AGEOD Forums

Here is a screenshot of the map after turn in question.

https://i.imgur.com/6H1IV6i.jpg

Here are the battle results for said battle.

https://i.imgur.com/PiEAQ00.jpg

As expressed in an earlier post I asked how could a force 1700 power strong enter into a region controlled by my forces 100% with a corps of mine about 1400 power strong and entrenched. I asked if he played any kind of cards on that turn like Cavalry Screen or Deep Recon or something and he said he did not.

For a few turns his corps just sits there not attacking and me not attacking either. He is losing supply each turn so he decides to retreat back to Grant across the river. With me still in defensive posture, he decides to move. He said he put his corps on Blue/Blue with evade. Mind you I have 100% military control the whole time. So with nobody on offensive posture a huge battle occurs pretty much all but annihilating Grants army. Now the amount of men on both sides are almost even. I am at a level 5 entrenchment but have the worst Army commander I have in command. He has the best Army commander he has which is Grant who has an attack value of 6 I believe and my guy Johnston has a defense value of 1. He has good subordinates as well. If you look at the link of the battle statistics I posted you will see he got wrecked.
What should he have done? Is his corps just stuck there(which what is left of it still remains there) I suggested he might have wanted to detach the corps and go green/green with evade and that way if he didn't escape at least the rest of his army wouldn't follow him to destruction. What is the proper way to do this because he insists that it is a game glitch but I think there must be something vital that he did wrong(although I'm not sure what).

Why was his corps allowed to cross without attacking a few turns ago in the first place?

Why cant he cross back over (if I'm not on offensive posture) without attacking me. I did have the entrenchment bonus so I'm certain I was on defense.

I did play the cavalry screen card that turn in the battle region just for hell of it to maybe test it. Not sure if that has anything whatsoever to do with it.

On a secondary note, even if whatever he did made him have to attack, why was it almost a fight till you die battle. He lost upwards of 80% of his men before retreating. My opponent was on blue/blue.

Although its a great victory for me, I really don't want to win this way. I have a pissed off opponent (which I understand) who just doesn't want to play anymore (at least for now) because of something we cant figure out.

By the way the terrain and weather are as follows. Clear/Storm/Harsh Weather. I'm not sure if that too has anything to do with it.

Somebody please help us to figure this out. We are willing to answer any questions.

Okay it's round 2 in the pic. Apparently, the MC finally kicked in when he tried to retreat and set his force to Offense. The Union troops have lost most of their cohesion, which would account for the cowardice icons and equate to not fighting well. Crossing a river is always a horrible idea in combat, so that would explain the Division getting destroyed.

Did he try to attack while on your side of the river without withdrawing? Did he try to bring in reinforcements?

Here is the full Battle log - I'm waiting for my opponent to get home so ill have more information later tonight probably.

https://i.imgur.com/jz3nHFp.jpg

https://i.imgur.com/BLvWAQN.jpg

https://i.imgur.com/qOs5Neo.jpg

https://i.imgur.com/ZWCYtFF.jpg

https://i.imgur.com/o6eoCUf.jpg

https://i.imgur.com/fHJ65aK.jpg

https://i.imgur.com/xGF1tnX.jpg

To get out of such a situation your opponent needed to do green/green with avoid combat selected. And, this is only a hope and a prayer.

Are you playing on Veteran Activation? If so I think the Union Corps commander never became active until the turn of the great battle. Gray Fox speculated that the military control kicked in and prompted the fight. I think this is the case also.

A detailed description of the battle exists as a temp file called !BattleLog. It's a long text file. It only exists until you fight another battle, so you may want to look at it if you didn't have any other combat. Mine end up in an additional folder under My Games/CW2/Logs. So it's not actually in the CW2 game folder. You may have to do a search to find it. Here's an exerpt from one I saved as a permanent file in another folder:

8:47:45 PM (Reporting) Battle in 225 Fauquier, VA Day: 11 Round: 0
8:47:45 PM (Reporting) Commit Chance 1008586 P. Kearny' Corps 100 %, Rolled: 73 Commited
8:47:45 PM (Reporting) GiveGroupsTargets 1008586 P. Kearny' Corps initiating fight against faction 1000002 engaging: 1002494 Manassas Guard
8:47:45 PM (Reporting) 1002494 Manassas Guard new target is 1008586 P. Kearny' Corps
8:47:45 PM (Reporting) Commit Chance 1003511 Army of the Potomac 100 %, Rolled: 52 Commited
8:47:45 PM (Reporting) GiveGroupsTargets 1003511 Army of the Potomac supports 1002494 Manassas Guard against 1008586 P. Kearny' Corps
8:47:45 PM (Reporting) Checking combat stance for United States of America in region 225 Fauquier, VA
8:47:45 PM (Reporting) Preliminary test: No need to retreat from this battle: OppPower: 1673 FacPower: 3443 .
8:47:45 PM (Reporting) Checking combat stance for Confederate States of America in region 225 Fauquier, VA
8:47:45 PM (Reporting) InCS %: 1 New Retreat Will %: 99
8:47:45 PM (Reporting) Ammo %: 100 New Retreat Will %: 99
8:47:45 PM (Reporting) Weighted average of Trench levels: 6
8:47:45 PM (Reporting) Power of opponents compared to us: 205.80 % Base Retreat Will: 40.00 %
8:47:45 PM (Reporting) Dice rolled: 59
8:47:45 PM (Reporting) => We want to stay in battle
8:47:45 PM (Reporting) 1008586 P. Kearny' Corps new target is 1002494 Manassas Guard
8:47:45 PM (Reporting) Commit Chance 1003511 Army of the Potomac 100 %, Rolled: 85 Commited
8:47:45 PM (Reporting) GiveGroupsTargets 1003511 Army of the Potomac supports 1002494 Manassas Guard against 1008586 P. Kearny' Corps
8:47:45 PM (Reporting) 1002494 Manassas Guard new target is 1008586 P. Kearny' Corps
8:47:45 PM (Reporting) <<<-----START----->>>

This gives a "Power of opponents compared to us: 205.80 %" line that the program works out. Numerically, Grant looks about equal to your force in the Battle Report, but this line of the !BattleLog file may show otherwise.

If you DL the CW2 DB from here...

viewtopic.php?f=340&t=35005

...you can check the Terrains file. Your defending force in Clear terrain gets a to-hit advantage for your regular infantry, artillery etc. Additionally, level 5 entrenchment gives your artillery a 1.6 factor in determining to hit, so +60% over Grant's guns. Basically, Grant should never have tried to cross the river and attack your entrenched force. I don't know why a battle did not happen right then, but the result most likely would have been the same.

Read your thread on Military Control. Your answer is there.

--
For everybody:

1. A stack in PP (Passive Posture) has in, most case, no influence on the game. They will not block supply, gain MC, fight (mostly they just take hits and flail with their arms while getting kicked in the ass), or block movement. They can be moved, see thing, retreat, use supplies (they will forage if they have to), and be killed.

2. Any non-stealthy stack (I believe it has only to do with stealth, so cavalry, raiders, etc) entering a region, where it has <5% MC will automatically gain 5%MC and be changed to OP, unless already in OP or AP.

This is why when you move your stack in DP into an enemy region it always goes to OP, which generally causes a battle.

A stack in PP, will of course not doe this, because it is in PP.

Everything you've described is WAD and not the least bit surprising. You've simply not be paying attention to what PP means, and what affects it can have.

Thanks Gray, i will have to start looking into these battle log reports.

Me and my opponent (redoing the turn many times) figured out the best course of action which is for him to go green/green evade to exit the region. Also detach the corps in case a battle does ensue, that way the rest of the army does not come up and also get annihilated. This never explained how he got into the region in the first place as that is still an ongoing discussion for me. Captain_Orso says you can enter an enemy controlled region on green/green without attacking. I know you can easily do this with cavalry but i didnt think you could do it with a whole Army.

This other discussion is labeled Military Control. Please join in.

I'm also certain that a large stack cannot evade contact. What can be done is to break the Corps down into many small independent stacks of a few elements each. Specifically, a brigade size of three to four elements. Set these to Green/Green Evade contact and get out of Dodge.

Using the cavalry screen and the disinformation RGDs can increase the hide value of all your regular troops by 4. These RGD's cannot be countered when played in hostile territory.

I've seen full sized corps escape using that combo of cards. Still, its best to break down the corps into smaller units like Fox says.

Seems like a good time for your opponent to have played the scorched earth RGD to regain some cohesion. I don't see a Union partisan in the area, but scorched earth does combo well with a partisan ambush: +20% cohesion to 45 of your elements and -33% cohesion to 15 enemy elements.

Captain_Orso wrote:Read your thread on Military Control. Your answer is there.

--
For everybody:


2. Any non-stealthy stack (I believe it has only to do with stealth, so cavalry, raiders, etc) entering a region, where it has <5% MC will automatically gain 5%MC and be changed to OP, unless already in OP

I'm a bit suprised that the corps could have gotten into a region with out forcing a fight because the infantry would force the rule above.

I did notice that weather was snowy. I have seen stacks get stuck when the time to retreat is more than 15 days. When the next turn rolls around the green stance goes orange because of the 5% rule. Its possible he entered the reagion, the stack tried to retreat but couldent escape in that turn. The next turn your opponent cancels the retreat, goes defensive. A few turns pass with him defensive, and then when he moved, a stance change was forced. Extra long movement times can make weard things happen.

Stacks in PP never start battles.

Movement is...weird.

Thought experiment

Imagine this real world situation. County A is a level open area, mostly open fields, with no large changes in the height of the landscape nor anything more than a small meandering creek. County B is equally as flat, but mostly swampy and marshy. Both are approximately 30x30mi square.

A corps is bivouacked about in the center of County A. In the morning it packs up and starts marching for County B where approximately in the center of County B it will stop to bivouac.

It takes a several days for the corps to reach the border between the two Counties, but once the lead formations enter the swamps of County B, the roads become smaller, softer, and fewer, slowing movement considerably.

Although it only took several days for the corps to march to the edge of County A, it took 2 times as long for it to march through the swampy landscape to reach its bivouac area in the center of County B.

So the most of the time marching the corps has spent in the swamps of County B.

Imagine these two Counties are two regions in the game.

A Corps Stack in A starts to move into B. The region's tool-tip tells us, it will take 18 days for your stack to make the move, At the end of the first turn, the Corps Stack is still moving, it has lost a small some cohesion for marching continuously for two weeks, and thus the move, which started off as 18 days of march is not listed as 24 days, and although the Corps Stack has been marching over clear terrain for 14 days, it sell hasn't started to enter the target hex.

After 24 days of marching, the corps arrives, and *BOOM* suddenly it is in the swamps and not clear terrain.

When it leave to return to County A, it will take 5 days, but these will all be spent in the swamp, and miraculously on the 5th day, *SNAP* the entire corps is in the center of County A.

Weird

Captain_Orso wrote:Stacks in PP never start battles.

I was going to disagree with this but then I decided to do an experiment before I put my foot in my mouth.

I went with the Gettysburg battle. Indeed, passive stacks do not start battles, and I was able to get Ewell, passive, into Yankee occupied Carroll MD without provoking a fight. (If I kept him on passive, with the Union on passive, I could extract him without a fight as well.)

The weird part, is that when I attacked with him....I got fight to the finish. He was wiped out. I'm attaching the battle log in an error report post, but the key part seems to be this : "Group 1001238 Stuart Div. will retreat in same region Carroll, MD." For all the rounds the CSA rolls retreat and then retreats in the same region. (Its also a bit weird that the battle happens at all, I was using hidden activation and Ewell was in reality inactive...so if you re run the turn half the time there is no battle, but if there is a battle, it is always to the death.)

Not quite sure what to make of it, but for whatever reason the retreat pathfinding couldn't see a way out (and there was a CSA 100%MC region right next door.)

pgr wrote:

Captain_Orso wrote:Stacks in PP never start battles.

I was going to disagree with this but then I decided to do an experiment before I put my foot in my mouth.

I do this often myself

pgr wrote:I went with the Gettysburg battle. Indeed, passive stacks do not start battles, and I was able to get Ewell, passive, into Yankee occupied Carroll MD without provoking a fight. (If I kept him on passive, with the Union on passive, I could extract him without a fight as well.)

This test is too extreme. You will never find this in a real game. Ewell is leading a corps? I'm going to say the Union stack is Reynolds, just to put a name to it.

Put Ewell's stack in PP and Reynolds' stack in DP and repeat. Basically there will be no difference, as long as we are only talking about Carroll MD. A stack in DP is as effective as if it were in PP, if the enemy stack entering its region is also in PP.

- So nobody does anything; everybody has become Dunkers? Well, no, not exactly. The defending stack in the region is still defending, and maintaining the MC in the region. And this has a profound affect on the enemy stack's ability to move out of that region into another enemy controlled region. Namely, it will likely not be able to do it. This is how ZOC (Zone Of Control) works in AGEod games.

pgr wrote:The weird part, is that when I attacked with him....I got fight to the finish. He was wiped out. I'm attaching the battle log in an error report post, but the key part seems to be this : "Group 1001238 Stuart Div. will retreat in same region Carroll, MD." For all the rounds the CSA rolls retreat and then retreats in the same region. (Its also a bit weird that the battle happens at all, I was using hidden activation and Ewell was in reality inactive...so if you re run the turn half the time there is no battle, but if there is a battle, it is always to the death.)

Not quite sure what to make of it, but for whatever reason the retreat pathfinding couldn't see a way out (and there was a CSA 100%MC region right next door.)

You make it sound like it would have been preferable for Stuart's division to leave the region.

I'm not exactly sure at the moment, what determines whether a stack retreats to out of a region or retreats within it. I do recall doing RC (Release Candidate - for patch releases) testing on the new retreat rules, and one of the major issues was retreating across a river, which basically, if the defending stack(s) was strong enough could easily lead to the retreating stack being completely destroyed in one turn; and I am talking about a fat corps stack.

So IIRC Pocus put some code in to not so easily throw a retreating stack to outside of a battle region.

The question which remains in my mind, as long as one side can put a stack into PP and move it across a river and into an enemy occupied region without causing a battle *cough*cough*cough*AI*cough*, what can you do to defend against this tactic?

I haven't experimented with this, but the only thing I can think, of would be to put a small stack into OP in the region, a kind of picket, so that if an enemy stack in PP does come across the river your are defending, it will get attacked by your picket.

Once the game has discovered that there is a stack in OP which is aware of an enemy stack in its region, this is when a battle starts.

Without going into too much detail, as in our example above, Reynolds' stack is in Carroll MD in DP with Buford leading a cav. in OP, when Ewell leads his stack into the region in PP. Buford's stack starts a battle with Ewell's stack, but all stacks in the region can be called up to fight. This is not like MTSG. This is BEFORE the battle starts, so any other stacks in the region picked to fight, fight right from the very first round of battle.

Since Buford's stack is much smaller than it's target stack (Ewell's stack) it check for other friendly stacks in the region, not already targeting or being targeted, and ofc finds Reynolds' stack. So now Reynolds' and Buford's stacks are fight against Ewell, who is still in PP, although Reynolds is in DP.

The biggest issue with this, is that Buford might be at higher risk during the battle of being targeted, but I'm not sure.

Captain_Orso wrote:This test is too extreme. You will never find this in a real game. Ewell is leading a corps? I'm going to say the Union stack is Reynolds, just to put a name to it.

Just open the Gettysburg Campaign, Ewell is indeed a corps commander... so I don't see your point.

Captain_Orso wrote:You make it sound like it would have been preferable for Stuart's division to leave the region.

In my experience, when a stack retreats, it generally retreats out of a region. In this case the power difference is so huge, the CSA force should defiantly choose to get out of there.

Orso, I've put the link at the bottom to the bug post which has the files attached, if you want to look at them. Again, I like to use the Gettysburg campaign to do tests, because armies are built and in close proximity.

The thread was started to ask 2 questions:
1. How can a stack enter a region where it has no MC and not provoke a battle?
2. Why did the resulting battle result in such lopsided causalities.

For #1, my thought was that it wasn't possible. Unless a stack was small and comprised of irregulars, cav, or partisans, it would be forced to go into an attack stance in, the presence of the enemy, to get 5% MC. (This was based on my memory of the ping-pong retreat problem from way back). So I moved Ewell around in passive, and he is indeed able to enter enemy occupied provinces without a battle, if they are on a DP. (Also he was able to move out, the next turn in passive without forcing a battle, so his posture didn't get flipped out of passive at any time...to force a battle I had to put him in offensive)

For #2, my understanding and experience is that a stack will retreat (or try to) if things are lopsided. In other words, if a force is getting thrashed, the battle won't go beyond 1-2 rounds, because the looser retreats. (Or the retreat happens before the battle starts... I'm looking at you JoJo). But both in my little test (Ewell's corps with approx. 1500 power and Stewart's cav division MTSG from outside vs the whole AoP approx. 8000) and the example posted above, the battle went the distance.

The only times before where I observed that behavior, was when retreat wasn't possible (failed amphibious assaults, or stacks in structures) and that was clearly indicated in the detailed battle log. Here, Ewell's force, wants to retreat, succeeds in retreating, chooses to retreat to the same region, and is then re-engaged by a Union stack (which is kinda odd considering they are all on defensive posture.)

An excerpt of the log for your reading pleasure:

10:12:24 AM (Reporting) *** End of action ***
10:12:24 AM (Reporting) Round 0 ended
10:12:24 AM (Reporting) Battle in 218 Carroll, MD Day: 4 Round: 1
10:12:24 AM (Reporting) 1001069 2nd Corps new target is 1001598 6th Corps
10:12:24 AM (Reporting) Commit Chance 1001740 Artillery Reserve 30 %, Rolled: 36 Not commited
10:12:24 AM (Reporting) Commit Chance 1001238 Stuart Div. 65 %, Rolled: 79 Not commited
10:12:24 AM (Reporting) Commit Chance 1001356 1st Corps 70 %, Rolled: 32 Commited
10:12:24 AM (Reporting) GiveGroupsTargets 1001356 1st Corps supports 1001699 12th Corps against 1001238 Stuart Div.
10:12:24 AM (Reporting) 1001598 6th Corps new target is 1001069 2nd Corps
10:12:24 AM (Reporting) Commit Chance 1001758 Cavalry Corps 60 %, Rolled: 85 Not commited
10:12:24 AM (Reporting) Commit Chance 1001496 3rd Corps 65 %, Rolled: 93 Not commited
10:12:24 AM (Reporting) Commit Chance 1001317 Army of the Potomac 65 %, Rolled: 31 Commited
10:12:24 AM (Reporting) GiveGroupsTargets 1001317 Army of the Potomac supports 1001699 12th Corps against 1001069 2nd Corps
10:12:24 AM (Reporting) Checking combat stance for United States of America in region 218 Carroll, MD
10:12:24 AM (Reporting) Preliminary test: No need to retreat from this battle: OppPower: 1591 FacPower: 9638 .
10:12:24 AM (Reporting) Checking combat stance for Confederate States of America in region 218 Carroll, MD
10:12:24 AM (Reporting) Auto retreat triggered, TotalHits remaining: 721 Avg Cohesion%: 86 Base AutoRet%: 20 Hits taken: 515
10:12:24 AM (Reporting) Auto retreat: Hits received altered by the CiC's ROE
will 1001071 Richard S. Ewell 100
10:12:24 AM (Reporting) Stuart Div. There is no enemy on offensive, retreat is automatic.
10:12:24 AM (Reporting) Stuart Div. succeeded in retreating
10:12:24 AM (Reporting) Stuart Div. will take 11 hits while retreating (though no hits can be done on round 0)
10:12:24 AM (Reporting) Group 1001238 Stuart Div. will retreat in same region Carroll, MD
10:12:24 AM (Reporting) 2nd Corps There is no enemy on offensive, retreat is automatic.
10:12:24 AM (Reporting) 2nd Corps succeeded in retreating
10:12:24 AM (Reporting) 2nd Corps will take 29 hits while retreating (though no hits can be done on round 0)
10:12:24 AM (Reporting) Group 1001069 2nd Corps has called off offensive Carroll, MD
(PGR: At this point, you would think the battle would be over... successful retreat/no enemy in offensive)
10:12:24 AM (Reporting) Commit Chance 1001356 1st Corps 55 %, Rolled: 69 Not commited
10:12:24 AM (Reporting) Commit Chance 1001758 Cavalry Corps 55 %, Rolled: 20 Commited
10:12:24 AM (Reporting) GiveGroupsTargets 1001758 Cavalry Corps initiating fight against faction 1000002 engaging: 1001069 2nd Corps (PGR: I don't get this at all, The Cav Corps is defensive...why commit? This keeps the battle going)
10:12:24 AM (Reporting) Commit Chance 1001317 Army of the Potomac 65 %, Rolled: 96 Not commited
10:12:24 AM (Reporting) Commit Chance 1001238 Stuart Div. 65 %, Rolled: 76 Not commited

So the battle ended up destroying the CSA force, because after retreating (and taking pursuit hits) the defensive Union stacks kept re-engaging. My supposition is that if the CSA force had not tried to retreat to the same region every time, it would have ended the battle right there. It should be noted that there was no river effect in this test. Ewell didn't need to cross a river to retreat. He had just entered Carroll MD passively and the next turn provoked a battle. (With Stewart Division MSTG in support from the outside). In any event, the result is similar to the problem encountered at the top of this thread. Something made the battle last longer than it should.

http://www.ageod-forum.com/viewtopic.php?f=342&p=396575#p396575

Edit: I did a bit more testing, and I think the passive posture is a bit of a red herring. I had Ewell attack into Caroll (again no river crossing) and he again was wiped out. The battle log had him retreating, but staying in the region. Again an odd decision, considering he was greatly outnumbered.. I then reset, and had Ewell fight Reynolds in a meeting engagement in Gettysburg, roughly equal force, and the battle lasted 4 rounds with Reynolds deciding to retreat to Fredrick on the last round and the retreat happening.
The crux seems to be in the battle log, from the new battle:

10:01:05 AM (Reporting) 1st Corps succeeded in retreating
10:01:05 AM (Reporting) 1st Corps will take 38 hits while retreating (though no hits can be done on round 0)
... (lots of lines as the engine evaluates retreat regions)
10:01:05 AM (Reporting) Region picked : 218 Carroll, MD
10:01:05 AM (Reporting) 1st Corps is retreating toward Carroll, MD
10:01:05 AM (Reporting) Group 1st Corps has retreated
10:01:05 AM (Reporting) Checking combat stance for Confederate States of America in region 200 Adams, PA
10:01:05 AM (Reporting) No need to retreat from this battle (enemy at 0 power).
10:01:05 AM (Reporting) Commit Chance 1001069 2nd Corps 80 %, Rolled: 74 Commited
10:01:05 AM (Reporting) Commit Chance 1000992 Army of Northern Virginia 0 %, Rolled: 82 Not commited
10:01:05 AM (Reporting) Commit Chance 1001238 Stuart Div. 95 %, Rolled: 99 Not commited
10:01:05 AM (Reporting) Commit Chance 1001156 3rd Corps 85 %, Rolled: 55 Commited
10:01:05 AM (Reporting) Commit Chance 1001004 1st Corps 95 %, Rolled: 27 Commited

When compared above, you'll note there is no line saying "Stuart Div. has retreated." In both examples, the non-retreating side's stacks roll to commit (which must mean that once the battle has started, defensive stacks can roll to commit). It looks like the "has retreated" check is key. With out it, the stack is still considered in the battle and can be committed against. In the second example, you can see that the stacks of the ANV commit...but there is no battle, because there is no-one to commit against.

So there seems to be a bug when a force cooses to retreat to the region it is in. The routine of "has retreated" never fires, the stack remains valid to be comitted against, and another round ensues. (Its even worse when you consider that each round the stack retreats and takes prusuit hits, in addition to combat hits, each round.)

It seems simple enough to fix, just make sure that if the stack chooses to retreat in same region, that it fires a "has retreated" check and is removed from the battle. (Or all retreats exit the region... that seems like the point of retreating.)

I wish I knew what made a stack retreat to the same spot it is in. No matter how I did it, Ewell never wanted to retreat out.

This seems more like the old bugbear Zone of Control.

http://www.ageod.net/agewiki/index.php? ... ction=edit

"Each element has a Patrol Value that represents the ability of the element to block (i.e. interrupt) enemy movement.
:The sum (modified by cohesion, prorated) of all the Patrol Values belonging to friendly elements is added to the Patrol value of any friendly fortifications in a region...
=== Entrenchments ===
Patrol is augmented by your entrenchment, x 1.1 per level of entrenchment...

==Evasion Values==
Each element has an Evasion Value that represents the ability of the element to avoid contact with enemy forces...

==Effectiveness of Zone of Control==
Once the strength of the Zone of Control (i.e. total modified Patrol Value) is determined, it is divided by the Evasion Value of the opposing Force. An opposing Force is prevented from ''retreating'' into any adjacent region where its level of military control is less than this number."

This rule has been subjected to several iterations of tinkering over the years, so I am not sure how much of this still applies verbatim. However, this rule is the only one I know of that prevents retreat.

Gray Fox wrote:This seems more like the old bugbear Zone of Control.

I like the thought, but I don't think that is it, because in all my little tests there were regions with 100% friendly MC next to the battle region. But who knows. How stacks choose where to retreat to when retreating out is pretty clearly described in the games defines. The logic of a stack choosing to retreat to the region it is in is totally opaque (and frankly, new to me.) I always thought if a stack succeeded in retreating, it automatically left. When I modded the retreat logic to reactivate the 5% rule (as in you need at least 5% to be able to retreat some place) my memory is that the logs showed the retreat as failing because there was no place to retreat to.

I checked some saved games. One stack of approximately 5k power had a Zone of Control of 102 and an evasion of 1. Entrenched stacks had a ZOC of over 1000. Either way, even the smaller 102 divided by 1 is 102, which is greater than the MC of 100. So if the rule still works, no retreat possible.

Gray Fox wrote:I checked some saved games. One stack of approximately 5k power had a Zone of Control of 102 and an evasion of 1. Entrenched stacks had a ZOC of over 1000. Either way, even the smaller 102 divided by 1 is 102, which is greater than the MC of 100. So if the rule still works, no retreat possible.

Except, there is a ZOC cap of 85%. If you have at least 85%MC, in the place you want to go to, you can always go. (Something that clearly pops on the map with red and not red). Perhaps the cap isn't being considered in the retreat logic, but I think people would be seeing the same thing a lot more if that was the case.

I've finally found time to look at the files you provided pgr. I'm still not sure exactly what happened. I only see Ewell with a move plotted, but apparently all Lee's corps were fighting in Carroll against all, or nearly all of Meade's corps, so the moves in the ord file are not the ones you are complaining about.

Who all MTSG'ed to the battle is probably irreverent. What losses are taken during the battle have nothing to do with retreat, so that is a mute point.

At the end of the battle, CSA stacks retreat, and take a whole lot of Pursuit damage, because the only CS stack with cavalry is Stewart's corps stack

Is your issue only to where the CS stacks are retreating? or what is the issue here?

Captain_Orso wrote: At the end of the battle, CSA stacks retreat, and take a whole lot of Pursuit damage, because the only CS stack with cavalry is Stewart's corps stack

Is your issue only to where the CS stacks are retreating? or what is the issue here?

My issue is that in the first exemple is that the CSA retreats but it does not end the battle. Round 0-5 all end the same way. So the cumulative effect of 6 rounds of combat and 5 rounds of retreat hits is a dead stack.

The second battle log exemple is a "normal" retreat. Union rolls retreat, picks a reagion to go to, and the system decides that that stack has retreated and it is gone and the battle end.

In the end, movement during the turn, and MTSG are not factors. It seems that somthing is bugging when retreating to the reagion that the stack is in preventing the battle from ending.

Long story short, retreating in reagion leads to a bad day. (It would be nice to know what makes a stack choose that option.)

Your assumption seems to be, that retreating within the region is causing additional battles. IIRC if a stack instigates a battle, in one region, it cannot instigate a second battle with the same enemy stack in the same region.

I would really need to see the turn of the battles to have any chance of understanding what's going on.

pgr wrote:

Captain_Orso wrote: At the end of the battle, CSA stacks retreat, and take a whole lot of Pursuit damage, because the only CS stack with cavalry is Stewart's corps stack

Is your issue only to where the CS stacks are retreating? or what is the issue here?

My issue is that in the first exemple is that the CSA retreats but it does not end the battle. Round 0-5 all end the same way. So the cumulative effect of 6 rounds of combat and 5 rounds of retreat hits is a dead stack.

The second battle log exemple is a "normal" retreat. Union rolls retreat, picks a reagion to go to, and the system decides that that stack has retreated and it is gone and the battle end.

In the end, movement during the turn, and MTSG are not factors. It seems that somthing is bugging when retreating to the reagion that the stack is in preventing the battle from ending.

Long story short, retreating in reagion leads to a bad day. (It would be nice to know what makes a stack choose that option.)

One of the challenges of most of the AGEOD games is avoiding what I call the Ping Pong battles. These are the kind of battles you describe where the losing force does not retreat, but keeps getting beat up.
It is not always possible to avoid such a battle.
But typically, if you know you are in deep do-do, you can go green/green and avoid combat and get away. This retreat does not always work. An alternative is to break your force into completely independent units, no corps or divisions, an have each retreat on its own.
As a radical option, leave one unit in normal defensive stanch and make all others avoid combat.
Takes some learning to retreat well. As someone who is often in trouble, just try some of these get out of jail free cards.

First off, pgr is using some terms incorrectly in describing the situation. A "Battle" is not all the combat taking place within one region during one turn, at least not in game terms.

A "Battle" is what occurs when in a region during turn execution:
• => 2 warring faction have => 1 stack each in that region
• => 1 of these stacks is in OP, and finds an enemy stack, and commits to attacking
• Battles are fought in rounds starting at the furthest range allowable, which depends on terrain and weather conditions
• After each round of battle, if all units of one sides have broken--are in retreat and/or were routed--the battle ends
• However, if both sides still have unretreated units, the range between the forces us reduced by 1 and the next round of battle is executed
• At range 0 only melee combat occurs and only two round of melee can occur per battle
• If the second melee round completes without one side having retreated, the winner of the battle is assessed on merits--I'm guessing best cohesion and power in some calculation unknown to me

This can happen several times within a region, but from my understanding, one stack can only commit against an enemy stack once per turn. So what cannot occur, per my understanding, is:
• US stack A and CS stack B are in a region, with stack A being in OP and stack B in PP
• Stack A commits to battle against stack B, and a battle if fought, at the end of which stack B retreats within the region
• The day* after the battle stack A finds stack B and commits--again--and a battle takes place--again--

* turns are resolved in 1 day steps

So what pgr is describing is confusing to me, because it goes against my understanding of the rules.

What patch level are you guys using?

Captain_Orso wrote:First off, pgr is using some terms incorrectly in describing the situation. A "Battle" is not all the combat taking place within one region during one turn, at least not in game terms.

A "Battle" is what occurs when in a region during turn execution:
• => 2 warring faction have => 1 stack each in that region
• => 1 of these stacks is in OP, and finds an enemy stack, and commits to attacking
• Battles are fought in rounds starting at the furthest range allowable, which depends on terrain and weather conditions
• After each round of battle, if all units of one sides have broken--are in retreat and/or were routed--the battle ends
• However, if both sides still have unretreated units, the range between the forces us reduced by 1 and the next round of battle is executed
• At range 0 only melee combat occurs and only two round of melee can occur per battle
• If the second melee round completes without one side having retreated, the winner of the battle is assessed on merits--I'm guessing best cohesion and power in some calculation unknown to me

This can happen several times within a region, but from my understanding, one stack can only commit against an enemy stack once per turn. So ]what cannot occur, per my understanding, is:
• US stack A and CS stack B are in a region, with stack A being in OP and stack B in PP
• Stack A commits to battle against stack B, and a battle if fought, at the end of which stack B retreats within the region
• The day* after the battle stack A finds stack B and commits--again--and a battle takes place--again--

* turns are resolved in 1 day steps

So what pgr is describing is confusing to me, because it goes against my understanding of the rules.

What patch level are you guys using?

I'm using the most current patch. (1.06.3)

I'm sorry for confusion, as I have gone from from a broad line of thought to a very narrow one over several posts. At this point, I am talking battle in the way Orso describes above. (So not multiple days, but something within a specific battle resolution moment). The bit I bolded above is what seems not to happen when a force rolls to retreat in the region it is in.

A normal end of battle log looks something like this:
Side A1st Corps succeeded in retreating
10:01:05 AM (Reporting) 1st Corps will take 38 hits while retreating (though no hits can be done on round 0)
... (lots of lines for choosing destination)
10:01:05 AM (Reporting) Region picked : 218 Carroll, MD
10:01:05 AM (Reporting)Side A 1st Corps is retreating toward Carroll, MD
10:01:05 AM (Reporting) Side A Group 1st Corps has retreated
10:01:05 AM (Reporting) Checking combat stance for Side B in region 200 Adams, PA
10:01:05 AM (Reporting) No need to retreat from this battle (enemy at 0 power).
10:01:05 AM (Reporting) Commit Chance 1001069 2nd Corps 80 %, Rolled: 74 Commited
10:01:05 AM (Reporting) Commit Chance 1000992 Army of Northern Virginia 0 %, Rolled: 82 Not commited
10:01:05 AM (Reporting) Commit Chance 1001238 Stuart Div. 95 %, Rolled: 99 Not commited
10:01:05 AM (Reporting) Commit Chance 1001156 3rd Corps 85 %, Rolled: 55 Commited
10:01:05 AM (Reporting) Commit Chance 1001004 1st Corps 95 %, Rolled: 27 Commited
battle ends

In my little testing, all the annihilation battles looked something like this:
10:12:24 AM (Reporting) Side A Stuart Div. There is no enemy on offensive, retreat is automatic.
10:12:24 AM (Reporting) Stuart Div. succeeded in retreating
10:12:24 AM (Reporting) Stuart Div. will take 11 hits while retreating (though no hits can be done on round 0)
10:12:24 AM (Reporting) Group 1001238 Stuart Div. will retreat in same region Carroll, MD
NO LINE saying "Has Retreated"
10:12:24 AM (Reporting) 2nd Corps There is no enemy on offensive, retreat is automatic.
10:12:24 AM (Reporting) 2nd Corps succeeded in retreating
10:12:24 AM (Reporting) 2nd Corps will take 29 hits while retreating (though no hits can be done on round 0)
10:12:24 AM (Reporting) Group 1001069 2nd Corps has called off offensive Carroll, MD
NO LINE saying "Has Retreated"
10:12:24 AM (Reporting) Side B Commit Chance 1001356 1st Corps 55 %, Rolled: 69 Not commited
10:12:24 AM (Reporting) Side B Commit Chance 1001758 Cavalry Corps 55 %, Rolled: 20 Commited
10:12:24 AM (Reporting) GiveGroupsTargets 1001758 Cavalry Corps Side B initiating fight against faction 1000002 engaging: 1001069 Side A 2nd Corps
10:12:24 AM (Reporting) Commit Chance 1001317 Army of the Potomac 65 %, Rolled: 96 Not commited
10:12:24 AM (Reporting) Commit Chance 1001238 Stuart Div. 65 %, Rolled: 76 Not commited
next round starts

In example 1, the battle is over. There is no one for side B to shoot at in the next round.
Example 2, side B finds someone to shoot at...who shouldn't be there, because they retreated...and the battle goes into the next round. (And at the end of that round, it happens again...and again). I think you and I Orso can agree, that is abnormal.

I'll PM you some battle logs Orso.

Are these examples all from the same turn execution?

The second battle is started by the Union's 2nd Corps. It also states the CS has no stacks in OP. Who started the first battle? If there are numerous Union stacks in the region, from my understanding each could start a battle, in which case, whether CS stacks are retreating to another region or within the same region it would make no difference, because it will take 4-5 days to leave the region, and if a Union stack starts a battle, all enemy stacks still in the region are eligible to be targeted, at least all stack not previously targeted, which is not to be confused with having been in a battle.

I did an analysis of the Wiki article on targeting in the beta forum IIRC. I never got a confirmation form Pocus if it is 100% correct, but I believe it is very close to it, if not completely. I will post it here later when I have more time.

Captain_Orso wrote:Are these examples all from the same turn execution?

No, the examples in my last post are diffrent.

Example 1, is a "control" example, with a normal retreat ending. (For specifics, CSA is attacking, USA defending. At round 4, all US stacks retreat, and the battle ends. I left out most of the log, and just forcused on the end of the log).

Example 2 is the "Strange example." Again, I focused on the retreat checks. Here a side rolls retreat, yet the battle dosen't end. You get 6 rounds of the same thing happening. (I only quoted then start of one round, but in the full log it is the same story round 0,1,2,3,4,5. Normaly, the battle should be over as soon as one side retreats) (Not really relevant, but it is CSA attacking, USA defending, and the CSA retreating all its stacks)

The only diffrence seems to be the "has retreated" line. In a normal log, at the very end of the retreat check for a stack (after the check to retreat, the retreat hits check, and destination check) stack X is validated as having retreated. (It's taken off the board if you will). The side that sticks around then rolls to see which of its stacks commit to the next round. The result does not really matter, because the other side is no longer there.

In the strange example, the retreating side stacks do everything (and the region they choose to retreat to is the one that they are in), but there is no validation that the stack "has retreated" at the end. The non-retreating retreating side then rolls for commitment. If one of their stacks commits, it initiates the next round of fighting against the retreating side...apparently because they never truly left.

10:12:24 AM (Reporting) Commit Chance 1001758 Cavalry Corps 55 %, Rolled: 20 Commited
10:12:24 AM (Reporting) GiveGroupsTargets 1001758 Cavalry Corps initiating fight against faction 1000002 engaging: 1001069 2nd Corps

Cav Corps USA should not be able to initiate a fight for the next fire round because 2nd Corps CSA had just retreated.

I cannot find my analysis of targeting stacks

I did find one tidbit of information though, which is pretty important. When a stack is forced to retreat, if there are no enemy stacks in the region in OP, the retreating stack will do a "tactical retreat" to within the region.

Captain_Orso wrote:I cannot find my analysis of targeting stacks

I did find one tidbit of information though, which is pretty important. When a stack is forced to retreat, if there are no enemy stacks in the region in OP, the retreating stack will do a "tactical retreat" to within the region.

That's intresting...has that always been the case?

No, it was part of the new retreat rules, in which auto-switch to OP and auto-gain of 5 MC was implemented.

Pocus's statement was something along the lines of, if one side has no stacks in OP and the other retreats, there's no reason for the retreating side to leave the region, because it's not being attacked.

It should basically help if you are trying to attack a region, and had an unfortunate loss, since you don't get kicked out, and your retreating stack can basically rest up the rest of the turn, although the lack of MC in the region will lessen cohesion regain, but better than if your stack were moving the entire time and using-up cohesion.