Battle Results

[runyan99]

Generally speaking, making the defensive march to the guns concentration 'riskier' seems like a reasonable idea, which would tend to discourage but not eliminate the possibility of spreading out corps.

However, if I understand your chart correctly, in my opinion I think you are leaning towards tweaking too low. 55% chance for a 6 strategic leader with a railroad seems awfully low. I'd like to think I could rely on General Jackson to reinforce General Lee a bit more often than that!

I'd rather see you make that ideal situation an almost assured reinforcement, but not quite, say a 90% chance, and then work down the chart from there for worse leaders and rough terrain.

[Jagger]

Here is an alternative with increased terrain costs and tweaked leadership

25% base chance

// -15% for each day of marching ---(inactive commanders have longer marching times)
// +10% if adjacent to army HQ
// +25% if the army HQ itself
// +15% for each pt of strat factor of the leader
// Every 5% of MC lacking gives -1% chance (both for start and end region)

Using the above data values, here are the adjusted percentage chance per each round of 6 rounds of combat.

--6----5----4----3----2----1--Leader Strategic Ratings + adjacent HQ
110---95---80---65---50---45--Railroad movement
-50---35---20----5----0----0--Clear movement
-35---20----5----0----0----0--Woods movement
--0-----------------------------Hills movement

I think I prefer the first table rather than this one. The arrival chance is a bit more level and even the lowly 1 strategic leader can show up once in a blue moon. Harder to predict results and we don't have the automatic reinforcement of the 6+ leaders.

[Jagger]

However, if I understand your chart correctly, in my opinion I think you are leaning towards tweaking too low. 55% chance for a 6 strategic leader with a railroad seems awfully low. I'd like to think I could rely on General Jackson to reinforce General Lee a bit more often than that!

Bear in mind, that is per round of combat with 6 potential rounds per battle.

So will Jackson be there on round one of combat---55% chance.
On round two of combat--2x55%=110%
On round three of combat --3x55%=165%
On round four of combat another 55% chance, etc through round 6 of combats

Jackson will probably arrive on round one of battle with a 55% chance. It is highly likely Jackson will be in combat by round two at the latest. Although with very bad luck, he may never arrive as a reinforcement if he keeps rolling greater than 55%, six straight times in a row.

If the attacker is lucky, he will beat Lee before Jackson arrives!!!

If you look closely at my first modified table, the odds favor arriving by round 4 of combat even with the poorest strategic leadership when moving by railroad. The strat leader of 1 has a 30% chance per round of arriving. Which means a cumulative 120% chance of arrival by round 4. Not guaranteed but likely arrival by turn 4. That isolated corps might be hammered for a few rounds of combat before McClellan arrives as reinforcements--if he arrives at all.

[Jabberwock]

That would make the odds for Jackson:
Round 1: 55%
1 or 2: 79.75%
1-3: 90.89%
1-4: 95.90%
1-5: 98.15%
1-6: 99.17%

[McNaughton]

Bear in mind, that is per round of combat with 6 potential rounds per battle.

So will Jackson be there on round one of combat---55% chance.
On round two of combat--2x55%=110%
On round three of combat --3x55%=165%
On round four of combat another 55% chance, etc through round 6 of combats

Technically, you cannot get more than a 100% chance... The math is a bit off (it isn't a multiplication of odds of success, as the failure odds are negated). While the chances that you get a positive result are greater the more rounds that you have (if a yes means that you end the roll and activate the march). However, it cannot surpass 100%, as with 6 rounds there still is that possibility that every time you end up on the 45% of no.

This 45% doesn't change, and every time you roll, you still have a 45% chance of failing the march. It just happens that you would get 6 chances to beat this 45%, which does affect the total odds, but, these odds do not reach beyond, or even at, 100%.

The total odds are closer to 90% given Jackson's 55% chance of success, and 6 rounds in which to take this 55% chance (remembering that every round there still is a 45% chance of failure every time). Still, 90% is very high.

I guess it is like playing a game of 'heads and tails' where if heads is ever landed on, that person wins, but, tails has to be landed on 6 times for the other person to win. Possible, but very low probability given these specific situations.

[Jagger]

Technically, you cannot get more than a 100% chance... The math is a bit off (it isn't a multiplication of odds of success, as the failure odds are negated). While the chances that you get a positive result are greater the more rounds that you have (if a yes means that you end the roll and activate the march). However, it cannot surpass 100%, as with 6 rounds there still is that possibility that every time you end up on the 45% of no.

This 45% doesn't change, and every time you roll, you still have a 45% chance of failing the march. It just happens that you would get 6 chances to beat this 45%, which does affect the total odds, but, these odds do not reach beyond, or even at, 100%.

The total odds are closer to 90% given Jackson's 55% chance of success, and 6 rounds in which to take this 55% chance (remembering that every round there still is a 45% chance of failure every time). Still, 90% is very high.

I guess it is like playing a game of 'heads and tails' where if heads is ever landed on, that person wins, but, tails has to be landed on 6 times for the other person to win. Possible, but very low probability given these specific situations.

Believe it or not, I took graduate level statistics about 24 years ago. I actually made an A as well. Which I was very proud of at the time as I didn't find statistics an intuitive subject. And now much of the knowledge is sadly gone forever. And I am too lazy to look up the formulas.

So what is the exact formula??

[Pocus]

Jabberwock one is correct, if your chance to arrive at a given round is 55%, then the chance to arrive at round 3 or less would be 1.0 - (0.45)exp3 = 0.908 (90.8% of arriving)

Say differently, if you have 45% chance of not arriving the first round. So the chance to not arrive at the first AND the second round is 45% of 45%, and so on. Then the opposite probability is 100% minus your result.

[Pocus]

The problem I see with your tweak, is that you are killing offensive march to the sound of gun. Take as an example a corps in a half controled region, trying to backup a corps which just advanced (0% MC) in another. The advancing corps will be alone for the fight... And this is a big problem for us.

[Jabberwock]

So what is the exact formula??

1 - f^n (*100 for %)

f = chance of failure
n = number of rounds

[runyan99]

The problem I see with your tweak, is that you are killing offensive march to the sound of gun. Take as an example a corps in a half controled region, trying to backup a corps which just advanced (0% MC) in another. The advancing corps will be alone for the fight... And this is a big problem for us.

If I understand correctly, that particular parameter is tweakable as well, isn't it? It's only a matter of choosing the right number to make offensive reinforcement possible again.

[Jagger]

The problem I see with your tweak, is that you are killing offensive march to the sound of gun. Take as an example a corps in a half controled region, trying to backup a corps which just advanced (0% MC) in another. The advancing corps will be alone for the fight... And this is a big problem for us.

Interesting. I don't think I have ever noticed offensive "march to the guns". Although I have never intentionally formed my attacks to take advantage of an offensive march to the guns.

I have noticed that military control is typically at 100% for the defending corps region. So the maximum impact of military control is -20% for the defending region. One possibility is to cut the impact of MC by half-negative 1% for every 10% of enemy MC. So an enemy region under attack at 100 MC would produce a -10% impact on offensive march to the guns instead of -20%.

Using that adjustment would produce the following results with these adjustment values.

25% base chance

// -5% for each day of marching ---(inactive commanders have longer marching times)
// +5% if adjacent to army HQ
// +15% if the army HQ itself
// +5% for each pt of strat factor of the leader
// Every 10% of MC lacking gives -1% chance (both for start and end region)
(For calculations, I am assuming 100% MC by defender and thus a -10% impact on results)

OFFENSIVE March to Guns

--6----5----4----3----2----1--Leader Strategic Ratings + adjacent HQ
-45---40---35---30---25---20--Railroad movement
-25---20---15---10----5----0--Clear movement
-20---15---10----5----0----0--Woods movement
--0---------------------------Hills movement

So a level 2 strat commander, adj to HQ, offensively marching to guns along railroad would have a 58% base chance to arrive by round 3 of combat. The same commander would have 82% chance of arriving by round 6 of combat. (Hey, the formula works!!)

Same situation but a strat level 4 commander marching in clear terrain would have a 39% base chance to arrive by round 3 of combat. The strat 4 commander would have 62% chance of arriving by round 6 of combat.

Although one important factor is the offensive PBEM commander has the ability to coordinate attacks to ensure the arrival of his troops and not rely on offensive March to the Guns. I think a human player would depend on his offensive plotting rather than rely on march to guns for reinforcements.

Another important factor to bear in mind is that I am only considering PBEM play as I only play PBEM. So the changes are intended to balance out play against the typical human strategy of heavily entrenched defensive lines. These types of tweaks may have major impact on AI play which may make them unuseable in a regular game against the AI.

An example of human play is my current game against Kilcavalry. It is now September 62. There has not been a major battle in the east since December 61. His defense is 5 major formations located at Clarke, Manassas, Stafford, Fredericksburg and army HQ/troops at Culpepper. All completely entrenched and in a semi-diamond formation with excellent mutual support to all points of defense. With march to the guns, Clarke is defended by 2, possibly 3, of his 5 formations. Manassas is defended by 4 of 5 formations. Stafford is defended by 3 of 5 formations. Culpepper by 5 of 5 formations. Frdericksburg is not on the front lines but still defended by 3 of his 5 formations.

As the Union, I have built up to 6.5 corps in the East. Considering the low strat ratings of my corps commanders, I might have two corps active on any one turn. Those 2 corps, even 3 corps, cannot successfully attack any segment of his line with an advantage in odds which is needed to attack entrenched troops. The last attack on Clarke, 2 corps, in December of 61 resulted in 20,000 Union casualties to 10,000 reb losses. My two corps attacked Clarke but ended up fighting the three formations at Clark, Manassas and Culpepper. I can't afford a repeat of those losses in troops with very little to gain.

Fortunately Kilcavalry and his rebs can't attack either. A railroad connects Louden and Fairfax. So an attack on either location would result in defense by 3.5 entrenched corps. Winchester is isolated as there isn't a railroad connection between Loudon and Winchester. Hills and a river prevent march to the guns. So Winchester could be attacked as an isolated position. However the corps at Winchester is strong, protected by a river, completely entrenched and has the army bonus. The rebs would lose far more troops than I without a guarantee of victory. In addition, he would have to abandon his entrenched defensive formation to attack. Which he is not about to do.

So we have had a completely static, entrenched, eastern frontline for the 9 months since Dec 61. The front is so static and defensively strong that I have withdrawn 2 of my 6.5 corps and sent them to the Gulf for invasions. My remaining 4.5 corps can hold the eastern front. In response, he has withdrawn most of his troops at Fredericksburg plus I think troops in the west. Although I can't really take advantage of that withdrawal at the moment as Fredericksburg is not on the front lines anyway.

I am trying to tweak the engine to prevent these static PBEM situations from developing by adjusting the command control, army bonus, march to the guns and maybe entrenchment values. I want to see isolated corps vulnerable so humans can advantageously attack them. I would also like to see more maneuvering space available. As it is now, there is very little space for maneuver once players spread their corps out and entrench. At that point, futile frontal attacks are required before space can be created. Thus static fronts develop. These situations don't develop in AI games but do develop in human PBEM games.

Humans can develop almost impregnable defensive lines.

[Jagger]

I have been experimenting against the AI with modded March to Guns (MTG) parameters combined with the command control mod reducing command range and increasing army bonus. In working with the MTG prameters, my primary objective was to prevent automatic MTG when adjacent corps moved by railroad to reinforce battles. With the modded parameters, there still remains a very good chance of MTG reinforcements when using railroad movement but far less likely to arrive on battle round one. Only good strategic corps leaders, or those in the same region with an Army HQ, have a chance of MTG when marching through clear or wooded terrain instead of using railroads.

Below are the modded MTG parameters and percentage chances of arriving as MTG reinforcements.

March to Guns (MTG)

20% base chance offensive posture
15% base chance Defensive posture
// -5% for each day of marching ---(inactive commanders have longer marching times)
// +6% if adjacent to army HQ
// +9% if the army HQ itself
// +3% for each pt of strat factor of the leader
// Every 10% of MC lacking gives -1% chance (both for start and end region)


Railroad(1day)-Independent Corps-defensive posture
Corps Leader---9---8---7---6---5---4---3---2---1--Strat rating
BattleRound1--37--34--31--28--25--22--19--16--13
BattleRound2--60--56--52--48--44--39--35--29--24
BattleRound3--75--71--67--63--58--51--47--41--34
BattleRound4--84--81--77--73--69--63--58--51--47
BattleRound5--90--87--84--81--76--71--65--58--50
BattleRound6--94--92--89--86--82--77--72--65--57

Clear Terrain(5days)-Independent Corps-defensive
Corps Leader---9---8---7---6---5---4---3---2---1--Strat rating
BattleRound1--17--14--11---8---5---2-----------
BattleRound2--31--26--21--15--10---4-----------
BattleRound3--43--36--30--22--14---6-----------
BattleRound4--53--45--37--28--21---8-----------
BattleRound5--61--53--44--34--24--10-----------
BattleRound6--67--60--50--39--28--11-----------

Woods Terrain(6days)-Independent Corps-defensive
Corps Leader---9---8---7---6---5---4---3---2---1--Strat rating
BattleRound1--12---9---6---3-------------------
BattleRound2--23--17--12---6-------------------
BattleRound3--32--25--17---9-------------------
BattleRound4--40--32--22--12-------------------
BattleRound5--47--38--27--14-------------------
BattleRound6--54--43--31--17-------------------

Adjusted percentages for other situations:
-Corps in same region as Army HQ---add 2 to strategic rating to find correct percentage column from above tables.

Example: Corps with strat leader of 3 in same region as Army HQ is equivalent to an independent corps with strat leader 5. So Jackson, (strat 6, with Army HQ, Using railroad to MTG), is equivalent to strat 8. He has an 92% chance of MTG by battleround 6.

-Army HQ---add 3 to strategic rating to find correct percentage column from above tables.

Example: Army commander McClellan HQ with strat rating of 1 is the equivalent to an independent corps with strat rating 4. Army HQ McClellan has a 11% chance of MTG reinforcements battleround 6 in clear terrain.

So General Lee as Army HQ, (strat 6, army HQ, MTG in clear terrain), is equivalent to a strat 9. Army HQ Lee has a 67% chance of MTG by battleround 6 from clear terrain.

A corps commanded by Stonewall (strat 6, HQ region, MTG clear terrain) in the same region as Lee HQ has a 60% chance of MTG by battleround 6. If he were not with Lee's Army HQ, he would be strat 6 and have a 39% chance of MTG by battleround 6. Corps with army bonus will also impact strategic rating.

A corps commanded by Bonham with strat rating of 2 in the same region as a CSA HQ is equivalent to strat 4 and has a 11% chance of MTG by battleround 6 in clear terrain. If Bonham were not in the same region as any CSA army HQ, he would not have any chance of MTG from clear terrain. Although the army bonus may adjust the strat rating of Bonham as well which might give him a different strat value.

These MTG parameters will work equally well with combat set at 4 or 6 rounds of combat. Although I have a second set of parameters which slightly increase MTG percentages when used with combat set at 4 rounds. I may post that in a few days.

I am attaching three files for anyone that would like to experiment against the AI. All files should be unzipped into the Settings folder of the ACW folder after making copies of your original Command&Leaders, Reserve and Combat files. The Command&Leaders file modifies command range and army bonus. The Reserve file modifies March to the Guns parameters. These two modded files are meant to be used together but can be used separately. I believe they are ideally used with 4 battlerounds instead of 6 battlerounds. The battlerounds can be adjusted in the "Combat folder" of the same settings folder. I will attach that file as well for any who want to test. Four rounds of combat reduce casualties in large battles which might normally go 6 rounds.

Any comments or thoughts on the MTG percentages? Anyone feel the changes don't feel right-too low or too high?

Also anyone interested in testing in a game, drop me a private message. I am ready to test vs a human at this point if someone is interested.

[runyan99]

Humans can develop almost impregnable defensive lines.

Um yeah, but you're not being very creative about maneuvering you opponent out of that rather short line either. Tweaking the march to the guns reliability sounds like a good idea, but on the other hand, you need to show a little more initiative as the commander in chief of the USA armed forces.

In that screenshot, you hold the entire Shenandoah valley, and there seem to be no defensive lines east or south of Fredericksburg.

Is there some reason you cannot post a screen of defensive corps in front of Washington, and then send a large force up the Shenandoah to Charlotesville, VA to flank the line and threaten Richmond?

Could you possibly load up a corps or two and float them down to a landing east of Fredericksburg?

Either of these strategies would make the CSA line at Manassas untenable. I mean, McClellan launched his Peninsular campaign for some reason....

The game is less flawed than your Union strategy.

[Jagger]

Double post-deleted.

[Jagger]

Um yeah, but you're not being very creative about maneuvering you opponent out of that rather short line either. Tweaking the march to the guns reliability sounds like a good idea, but on the other hand, you need to show a little more initiative as the commander in chief of the USA armed forces.

In that screenshot, you hold the entire Shenandoah valley, and there seem to be no defensive lines east or south of Fredericksburg.

Is there some reason you cannot post a screen of defensive corps in front of Washington, and then send a large force up the Shenandoah to Charlotesville, VA to flank the line and threaten Richmond?

Could you possibly load up a corps or two and float them down to a landing east of Fredericksburg?

Either of these strategies would make the CSA line at Manassas untenable. I mean, McClellan launched his Peninsular campaign for some reason....

The game is less flawed than your Union strategy.

I considered an end around through the Shenandoah Valley even though it is out of supply range. So I sent cavalry towards Charlottesville, Va to recce and found a couple of entrenched divisions there. Very easy and quick for him to move even more troops there if needed. Remember Kilcavalry and I are both maxing out recruitment. He has a lot of troops. I have almost maxed out on divisions and probably will be by the end of 62. I suspect he is in the same situation.

I have gone completely defensive in the East with minimal forces which can't be defeated and have been sending everything to the Gulf. I have most of Texas and gradually taking over Louisiana. There is real maneuver room in the Gulf that doesn't exist in the East.

I find I am getting more VP's going deep south than battering against the wall in the East. Although we are in another recruiting cycle now. He may be able to stabilize the situation in the Gulf shortly if he can produce enough troops.

Is taking the Gulf states creative enough for you?

[runyan99]

Is taking the Gulf states creative enough for you?

Well, I feel the Union won the war in the west anyway...

[Jagger]

Well, I feel the Union won the war in the west anyway...

Lots of rivers to push quickly into various areas along the Gulf. Although I see he is entrenching divisions with artillery along the mouths of rivers and various points along rivers to prevent river supply. I will have to assault his entrenched divisions to allow supply upriver. Fortunately, I can quickly concentrate using my navy for assaults and he has to defend everywhere.

[runyan99]

By the way, if you would build a depot at Harrisonburg, I'm not at all sure you would not have sufficient supply at the south the end of the valley to sustain a flanking offensive.

[Jagger]

By the way, if you would build a depot at Harrisonburg, I'm not at all sure you would not have sufficient supply at the south the end of the valley to sustain a flanking offensive.

Do you mean Winchester?

I have never seen a large or even medium army supported beyond one region from a supply depot. Either they capture Charlottesville in one, maybe two, attacks or they must retreat due to lack of supply.

Even my cavalry runs out of supply two regions from Winchester.

[runyan99]

Do you mean Winchester?

No. You can build a depot anywhere, even if there is not a game-city in the region.

[Pocus]

Do you mean Winchester?

I have never seen a large or even medium army supported beyond one region from a supply depot. Either they capture Charlottesville in one, maybe two, attacks or they must retreat due to lack of supply.

Even my cavalry runs out of supply two regions from Winchester.

Only supply wagons can 'call' supply from depots. If you lack them, or have only one, then it is normal that you can only supply yourself in or adjacent to the supply depot region. Combat troops don't have this 'call' feature.

[GShock]

It's a solid thread, probably one of the most important ones in the forums about how to improve things in the game.

On my opinion, Mc Naughton has the upper edge with his point about helping the side who's offensive and hampering the side who's on the defense.

Now, historically, the South seceded and it was annhilated by the North when Lee stopped defending and went offensive. The goal of the seceding state is survival, the goal of the Union is to subdue the seceding States and reunite the nation. So the goal of CSA is to tire down the Union (as proper and historical) while the goal of USA is to make CSA surrender...and this is achieved by destroying armies, economics, and thus, the morale.

Maybe trenches are too powerful or too easy to achieve but that's not the point imho, the point is that USA shouldn't really need to bleed in VA even though it's the fastest way to win.

NM should become a critical factor beating the entrenchment bonus of CSA.
Prize more the NM, increase neg. modifiers for the side whose morale is low and players will need to go offensive.

This way CSA has to counter different threats in multiple states to defend its strategic towns and it won't be so easy to build the Maginot line in VA we see in the SS.

Something i was thinking about: VP are assigned turn after turn to the side who owns obj cities. If these VP are converted into NM turn after turn (i.e. 10VP=1NM)?

NM gives bonus and penalty to combat and other factors. The side with less than X NM must convert VP to NM (and we know with few VP your side can can raise less funds) at that point, it's obvious the side who owns objs will be advantaged against the side who's just entrenching both militarly and economically...thus prizing more objs and less the entrenching and conservative play.