragg wins another battle at Mead Ky. against a weak Div. and then moves to assault and siege Louisville after winning a battle there. He has covered 168 real map miles, crossed 2 rivers, and fought 3 major battles in one Game turn of 2 weeks.

I guess the hard marching Reb Infantry might be able to do that.....but why would the otherwise smart AI (previous to 1.06 patch) do that unless it thinks he`s a Cav. raid..???

He`s isolated, can`t take Louisville, and I will destroy him with Shermans Division from Grant`s Corps and Porter`s Divsion from Lexington.
The tool tip indicates he has 2 divisions with him, however I wonder if it`s just his HQ sitting there, which seems to be the AI`s newest best trick, disappearing the line units in a stack after a deep advance.

It did this around Harpers Ferry with Jackson Corps in late 1861, when I cought him North East of there, there was just a HQ ?

The AI of 1.06 was supposed to be better, but it seems just more aggressive.

Playing my usual careful, slow, historical Game, without huge killer stacks, or taking advantage of AI weakness, at July 1862 the Reb losses are 41,186. Union 21,700.

Something is not right here, and I think it`s the "new' 1.06 AI..

[PBBoeye]

Interesting. Just bought it, so I'd be interested to hear if Bragg's command was just the HQ left, as you mention as a possibility.

[McNaughton]

Well, what are your difficulty stats? What is the FOW setting? This information would actually be useful! Try a more passive AI setting if this one is too aggressive, try very minimal, or even no FOW bonus (try one, then the other, then both to see if it improves things). Maybe the AI has reached the point where it needs no FOW bonus?

[Black Cat]

Well, what are your difficulty stats? What is the FOW setting? This information would actually be useful! Try a more passive AI setting if this one is too aggressive, try very minimal, or even no FOW bonus (try one, then the other, then both to see if it improves things). Maybe the AI has reached the point where it needs no FOW bonus?

Well I think it needs something to stop those tactics, which as I said did not seem to occur in 1.05 vers. which I played with the *same settings* for at least 50+ hours and 4 re-starts of the April Grand Campaign.

Here you go.

AI Difficulty is: Normal
AI Aggresiveness is: Normal
Allow Retreat is: On
Give AI more time: On
FOW is: On low

Are you suggesting I lower those settings ? Wouldn`t that just put the AI to sleep ?? If your beta testing the Game, or have put in many hours, then I`m sincerely interested in your freedback on this issue.

I`d like to see this project work well.

[McNaughton]

Well I think it needs something to stop those tactics, which as I said did not seem to occur in 1.05 vers. which I played with the *same settings* for at least 50+ hours and 4 re-starts of the April Grand Campaign.

Here you go.

AI Difficulty is: Normal
AI Aggresiveness is: Normal
Allow Retreat is: On
Give AI more time: On
FOW is: On low

Are you suggesting I lower those settings ? Wouldn`t that just put the AI to sleep ?? If your beta testing the Game, or have put in many hours, then I`m sincerely interested in your freedback on this issue.

I`d like to see this project work well.

Since you have tried these settings, with the same results, try changing them (maybe then you will see changes).

The FOW application was given to boost the AI's aggressive behaviour in order to have it exploit holes in a player's defense. This was pioneered for an early version of the ACW AI.

Indeed, it does appear that the AI and FOW application is working as designed, as the AI does fully exploit a hole in your defenses, and makes it straight for a target. Unfortunately, the AI works on a turn-by-turn basis, and does not assume that you will move your force to cut them off. Basically, the AI moves in through a hole, moves in with force, then you move to cut them off. The AI is now cut off, but could not predict that it would have happened (it just saw an opening and went for it).

With the increased efficiency of the AI, forming better and larger forces, their 'raids' have become more effective, and larger.

I am 'guessing' that with the improved state of the AI, it is pretty much fighting like before, but, its attacks are more effective, but, still lacks the ability to 'deal' with being suddenly cut off (it cannot predict this possibility).

The accepted FOW setting, which was 'designed' to enhance an older verion of the AI, and may actually be making it too wreckless, exploiting holes that are easy for a human defender to plug. The AI makes larger formations, and when they get cut off and destroyed, it is no longer a lone cavalry regiment, but a division worth of troops.

Lower FOW to zero, the AI will 'probably' move more cautiously, moving as far and fast as their visible intelligence will allow them (instead of rushing for a vacant city behind your lines, they will have to probe more).

Try this, see what happens. If the AI is too passive in their own territory (i.e., takes NO risks), try increasing their aggressiveness one step (they should attack and defend at lower odds, but not run off behind enemy lines to get cut off). If you find that they are too weak, increase the AI bonus a step (so their reinforcements are greater, and fight a bit more efficiently).

I just think that players need to find a new set of personal settings to maximize the new AI. It is better at making divisions, and corps, but I think the same behaviour being seen now, was always there, just done less efficiently (thereby more devestating when the AI gets itself cut off and destroyed).

AI Difficulty is: (change this to the first bonus above normal)
AI Aggresiveness is: (change this to slightly more aggressive than normal)
Allow Retreat is: On
Give AI more time: On
FOW is: (No FOW bonus)

[Bodders]

You know I'm going to try this - I've been playing with 'very high' on detect because I assumed that would help the ai the most but I do see a lot of these suicidal attacks, even with aggression on low. Perhaps you're right and if it doesn't know what you have then it's less inclined to advance.

Not sure I can go to 'reckless' just yet but I'll give it a go on hard/no fow advantage and normal aggression and see my experiences. Perhaps between us we can get a consensus suggested setting

[joram]

Indeed, it does appear that the AI and FOW application is working as designed, as the AI does fully exploit a hole in your defenses, and makes it straight for a target. Unfortunately, the AI works on a turn-by-turn basis, and does not assume that you will move your force to cut them off. Basically, the AI moves in through a hole, moves in with force, then you move to cut them off. The AI is now cut off, but could not predict that it would have happened (it just saw an opening and went for it).

I find that quite shocking. If the AI has no predictive ability, it will get slaughtered on any setting that it isn't outright cheating on (with significant morale/combat bonuses). I don't expect an immediate answer on this as they are all off on vacation but just knowing this simple fact has given me all sorts of ideas to destroy the opposition! Thanks!

[Winfield S. Hancock]

I agree with McNaughton's assessment. The issue of AI recklessness and how it relates to FOW has been discussed previously as well. The bottom line seems to be that for the best game, you dont want the AI to have a FOW advantage as they will make too many reckless attacks.

I am getting ready to start a new April 61 campaign as the Union with my mod, and I am planning on not giving the AI any FOW bonus, giving it a difficulty bonus, and perhaps even setting it at lower than normal aggressiveness. My thoughts are that this will give the CSA the best chance of preserving large forces in being that will compel me as the North to take the war to them, and make it more difficult to secure key objectives. IMHO, this is at it should be, with a CSA AI most interested in force preservation and retaining its existing territory -- which were in essence the goals of the Confederate government during this conflict.

[Black Cat]

I agree with McNaughton's assessment. The issue of AI recklessness and how it relates to FOW has been discussed previously as well. The bottom line seems to be that for the best game, you dont want the AI to have a FOW advantage as they will make too many reckless attacks.

I am getting ready to start a new April 61 campaign as the Union with my mod, and I am planning on not giving the AI any FOW bonus, giving it a difficulty bonus, and perhaps even setting it at lower than normal aggressiveness. My thoughts are that this will give the CSA the best chance of preserving large forces in being that will compel me as the North to take the war to them, and make it more difficult to secure key objectives. IMHO, this is at it should be, with a CSA AI most interested in force preservation and retaining its existing territory -- which were in essence the goals of the Confederate government during this conflict.

Very Good ! The historical Union approach. I think it will be very helpful, please let us know how it plays out.

[Black Cat]

Interesting. Just bought it, so I'd be interested to hear if Bragg's command was just the HQ left, as you mention as a possibility.

Well the good news is that Bragg`s stack did include at least a Small Corps+ .

I tried to trap him north of Bowling Green, but he hit McClernad`s Div. again on the way down, and got away, although losing about 3000+ to 1200 this time.

I`ll try McNaughton`s settings and see how it plays out from here.

[McNaughton]

General Hancock, One problem with low aggression is that the CSA may never end up counterattacking your forces, meaning you are always attacking. I would advise some small testing changes (i.e., change one factor, like FOW) and seeing what happens rather than drastic changes all in one direction (i.e, no FOW, and low aggression, you may find the AI too passive). You may find that without the FOW bonus, the AI isn't making suicidal attacks getting itself surrounded, but, it still makes well timed and executed attacks.

In the end, it may be the AI needs more tweaking, like to secure a line of supply when attacking (instead of putting all forces in the front, try spreading things out).

Also, what players may want to do, is to create some personal house rules regarding the AI. Never surround the AI (i.e, always leave routes for supply), but attack it head on instead. Or, if you are to 'surround' the AI, do what the Union and Confederates did, use your cavalry to ride around your opponent, and disrupt their supply (but not surround it with fighting forces).

In my opinion, sometimes players play to the AI's weaknesses, and then have issues about the results. I know this isn't perfect, but maybe games will be more interesting if house rules are in place until the AI catches up?

[Bodders]

I am getting ready to start a new April 61 campaign as the Union with my mod, and I am planning on not giving the AI any FOW bonus, giving it a difficulty bonus, and perhaps even setting it at lower than normal aggressiveness. My thoughts are that this will give the CSA the best chance of preserving large forces in being that will compel me as the North to take the war to them, and make it more difficult to secure key objectives. IMHO, this is at it should be, with a CSA AI most interested in force preservation and retaining its existing territory -- which were in essence the goals of the Confederate government during this conflict.

That's exactly what I'm doing General Hancock, except I'm going to try with normal aggression now I've changed fog of war down and see how that goes. I'll try 'low' if I'm still finding it too aggressive

The mod makes it a lot of fun - to get a corps in to the valley, I end up having to use Banks over there for liberation of Harper's Ferry/Winchester if possible. It feels so right

[Winfield S. Hancock]

Thanks Bodders. Let me know how it goes on normal aggression, I will run my game with CSA on low aggression, and then lets compare notes.

[Black Cat]

In the end, it may be the AI needs more tweaking, like to secure a line of supply when attacking (instead of putting all forces in the front, try spreading things out).

Also, what players may want to do, is to create some personal house rules regarding the AI.

In my opinion, sometimes players play to the AI's weaknesses, and then have issues about the results. I know this isn't perfect, but maybe games will be more interesting if house rules are in place until the AI catches up?

I agree with all of that and indeed do play that way, with my own " house rules ". However it does seem, under "tweeking" above, that the AI fixates on " Victory Cities " ( Louisville in this case ) and if there is nothing strong **right in front of it** off it goes.

I know it`s tough to program for that, but perhaps the Devs may want to think on it a bit ?

Thanks

[bloodybucket]

An interesting topic. Given the overall strategic situation that South was in, the idea of making risky attacks and having a go for broke attitude isn't completely unwise.

I bet having the AI make tactical choices that consider the most likely course of action of the player is a pretty tall order.

It is also true that the player, with knowledge of what was, the safety and comfort of an armchair and a full belly, and the luxury of knowing that defeat is undone by reloading the last save with no death and destruction to nag his conscience, is unlikely to make the same "errors" that the historical commanders did. Perhaps the answer to a more simulative match with the AI would be to hobble the human player instead of making the AI smarter.

[Sheytan]

im playing 1.6 all settings are normal except for delay which is medium. confed AI banzai charges thru the river gap between HF and alexandria/washington. I was of course delighted the AI decided to give me this victory. at any rate after running around by baltimore/annopolis area a few turns these corp/army stacks then tried to retreat back across the potomac headed south to apparently retreat? at any rate im poised to overrun richmond, mind you this is like turn 6(or 8 lol I forget) .... :} BUWHAHAHAH!

something needs to be done :} imo. as a aside the irony was in this game as opposed to others ive played confed AI didnt blitz kentucky sp? right off. so my stacks of divisions from paduch to cinci sat and waited.

Mc I agree AI needs tweeking...

anyway irrespective of all the goofyness this is still a very outstanding game AND game system(even though I own forge of freedom I still used to whimsically recall how much fun playing the old SSI game was on the civil war, this is mucho grande imo ageodacw that is), heck using this model all sorts of wargames can be made, and the "franchise" of this company IE its games can only get better and improve. Also as folks have noted the support here is outstanding. I compare what happened to Mark of Chaos and what i see here and its simply a tremendous difference. (MOC died of neglect...sure hope Warhammer Online gets off the ground! cant wait for that one!)

General Hancock, One problem with low aggression is that the CSA may never end up counterattacking your forces, meaning you are always attacking. I would advise some small testing changes (i.e., change one factor, like FOW) and seeing what happens rather than drastic changes all in one direction (i.e, no FOW, and low aggression, you may find the AI too passive). You may find that without the FOW bonus, the AI isn't making suicidal attacks getting itself surrounded, but, it still makes well timed and executed attacks.

In the end, it may be the AI needs more tweaking, like to secure a line of supply when attacking (instead of putting all forces in the front, try spreading things out).

Also, what players may want to do, is to create some personal house rules regarding the AI. Never surround the AI (i.e, always leave routes for supply), but attack it head on instead. Or, if you are to 'surround' the AI, do what the Union and Confederates did, use your cavalry to ride around your opponent, and disrupt their supply (but not surround it with fighting forces).

In my opinion, sometimes players play to the AI's weaknesses, and then have issues about the results. I know this isn't perfect, but maybe games will be more interesting if house rules are in place until the AI catches up?

[Bodders]

Thanks Bodders. Let me know how it goes on normal aggression, I will run my game with CSA on low aggression, and then lets compare notes.

Well, I didn't get that far before I got a bit dis-satisfied with the behaviour on normal aggression.

I definitely like what turning the Fog of War bonus off seems to do - the AI has kept the majority of its forces around the Fredericksburg/Richmond area. However, towards the end of 1861 as I was in Manassas, they advanced in to Stafford, VA where I was then able to move and fight the whole of the CSA army of Potomac (J Johnston and EK Smith had corps so the ai had done quite well command wise).

I won but the ai keeps coming back, I won again then moved back to Manassas to rest and saw they were in Stafford again! This time I moved back to Alexandria and the small 'army stack' under Beauregard has reached Washington, DC while the bulk of the army is in Stafford, presumably moving to Alexandria where I can beat it again.

It seems to have gone in to suicidal 'get to Washington' mode for some reason - losing more than 10 elements so far in the attempt to my 0.

It also seems to be obsessed with the Stafford, VA region and 'slipping by' my army in Manassas whichever side it is! This is simply not possible so it should get less obsessed with doing this

That's enough for me to think 'low' is the only way to go although I expect still to see it in Stafford at some point

[Winfield S. Hancock]

I must say that as of right now, I am pretty happy with the results of my game with the CSA on low aggression.

I just made it to January 1862 last night. So far, the CSA AI has done no suicidal raids or attacks into the North. There was one major cavalry raid, with a couple of Tennessee regiments that went into the central Ohio area, tearing up track and occupying the town of Lima for awhile. Eventually, they retreated after some Union counterattacks.

In the West, possession of Ft. Smith has gone back and forth, with Stand Watie recently retaking it for the CSA in a raid. No rebel invasions of Iowa or the like. Sterling Price made threatening moves toward Rolla, but backed off when he saw Lyon had superior force.

Same thing in the central theater. I control Bowling Green, Rebs control Lexington. Earl Van Dorn led a large force North to the Ohio river out of Lexington in early Nov 61, then turned around when the weather got bad and he saw that Cincy was occupied. He is now in winter quarters at Lexington, and both that town and Ft. Donelson have been garrison strongly.

In the east, no raids up into PA or NY as in the past at all. Instead, the AI has built a strong army with 3 corps (although according to my intel, many of the units are way understrength -- dont know if this is FOW, poor spy network, or most of the units really are at 15% TOE -- if so, perhaps Pocus could address in a future patch). Anyway, that Army has alternated between trying to cover Manassas and Winchester. At one point, it massed to threaten Harpers Ferry, where I had built a fort, but then backed away when I reinforced the area with McDowells strong corps. My only problem with AI behavior in the east is the fact that I think the Reb AI values Winchester too highly. It usually has the bulk of its army there, even when I control Manassas and have the bulk of my army there. I would think that once the Union controls Manassas, the logical place for the CSA army to be would be Fredericksburg, to block any March on Richmond. Instead, it seems to be perennially in Winchester threatening Harper's Ferry, and the road to Richmond seemingly is open. Perhaps this could be changed by weighting the value of Harpers Ferry less, or removing it as a strategic city, or adding Fredericksburg as a strategic city, to make the Confederate AI value it more and Harper's Ferry less.

In the South, Bragg and McCulloch seem obsessed with retaking Ft. Pickens, attacking nearly every turn. I have sent reinforcements, and the Rebs are making no headway. I think this is a problem for the Confederate AI as well, as they are bleeding themselves too much in this effort. Perhaps a tweak to the AI which makes them less likely to try to regain coastal fortifications would help.

Overall, I am very satisfied with this game so far, and it with CSA aggression on low with no FOW bonus and the 'hard' setting for CSA AI, it seems to produce the most historical game I have seen so far with AACW.

[Jacek]

My only problem with AI behavior in the east is the fact that I think the Reb AI values Winchester too highly. It usually has the bulk of its army there, even when I control Manassas and have the bulk of my army there. I would think that once the Union controls Manassas, the logical place for the CSA army to be would be Fredericksburg, to block any March on Richmond. Instead, it seems to be perennially in Winchester threatening Harper's Ferry, and the road to Richmond seemingly is open. Perhaps this could be changed by weighting the value of Harpers Ferry less, or removing it as a strategic city, or adding Fredericksburg as a strategic city, to make the Confederate AI value it more and Harper's Ferry less.

I have seen this also in my games. Too much importance of CSA AI on Harper's Ferry and Winchester. Manassas, Fredericksburg or Alexandria could be named Strategic cities instead. The only vital axis in the East is Richmond -Philadelphia. Putting stacks in Harper Ferry or Winchester is USELESS - overstretched supply line which can be cut anytime; units away from the main battle field. If CSA holds Manassas or Alexandria it pretty much blocks Union progress.

[Bodders]

I have seen this also in my games. Too much importance of CSA AI on Harper's Ferry and Winchester. Manassas, Fredericksburg or Alexandria could be names Strategic cities instead. The only vital axis in the East is Richmond -Philadelphia. Putting stacks in Harper Ferry or Winchester is USELESS - overstretched supply line which can be cut anytime; units away from the main battle field. If CSA holds Manassas or Alexandria it pretty much blocks Union progress.

I must admit I've not seen this, perhaps because I take Winchester in 1861 when I take Manassas usually. To get to Winchester, the AI generally needs to go via Manassas so it doesn't seem to make an inordinate effort to do that.

I agree with the pointless attacks against Fort Pickens, though - the AI loves to throw units away there and that needs a tweak. I also think its tendency to always be the one that invades Kentucky could do with tweaking.

Overall though, I look forward to restarting with the mod and on hard/no fow/low agression per Hancock's testing (and mine on medium finding it too aggressive still).

[McNaughton]

@Winfield, does the AI attack on your aggressive setting? I didn't read of any examples of the AI attacking you, but most engagements with the AI pulling back because of enemy resistance sighted, or you attacking.

@Bodders, did you have any of your forces left at Washington or Alexandria? Historically, a corps worth of troops (eventually being XXII Corps by 1863) was stationed around the forts of Washington.

It may be that players aren't defending Washington enough to keep the Confederates away even at 'basic' aggressive levels. Also, historically, before someone could move North into Maryland, or South into Virginia, control over the Shenandoah Valley was acquired.

How about trying the following to see what happens on normal aggression (we want the AI to be a threat, just not a suicidal threat).

#1. Move your capitol, play on normal aggression. See if the AI will still make moves for Washington.

#2. Keep your capitol at Washington, play on normal aggression, but keep a corps (of 2-3 Divisions) garrisoning it. See how the AI moves against your force once you leave Washington with the main army.

#3. Play historic, on normal aggression (play according to rules #2). Have two Confederate and Union focus' in Virginia/Maryland like was historically the case. Have an army based around Washington/Manassas, and another one based in the Shenandoah Valley. Before you move south against Fredricksburg or Manassas or Alexandria or Washington, you must secure then hold the Shenandoah Valley. Leave a force (corps) guarding the valley (historically was Army of Western Virginia, VII Corps, Johnston's Army of the Shenandoah, etc.).

#4. (in combination with any of the above) During winter months chainge AI aggression to low. Since the game works well at changing bonus levels, aggression levels, etc., mid-game and we want the AI to be passive during the winter (for their benefit), this may make their summer campaigns more effective (if they rest and build up in the winter).

[Bodders]

Yes, I had Keyes division, which was originally Runyon's light division (so not that much I admit) in Alexandria at all times and later moved a new division under Burnside to defend either in Alexandria or Manassas depending where my main army was.

Effectively, though, the entire Army of Northeastern Virginia was defending Washington as I could see the move in to Stafford and then rail back to Alexandria immediately to defend. You simply can't move up via Stafford and slip past an army in Manassas, which the ai always seems to try.

I also always have a division or two in a corps in the Shenandoah valley - under Banks originally and Butler once he's no longer fixed.

I'm not going to try on 'normal' because overall on prior games as well I'm happier with the behaviour on 'low'. It does still attack - Kentucky will be attacked come June/July as always for example and the Fort Pickens attacks occur even on low.

[McNaughton]

How about troops directly in Washington? Because of the value placed on Washington itself, the AI will probably disregard if troops could get their quickly, as it sees what is only there at the moment. How about trying to have two good quality (i.e., 200 or so power) divisions in a corps in Washington?

Maybe, if high victory cities are heavily garrisoned, the AI will see potential threats as greater opportunities?

Also, does the AI actually use its army to attack your army, or does it attack small garrison and militia forces? I am just worried that at low aggression the AI is giving up opportunities of field battles.

After everything, I do belive, like others too, that it comes down to victory point cities. The AI is driven to move toward them, right or wrongly (their goal is to win the game, this is how its done!), but I think two things need to be done.

#1. Change VP locations to replicate strategic crossroads and towns that the player sees as valuable (Manassas, for example), but the AI tends to ignore.

#2. Change the AI so it sees greater importance in dealing with potential threats to their VP areas, or to destroying enemy armies vs conquering territory.

[Black Cat]

After having the AI take Washington in late 1861 in an early, and so far my most interesting & exciting Campaign Game, at the default settings , I`ve kept 1 small Division starting in late `61 in D.C.

The AI has yet to again exhibit the excellence it did in that Game, where it did everything right AND for the right reasons, and which left me staring slack jawed at the computer screen.

It now never tries to take DC, or assault Alexandria, where the HQ of the AEV is dug in to 7 with 2 large Divisions.

In 1861 Banks Corps is at Harpers Ferry, also dug in with Hooker`s big Division and some Brigades. They will come up and look around and laugh at Banks not being able to activate, but that`s it. Sometimes it leaves a Brigade at Manassas, then pulls it back. So I`ve learned how to cripple the AI, but produced a boring Game in the process.....

If I do the historical move to Manassas in July`61, I usually get the historic outcome, either when defending or attacking.

By 1862 you have too many units and good Generals for the AI to do much around Washington - Harpers Ferry - Winchester if your careful, however if your careless around Manassas, it can and will give you a nasty lesson.

By summer of 1862 your just too strong in the East for the AI, and really need to implement some house rules on attacking if you want the Game to last, especially if you have re-taken Norfork in `61 ( you did do that, didn`t you all )and have a large force moving up the Va. pennisula.

I think you all are on the right track as far as Option settings, and hopefully the Devs. will re-think some of the AI`s priorities, but IMO the excessive Weather thing needs fixing before you start ajusting the AI movements, ( the AI doesn`t "know" it going to lose 75% of it`s force because it`s snows for 4 months in Ky. ) some of which actually make military sense.

The other thing is the CSA and the Union I would guess opperate under the same AI programing, with all that that implies.

Thanks and sorry for the long post

[Bodders]

Good questions - I could try leaving some in Washington as well if it still seems to be making a bee-line even on low.

I've never been attacked as the Union once in Manassas with my main force on any aggression level. Occasionally, the division I've left there when I've moved to Stafford has been attacked and my army in Stafford did get attacked on 'normal'.

I'd rather give up the ai attacking armies to be honest, it never has enough combat value to win so it's a losing proposition for it, it just needs to pick on garrisons where appropriate. That's one reason I do want to go to 'low' - it's not generally a good idea for it to attack my armies because of its command issues and generally lower CV.

[McNaughton]

Here's some thinking...

#1. Alexandria Virginia should be left empty at the April 1861 scenario. It was invaded by the North in May 1861 (colonel of a zouave regiment killed over a flag).

#2. The 'Union Army' of McDowell's should appear in Washington, not Alexandria.

I think that Alexandria is too fortified for the Union in 1861 (April especially, when it should be under Confederate control), and gives them a too large of a bridghead accross the Potomac, with McDowell's army appearing there (along with having as strong of a garrison as Washington). This might lead toward the possible trouble the Confederates are having in the East, due to the already strong force in Alexandria in 1861 (two brigades, plus artillery and support).

[McNaughton]

After having the AI take Washington in late 1861 in an early, and so far my most interesting & exciting Campaign Game, at the default settings , I`ve kept 1 small Division starting in late `61 in D.C.

The AI has yet to again exhibit the excellence it did in that Game, where it did everything right AND for the right reasons, and which left me staring slack jawed at the computer screen.

It now never tries to take DC, or assault Alexandria, where the HQ of the AEV is dug in to 7 with 2 large Divisions.

In 1861 Banks Corps is at Harpers Ferry, also dug in with Hooker`s big Division and some Brigades. They will come up and look around and laugh at Banks not being able to activate, but that`s it. Sometimes it leaves a Brigade at Manassas, then pulls it back. So I`ve learned how to cripple the AI, but produced a boring Game in the process.....

If I do the historical move to Manassas in July`61, I usually get the historic outcome, either when defending or attacking.

By 1862 you have too many units and good Generals for the AI to do much around Washington - Harpers Ferry - Winchester if your careful, however if your careless around Manassas, it can and will give you a nasty lesson.

By summer of 1862 your just too strong in the East for the AI, and really need to implement some house rules on attacking if you want the Game to last, especially if you have re-taken Norfork in `61 ( you did do that, didn`t you all )and have a large force moving up the Va. pennisula.

I think you all are on the right track as far as Option settings, and hopefully the Devs. will re-think some of the AI`s priorities, but IMO the excessive Weather thing needs fixing before you start ajusting the AI movements, ( the AI doesn`t "know" it going to lose 75% of it`s force because it`s snows for 4 months in Ky. ) some of which actually make military sense.

The other thing is the CSA and the Union I would guess opperate under the same AI programing, with all that that implies.

Thanks and sorry for the long post

How about varying the aggression level of the AI based on seasons?

Set the AI to the most passive of behaviours in winter months, and the seasons building up to winter. This will hopefully keep them preoccupied with building up forces and resting. However, as long as the AI is in 'passive mode', so to must you be (to be fair).

During the good weather months, raise the aggressive level to be above normal. Their forces should be in a better state after resting up for the winter.

[McNaughton]

Good questions - I could try leaving some in Washington as well if it still seems to be making a bee-line even on low.

I've never been attacked as the Union once in Manassas with my main force on any aggression level. Occasionally, the division I've left there when I've moved to Stafford has been attacked and my army in Stafford did get attacked on 'normal'.

I'd rather give up the ai attacking armies to be honest, it never has enough combat value to win so it's a losing proposition for it, it just needs to pick on garrisons where appropriate. That's one reason I do want to go to 'low' - it's not generally a good idea for it to attack my armies because of its command issues and generally lower CV.

The problem with having the AI totally inactive is that it is just a matter of time before you conquer them city after city. They don't counterattack, and you don't have to worry as much about garrisoning your forces.

Personally, I think a different response would result in a more fun experience.

As I said, keep the AI passive during winter months (set aggression to the lowest), and don't attack it (or exploit holes they make).

Increase the aggression of the AI during the summer months (so they can attack, but their forces should be better rested and organized due to winter settings).

If the AI makes a breakthrough, don't do what Grant did in every situation, do what McClellan did. Chase after the army, and confront it head on. Don't go for Richmond, don't go and cut off the enemy army from their supply lines, attack it head on instead.

[Bodders]

I just don't find 'low' to be that passive - it's just about the right amount of aggression for me. Oh yes and I don't attack in winter at all or have a second army attacking down the peninsula - a force in the valley and then the main army attacking down the Alexandria/Manassas/Fredericksburg etc. axis. That will hopefully give me an enjoyable game - it's the settings I've come the closest with in the past, though I gave the AI very high detect bonus in the past.

Others mileage may of course vary

[Winfield S. Hancock]

@Winfield, does the AI attack on your aggressive setting? I didn't read of any examples of the AI attacking you, but most engagements with the AI pulling back because of enemy resistance sighted, or you attacking.

Yes, the CSA has attacked me on low aggression. They went after Bowling Green quite hard after I sent a few brigades by River to capture it. They were laying siege to the city, and the siege only was lifted after I sent a strong (350) division to the relief of Bowling Green, which was just in the nick of time. There have been plenty of attacks against Ft. Pickens (too many) and smaller reb units in the TransMississippi have been active attacking the Indian villages and smaller forces.

I have yet to see the main Reb armies actually attack my main armies in a force on force battle. This could be because of low aggressiveness, or just the fact that in 1861, we are both simply building up our forces at this point and are not quite ready for major operations.

Overall, I will not be disturbed if the Confederate AI does not attack me much. In my opinion, the Confederates need to be on the strategic defensive at all times. Their goal is to simply exist as an independent state, not capture territory in the North to expand their borders. When the CSA did undertake strategic offensives, such as Antietam, Gettysburg, and Perryville, the results were not good for them. One can only wonder what difference the men lost in these campaigns could have made to the defense of Richmond or Atlanta later in the war. I

However, it is appropriate for the CSA to take the tactical offensive at times. I think it can be debated whether this tactical offensive should be modeled by the CSA attacking Union armies by moving from one region to the next (as happened at Shiloh or Chickamauga, for example) or whether it is best modeled within the combat model when combat occurs in a region (such as at Chancellorsville).