Nothing but Battle Stuff Thread

[GraniteStater]

OK, I posted this in the History forum and also on ACG's Layman's Guide thread:

[color="#40E0D0"]Maps and Mapmakers of the Civil War[/color] has a captioned commentary on one map that states (paraphrase): "Bayonet wounds were rare and hand-to-hand fighting was uncommon." Page 183, somewhere around that point.

If combat always ends in a 'hand-to-hand'...maybe the modelling isn't entirely congruent to historical reality. Not that it has to be.

This is kind of the starting point of this thread. Some guidelines:

* This thread is chiefly about about what happens in a Region in a battle. In one Region. MTSG is relevant - so is Bombardment, but my intent is battles in Regions, including naval battles, a sadly neglected topic.

* Why? Well, some of you here know I talk about modelling, the model. When we talk of 'Newton's model' or "Einstein's model' we are aware that the model is not reality. The map is not the territory. Every map betrays the prejudices of the mapmaker; every model is a sieve; the operator makes choices and has priorities and beliefs. Indeed, Kant, if he's your guy, said we can't (ha!) know reality.

* In CW2, what is 'reality'? Reality is Historical Reality (HR). Regiments were the tactical unit, not brigades; brigades and divisions were conveniences, really, and most certainly not anything, conceptually or otherwise, like a 20th century division, for instance. We shall posit that we can know Historical Reality for the purposes of this thread, but we shall use it as a referent, not a subject.

* Then we have the Game, or Game Reality (call it GR); the model. How does Combat work? We shall not recreate ACG's fine post-doctoral work in The Layman's Guide, but use it we may. Can we apprehend GR? Can we understand this model, IOW? What is happening in the model when scouting Cav withdraws successfully? Do Sharpshooters apply their attribute, their bonus, to a stack? Are 20 lbers and their heavier cousins the most preferred loose Arty in a stack or Corps, or are 10 lbers good enough? Should 10 lbers go in a Div instead of loose? Is integrated Cav in a Corps or stack really doing its scouting well enough, or should one have some loose Cav in a Corps or stack?

* I write this so that we may have some Knowledge of what is nearest and dearest to our warring, nerdish hearts: brawling. I wish to Know. I want to win my Battles. I want to know when discretion is better and when a steely resolve is called for. That's what this thread is for.

* This thread is not intended to repeat or engage in Analytical Dissections, though, of course, feel free to apply it as you wish.

* In that vein, I hope we may arrive at some bits and pieces of Knowledge; perhaps, if lucky, some Facts. I hope we may establish some, and record them here, in suitable forms. I hope to establish this thread as a repository and a reference for Battle Stuff.

Enough of a preface. We may commence with the Battle Screen: what is it telling us? In more than one way, I wish we had some elements of the AACW BS still in place. Nevertheless, in AACW, the numbers of men, horses and cannon were 'flavor numbers,' although not entirely without use. Are the M/H/C Numbers in the CW2 BS screen authoritative? How about when, as I see it, most often an Army Cdr in a neighboring Region's face is displayed: does he contribute to the Battle? The BS may be a good starting point.

[GraniteStater]

Or things like this:

(He is inactive, so no MTSG behaviors for this battle). Since McClellan was certainly threatened, McDowell switched postures and joined on the next round...

Quoting ACG from an AAR thead.

Is this right? AFAIK, an inacive status reduces MTSG probabllites, perhaps greatly, but does not rule it out - or am I getting this confused with Posture?

And the use of "battle" and then "round": what is ACG saying here? Inactive precludes MTSG in the battle, or a round? McD "switched postures" and joined - but if he's inactive...???

See? This is why I started this thread. We have lots of good info about Battle, but it's spread out hither and yon. I'd like to gather it.

[minipol]

I already made a few summaries of battle stuff in other threads a while back.
All info I collected from discussions. It definitely would be interesting to put it all in a thread and eventually the wiki.

[ArmChairGeneral]

There were several set of question in GS's posts, so I am going to break them into a couple of posts.

Inactivity, posture, MTSG:

When I did the Combat Testbed I probably set up and watched about 60 or 70 battles, of which about six had either a Corp or an Army stack that was inactive. In no case among these battles did any inactive formation MTSG on any round; ~6 battles, avg ~2 attempt per battle to MTSG (the logfiles seems to show that formation has to roll to MTSG on every round) so 0 out of around 12 times. The time to march was 8 days in all the battles, all were in defensive formation so had a malus, and the two formations involved were an army and its corps, so they got some bonus to want to march to each other. Back of the envelope calculation of the odds of MTSG under these conditions (assuming that they can do it when inactive) are like 25-30%. You would expect to have three or four MTSGs but there were zero. Going by my estimate of the MTSG chance at 25%, then there would be ~3% probability that all twelve would fail to MTSG (Binomial distribution: P[12 failures out of 12 tries] = .75^12 = .0317). When they were both active they MTSG two of three eligible rounds. I concluded that, based on these numbers, brownies either do not MTSG at all or there is a penalty that makes the probability so low that I just assume it won't happen.

(Caveat: The analysis relies on the estimate of the percentage chance. If the percentage were 10%, then twelve failures in a row has a ~28% chance of happening, which is actually fairly likely. 12 isn't a very big n, so it is possible that there is a very small chance to MTSG when inactive and I just happened not to get one. Either way, it is WAY less common than when activated)

Formations will not MTSG in passive posture according to the wikis. In Defensive posture they can but have a smallish penalty to their MTSG roll. THE overriding factor in MTSG % chance is days march, -10% chance (starting from base 100% chance) per day's march. Terrain, weather, rivers, roads all count (but even if you give them rail pool access, the probability is based on how long it take to march, not ride). If you succeed in MTSGing across a river, the formation does not suffer a river crossing COMBAT penalty, nor is there any cohesion loss in getting to the battlefield. I order the stack to march to the region in question to find out how many days the MTSG % is based on, and then delete the move. (If the stack is already marching toward the battle region, supposedly the % is based on the days remaining, so being partway there increases MTSG chance.)

MTSG does not occur on the first round (At least not in any of the 60+ battles I looked at in the Testbed). MTSGing only occurs on the second and later rounds. More shots are fired in the first round (see next post) so the stack that is starting the battle must be strong enough to survive the first round on its own. When MTSGers arrive they do not benefit from entrenchments, and since they are fresh and the original stack is weakened and less likely to be chosen to fill the frontage, the opposing force will be very likely to target the newly arrived and un-entrenched MTSGers.

If the battle is won, the MTSGers return to their original region in the same magical way they got there (no cohesion loss for movement, etc.) and re-enter their original entrenchments. If an MTSGer withdraws from the battle however, (it has a chance to withdraw by choice at the start of each round) its retreat path is mapped from the region the battle takes place in; since it didn't technically "move" to the battle region, it has no special preference in the retreat algorithm to want retreat to its original region so will often retreat to a "dumb" place instead.

Withdrawal decisions are made by the Army commander, but each stack has to roll separately to see if it can actually disengage or if it must still fight that round. Sometimes one stack withdraws and the other fights all by itself (this is usually bad). ROE effects withdrawal chances, so I try to keep ROEs uniform among potential MTSGers so that none of the stacks is extra prone to withdrawing. It is harder to make withdrawal rolls as the rounds go by.

[ArmChairGeneral]

Definition of terms:


Battle: One event that generates a Battle Report Screen. May have multiple stacks, MTSGing, whatever, but it takes place on a single day, listed at the top of the Battle Report Screen. It ends either in stalemate or with one side retreating. Battles are made up of Rounds, which are in turn made up of Phases. More than one battle can happen in a region on different days of the turn; this usually happens when another stack arrives later in the turn.

Round, AKA hour: Found by clicking the big E on the Battle Screen. Frontage, initiative, withdrawal rolls and stack targeting are calculated each round. Unit-to-unit targeting is probably also done by round, but I'm not 100% sure (it could be re-selected at each phase).

Phase: Where actual hit rolls are made. Each round is made up of one or more Fire (ranged) Phases that occur at each available distance and exactly one Assault Phase that occurs at 0 distance. The first round begins at max range with a Fire Phase for each distance until the range is zero and then one assault phase: Max Range 4 = 4 Fire Phases, 1 Assault Phase. For subsequent rounds there are always two phases, one Fire Phase at rage 1, and one Assault Phase at range zero. Notice on the Battle Report Screen that round 1 always results in more hits (hearts) than subsequent rounds; this is because there are more Fire Phases in the first round, and so more shots taken.

[GraniteStater]

THXX!!!

*wets thumb & starts leafing thru pages*

[Mickey3D]

Formations will not MTSG in passive posture according to the wikis. In Defensive posture they can but have a smallish penalty to their MTSG roll. THE overriding factor in MTSG % chance is days march, -10% chance (starting from base 100% chance) per day's march. Terrain, weather, rivers, roads all count (but even if you give them rail pool access, the probability is based on how long it take to march, not ride). If you succeed in MTSGing across a river, the formation does not suffer a river crossing COMBAT penalty, nor is there any cohesion loss in getting to the battlefield. I order the stack to march to the region in question to find out how many days the MTSG % is based on, and then delete the move. (If the stack is already marching toward the battle region, supposedly the % is based on the days remaining, so being partway there increases MTSG chance.)

In AACW (and I suppose it is the same in CW2) this is the list of factors modifying success % of MTSG :
-10% for each day of marching (all normal factors affecting the stack's movement apply)
+10% if adjacent to army HQ
+25% if the army HQ itself
+5% for each point of strategic rating of the leader
Every 5% of military control lacking gives -1% chance (both for start and end region)
-10% if in defensive posture

[GraniteStater]

So you pretty much have nailed it down that HQ -> Corps is more probable than Corps -> HQ. All else equal, of course, and assuming no major changes from AACW.

Great! We have a Game Fact!

[GraniteStater]

And something from ACG's remarks above - a river crossing does not inhibit MSG unduly, or perhaps at all.

But...there are navigable and non-navigable rivers on the map. Defending behind the upper Rappahannock, well, you get the defending bonus - for all non-navig. rivers, in fact.

How about a navigable stretch? Does that inhibit MTSG?

I would want to say, Yes, at least to some degree. We also don't know the back-end code for some places: across Mobile Bay, for example, is certainly across a 'navigable' stretch, but the code might regard it as more than just a navigable crossing, if I'm clear here.

[ArmChairGeneral]

A river crossing increases the number of days it would have taken to march there, so if the crossing adds 4 days it is -40% to MTSG chance. Mobile bay takes a long time to cross IIRC, so that alone would inhibit MTSG if the stretch is not MTSG verboten in the code. If it succeeds at MTSGing anyway, it does not fight as if it crossed a river and doesn't suffer the cohesion loss it would have had if it had to actually march to the battle into the battle.

There are different sizes of river crossings. If it is too large I think MTSG is either hard coded to not be possible or the extra time makes it a really low percentage chance. I don't remember seeing anyone MTSGing across the Mississippi, so there must be an effective or explicit limit of some kind. I have definitely MTSGed across smaller rivers, but try to avoid it to keep the chances as high as possible. I certainly wouldn't bet my Corps on relying on support from across a small river, much less a navigable one.

You can defend behind the river, and the attacking stack who marches across the river pays the penalty. Based on my reading of the source material, a Corp MTSGing to support the attack from across the river would not suffer the crossing penalty, but would have a lower probability of MTSGing in the first place. In any case, because MTSGers can only fight on the second and later rounds, river crossing combat penalties wouldn't apply anyway (only the first round has a malus, after that the attacker has forced a crossing and is fighting on the other side).

That's my take anyway.

[ArmChairGeneral]

Multi-stack Behavior Part 1


This is the part where I long-windedly explain what I meant in my vague post in pgr's AAR (be careful what you ask for GS ). Unfortunately it is kind of complex. Skip to Part 3 for a simplified but usable takeaway (that you already probably know because it conforms to common sense and experience). The first parts are groundwork.

I am just now starting to get a handle on deciphering multi-stack behavior, these kinds of battles have really burned me in the past. I have come up with a general framework for thinking about them that has improved my results since I began using it. I am sure that this is not the exactly correct explanation and that some of the details are misunderstood or outright wrong, but this conceptual model seems to provide a rational explanation for some of the strange behaviors and has made these types of battles more predictable (and thus winnable) in my recent games.

Stack targeting and participation (unit-unit targeting once the frontage has been filled is a different topic) depends primarily on posture. Some notation for precision:

Posture Color Abbreviation / ROE Color Abbreviation
examples:
R/R = Assault posture (Red) / All out attack (Red)
O/* = Orange (regular attack) posture / Any ROE
B/O = Defensive posture (Blue) / Normal Engagement (Orange)
G/G= G/* Passive posture only has Green ROE

In terms of stack behavior there are two kinds of postures, the R/* and O/* "attacking" postures and the B/* and G/G "defensive" postures. R/* and G/G are special cases of O/* and B/* respectively. They can be thought of as subsets of the more general O/* and B/* postures for the Stack participation and targeting determinations under consideration here, so I will not continue to repeat that O/* stack behaviors apply to R/* postures as well.

Accidentally storming a structure could be bad, so you have to specifically give leave to attack the structure with R/* instead of O/*. Otherwise the behavior is the same, the attacker chooses the targets, but R/* allows targeting stacks inside structures instead of just the stacks in the region like with O/* postures.

B/*'s are automatically eligible for first round frontage-filling if targeted by an attacking stack. If not, they do not fill first round frontage. They were manning their positions but no one charged them so they didn't have anyone to fight. If a friendly B/* was targeted on the first round then the ones who didn't fight the first round are unlocked and their elements can be selected for frontage on the second and subsequent rounds to support the other stack(s) that were attacked in the first round. They do not always 100% of the time engage on the second round, though, they are just allowed to. Also just because they are now available to fill frontage does not guarantee that they will be picked, there are other factors.

G/G represents the special case where a stack has moved to the rear of the region, putting friendly stacks in other postures between them and enemies to do the fighting while the G/G stack regroups. To model this, attackers prioritize targeting other movement and posture combinations over stationary G/G stacks. So if one of your stacks is B/* and the other G/G the attacker will tend to target the B/* stack first. If their stacks outnumber yours or if all the stacks in the region are G/G then the G/Gs can be targeted. If targeted, they fight back but try to withdraw even before the first round starts. If they fail their withdrawal roll (and if they are G/G and fighting then they did by definition) there is some kind of combat penalty for the rounds they remain on the field.

Continued...

[ArmChairGeneral]

Multi-Stack Behavior Part 2

Cases:

If you have have two B/* stacks and the opponent has one O/* the attacker will choose (random-weighted by size in hits) one of them as a target. Those two stacks engage the first round. The other B/* stack sits out the first round because it was not targeted. On the second round your untargeted B/* stack may now Join In Support of its buddy by becoming eligible for the next round's frontage allocation. If the first stack took a pounding then the second stack elements have a good chance of being "frontaged" and engaging. This is the outcome you were shooting for, your first stack wears them down in the first round, and your fresh second stack then engages the already weakened troops.

It is very important to make sure that a stack can survive the first round. If it is too small it can easily be overwhelmed before the second round when its friend(s) can join. Even an early withdrawal can be problematic for the same reasons as in the MTSG post above. Just because it is small does not mean it won't get picked.

What about when you have an O/* and B/* stack and you opponent has one B/*? Your O/* is the attacker and so targets the only choice, the lone enemy B/* . Your O/* and his B/* are eligible for first round frontage and they fight each other for the first round while your B/* sits out the round. Your B/* has orders to stay in its entrenchments and let the other stack do the fighting. They will do this as long as the O/* is doing reasonably well. If the O/* is in over its head, the B/* can Join in Support and fill frontage starting on the second round. If the O/* has it under control the B/* stack will sometimes never even get into the fight letting the O/* do all the work.

A subtle but important distinction is that in actual shooting terms, the attacker is the one whose side has an O/* stack. A battle can have two attacking sides but can't have two defending sides: no battle would occur. Everyone on the "attacking" side(s) fights using offensive stats, modifiers, damages etc. Only the stack selection process cares what the original posture was, not the battle engine. If the B/* engages for any reason, its elements will use offensive numbers in the combat calculations even though for stack-choosing purposes it was considered a B/*. This is what I meant by "switching" posture, though it isn't technically switching, it was always going to be fighting using offensive stats and no entrenchments because of its teammate in O/* posture. If they are inactive, they suffer the same combat penalties they would have if they were forced into O/* by 100% enemy MC.

Say you have two O/* stacks that you are lucky enough to have in the same region on the same day (they were stationary together or there was a synchronized move or you just caught a break). The first one chooses a target stack. Unless the first target stack has 150% or more hits (or maybe power, unclear) than the first O/* stack, the second O/* stack will usually target a different stack if one is available. If there were a third enemy B/*stack, it wouldn't be targeted by your two O/* (they already have targets) so the third B/* would not fight in the first round.

Continued...

[ArmChairGeneral]

Multi-Stack Behavior Part 3

So, applying this model of the stack choosing system to explain what happened to McClellan in Post #15 of pgr's [url="http://www.ageod-forum.com/showthread.php?34684-Seeing-the-Elephant-(Pgr-Vs-Highlandcharge)"Seeing the Elephant]Seeing the Elephant AAR with Highlandcharge.[/url] My explanation of what happened was:

Bee had one division entrenched in the structure at Manassas. McDowell had a three division Corps beseiging him but was inactive. In order to force McDowell into Assault posture, McLellan, with an Army stack consisting of one brigade, a Sharpie and a cannon, entered Manassas from Alexandria in Assault posture and assualted Bee's division. The only Union Stack in an attack posture was McClellan; McDowell was inactive, so under B/* orders. Therefore McDowell can't do anything the first round and sits idly by while McClellan's small stack targets and is completely destroyed by Bee's full Division.

Because McClellan's stack was overmatched, McDowell can be drawn into the battle on the second round, "changing" posture to Assault and storming the structure. He randomly chooses to target the lone garrison unit rather than Bee (which could have happened either as a result of stack-targeting or through the unit targeting process), and wipes it out without damaging Bee at all. Then, on the third round McDowell succeeds in making the withdrawal roll that McClellan probably called for at the start of the second round and calls off the attack (he must have failed the withdrawal roll at the beginning of the second round when McClellan would have ordered it.)

The bottom line is that if you are coordinating a multi-stack battle, stacks that are too small to survive the first round of combat on their own are a liability and can be wiped out/initiate disadvantageous retreats. I have lost many NM points or retreated from positions I needed to hold in battles exactly or very much like the one in the AAR.

[GraniteStater]

Wow.

[Mickey3D]

HQ -> Corps is more probable than Corps -> HQ

Indeed, the army HQ should be used as reserve for your corps. Moreover if Lee is the army commander (6 strategic rating) you will have a high probability to MTSG.

[GraniteStater]

The bottom line is that if you are coordinating a multi-stack battle, stacks that are too small to survive the first round of combat on their own are a liability and can be wiped out/initiate disadvantageous retreats. I have lost many NM points or retreated from positions I needed to hold in battles exactly or very much like the one in the AAR.

ACG's remarks should serve as a warning for those new to the game, and AACW vets, too : CW2 is not AACW. It isn't and we're all learning how so.

Per the remarks above: early, early war, before Divs - McDowell in a large stack and Hunter is a smallish, albeit respectable stack, in Alexandria. My intel was saying, we're OK, JJ in Manassas is proportionate. Well, JJ attacked (AI) and wiped Hunter right the **** out, Hunter WIA. McD retreated to DC.

Another one: three Div Corps, 840 Pwr, arty, support, a good Corps. How much stuff could Esteemed Opponent have in Little Rock - we don't need no stinkin' intelligence, we'll catch 'em napping. Well, enough to wipe the whole thing into cyberdust (a river landing). What was enough? One Division, with little arty and very modest trenches.

Call this example Beware of Landings - AACW wasn't so nasty about them, if you had 3:1, you could usually make a respectable showing and hang around for reinforcements coming the next Turn.

You've been warned - that's why I started this thread. CW2 is not AACW.

[GraniteStater]

Early Early War, Pre-Div: stay together if in the neighborhood of Many Unfriendlies. 35% is the max hit on CPs, so once you've reached it, there's no further penalty.

Oct 61 - Mar 62 - better, Divisions help a lot. Keep the above rule of thumb in mind, though.

Mar 62 -> Now, with MTSG, you're in a different landscape. Intelligence is crucial - you can't know the Oppo's command structure, but it will say "Corps" on it, unless it's retaining the designation from the formation procedure, OR, your human opponent is deliberately naming them 'Jim's Pancake House', or something - which has happened, but is frowned upon. Hey, I don't bother with nomenclature too much myself, I know I had a Corps called '61 Division Generals', but I don't alter them on purpose. Anyhow, Scoping things Out is Very Important - havi in our PbeM had two Army Cdrs in the Rappahannock/Richmond/C-ville area, PGT & Lee - whose Corps is whose? You don't know for a fact, but observe, read the Battle Screen, try to figurre it out, if you can - if you can, it's not easy.

Anyhow, if mutual Corps are balanced and built well, then one need not fear a Wipeout as much, MTSG is a different paradigm. Non-Corps actions bear scrutiny, though - and remember - individual Divisions, or loose stacks cannot MTSG - keep that in mind when you're guarding something, Longstreet is not coming to the rescue, you ain't a Corps, so tough noogies.

[ArmChairGeneral]

Hah, renaming! That's a pretty clever PBeM disinformation trick! I only rename stuff in single player to help me keep things straight.

[loki100]

A river crossing increases the number of days it would have taken to march there, so if the crossing adds 4 days it is -40% to MTSG chance. Mobile bay takes a long time to cross IIRC, so that alone would inhibit MTSG if the stretch is not MTSG verboten in the code. If it succeeds at MTSGing anyway, it does not fight as if it crossed a river and doesn't suffer the cohesion loss it would have had if it had to actually march to the battle into the battle.

There are different sizes of river crossings. If it is too large I think MTSG is either hard coded to not be possible or the extra time makes it a really low percentage chance. I don't remember seeing anyone MTSGing across the Mississippi, so there must be an effective or explicit limit of some kind. I have definitely MTSGed across smaller rivers, but try to avoid it to keep the chances as high as possible. I certainly wouldn't bet my Corps on relying on support from across a small river, much less a navigable one.

You can defend behind the river, and the attacking stack who marches across the river pays the penalty. Based on my reading of the source material, a Corp MTSGing to support the attack from across the river would not suffer the crossing penalty, but would have a lower probability of MTSGing in the first place. In any case, because MTSGers can only fight on the second and later rounds, river crossing combat penalties wouldn't apply anyway (only the first round has a malus, after that the attacker has forced a crossing and is fighting on the other side).

That's my take anyway.

In Rise of Prussia (so one assumes exactly the same engine), if you have a corps separated by a large river the chance for MTSG is seriously reduced unless there is a bridge between the province of the engaged stack and the re-acting stack. Have, sometimes seen corps support despite having the Elbe between them, but more often they will not MTSG. So its one of those situations where you really cannot rely on MTSG to save you.

My understanding is that if you react via MTSG, you pay NO penalty for the river (if a reaction occurs), so this is unlike the rule for attacking across a river.

The key modifiers are:

resBaseChanceOff = 100 // Base chance if in Offensive posture
resBaseChanceDef = 90 // Base chance if in Defensive posture
resCostPerDay = 10 // -10% for each day of marching
resModAdjGHQ = 10 // +10% if adjacent to army HQ
resModIsGHQ = 25 // +25% if the army HQ itself
resModLeaderStrat = 5 // +5% for each pt of strat factor of the leader
resControlChunkMod = 5 // Every 5% of MC lacking gives -1% chance (both for start and end region)
resCohCostPerDay = -3 // -3 cohesion for each day of marching

[Pocus]

each day of travel counts. If the bridge suppresses completely the extra cost of moving through a river, then MTSG is not penalized indeed.

[ArmChairGeneral]

Do we have bridges in CW2?

[Gray Fox]

Yes. A bridge links Alexandria and D.C. You can also see RR lines going across rivers.

[GraniteStater]

Some RRs across the Mississippi & Ohio...not as a distinct game feature, I don't think.

[ArmChairGeneral]

So what effect do those bridges provide to troops on the march (rather than by rail, since MTSG % doesn't use rail movement rates)? Act like a road so the time to march is for MTSG purposes is reduced? I have not noticed foot speeds to be reduced when crossing at a railbridge, maybe I just didn't notice?

[Gray Fox]

If one moves from a friendly region to another friendly region over a river, then you get any bridge movement speed. Move into an enemy region and you must fight your way over the river. MTSG movement doesn't get a river crossing penalty of any kind. The force just shows up across the river.

[ArmChairGeneral]

Aha, they are reducing the marching time. They basically allow you to cross the river as if it weren't there if you have at least 10% MC on both sides. So it takes no more than (I think 125% of) the terrain cost to move into Clear terrain plus any weather related time, while the time to cross the river gets essentially ignored. (All RR count as roads for marching and non-RR supply movement purposes.) If you don't have the MC on both sides you cross the river.

So with a RR bridge your Corps actually have a pretty good chance of MTSGing to each other if I am interpreting this all correctly.

Disambiguation: River crossing (combat) penalty versus river crossing movement cost. Crossing rivers in MTSG reduces the chances of the MTSG occurring due to the increased (notional) movement time, which can be reduced by a bridge. When the MTSG stack arrives (instantaneously) on the other side of the river it is not considered to have crossed the river in terms of combat penalties (plus it arrives after the first round, so no amphibious combat penalties anymore anyway).

A stack with Pontoons has a better chance of MTSGing across rivers than the same stack without them.

[Mickey3D]

I don't think the bridges between Alexandria/Washington and Louisville/<Put here the name of the city whose name I forgot > bring any movement advantage except if you use railroad movement. So it should not impact MTSG.

By the way, in AACW, giving RR movement order to your force (even if you don't move actually) was providing an advantage if the place where you were MTSG had RR (movement cost being lowered to 1 day) : see post #2 here.

[ArmChairGeneral]

Mickey3D,

I didn't think so either, but I can move a Division from Louisville to New Albany quickly by foot but it takes longer to cross other places along the Ohio, so that's why I concluded that it is helping.

In the Corps Combat testbed I specifically tried to induce MTSG rail effects a la the link you referenced, but the MTSG roll shown in the battle log file showed the same MTSG % as when no rail move was available, so if this is still the intended rule in CW2 it is either not working or I misunderstood what I was seeing in the log file (very possible) or otherwise made some other observational or experimental error. I didn't test it under all conditions, once it appeared that MTSG didn't use rail movement rates I quit trying in later testing.

[ArmChairGeneral]

This came up in another thread and is related to MTSG combats and Stack Behavior.

When on the defensive in MTSG situations, it has already been discussed that the initial stack must be large enough to survive a round of combat on its own. Almost as important is whether or not it retreats. Because it is probably outnumbered (your forces are split) and takes lots of hits due to the extended range in the first round, the leader often orders a retreat at the beginning of the second round, even in Orange (Normal) ROE. This can result in your MTSGers showing up and having to fight by themselves (with penalties) for a round or two until they can make their own retreat rolls, or retreating before ever actually seeing combat. Even though (once the MTSGers got there) you had enough strength to win the battle, you withdraw and lose the battle and leave the position.

The MTSGers will withdraw from the combat region to wherever the retreat algorithm sends them, not necessarily to their initial region, so in many cases you end up abandoning not one but two good defensive positions.

By setting the initial stack's orders to B/R (Defend/Defend-at-all-Costs) the overall leader (the Army commander, who may not be physically present for the battle) will not be able to order withdrawal until the beginning of the third round. This will give your MTSGers a chance to arrive and fight a round, evening the odds before the top-leader makes his decision (and his affinity for retreating in the later rounds is greatly reduced compared to Orange ROE). When MTSG is successful, this leads to much better combat results than under Orange ROE (assuming that your combined forces are big enough to win the battle) since everyone stays on the field.

Of course the downside is that MTSG may not succeed. In this case your initial stack will be fighting for several rounds on its own, without being able to withdraw, which can lead to serious losses. Only use Red posture when you think you have a high probability of MTSG, otherwise your initial stack may be hung out to dry.

It appears that B/R orders do allow voluntary retreats (at a low percentage chance) in later rounds, though the manual and tooltips contradict this.

The above is mainly for defensive situations. Red ROE in either of the offensive postures similarly reduces the leader's affinity for retreat (although slightly less than when in defensive) but this is not as often a Good Thing. On offense you really need to preserve hits and cohesion as much as possible to keep momentum going, so it is almost always better to withdraw and recover than keep fighting. O/R and R/R orders also give a noticeable combat advantage to the enemy increasing the danger. B/R orders do not give the opponent outright combat bonuses.

If you have more than one (non Army/Corp) stack in a region, B/R posture can help the first stack targeted to not retreat early so the second stack can join in on round 2, but is much less effective than combining everything into a single stack, even if you suffer command penalties.

[ArmChairGeneral]

Unit-to-Unit Targeting:

So far we have looked at Stack Behavior in multi-stack combat: Stacks target Stacks.

But once Stacks have their targets, the units in those stacks choose enemy units to target. Once a unit target is chosen a unit will (mostly) only fire on that unit. The hits a unit takes are (essentially) evenly distributed among its component elements. This means that if one unit is larger than the other, the larger unit has a significant advantage. Not only does it potentially make more hit rolls than the enemy, (depending on how many of its elements were selected for frontage) when it takes damage the hits are spread out among more elements so has a lower likelihood of having its elements routed or destroyed outright (counter-intuitively, an element does not have to be in the frontage to get hits spread to it from the rest of its unit).

Say you have a division on one side that targets a loose brigade on the other side. Almost always a unit will target back a unit that has scored a hit against it, so the Brigade will be engaging the Division for at least one round, maybe more. The Division has a good chance of destroying elements in the Brigade since all the hits go to a few elements, while the Brigade has almost no chance of seriously damaging any enemy elements since its hits are dispersed across a division's worth of elements.

This is readily apparent in the Battle Results screen. Loose brigades often get wiped out even in otherwise evenly matched combats. It is common for single element units like loose Cavalry to be destroyed, since they succumb quickly when targeted by Divisions or large brigades. The targeting mechanism weights toward choosing larger enemy units, but if a single element unit lands a hit on a Division that has not already selected a target, the Division almost automatically targets it back. Loose Sharpshooters tend to land hits early in a battle (because of their Initiative) so are prone to drawing fire from large enemy units that do not yet have a target.

Loose artillery, since they are support units, do not suffer from this phenomenon. The unit targeting mechanism strongly prefers targets with combat elements, so hitting an enemy Division with one of your loose cannons does not necessarily cause the victim to target back. OTOH, if the cannon is integrated in a division, then its unit DOES have combat elements, so hits landed by a Division's artillery WILL attract the unit it hits to target-back at the Division. (But this is fine, hits go to combat elements first, so the arty is basically shielded from hits in the early going. Too many artillery elements in a Division concentrates hits taken onto fewer friendly combat elements.)

Because of this, I find that leaving loose artillery in a Corps or Army stack is perfectly safe, but that loose combat brigades should be rolled into divisions whenever possible. I even put Rangers, Partisans and other zero CP units into Divisions when preparing for major battles: their survivability goes WAY up when they are not their own units.

On the Battle Report Screen, it is easy to differentiate between stacks, units and elements. Elements get an icon. Every element on a single line comprises a unit (sometimes it is a single element unit). Units in a stack are grouped together, so you have the leader on one line, the divisions on the next lines, and then the loose brigades, artillery, medics and wagons on their own lines. The leader that comes after the support units is the commander of the next stack.