IS failing to MTSG possible?

[Cardinal Ape]

Its never been my experience in this game that stacks failing to MTSG is something one needs to pay any attention to. For as long as I have played this game MTSG'ing has always seemed to work all of the time. At some point I just stopped paying it any heed at all because it never failed.

Is this not everyone's experience? Do you all have stories about how big battles went wrong because half your army was too drunk to fight? I don't have any.

---

Anyhow, I tried to test it before posting. Should be easy to get a corps to fail an MTSG roll, right? Wrong. I tried chasing this rabbit a good long while and couldn't catch it.

According to the values in the gamelogic.opt file it should be possible to set up a situation where a corps is guaranteed to fail it's MTSG roll. Or at least close enough to confirm that failing to MTSG is possible. I couldn't pull it off.

I tried to create a battle where most corps would have a low chance to MTSG. I used a muddy mountain region. The worst corps was led by an inactive general with the slow mover trait and a strat rating of 1. His arrival time was 45 days. The MC in the battle region was 95% enemy controlled. According to the rules, assuming this corps has all positive modifiers, its chance to MTSG is beyond negative 300%. Not a single corps fails to join the battle, even the worst one joins without fail, no matter how many times I re-run the battle.

I also spent some time changing the values in the gamelogic.opt file thinking I could simply up the values high enough to essentially turn off MTSG'ing. It doesn't work. No matter what I do I can not get a corps to fail MTSG'ing.

Is it broke? Intentionally broke perhaps? Those values in the gamelogic.opt file would imply that failures would be more common than success in any battle where corps have a travel time of more than 7 days. That would make for quite a different game... Maybe it was like that before I joined the CW2 party, sometime around patch 1.4, either way I'm baffled.

[loki100]

I'll freely confess to less experience with CW2 than some of the earlier games, but yes I have seen failures.

Some complete - ie a corps never turns up but these most usually due to one of the major malus factors - like being on the wrong side of an unbridged major river.

Most often in games such as RoP and PoN you see formations drop in and drop out of a long battle. To unpick this you do need to go into the very detailed logs - as this can happen either due to failing MTSG or a corps losing so much cohesion that it will no longer commit to battle - you can't just rely on changes to the indicator of numbers engaged.

[principes romanes]

An interesting question.

From my experience, where I've noticed MTSG failing is when there is a very short battle. I've had a 4 element cavalry 'corps' be wiped out by Longstreet on the first round of battle before any other corps could MTSG. I admit I haven't noticed if corps have failed to MTSG in larger battles that last past the first or second round of combat.

[pgr]

Ive seen failure to MTSG. Try having Little Mac as an Army commander and he'll nerf his own corps commanders into inactivity.

[Durk]

I too have send MTtSG fail. I always encourage players to replay battle to see the range of actions. While this might not approach the range of research Cardinal Ape has done, it is enlightening to see that no battle has a predetermined outcome.
In this type of exercise I have seen unites march to support and decline support.

[pgr]

An interesting question.

From my experience, where I've noticed MTSG failing is when there is a very short battle. I've had a 4 element cavalry 'corps' be wiped out by Longstreet on the first round of battle before any other corps could MTSG.

Of course there is no MTSG in the first round. The stack better be strong enough to stand on its own or else you have problems. (Like premature retreats

[principes romanes]

Of course there is no MTSG in the first round.

Makes sense. I'm sure I was aware of that at one point but had forgotten it. That complexity is one of the joys of this game

[Rod Smart]

The original post is mostly right, corps will support defending stacks even across muddy hills in the snow. Mostly.

A couple of examples off the top of my head:
- Army HQ stacks. Even in the same region. That's a discussion we've had a couple times.
- Locked/fixed units.
- As mentioned, if the battle doesn't go on long enough. Not just day one, I've seen corps not show up until round 4 or later.
- I've had trouble with Leesburg supporting Harpers Ferry, and vice versa. Something about a large mountain being in the way perhaps?
- If the river is blocked by a navy, the supporting stack won't cross the river.


I've also had trouble with corps who marched to the sound of the guns, and adversely affected the battle. A corp that has been beat down to nothing the first 12 days of the turn contributes nothing but an easy rout on the final battle. Similarly, a tiny, weak stack in the care of a two star (like a garrison behind the lines, supporting the front line) contributes a net-negative.

[Cardinal Ape]

Huh. So it appears that the delayed commitment option has a rather large impact on MTSG. That caught me by surprise; I don't remember seeing that documented anywhere.

Anyhow, the reason why I couldn't get a single MTSG failure in my first test was that my delayed commitment option was set to 'no delay'. From what I can tell it seems that the delay option sets hard caps on the % chance to MTSG. On the 'no delay' commitment option, MTSG appears to be 100% guaranteed.

I searched the forum and the wiki for more details on this option but all the explanations are rather vague. Though, I did find a thread in the To End All Wars forum where multiple people expressed that they chose to play on the 'large delay' setting specifically to get more MTSG failures.

I re-ran the same battle on each delay setting and recorded the MTSG chances. Each setting looks like it has its own bracket of minimum and maximum chances to MTSG. Essentially, the delay setting is the MTSG setting, I think..

The min/max commit chance of eligible corps:
-No Delay: Range = 100-100
-Small Delay: Range = 65-80
-Medium Delay: Range = 55-70
-Large Delay: Range = 40-60

That's for the first round of MTSG, for each additional round of combat the base chance will increase by 10%.

That's not what I expected find. Every corps that is eligible to MTSG will have a commit chance that falls in those ranges. I can't figure out what the math is behind the individual corps commit chances, whatever it is it bears no resemblance to what the wiki and manual say. The stated '-10% chance for each day spent marching simply can't be true.

Another thing I learned is that a corps doesn't get just one roll or only one chance to MTSG per round of combat; its a lot more complex than that. Seldom does it happen where a stack gets only one chance to join the battle. If a corps fails to commit against one target it can be called upon in the same round to roll against another target.

Nor does a corps need to actually succeed on a commit roll to join the battle; if an opposing stack successfully targets and commits against a stack not yet in battle then that stack will automatically join the battle. Technically, Jackson can go chase down 'lil'Mac and drag him into battle even if Mac failed all his commit rolls. So in order for Mac to not MTSG he would need to fail all his commit rolls and all CSA corps would need to fail their commit rolls against him.

I'm guessing that in order to get a battle where the majority of the eligible corps failed to MTSG it would require mass strategic incompetence on both sides.

[pgr]

Intresting observation about delayed commitment. I generally play with it on at a medium to high setting, which might help explain why I frequently see stacks failing to MTSG.

I also see failures within battles. A stack will MTSG one round and not the next.

Of course saying MTSG failures only tend to happen in short battles (1-4 rounds) is kind of a chicken and egg issue. Id say the battles go short because MTSG failed!

[Captain_Orso]

I will not bang my head on the table I will not bang my head on the table I will not bang my head on the table I will not bang my head on the table I will not bang my head on the table ....

Please read Combat Explained.

There is no such thing as "rounds of MTSG". There are rounds of battle. At the start of each, MTSG might occur. Just because a Corps which MTSG in one round of battle, it does not guarantee, it will MTSG for the next round of battle. (I have no idea why it should work this way, but that's what's in the Wiki.)

Given that you have a 15 day turn, and that it takes at about 5 days, in good weather, on dry ground, in good terrain, to move into a region, that means that, if one side marches into the battle region, the earliest a battle might start would be on day 5 or 6.

Now, depending on game settings, there might be a delay in commitment. So there could be several days, once one enemy stack in OP detects an enemy stack, before stacks actually start targeting each other and a battle can begin.

Since MTSG takes the number of days a Corps stack needs to march to the battle region into account, if moving into the region, and delayed commitment push the battle toward the end of the turn, if the an MTSG'ing stack's hypothetical arrival day goes beyond the end of the turn, it cannot successfully MTSG.

So if the battle starts on day 11 or later, there is almost no way MTSG can succeed.

[loki100]

...
Since MTSG takes the number of days a Corps stack needs to march to the battle region into account, if moving into the region, and delayed commitment push the battle toward the end of the turn, if the an MTSG'ing stack's hypothetical arrival day goes beyond the end of the turn, it cannot successfully MTSG.

So if the battle starts on day 11 or later, there is almost no way MTSG can succeed.

with respect for your understanding of the game system, this is, I believe, wrong.

Its relatively unlikely that a battle will start late into the 15 day turn but it can in those games either with relatively large provinces (eg PoN) or poorly developed terrain (some regions in WoN/RoP etc). I'm assuming in CW2 that most MTSG influenced battles are taking place in the relatively well developed regions and the two armies are well designed in terms of CP malus etc - if so even if you march into a battle its unlikely to push into the late turn unless you are using the harshest of the traffic rule options. But in other games I have seen MTSG even for late turn battles.

A problem with MTSG is the mechanism is contrived. Time to arrive is not time elapsed in the sense that you are presenting it. Its simply becomes a factor that may limit/reduce the likelihood of it happening.

To me, my understanding is that it is not the French II Corps wandering between Quatre Bras and Ligny chasing cannonfire like my cats chasing a butterfly. Instead, one or both of the commanders spot/fear an emerging threat and prepare to respond. But a corps arriving via the means does not spend any real time in transit - distance simply reduces probability of success.

[Captain_Orso]

Yes, MTSG is very contrived, and in no way resembles physical reality, if you are looking at stacks in regions and distances. Two very poignant examples:

- First Situation: the CS has Lee (Army stack with several divisions) in the middle with a Corps each left and right (Longstreet and Jackson), moving forward line abreast.

The action:
First battle: Longstreet and Jackson both encounter enemy stacks in their target regions, and battle ensues in both. Jackson, being the faster mover, arrives earlier in his target region, and battle starts there first, and Lee MTSG's to Jackson's aid.
Second battle: One or two days later, Longstreet goes into battle, and again Lee MTSG's to the rescue.

Comparing game results with reality: Lee battled twice in one turn, in both regions to his left and right. In reality, this would mean that Lee marched first into Jackson's region and took part in his the battle, and then in the opposite direction, into the neighboring region on the other side of his own region, to take part in Longstreet's battle, all within the span of several days.

- Second Situation (Jackson is in Tennessee between Memphis and Corinth, with orders to cross the Mississippi north of Memphis and attack the enemy there. He's a fast mover, his Corps is in good condition, and the weather is good, so he will arrive just in time to kick some Yankee butt. A Corps from the same CS army is sitting in Corinth defending, trying to buy time and delay the Union's advance. (In a strange set of circumstances, the Union is advancing from the south in this case).

The action:
Jackson is one day away from arriving in the first region outside his starting region, north of Memphis, when battle starts in Corinth, and his Corps is called in to help, which it does. After the battle, Jackson 'continues' with his move.

Comparing game results with reality: Jackson, on his way westward, has just taken part in a battle in the region to the east of his starting region: In reality, that would mean that one day before being outside MTSG range, he counter-marched into the region in the opposite direction of of his march, took part in a battle, and nearly managed to arrive in his target region at the end of two weeks. (Actually, in the game this occurred in, he failed to reach his target region on time, because of the cohesion his Corps lost in the battle, plus IIRC the ground might have been muddy too. If successful, it would have been a brilliant Jackson-maneuver, and saved Memphis.)

In reality, Jackson's Corps would have march westward four days, then counter marched to the east eight days, done battle for a day, then counter marched again 8 days, and continued on for another eleven days to nearly reach their goal. That's 28 days in one turn... 28 Days... 28 Days Later!... Jackson's Corps is comprised of fast Zombies!

MTSG doesn't attempt to represent reality in days-of-movement. Turns are 15 days, and withing 15 days campaigns culminate, major battles, with large movements of troops, are fought and won. In the game, the player can only sit and watch as his plans for maneuver from the start of these 15 days, are carried out. There is no opportunity to react to the opponents movement within these 15 days, where in reality, many decision would be made, changes of orders carried out as the situation develops. In reality, an army maintained its Corps, not within days march of each other, but within an hour or less. But in reality, on realizing the enemy's maneuver, plans were overturned, marching orders changed, and all plans from the previous days forgotten. In the game, this is not possible. If in the game, we didn't have MTSG, Corps would skirt past each other with ease making a defense impossible, and the attacker could concentrate an attack with all an army's corps, onto a single enemy corps, without the defending player having a way to react. This is why we MTSG.

It is not at all unusual for battles to start late in a turn. Enemy stacks do not always start in neighboring regions, and poor weather is a fact of game-life. It may not be the most common of cases, but it is far from rare.

[Captain_Orso]

Yes, MTSG is very contrived, and in no way resembles physical reality, if you are looking at stacks in regions and distances. Two very poignant examples:

- First Situation: the CS has Lee (Army stack with several divisions) in the middle with a Corps each left and right (Longstreet and Jackson), moving forward line abreast.

The action:
First battle: Longstreet and Jackson both encounter enemy stacks in their target regions, and battle ensues in both. Jackson, being the faster mover, arrives earlier in his target region, and battle starts there first, and Lee MTSG's to Jackson's aid.
Second battle: One or two days later, Longstreet goes into battle, and again Lee MTSG's to the rescue.

Comparing game results with reality: Lee battled twice in one turn, in both regions to his left and right. In reality, this would mean that Lee marched first into Jackson's region and took part in his the battle, and then in the opposite direction, into the neighboring region on the other side of his own region, to take part in Longstreet's battle, all within the span of several days.

- Second Situation (Jackson is in Tennessee between Memphis and Corinth, with orders to cross the Mississippi north of Memphis and attack the enemy there. He's a fast mover, his Corps is in good condition, and the weather is good, so he will arrive just in time to kick some Yankee butt. A Corps from the same CS army is sitting in Corinth defending, trying to buy time and delay the Union's advance. (In a strange set of circumstances, the Union is advancing from the south in this case).

The action:
Jackson is one day away from arriving in the first region outside his starting region, north of Memphis, when battle starts in Corinth, and his Corps is called in to help, which it does. After the battle, Jackson 'continues' with his move.

Comparing game results with reality: Jackson, on his way westward, has just taken part in a battle in the region to the east of his starting region: In reality, that would mean that one day before being outside MTSG range, he counter-marched into the region in the opposite direction of of his march, took part in a battle, and nearly managed to arrive in his target region at the end of two weeks. (Actually, in the game this occurred in, he failed to reach his target region on time, because of the cohesion his Corps lost in the battle, plus IIRC the ground might have been muddy too. If successful, it would have been a brilliant Jackson-maneuver, and saved Memphis.)

In reality, Jackson's Corps would have march westward four days, then counter marched to the east eight days, done battle for a day, then counter marched again 8 days, and continued on for another eleven days to nearly reach their goal. That's 28 days in one turn... 28 Days... 28 Days Later!... Jackson's Corps is comprised of fast Zombies!



MTSG doesn't attempt to represent reality in days-of-movement. Turns are 15 days, and withing 15 days campaigns culminate, major battles, with large movements of troops, are fought and won. In the game, the player can only sit and watch as his plans for maneuver from the start of these 15 days, are carried out. There is no opportunity to react to the opponents movement within these 15 days, where in reality, many decision would be made, changes of orders carried out as the situation develops. In reality, an army maintained its Corps, not within days march of each other, but within an hour or less. But in reality, on realizing the enemy's maneuver, plans were overturned, marching orders changed, and all plans from the previous days forgotten. In the game, this is not possible. If in the game, we didn't have MTSG, Corps would skirt past each other with ease making a defense impossible, and the attacker could concentrate an attack with all an army's corps, onto a single enemy corps, without the defending player having a way to react. This is why we MTSG.

It is not at all unusual for battles to start late in a turn. Enemy stacks do not always start in neighboring regions, and poor weather is a fact of game-life. It may not be the most common of cases, but it is far from rare.

[Cardinal Ape]

A little follow up on the 'delayed commitment' option:

When two opposing forces meet the chances per day for a battle to start are:
-Large = 30
-Medium = 40
-Small = 50
-None = 100

However, those numbers may be misleading. If both opposing forces are in offensive posture then they each get to make their own commit roll.

Also, it appears that each force gets to make two rolls per day. Not sure whats up with that. I thought there might be a chance that the first roll was for a full duration battle and the second for a short duration/late in the day battle but I can't find any difference between the two.

Here is a snippet of Hancock taking four days to engage on the 'large delay' setting:

7:24:28 PM (Reporting) Battle in 239 Augusta, VA Day: 7 Round: 0
7:24:28 PM (Reporting) Commit Chance 1007837 W. Hancock' Corps 30 %, Rolled: 35 Not commited
7:24:28 PM (Reporting) Commit Chance 1007837 W. Hancock' Corps 30 %, Rolled: 45 Not commited
7:24:28 PM (Reporting) Battle in 239 Augusta, VA Day: 8 Round: 0
7:24:28 PM (Reporting) Commit Chance 1007837 W. Hancock' Corps 30 %, Rolled: 66 Not commited
7:24:28 PM (Reporting) Commit Chance 1007837 W. Hancock' Corps 30 %, Rolled: 51 Not commited
7:24:28 PM (Reporting) Battle in 239 Augusta, VA Day: 9 Round: 0
7:24:28 PM (Reporting) Commit Chance 1007837 W. Hancock' Corps 30 %, Rolled: 97 Not commited
7:24:28 PM (Reporting) Commit Chance 1007837 W. Hancock' Corps 30 %, Rolled: 87 Not commited
7:24:28 PM (Reporting) Battle in 239 Augusta, VA Day: 10 Round: 0
7:24:28 PM (Reporting) Commit Chance 1007837 W. Hancock' Corps 30 %, Rolled: 84 Not commited
7:24:28 PM (Reporting) Commit Chance 1007837 W. Hancock' Corps 30 %, Rolled: 12 Commited
7:24:28 PM (Reporting) TGroup.SetTarget W. Hancock' Corps is targetting A.P.Hill' Corps
7:24:28 PM (Reporting) TGroup.SetTarget A.P.Hill' Corps is targetting W. Hancock' Corps
7:24:28 PM (Reporting) GiveGroupsTargets 1007837 W. Hancock' Corps initiating fight against faction 1000002 engaging: 1008043 A.P.Hill' Corps

Finily, Athena kan't spellz 'committed.' Drives my spellchecker nuts.

[ArmChairGeneral]

I am beginning to suspect that the log uses the word commit in multiple contexts. There is one commit roll (the ones you have in your snippet) to see whether battle occurs at all and on which day.

Then once the battle actually starts, there are two more "commitment" rolls that have more to do with in-battle stack behavior and targeting, but are referred to as commitment anyway. Why there are two I have no idea, although only the second roll seems to actually determine in-battle behavior. It is not out of the question that the first roll is just junk, (as far as I can tell there are junk entries in the !mainlog, so why not in the !battlelog?) but it may have to do with withdrawal-before-battle rolls.

[Gray Fox]

In the battle logs I scrutinized, the second roll was to determine if the stack retreated from combat.