One more on MTSG.
Pocus will be changing returns to the region the MTSG was in when it MTSG'd.
The existing code does allow both situations - returning to own region or remaining with the stack being reinforced.
The game between Felixz and Pgr has had a lot of successful MTSG. There was one MTSG where a Union Corps attacked from Hickman TN into Maury TN. The Union Corps was supported by MTSG's from Murfreesboro and Hickman TN. The defending Confed was supported by a CSA Corps MTSG from Coffee TN. The Union won the battle pushing the defending CSA Corps into Pulaski TN along with the CSA Corps from Coffee. This particular retreat was adventeous to the CSA because there was a Union partisan trying to damage or sabotage rails in Stevenson - both efforts unsuccessful. But it could have hindered CSA movement on the next turn.
Pocus's CW2 change to match TEAW will eliminate this possible result.
I believe MTSG reinforcements have remained in the battle region if there was insufficient time (days) remaining to make the return move.
I've looked into save game CptOrso posted, rewinded a turn and looked at the hst file via notepad editor. This is what I've found:
The line containing text 233 is Beauregard current region ID (Fredericksburg). But the line containing line 232 is the last region he or some unit in his stack occupied (Culpepper). That is why he retreated there.
Obviously, this is not the desired behavior and he should not retreat there because several turns ago, he was there.
On a minor note. When I mentioned MTSG Corps returning to their own province, I was thinking about some players complaining while ago why his heavily entrenched Army in Mannassas was pushed to Fredericksburg after helping to defend small Corps in Fallmouth. I am only afraid that the defense in the game will become too strong if the MTSG Corps would use their current entrench like in TEAW. If the Corps is MTSGing, his entrench value should be 0 for this era like it was always in AACW/CW2 games.
Group}
1009428
-1
50001
233
-1
1000002
14930
4770
50001
233
-1
-1
-1
331
14930
4770
1
232
0
5
0
-1
-1
1
0
1
0
0
-1
0
100
100
0
0
0
P. Beauregard' Corps
The line containing text 233 is Beauregard current region ID (Fredericksburg). But the line containing line 232 is the last region he or some unit in his stack occupied (Culpepper). That is why he retreated there.
Obviously, this is not the desired behavior and he should not retreat there because several turns ago, he was there.
On a minor note. When I mentioned MTSG Corps returning to their own province, I was thinking about some players complaining while ago why his heavily entrenched Army in Mannassas was pushed to Fredericksburg after helping to defend small Corps in Fallmouth. I am only afraid that the defense in the game will become too strong if the MTSG Corps would use their current entrench like in TEAW. If the Corps is MTSGing, his entrench value should be 0 for this era like it was always in AACW/CW2 games.
-
pgr
- General of the Army
- Posts: 671
- Joined: Fri Jan 24, 2014 9:33 pm
- Location: Paris France (by way of Wyoming)
Thanks for the breakdown ace. As I mentioned, Beaureguard wasn't in Culpeper (except for something like 10 or 15 years previously, and he had moved all over the place since then.) the other possibility is that when the stack MTSG for the defense of Cupeper about 6 turns before, the region was flagged as a previous region. (the stack didn't move after that MTSG.)
-
loki100
- AGEod Guard of Honor
- Posts: 2401
- Joined: Sun Aug 14, 2011 4:15 pm
- Location: Caithness
- Contact: Website Twitter
Ace wrote:On a minor note. When I mentioned MTSG Corps returning to their own province, I was thinking about some players complaining while ago why his heavily entrenched Army in Mannassas was pushed to Fredericksburg after helping to defend small Corps in Fallmouth. I am only afraid that the defense in the game will become too strong if the MTSG Corps would use their current entrench like in TEAW. If the Corps is MTSGing, his entrench value should be 0 for this era like it was always in AACW/CW2 games.
I don't think this is a minor point. I can see the logic for the new concept in EAW but not in other games.
My logic is that in the sequence of games that uses the 'end up in the battle province' logic from RoP to ACW (& PoN up to say 1900), what that is reflecting is that to support a battle you have to go to the relatively small space of ground that it takes place in. With an effective killing range of 50-200m that is inevitable.
EAW is into the period when a battle stretched over a huge area, so units could support the core offensive from where they were - aided by the importance of artillery to both kill and deny the opposition unchallenged movement. Also, EAW has to simulate precisely the effect noted - that defenses built up incrementally over time and became independent of the unit that had created them.
AJE The Hero, The Traitor and The Barbarian
PoN Manufacturing Italy; A clear bright sun
RoP The Mightiest Empires Fall
WIA Burning down the Houses; Wars in America; The Tea Wars
PoN Manufacturing Italy; A clear bright sun
RoP The Mightiest Empires Fall
WIA Burning down the Houses; Wars in America; The Tea Wars
-
pgr
- General of the Army
- Posts: 671
- Joined: Fri Jan 24, 2014 9:33 pm
- Location: Paris France (by way of Wyoming)
Pocus wrote:By the way I can't check anything without the order files... I tried playing the turn but got nothing worthy, Beauregard is blocked in Spotsylvania, can't retreat and get butchered. Please provide save. Even better, a small scale engagement with stupid retreat...
Quick question about him being blocked when you played with it, was it because he intended to go to Culpeper and ZOC prevented him?
The main question is why didn't he choose to go to Caroline, Lousia, or Albemarle for his first retreat destination?
From what Ace looked at, it's because Culpeper was considered the spot where Beaureguard had come from. (despite the fact that the stack was defending and hadn't moved for several turns)
-
Captain_Orso
- Posts: 5766
- Joined: Tue Sep 01, 2009 5:02 pm
- Location: Stuttgart, Germany
There are a number of things to talk about since yesterday.
First, Ace, that's a great find. It will certainly be important to know this when finding a solution.
pgr: MC--Military Control--is a measure of a faction's confidence to conduct communications--the transportation of troops, material and information--through a region. Without MC in the game the player might attempt to communicate through region, in which in reality a real leader would not have such confidence.
During retreats, if a commander is in the situation of either maintaining his position, where his force is in imminent danger of destruction--retreat results through combat--or retreat through a region in which he would otherwise have no confidence to safely communicate, he will certainly weight the dangers of such a move versus maintaining position. In other words, if the danger of not retreating is greater than retreating through a region where he has 0 MC, I think he must be allowed to choose the lesser of two evils. Anything else makes no sense to me.
--
With regard to the "ctlRetreatPrevSubSpaceCoeffH" parameter. It can only make sense within the framework of a stack moving during the current turn which encounters enemy an stack(s). This parameter is based on the idea that the moving stack is entering the battle region from one side and if it encounters the enemy, it will likely be situated between the enemy and the region from which it entered the battle region. Thus during retreat it would likely retreat "back the way it came".
On subsequent turns however this makes no sense anymore as the force would situate itself in accordance with its needs. All other factors--MC of surrounding regions, enemy and friendly presence in surrounding regions, nearby facilities (cities, forts, depots)--will however still be in affect. So it is generally unlikely that if forced to retreat, will not retreat in an illogical direction.
--
With regards to am MTSG stack retreating, loki100, your argument does make realistic sense. However, there are some issue with this in CW2.
Firstly, the way the game is played, is in general, is very unrealistic. It is very dangerous to divide your forces in the face of the enemy. Spreading corps out in a battle line over several regions, would in reality be insane.
In reality, you had your corps gathered into one region in front of the enemy, if the enemy tried to move around your flank, if you realized that he was doing this, you would either move your force over to in front of the enemy force again to thwart his move, or attack the enemy while in motion, when it is very vulnerable to attack or can be forced to halt its movement. This cannot be done in the game. There is no such SO--Special Order--or method to do this. The only thing you can do is to position one corps each to the left and right of the enemy's axis of approach to make a move around your flank to a target region in your rear so long that it will not complete within one turn, thus allowing the defending player to attack the flanking force before it has reached his goal.
If the defending corps which MTSG'ed to partake in a battle all retreated with the force which started in the battle region, there would be a wide swath of empty regions suddenly appearing in the front line, with all the retreated corps and army stacks having been retreated into one region somewhere behind the front.
Thus, retreating an MTSG force back into its original region might not be completely realistic, but neither is the way a defender must spread out his forces to defend his hinterland. Putting both together however does come close overall to the historical situation.
First, Ace, that's a great find. It will certainly be important to know this when finding a solution.
pgr: MC--Military Control--is a measure of a faction's confidence to conduct communications--the transportation of troops, material and information--through a region. Without MC in the game the player might attempt to communicate through region, in which in reality a real leader would not have such confidence.
During retreats, if a commander is in the situation of either maintaining his position, where his force is in imminent danger of destruction--retreat results through combat--or retreat through a region in which he would otherwise have no confidence to safely communicate, he will certainly weight the dangers of such a move versus maintaining position. In other words, if the danger of not retreating is greater than retreating through a region where he has 0 MC, I think he must be allowed to choose the lesser of two evils. Anything else makes no sense to me.
--
With regard to the "ctlRetreatPrevSubSpaceCoeffH" parameter. It can only make sense within the framework of a stack moving during the current turn which encounters enemy an stack(s). This parameter is based on the idea that the moving stack is entering the battle region from one side and if it encounters the enemy, it will likely be situated between the enemy and the region from which it entered the battle region. Thus during retreat it would likely retreat "back the way it came".
On subsequent turns however this makes no sense anymore as the force would situate itself in accordance with its needs. All other factors--MC of surrounding regions, enemy and friendly presence in surrounding regions, nearby facilities (cities, forts, depots)--will however still be in affect. So it is generally unlikely that if forced to retreat, will not retreat in an illogical direction.
--
With regards to am MTSG stack retreating, loki100, your argument does make realistic sense. However, there are some issue with this in CW2.
Firstly, the way the game is played, is in general, is very unrealistic. It is very dangerous to divide your forces in the face of the enemy. Spreading corps out in a battle line over several regions, would in reality be insane.
In reality, you had your corps gathered into one region in front of the enemy, if the enemy tried to move around your flank, if you realized that he was doing this, you would either move your force over to in front of the enemy force again to thwart his move, or attack the enemy while in motion, when it is very vulnerable to attack or can be forced to halt its movement. This cannot be done in the game. There is no such SO--Special Order--or method to do this. The only thing you can do is to position one corps each to the left and right of the enemy's axis of approach to make a move around your flank to a target region in your rear so long that it will not complete within one turn, thus allowing the defending player to attack the flanking force before it has reached his goal.
If the defending corps which MTSG'ed to partake in a battle all retreated with the force which started in the battle region, there would be a wide swath of empty regions suddenly appearing in the front line, with all the retreated corps and army stacks having been retreated into one region somewhere behind the front.
Thus, retreating an MTSG force back into its original region might not be completely realistic, but neither is the way a defender must spread out his forces to defend his hinterland. Putting both together however does come close overall to the historical situation.
Hi,
Revision done in the code, there was a bug about the previous region bonus that removed entirely all penalties from enemy presence.
With the changes, Beauregard, when he manages to roll for retreat, will go south:
You can download an 'avant-premiere' of the new code here (this is only CW2.EXE):
http://we.tl/YLYspp9W4A
Also, additions to the code:
a) penalty for each day of travel added
b) bonus for friendly presence (before only penalty for enemy presence)
c) all changes in the fitness value for each region evaluated are now logged in the battle log, like this:
12:01:01 (Reporting) => We will try to retreat
12:01:01 (Reporting) Retreat Chance after ROE: 39 %
12:01:01 (Reporting) The retreat chance for 1009428 P. Beauregard' Corps is: 39 %
12:01:01 (Reporting) Dice rolled: 36
12:01:01 (Reporting) P. Beauregard' Corps succeeded in retreating
12:01:01 (Reporting) P. Beauregard' Corps will take 14 hits while retreating (though no hits can be done on round 0)
12:01:01 (Reporting) Retreat Path Inspecting 2247 Rappahannock River
12:01:01 (Reporting) Can't move into, discarded
12:01:01 (Reporting) Retreat Path Inspecting 234 Caroline, VA
12:01:01 (Reporting) Initial value from MC: 100
12:01:01 (Reporting) Value after move cost: 70
12:01:01 (Reporting) Value after structures ownership: 70
12:01:01 (Reporting) Value after enemy presence: 70
12:01:01 (Reporting) Value after friendly presence: 70
12:01:01 (Reporting) Value after bottleneckness / Passage hub : 105
12:01:01 (Reporting) Not coming from this region, no change
12:01:01 (Reporting) ---> New best region: 234 Caroline, VA Value: 105
12:01:01 (Reporting) Retreat Path Inspecting 243 Hanover, VA
12:01:01 (Reporting) Initial value from MC: 100
12:01:01 (Reporting) Value after move cost: 70
12:01:01 (Reporting) Value after structures ownership: 70
12:01:01 (Reporting) Value after enemy presence: 70
12:01:01 (Reporting) Value after friendly presence: 70
12:01:01 (Reporting) Value after bottleneckness / Passage hub : 119
12:01:01 (Reporting) Not coming from this region, no change
12:01:01 (Reporting) ---> New best region: 243 Hanover, VA Value: 119
12:01:01 (Reporting) Retreat Path Inspecting 242 Louisa, VA
12:01:01 (Reporting) Initial value from MC: 100
12:01:01 (Reporting) Value after move cost: 70
12:01:01 (Reporting) Value after structures ownership: 70
12:01:01 (Reporting) Value after enemy presence: 70
12:01:01 (Reporting) Value after friendly presence: 102
12:01:01 (Reporting) Value after bottleneckness / Passage hub : 193
12:01:01 (Reporting) Not coming from this region, no change
12:01:01 (Reporting) ---> New best region: 242 Louisa, VA Value: 193
12:01:01 (Reporting) Retreat Path Inspecting 241 Albemarle, VA
12:01:01 (Reporting) Initial value from MC: 100
12:01:01 (Reporting) Value after move cost: 70
12:01:01 (Reporting) Value after structures ownership: 70
12:01:01 (Reporting) Value after enemy presence: 70
12:01:01 (Reporting) Value after friendly presence: 70
12:01:01 (Reporting) Value after bottleneckness / Passage hub : 135
12:01:01 (Reporting) Not coming from this region, no change
12:01:01 (Reporting) Retreat Path Inspecting 232 Culpeper, VA
12:01:01 (Reporting) Initial value from MC: 0
12:01:01 (Reporting) Value after move cost: -30
12:01:01 (Reporting) Value after structures ownership: -30
12:01:01 (Reporting) Value after enemy presence: -498
12:01:01 (Reporting) Value after friendly presence: -498
12:01:01 (Reporting) Value after bottleneckness / Passage hub : -498
12:01:01 (Reporting) Not coming from this region, no change
12:01:01 (Reporting) Retreat Path Inspecting 226 Stafford, VA
12:01:01 (Reporting) Initial value from MC: 0
12:01:01 (Reporting) Value after move cost: -30
12:01:01 (Reporting) Value after structures ownership: -30
12:01:01 (Reporting) Value after enemy presence: -262
12:01:01 (Reporting) Value after friendly presence: -262
12:01:01 (Reporting) Value after bottleneckness / Passage hub : -262
12:01:01 (Reporting) Not coming from this region, no change
12:01:01 (Reporting) Region picked : 242 Louisa, VA
12:01:01 (Reporting) P. Beauregard' Corps is retreating toward Louisa, VA
12:01:01 (Reporting) Group P. Beauregard' Corps has retreated
Revision done in the code, there was a bug about the previous region bonus that removed entirely all penalties from enemy presence.
With the changes, Beauregard, when he manages to roll for retreat, will go south:
You can download an 'avant-premiere' of the new code here (this is only CW2.EXE):
http://we.tl/YLYspp9W4A
Also, additions to the code:
a) penalty for each day of travel added
b) bonus for friendly presence (before only penalty for enemy presence)
c) all changes in the fitness value for each region evaluated are now logged in the battle log, like this:
12:01:01 (Reporting) => We will try to retreat
12:01:01 (Reporting) Retreat Chance after ROE: 39 %
12:01:01 (Reporting) The retreat chance for 1009428 P. Beauregard' Corps is: 39 %
12:01:01 (Reporting) Dice rolled: 36
12:01:01 (Reporting) P. Beauregard' Corps succeeded in retreating
12:01:01 (Reporting) P. Beauregard' Corps will take 14 hits while retreating (though no hits can be done on round 0)
12:01:01 (Reporting) Retreat Path Inspecting 2247 Rappahannock River
12:01:01 (Reporting) Can't move into, discarded
12:01:01 (Reporting) Retreat Path Inspecting 234 Caroline, VA
12:01:01 (Reporting) Initial value from MC: 100
12:01:01 (Reporting) Value after move cost: 70
12:01:01 (Reporting) Value after structures ownership: 70
12:01:01 (Reporting) Value after enemy presence: 70
12:01:01 (Reporting) Value after friendly presence: 70
12:01:01 (Reporting) Value after bottleneckness / Passage hub : 105
12:01:01 (Reporting) Not coming from this region, no change
12:01:01 (Reporting) ---> New best region: 234 Caroline, VA Value: 105
12:01:01 (Reporting) Retreat Path Inspecting 243 Hanover, VA
12:01:01 (Reporting) Initial value from MC: 100
12:01:01 (Reporting) Value after move cost: 70
12:01:01 (Reporting) Value after structures ownership: 70
12:01:01 (Reporting) Value after enemy presence: 70
12:01:01 (Reporting) Value after friendly presence: 70
12:01:01 (Reporting) Value after bottleneckness / Passage hub : 119
12:01:01 (Reporting) Not coming from this region, no change
12:01:01 (Reporting) ---> New best region: 243 Hanover, VA Value: 119
12:01:01 (Reporting) Retreat Path Inspecting 242 Louisa, VA
12:01:01 (Reporting) Initial value from MC: 100
12:01:01 (Reporting) Value after move cost: 70
12:01:01 (Reporting) Value after structures ownership: 70
12:01:01 (Reporting) Value after enemy presence: 70
12:01:01 (Reporting) Value after friendly presence: 102
12:01:01 (Reporting) Value after bottleneckness / Passage hub : 193
12:01:01 (Reporting) Not coming from this region, no change
12:01:01 (Reporting) ---> New best region: 242 Louisa, VA Value: 193
12:01:01 (Reporting) Retreat Path Inspecting 241 Albemarle, VA
12:01:01 (Reporting) Initial value from MC: 100
12:01:01 (Reporting) Value after move cost: 70
12:01:01 (Reporting) Value after structures ownership: 70
12:01:01 (Reporting) Value after enemy presence: 70
12:01:01 (Reporting) Value after friendly presence: 70
12:01:01 (Reporting) Value after bottleneckness / Passage hub : 135
12:01:01 (Reporting) Not coming from this region, no change
12:01:01 (Reporting) Retreat Path Inspecting 232 Culpeper, VA
12:01:01 (Reporting) Initial value from MC: 0
12:01:01 (Reporting) Value after move cost: -30
12:01:01 (Reporting) Value after structures ownership: -30
12:01:01 (Reporting) Value after enemy presence: -498
12:01:01 (Reporting) Value after friendly presence: -498
12:01:01 (Reporting) Value after bottleneckness / Passage hub : -498
12:01:01 (Reporting) Not coming from this region, no change
12:01:01 (Reporting) Retreat Path Inspecting 226 Stafford, VA
12:01:01 (Reporting) Initial value from MC: 0
12:01:01 (Reporting) Value after move cost: -30
12:01:01 (Reporting) Value after structures ownership: -30
12:01:01 (Reporting) Value after enemy presence: -262
12:01:01 (Reporting) Value after friendly presence: -262
12:01:01 (Reporting) Value after bottleneckness / Passage hub : -262
12:01:01 (Reporting) Not coming from this region, no change
12:01:01 (Reporting) Region picked : 242 Louisa, VA
12:01:01 (Reporting) P. Beauregard' Corps is retreating toward Louisa, VA
12:01:01 (Reporting) Group P. Beauregard' Corps has retreated
Hofstadter's Law: "It always takes longer than you expect, even when you take into account Hofstadter's law."
Additions:
the message about Culpeper is misleading:
12:01:01 (Reporting) Retreat Path Inspecting 232 Culpeper, VA
12:01:01 (Reporting) Initial value from MC: 0
12:01:01 (Reporting) Value after move cost: -30
12:01:01 (Reporting) Value after structures ownership: -30
12:01:01 (Reporting) Value after enemy presence: -498
12:01:01 (Reporting) Value after friendly presence: -498
12:01:01 (Reporting) Value after bottleneckness / Passage hub : -498
12:01:01 (Reporting) Not coming from this region, no change
The truth is that Beauregard was coming from it, but the current value being negative, no bonus was granted on that. I'll differentiate the message here.
I'll also see how to remove the 'coming from' after one turn, do we agree on that?
the message about Culpeper is misleading:
12:01:01 (Reporting) Retreat Path Inspecting 232 Culpeper, VA
12:01:01 (Reporting) Initial value from MC: 0
12:01:01 (Reporting) Value after move cost: -30
12:01:01 (Reporting) Value after structures ownership: -30
12:01:01 (Reporting) Value after enemy presence: -498
12:01:01 (Reporting) Value after friendly presence: -498
12:01:01 (Reporting) Value after bottleneckness / Passage hub : -498
12:01:01 (Reporting) Not coming from this region, no change
The truth is that Beauregard was coming from it, but the current value being negative, no bonus was granted on that. I'll differentiate the message here.
I'll also see how to remove the 'coming from' after one turn, do we agree on that?
Hofstadter's Law: "It always takes longer than you expect, even when you take into account Hofstadter's law."
-
Captain_Orso
- Posts: 5766
- Joined: Tue Sep 01, 2009 5:02 pm
- Location: Stuttgart, Germany
Okay, I will summarize what I understand.
- The "ctlRetreatPrevSubSpaceCoeffH" parameter will only affect retreats in the turn the a retreating stack has actually moved into the battle region, but will affect all retreats during that turn.
- The retreat-region will be picked from a list of all regions into which the retreating stack could retreat, for being the region with the highest points/least negative points.
I see that there is also a "ctlRetreatLandLink" parameter set at "10", but I don't see it addressed in the log.
If what I've summarized is correct, and works, then I think we're good to go
!!
- The "ctlRetreatPrevSubSpaceCoeffH" parameter will only affect retreats in the turn the a retreating stack has actually moved into the battle region, but will affect all retreats during that turn.
- The retreat-region will be picked from a list of all regions into which the retreating stack could retreat, for being the region with the highest points/least negative points.
- Positive points are accumulated for a region for:
- Friendly MC -- 1 per MC% point.
- Friendly Owned Stuctures -- 5 points per town/city level (ctlRetreatAdjCity), 25 per depot level(ctlRetreatAdjDepot), 30 per fort level (ctlRetreatAdjFort).
- Friendly SU's -- 4 per friendly combat (leaders included) and supply SU present (converse of cltRetPenaltyPerNmySU).
- Per "Value after bottleneckness / Passage hub". I do not understand this. It did not accumulate for enemy held regions, but it did for friendly held regions.
- Negative points are accumulated for a region for:
- [ Points are not subtracted for Enemy MC ].
- [ Points are not subtracted for Enemy Owned Structures ].
- Enemy SU's -- 4 per enemy combat (leaders included) and supply SU present (cltRetPenaltyPerNmySU).
- Cost to enter the region -- I could not figure out how this was assessed, as all regions in the example cause -30 points.
- Modifiers:
- IF the retreating stack has moved during this turn, 2.5 x the other accumulated points (250 Coefficient) for the regions, from which the stack entered the current region (ctlRetreatPrevSubSpaceCoeffH).
I see that there is also a "ctlRetreatLandLink" parameter set at "10", but I don't see it addressed in the log.
If what I've summarized is correct, and works, then I think we're good to go
!!-
pgr
- General of the Army
- Posts: 671
- Joined: Fri Jan 24, 2014 9:33 pm
- Location: Paris France (by way of Wyoming)
Pocus wrote:Additions:
I'll also see how to remove the 'coming from' after one turn, do we agree on that?
Oh yay! that was quick!
I'll let you confirm if Cpt_orso's summery is correct. But is sure sounds good.
And I am all for limiting the "coming from" to one previous turn. (Rather than the system somehow tracking all movement)
@ Cpt_Orso, thanks for your effort on the thread and sorry for the insistence on the MC bit. Looks like I can shut up about minimum MC...
-
Captain_Orso
- Posts: 5766
- Joined: Tue Sep 01, 2009 5:02 pm
- Location: Stuttgart, Germany
-
pgr
- General of the Army
- Posts: 671
- Joined: Fri Jan 24, 2014 9:33 pm
- Location: Paris France (by way of Wyoming)
Captain_Orso wrote:I tried the battle with the new engine.
E. Johnson retreats back to Albemarle and Beauregard to Louisaompom:
Just on a side note,I do wonder why Lee didn't MTSG along with Johnson
Anybody have an idea if it might be because Lee as no units with him?
Ya I think it was because he was all alone. Even if he had had a supply wagon, he probably would have participated.
That's almost that, retreat will only use once the previous region the stack came from, and this memory will be erased after one turn or if the stack retreats once.
As for the parameter ctlRetreatLandLink it is used in the "bottleneckness", if a region is friendly, it gets a bonus if there is a lot of escape possibilities from it (and added bonus if on working railways). Being cornered in a peninsula is not good generally!
As for the parameter ctlRetreatLandLink it is used in the "bottleneckness", if a region is friendly, it gets a bonus if there is a lot of escape possibilities from it (and added bonus if on working railways). Being cornered in a peninsula is not good generally!
Hofstadter's Law: "It always takes longer than you expect, even when you take into account Hofstadter's law."
-
Captain_Orso
- Posts: 5766
- Joined: Tue Sep 01, 2009 5:02 pm
- Location: Stuttgart, Germany
pgr wrote:Of the top of my head in a current tournament game I am hosting, Jim-NC and Ace me thinks, CSA troops in Winchester retreated during a battle and chose 100% Union controlled Harper's Ferry (with a garrison if memory serves) over CSA controlled Strasbourg (with its friendly depot).
I also seem to recall AI games where a Union stack next to Richmond was attacked by a corps coming from Richmond and chose to retreat to my waiting army in Richmond rather than move away in the direction of FT Monroe. Of course the memory could be a bit fuzzy
Something similar just happened and caused a current PBEM to implode. And I really can't fault my CSA opponent for being upset. Weird retreats have been the bane of the game system for ages. It happens again and again that forces retreat away from paths of safety into the jaws of destruction. Or away from open lines of supply and waiting stocked depots into barren wilderness.
I suggest an elegant solution. Add a "Preferred Retreat" order option to the unit commands menu. If enabled, it would allow the owning player to click on an adjacent eligible area (which would blink to acknowledge the command, like amphibious landing selections do) which met certain baseline parameters, e.g. could not choose areas with sighted enemy forces with equal/greater combat power and/or 100 MC. Then, during any ensuing combat, if the force had to retreat, it would go to the preferred area. It may still have to encounter enemy forces that moved in during the turn, but that would be the breaks. I am not sure the system allows it, but ideally the order should allow you to "chain" Preferred Retreats to two regions. If you wanted to make this second region ability only available to forces with leaders who have "4" or above command ability, that might be a nice extra touch.
If you wanted to allow for deep raiding forces, then tweak the Preferred Retreat setting command so that mobile forces (e.g. cavalry, horse artillery) led by a leader with a deep raider symbol could also choose Preferred Retreats into areas of enemy 100 MC, subject during the turn itself to making a command roll (which would retain some element of chance).
Adding this function adds some complexity and another layer of micro-management. But I think it would be welcomed by most players, who'd at least use it for key forces.
-
charlesonmission
- Posts: 781
- Joined: Wed Mar 09, 2011 5:55 am
- Location: USA (somewhere)
-
Jagger2013
- General of the Army
- Posts: 641
- Joined: Sat Apr 27, 2013 2:14 am
elxaime wrote:Something similar just happened and caused a current PBEM to implode. And I really can't fault my CSA opponent for being upset. Weird retreats have been the bane of the game system for ages. It happens again and again that forces retreat away from paths of safety into the jaws of destruction. Or away from open lines of supply and waiting stocked depots into barren wilderness.
I suggest an elegant solution. Add a "Preferred Retreat" order option to the unit commands menu. If enabled, it would allow the owning player to click on an adjacent eligible area (which would blink to acknowledge the command, like amphibious landing selections do) which met certain baseline parameters, e.g. could not choose areas with sighted enemy forces with equal/greater combat power and/or 100 MC. Then, during any ensuing combat, if the force had to retreat, it would go to the preferred area. It may still have to encounter enemy forces that moved in during the turn, but that would be the breaks. I am not sure the system allows it, but ideally the order should allow you to "chain" Preferred Retreats to two regions. If you wanted to make this second region ability only available to forces with leaders who have "4" or above command ability, that might be a nice extra touch.
If you wanted to allow for deep raiding forces, then tweak the Preferred Retreat setting command so that mobile forces (e.g. cavalry, horse artillery) led by a leader with a deep raider symbol could also choose Preferred Retreats into areas of enemy 100 MC, subject during the turn itself to making a command roll (which would retain some element of chance).
Adding this function adds some complexity and another layer of micro-management. But I think it would be welcomed by most players, who'd at least use it for key forces.
I agree. Your suggestion is very similiar to my suggestion. The main difference is I use a town or depot to represent the line of communications rather than having the player select an adj region. Check out some of my posts earlier in the thread. It would have to be coded but it seems all the components are already in the game.
It is very rare for an army not to retreat along their lines of communications. So it seems the solution is player designated lines of communication with a few rules for special situations or to prevent abuse.
-
Captain_Orso
- Posts: 5766
- Joined: Tue Sep 01, 2009 5:02 pm
- Location: Stuttgart, Germany
elxaime wrote:On the other hand, it could be the code changes above will make things satisfactory.
It sounds like this is an outgrowth of changes in the retreat algorithm for To End All Wars. Does this likely mean that the eventual fix will be a solution common to all games using this engine?
Probably.
Here's a fun one for you guys:[ATTACH]31917[/ATTACH]
Feel free to run the turn and see if you can get Longstreet to go anywhere but Manassas.
Note TJ and Lee are set to evade because if they're allowed to MTSG, they'll end up in Manassas too. Longstreet can't retreat across the Rappahannock, but TJ can MTSG across it.
Feel free to run the turn and see if you can get Longstreet to go anywhere but Manassas.
Note TJ and Lee are set to evade because if they're allowed to MTSG, they'll end up in Manassas too. Longstreet can't retreat across the Rappahannock, but TJ can MTSG across it.
- Attachments
-
- Bad Retreat.zip
- (495.9 KiB) Downloaded 225 times
-
Captain_Orso
- Posts: 5766
- Joined: Tue Sep 01, 2009 5:02 pm
- Location: Stuttgart, Germany
I ran this a couple of times. There are couple of things I'm seeing. This was done with the temporary fixed engine from 2014-10-16.
1. You cannot plot a move across the Rappahannock, because of the fleet in the Rappahannock River region.
I think actually you should be able to plot the move, but if you don't succeed in your river-crossing dr, your stack will simply stop at the river.
2. MTSG doesn't seem to take the blocking fleet into account, or because of the way the random numbers are generated used for the dr, I always got the same dr for the crossing and could always cross.
Lee and Jackson with Evade Combat seem to have less of a chance to MTSG. Beside, even if they did, they would still be set to EC, so what would be the point to that? just for morale support?
3. Longstreet's retreat to Manassas is absolutely valid.
The battle takes place on day 1. On day one the Union corps stack in Culpeper is still in Culpeper, even if it is rail-moving to Manassas. On day 1 it's still in Culpeper. In Manassas there is only a division under Howe and in Alexandria, well...
Whether Longstreet, when his retreat location is decided, has to role for crossing the Rappahannock--10% chance vs 4 naval combat elements--or it is simply forbidden, it seams that he cannot retreat to Fredericksburg. The next most likely retreat location is Manassas.
When Lee and Jackson MTSG'ed, they both returned to their original location, which is part of the engine's code. Whether the MTSG code takes the blocking fleet into account when deciding if they can MTSG I don't know, but I would guess that the engine is coded that if they MTSG'ed into Stafford, they will be returned to where they came from, as long as time allows.
1. You cannot plot a move across the Rappahannock, because of the fleet in the Rappahannock River region.
I think actually you should be able to plot the move, but if you don't succeed in your river-crossing dr, your stack will simply stop at the river.
2. MTSG doesn't seem to take the blocking fleet into account, or because of the way the random numbers are generated used for the dr, I always got the same dr for the crossing and could always cross.
Lee and Jackson with Evade Combat seem to have less of a chance to MTSG. Beside, even if they did, they would still be set to EC, so what would be the point to that? just for morale support?
3. Longstreet's retreat to Manassas is absolutely valid.
The battle takes place on day 1. On day one the Union corps stack in Culpeper is still in Culpeper, even if it is rail-moving to Manassas. On day 1 it's still in Culpeper. In Manassas there is only a division under Howe and in Alexandria, well...
Whether Longstreet, when his retreat location is decided, has to role for crossing the Rappahannock--10% chance vs 4 naval combat elements--or it is simply forbidden, it seams that he cannot retreat to Fredericksburg. The next most likely retreat location is Manassas.
When Lee and Jackson MTSG'ed, they both returned to their original location, which is part of the engine's code. Whether the MTSG code takes the blocking fleet into account when deciding if they can MTSG I don't know, but I would guess that the engine is coded that if they MTSG'ed into Stafford, they will be returned to where they came from, as long as time allows.
I had Lee and Jackson set to evade because that stops them from MTSGing into Falmouth. If you run it with them set any other way, they MTSG, cannot retreat back to Fredericksburg, and end up in Manassas with Longstreet.
What bothers me is the total lack of consideration for Warsaw, which is where I expected Longstreet to end up. I'm aware Manassas is a valid retreat location, but it's enemy held and 100% enemy MC. Warsaw has no supplies, but is friendly and 100% friendly MC. Warsaw should be the first priority, Manassas the second.
Also, the ability to MTSG across a blocked river region is highly problematic and shouldn't happen. As I see it, right now MTSG is not affected by a river blockade, but retreat is. This can be disastrous and even be abused, I think.
What bothers me is the total lack of consideration for Warsaw, which is where I expected Longstreet to end up. I'm aware Manassas is a valid retreat location, but it's enemy held and 100% enemy MC. Warsaw has no supplies, but is friendly and 100% friendly MC. Warsaw should be the first priority, Manassas the second.
Also, the ability to MTSG across a blocked river region is highly problematic and shouldn't happen. As I see it, right now MTSG is not affected by a river blockade, but retreat is. This can be disastrous and even be abused, I think.
-
Captain_Orso
- Posts: 5766
- Joined: Tue Sep 01, 2009 5:02 pm
- Location: Stuttgart, Germany
Merlin wrote:I had Lee and Jackson set to evade because that stops them from MTSGing into Falmouth. If you run it with them set any other way, they MTSG, cannot retreat back to Fredericksburg, and end up in Manassas with Longstreet.
??? You had 94% MC in Stafford, which means that the Union Corps there must go to OP unless set to PP, which means he will attack and call in McDowell and co., which means Longstreet will be fighting something like 12 divisions with his 3?, 4?, plus he's set to hold-at-all-costs. Suicide. But that's beside the point.
Merlin wrote:What bothers me is the total lack of consideration for Warsaw, which is where I expected Longstreet to end up. I'm aware Manassas is a valid retreat location, but it's enemy held and 100% enemy MC. Warsaw has no supplies, but is friendly and 100% friendly MC. Warsaw should be the first priority, Manassas the second.
Warsaw is avoided because it's on a peninsula. The code looks at how many adjacent land regions a retreat-target has. The few adjacent land regions, the less likely for it to retreat there.
Merlin wrote:Also, the ability to MTSG across a blocked river region is highly problematic and shouldn't happen. As I see it, right now MTSG is not affected by a river blockade, but retreat is. This can be disastrous and even be abused, I think.
I think the code should take blocking into account. Pocus will have to have a look at that aspect.
Captain_Orso wrote:??? You had 94% MC in Stafford, which means that the Union Corps there must go to OP unless set to PP, which means he will attack and call in McDowell and co., which means Longstreet will be fighting something like 12 divisions with his 3?, 4?, plus he's set to hold-at-all-costs. Suicide. But that's beside the point.
By the time I uploaded the game, I was just experimenting with ROEs. Longstreet was initially set to B/G to hopefully retreat before things got worse because I knew they were already bad. Lee, TJ and company were B/R from a previous turn and Evade was an attempt to salvage the game.
Warsaw is avoided because it's on a peninsula. The code looks at how many adjacent land regions a retreat-target has. The few adjacent land regions, the less likely for it to retreat there.
I didn't know that. Manassas makes more sense with that particular bit in mind.
I think the code should take blocking into account. Pocus will have to have a look at that aspect.
Yeah, that's the most glaring thing right now, and the only one which can mean the difference between a bad loss and losing an army.
In next release candidate, a stack won't MTSG if it has to cross a river which is being blockaded by ships, or if the ZOC in its own region (generated by enemy) would prevent it from entering it (its own and current region), meaning it can't reach the battle region.
Hofstadter's Law: "It always takes longer than you expect, even when you take into account Hofstadter's law."
Pocus wrote:In next release candidate, a stack won't MTSG if it has to cross a river which is being blockaded by ships, or if the ZOC in its own region (generated by enemy) would prevent it from entering it (its own and current region), meaning it can't reach the battle region.
Oh god, yes! You guys are fecking awesome.
Who is online
Users browsing this forum: No registered users and 2 guests