mod J & P Rebalance Mod by JonnyH13 and Pandakraut 05/06/2023 1.28.4

[pandakraut]

27 minutes ago, zyszczak said:

Hi, I was playing a lot with mod lately and want to give you guys some feedback. I mainly played csa campaign with only touching union side.

Thanks for the feedback. I agree with most of your criticisms of the current state of the mod, especially those around perks and 0* units. Many of these will hopefully be getting resolved in the future. Getting the difficulty right for all players has certainly been a challenge. We've got players who are struggling on BG and others who can mostly cruise through on Legendary so past a point we've just tried to provide configurable options to adjust the difficulty to player preference.

The variance mode is definitely intended to generate situations that are not winnable if your RNG is bad enough. I went back and forth a lot on enabling it by default or not. I ended up enabling it because I think the positives in mixing up the battles a bit is worth the trade off of the occasional ridiculous situation. You can also tweak the probabilities as well if you don't want to disable it entirely. For battles other than Shiloh I would probably recommend increasing the duplicate chance a bit if things are to easy. The extra brigades seem to bring out the most positive improvements to AI performance compared to anything else we've tried. I wouldn't recommend going above 50% though. Some of the larger battles will crash if to many units end up on the field.

My goal at CSA Shiloh is always to manipulate the battle in such a way that the landing can be taken almost unopposed. I tend to end up with fairly stretched out lines in an effort to keep the AI spread thin and unlikely to charge. Unlike in the base game I'm not sure you have the ability to actually surround and destroy the AI units and still take the landing in time. Maybe someone can pull it off though. In some ways Shiloh is the worst battle in the mod due to having to deal with huge AI brigades when you haven't had a chance to build your own. In other ways it's the best battle in the mod, because the player can't just rely on completely rolling over the enemy force. Because of this there is actually a chance of barely getting a win, getting stuck with a draw, etc. 

I'm not really sure what the solution to the full force destruction is. That is also possible in nearly every battle in the vanilla game as well, though you'll take pretty heavy casualties in a few of the CSA battles trying it. The route we are currently experimenting with is lower damage but higher morale impact. Early results are promising, but still needs a lot of work to see if it will actually improve things. We would probably make AI units rout off the field if we could, but that isn't something that I've figured out how to implement. Making units shatter or surrender easier sort of accomplishes the same thing, but each comes with it's own downsides.

If you're interested, something you could try is reducing the battle timer changes. Perhaps endOfDayMultiplier set to 1.05 and timerRecommendedMultiplier and timerMandatoryMultiplier to 1.2. The defaults are set towards allowing players more time than they need to play the battles at the pace they want, but this also allows experienced players plenty of time to clear the map.

The topic of the AI logic is a tricky one. It's one of the most complex systems in the game and we've only figured out how parts of it actually work. I usually deliberately influence the AI to bunch up as that works best with how I build my armies. But I've also heard Jonny complain that they spread out to much and he just punches straight through them with his setup. 1.3 currently has a few changes to cannon AI so that the AI should be a bit less incompetent with them in most cases. We could make the AI a bit more likely to charge all at once, but that leads back to the problem of if the AI just mass charges you every time newer players will get overwhelmed and frustrated while an experienced player will just change the type of traps they set up.

[pandakraut]

2 minutes ago, nik_nv said:

Not only, the skirmishers now are more fast than cavalry! pandakraut please fix it 😓

I've seen this complaint multiple times but every time I've tried to reproduce it the cavalry is faster unless you're in a scenario where perks, terrain, condition are not in your favor. I'd call this working as designed out side of the larger issue that it's possible to stack speed bonuses to many times so that you can get pretty big disparities in how fast units move.

However, if you'd like to reduce skirmisher speeds the unitModifiers.csv has a skirmishSpeedModifier that can be changed. Maybe try 1.45 or 1.4?

[pandakraut]

1 hour ago, zyszczak said:

Also in this mission to win you have to lose less then 75% of your armies but duplicated brigades start with around 1k losses at the start (bug)  In mt case it is around 3k men.

I've tried to reproduce this multiple times, but I'm fairly sure this is a display only bug. The numbers are correct when the post battle results are calculated.

[Kristoph420]

1 hour ago, zyszczak said:

 

 

• AI is dumb like in vanilla except from charging decisons. It would be great if you guys were able to fix it .This is simulation of AI's thought process: 
  • Use cover - Hell no. Cover is for weak and I dont need weak mean in my army anyways.
  • How about flanking general, we have more troops then the enemy. - Flanking is stupid and gay.
  • So what is the best tactic? - We are just gonna attack in short line and the rest of our 40k army will stand in big spaghetti formation just behind our fighting troops. (But not always sometimes it actually can get into a nice line with multiple attacks, mainly at start of the battle but after breaking AI lines it have big problems with reforming it).

 The part about the AI being dumb is just to funny. I agree with a lot of your pros and cons. I am so looking forward to the newer version.

 

 

Seriously, the mod is awesome and a great thank you to J&P for taking their time to making this great game even greater. The mod will still have some issues but eventually everything will be perfect

 

Edited March 25, 2019 by Kristoph420

[civsully1]

2 hours ago, Kristoph420 said:

 

 

Seriously, the mod is awesome and a great thank you to J&P for taking their time to making this great game even greater. The mod will still have some issues but eventually everything will be perfect

 

Absolutely agree!

[civsully1]

So PK, It's been well over a year ago when I finished playing UGCW until I recently found your mod. I haven't watched your video from a couple of months ago when you did 2nd Bull Run yet. But I was reading my notes from a year ago and saw a quote that Col Kelly had posted on his excellent series that " there is this bug preventing the units deployed in the 1st day to deploy on the 2nd".  And so he never did the first day battle.  Does his comment on this bug occur in your mod with this battle?  Thx in advance.

Edited March 25, 2019 by civsully1

[pandakraut]

13 minutes ago, civsully1 said:

So PK, It's been well over a year ago when I finished playing UGCW until I recently found your mod. I haven't watched your video from a couple of months ago when you did 2nd Bull Run yet. But I was reading my notes from a year ago and saw a quote that Col Kelly had posted on his excellent series that " there is this bug preventing the units deployed in the 1st day to deploy on the 2nd".  And so he never did the first day battle.  Does his comment on this bug occur in your mod with this battle?  Thx in advance.

I've never encountered that bug so perhaps it was fixed since Col Kelly recorded.

[Minas Moth]

From my experience so far with this mod and I usually play every battle two or three times just to see what will work and what will not work it seems that most of the problems for the AI comes if player isn't playing by the script or historically how the battle objectives were set, units placed etc. Case and point is the battle of Gaines Mill... If you play it historically and try to hold the breastworks you are doomed. However if you form your defensive line on the creek that runs north-south on the western part of the map the AI will kill itself trying to force a crossing. A boggy terrain, water crossing and forest really hamper the AI... But guess what, Lee would also be extremely hampered had McClellan formed his line there. What I'm trying to say is... the moment the AI gets really "better" most people will not bother with the game anymore. As it is now, I find a lot of stuff placed well with some tweaks here and there that will do the trick. At the moment the replay-ability of the game comes from this mod forcing you to find new ways on how to fight battles. I get very mixed results... I had land slide victories with 10 to 1 casualties in my favor, I had narrow victories with more losses on my side, I had defeats snatched from the jaws of victory when AI managed to take the objective in last second. 

Perk system could be tweaked a bit... especially the "scale of bonuses" that perks grant to a certain units.

@pandakraut how does the damage dealing system work? are all parameters tied to a certain weapon or are there also some values that affect the ammunition (bullet etc)?

[ErikS1998]

Quote

That's a bug from the base game that gets reported every now and again. Unfortunately once it happens your save file is basically broken. Going back to a prior save and replaying the last battle before the bug occurred is probably the only solution.

I've also seen it suggested that the game needs to be reinstalled to fix this issue, but I don't know for sure if that's the case. If you find a way to resolve the issue, please let me know.

Loading the previous save did it for me.

Edited March 26, 2019 by ErikS1998

[pandakraut]

2 hours ago, ErikS1998 said:

Loading the previous save did it for me.

Great, glad you were able to get around the issue.

[pandakraut]

4 hours ago, Minas Moth said:

From my experience so far with this mod and I usually play every battle two or three times just to see what will work and what will not work it seems that most of the problems for the AI comes if player isn't playing by the script or historically how the battle objectives were set, units placed etc.

This matches my experience as well. Where the AI really falls apart is when you break the sequence of the battle and it doesn't find units where it is expecting them. Even if you follow the script, if you get to a point where your units are all hidden that's when you'll see the AI standing around or scouting for you with their artillery. 

Rather than bothering me I look at it is another aspect of player skill. Can you play the AI against itself so it's logic breaks and you can take advantage of it. The same tricks don't always work from battle to battle so there is enough variety to stay interesting.

5 hours ago, Minas Moth said:

how does the damage dealing system work? are all parameters tied to a certain weapon or are there also some values that affect the ammunition (bullet etc)?

The basics is that once an attack animation starts the attack will hit. Base damage is set by the weapon * then everything else applies a multiplier to that base. Cover, size, range, firearms, efficiency, unit type, etc.

[Minas Moth]

@pandakraut and what about ammunition or projectile type? when I used to mod for American Civil War Empire Total War we used different values for projectiles to even further the detail and unique features of certain weapons.

[pandakraut]

18 minutes ago, Minas Moth said:

and what about ammunition or projectile type? when I used to mod for American Civil War Empire Total War we used different values for projectiles to even further the detail and unique features of certain weapons.

In UGCW only artillery ammunition types apply additional properties to the damage calculation. For non-artillery weapons any inherent differences in the effects of different calibers are included in the base damage values.

[GioMoreno]

Have been playing this mod for some time. Thank you for your great work! Here are my two cents:

1. Confederate Shiloh on MG and Legendary are insane (those 5k+ union brigade very strong when charging/defending, and they have to be beaten back multiple times). I ended up tying up as many Union force with right-wing AI and sweep into the landing with my left wing force and barely make it after regular time. What makes it worse is the AI force has insufficient ammo in the supply unit! It seems to hold ~20k and they just ran out of ammo attacking Devil's Den! It would help if the AI has more supplies.

2. THIS IS A QUESTION: does accuracy has lower marginal utility when shooting skill is already high? should i pick accuracy perk if my unit's shooting is 100 and the first two perks are on accuracy? Or should I pick a melee/discipline perk because accuracy has a soft/hard ceiling?

3. Artillery seems too strong compared to infantry weapon, even 12pdr howitzers have mass killing potential in this mod (and very available). Considering ~90% casualties are caused by infantry weapons historically (Earl Hess, 2017) the 0*-1* units cause way too less damage, especially when both the defender are in forest and a volley only deal 3~5 kills.

4. Melee is better than base game, but still historically hand to hand combat rarely occurs -- melee should deal more morale damage than actual damage --- while range attack more actual damage than current, to keep up pace with arty damage.

5. is it intentional that Enfield and Lorenz are more expensive than US made light arms of similar quality? Historically Lorenz should be on par with HF1841 while Enfield similar to SF1861. But I'd agree to keep it as is for game design.

[pandakraut]

11 hours ago, GioMoreno said:

1. Confederate Shiloh on MG and Legendary are insane (those 5k+ union brigade very strong when charging/defending, and they have to be beaten back multiple times). I ended up tying up as many Union force with right-wing AI and sweep into the landing with my left wing force and barely make it after regular time. What makes it worse is the AI force has insufficient ammo in the supply unit! It seems to hold ~20k and they just ran out of ammo attacking Devil's Den! It would help if the AI has more supplies.

Addressing supply amounts and allied units in general is on the list of things to adjust. Shiloh is definitely one of those battles where both sides are starved for supplies.

11 hours ago, GioMoreno said:

2. THIS IS A QUESTION: does accuracy has lower marginal utility when shooting skill is already high? should i pick accuracy perk if my unit's shooting is 100 and the first two perks are on accuracy? Or should I pick a melee/discipline perk because accuracy has a soft/hard ceiling?

There is no cap on accuracy as far as I know, it's just a straight multiplier so the more you have the better. Firearms has a comparatively small effect on accuracy low and reload speed. So you can pick perks based on what you'd like to do rather than worrying about anything hidden. I tend to go with a balanced selection of perks to make my infantry more generically capable, but stacking one stat is incredibly effective(will be changed in the future).

11 hours ago, GioMoreno said:

3. Artillery seems too strong compared to infantry weapon, even 12pdr howitzers have mass killing potential in this mod (and very available). Considering ~90% casualties are caused by infantry weapons historically (Earl Hess, 2017) the 0*-1* units cause way too less damage, especially when both the defender are in forest and a volley only deal 3~5 kills.

See my comments on perks vs firearms in terms of the reasons that 0-1* units have issues. For a more general answer, our opinion is that rifle fire is to strong in the base game and artillery to weak. However, in the mod rifle fire only really becomes effective once accuracy perks are added. Artillery by contrast does pretty well even without accuracy, but once you stack the accuracy it will win battles by itself as long as you have enough supply. My current opinion is that the low end of rifle damage needs to come up and the high end of artillery damage needs to come down a bit.

11 hours ago, GioMoreno said:

4. Melee is better than base game, but still historically hand to hand combat rarely occurs -- melee should deal more morale damage than actual damage --- while range attack more actual damage than current, to keep up pace with arty damage.

We are currently experimenting with higher morale damage and lower actual damage across the board. Fun as it can be, if we can bring down the ridiculous casualty rates that experienced players can inflict(base game and mod) a bit I think that would be a positive change.

While we want melee to be effective, it's current ability under optimal conditions to mow down men faster than point blank canister certainly has an odd feel to it. The big question in both cases is does the game still feel good to play if we take steps to reduce casualties. Just for an example, I'm not sure the game gets better if instead of full clearing maps every battle ends with the AI bottled up in the corner infinitely routing but dieing very slowly. Will have to see how it turns out.

11 hours ago, GioMoreno said:

5. is it intentional that Enfield and Lorenz are more expensive than US made light arms of similar quality? Historically Lorenz should be on par with HF1841 while Enfield similar to SF1861. But I'd agree to keep it as is for game design.

Yes, import pricing is the idea behind the current costs. The current pricing structure we are using does not always result in a direct relation of price and performance. 

Thanks for the feedback. Hope you continue to enjoy the mod

[Minas Moth]

@GioMoreno & @pandakraut

Regarding artillery and casualty rates both in American Civil War and this mod. Artillery of the time was devastating. Every weapon from 6pdr to big siege guns had an immense potential to deal carnage. The most of that potential was fulfilled when:

a) capable men like Hunt on the Union side and Alexander Porter on the Confederate side commanded artillery units

b) ammunition and here I especially refer to shrapnel shot was not faulty (which was a big problem for CSA) and time fuses were set correctly and actually burned for the time they were set. in a computer game you obviously can recreate such faults in ammunition but it is much easier and elegant to use other mechanics. In this game artillery all uses same A standard ammo that never fails, and when you add to the mix more and more experienced commanders that "know what they are doing" you get really good artillery units that indeed can and  should be able to win battles.

On the fact of most of the casualties being inflicted by small arms fire. This is very very true but the thing is in American Civil War from around 10 000 engagements not many were fought in places where artillery could be used effectively. You have to keep in mind that at that point in time most of artillery was what we would call field artillery in a sense they needed direct line of sight to engage the enemy. Indirect fire was rare. Now imagine making a 6 gun battery maneuver through thick vegetation of Shiloh, Wilderness, Chancellorsville or complete area of Vicksburg campaign. On the other hand in places where artillery could actually use its full potential it proved murderous. Point and case would be battles of Malvern Hill, Fredericksburg, Gettysburg etc... In fact, some scholars claim that even at Gettysburg most of guns never fired at their maximum effective range as they either couldn't see what they were supposed to shoot or they couldn't reach it because of the terrain (most of the guns had a maximum effective range of 1.600 to 2.000 yards or 1.463 meters to 1.829 meters) . And Gettysburg is one of if not the only battle where long range artillery duels happened. The entire point why Lee decided to fight at Wilderness during Overland Campaign was to rob Grant of two things Union bested Confederates at: 1) use of artillery and 2) superior range and accuracy of Union muskets.

Now as you know, the base game doesn't work like that. Artillery can fire from weird angles as long as some units are spotting the targets. There are no elevation or terrain penalties for the artillery or areas where you can't even move it because it will not be effective. Also, one of the most exploited mechanics in this game is players putting their batteries in densely forested area. That thing very, very rarely happened in Civil War as the logistics involved were just not worth the end result. 

Bottom line, as a former modder, there is not much you can do with certain aspects of any game. And complexity which would involve proper artillery representation of the era simply outweighs the cons of the current system of the base game.

Edited March 27, 2019 by Minas Moth

[pandakraut]

7 hours ago, Minas Moth said:

Now as you know, the base game doesn't work like that. Artillery can fire from weird angles as long as some units are spotting the targets. There are no elevation or terrain penalties for the artillery or areas where you can't even move it because it will not be effective. Also, one of the most exploited mechanics in this game is players putting their batteries in densely forested area. That thing very, very rarely happened in Civil War as the logistics involved were just not worth the end result. 

While very rare, there are a few maps that have enough elevation that they do block artillery fire. Overall the game is incredibly generous in what it lets you fire over though. I only played it a bit myself, but I never really figured out how artillery worked in UGG where artillery required more realistic lines of fire. Though I feel like that was more a UI feedback issue that persists into UGCW where in the few situations where artillery can't fire you get no indication that it's sitting idle or needs to move unless you have a direct fire order issued.

Regarding dense forests, all of the damage penalties from terrain are currently not working in the game. So you'll do just as much damage in the open field as in dense forest or the middle of a river. The next version will be fixing these. I'm expecting that sitting in dense woods will still be very effective, but not quite as much as it currently is. Though that also just means the player will sit most of their artillery behind the dense woods and fire at units on the other side of them. At least  they will be more exposed to counterbattery fire which I've added logic for the AI to actually try to do when artillery is visible even if there is a slightly closer target.

[GioMoreno]

Thank you both for your reply, I think they make a lot of sense. Historically the complex topological features especially in the Eastern Theater makes ideal artillery positioning very difficult, since the rifle themselves can be fired to at least 400 yds with some accuracy and you cannot position them on level ground directly behind your infantry position and fire safely with canister shots. Also positioning them in dense forest is impossible since they limit firing angle and visibility.

@pandakraut

Since we cannot change innate game mechanics, fixing the former is impossible... but Pandakraut, have you considered implementing different damage penalties for infantry vs. artillery? For example, forest make infantry output damage lower by 40% but artillery lower by say 60-70%.

Edited March 27, 2019 by GioMoreno

[nik_nv]

14 hours ago, pandakraut said:

I'm not sure the game gets better if instead of full clearing maps every battle ends with the AI bottled up in the corner infinitely routing but dieing very slowly. Will have to see how it turns out.

Why to do not let AI to completely retreat? Why the player can do it for his units but AI can not?

If AI feels it is loosing the battle the best strategy is to retreat right, not stack in the corner, but move out from the map. Is there an issue in the base game? I never saw AI completely removing its units, I saw it only for captured ones..

AI retreating will resolve many realism problems with too many kills/captives.

Edited March 27, 2019 by nik_nv

[pandakraut]

1 hour ago, nik_nv said:

Why to do not let AI to completely retreat? Why the player can do it for his units but AI can not?

If AI feels it is loosing the battle the best strategy is to retreat right, not stack in the corner, but move out from the map. Is there an issue in the base game? I never saw AI completely removing its units, I saw it only for captured ones..

AI retreating will resolve many realism problems with too many kills/captives.

AI Supply wagons can retreat off the field when they are empty so theoretically this should be possible to achieve. But I've tried to enable this multiple different ways and nothing has worked. The closest I've been able to get is that I think the retreating units were able to cross the border, but their pathing algorithm just has them run along the edge instead of actually leaving.

Hypothetically, if this could ever be enabled I can still see some serious abuse cases. Spawn camping the AI, pinning it in corners, and getting a rolling rout going where you can drive them wherever you like are all currently very easy to do. If you can just drive them off the board I think many battles would become very easy. Larger maps would mitigate this issue a little, but that isn't possible either at the moment.

Edit: have a few new ideas to try, will see where they lead.

Edited March 27, 2019 by pandakraut

[Minas Moth]

@pandakraut

Regarding the artillery in the mod... I think Napoleon 12 pdr and 3-inch Ordnance Rifle should be reworked (buffed a bit).

1. The Model 1857 gun (or Napoleon) was the most popular and dependable gun. its best characteristics were that it was effective, reliable and very easy to maneuver. It had a range o 1.600 yards (1.463 meters) at five degrees elevation and optimal effective range was 1.200 yards (1.097 meters). It was also classified as gun-howitzer because of its shorter than standard barrel and light weight. For example, Spherical case shot at 3,45 degrees elevation could reach effectively to 1.135 yards (1.038 meters) and Shell at same elevation could reach up to 1.300 yards (1.189 meters). Also, a solid shot was quite effective as the "ball" would bounce of the ground and continue to injure and kill people even beyond it's effective range. Especially it has to be considered that it was the most popular gun of the war as it had best performance for its cost.

2. The Model 1861 3-inch wrought iron rifle (or 3-inch ordnance rifle) was probably second in popularity (right after Napoleon 12 pdr). It fired Schenkl and Hotckiss shells to as far as 2.000 yards (1.829 meters) using a one pound powder charge. it could achieve maximum elevation of 16 degrees and could fire up to 4.180 yards (3.822 meters). And it was famous for its accuracy. It also wasn't the most expensive artillery piece out there and it was a great at counter battery fire. 

Basically what I mean is... increase the performance of both the guns a bit. Increase the prices and lower availability of other artillery pieces.

 

Edited March 28, 2019 by Minas Moth

[pandakraut]

6 hours ago, Minas Moth said:

Regarding the artillery in the mod... I think Napoleon 12 pdr and 3-inch Ordnance Rifle should be reworked (buffed a bit).

It's on our list. The problem in both cases is the range degrade curves are terrible for these weapons. In the base game the curves were only acceptable because of the long range(compared to other smoothbores) of the Napoleon and the huge damage range compared to other cannon of the 3in. I'd actually argue that both cannon are bad in the base game as well. Outside of a few very large or early battles where you don't have enough there is almost no reason to bother using them in place of 24pdrs and 20pdrs. As the CSA you'll need to use something else as you won't have enough 20pdrs but the Tredegar and 10pdr parrot are going to perform better in most cases as they share the 3in's huge damage range but have more useful curves. 

We're probably going to end up swapping curves around to fix this. For example, swapping the 6in and napoleon curves and swapping the 10pdr parrot and 3in ordinance.

[Minas Moth]

@pandakraut so there is no way to change the curves themselves?

 

[pandakraut]

23 minutes ago, Minas Moth said:

so there is no way to change the curves themselves?

Not without access to the original unity project. They can be replaced with a formula which I've been trying to limit for a variety of reasons. If swapping the curves doesn't work, as it sometimes won't for technical reasons, then the last resort is creating a new curve in excel.

[nik_nv]

21 hours ago, pandakraut said:

I've tried to enable this multiple different ways and nothing has worked

I have an idea. Its supposed to be there a cap stating if when the unit is destroyed (and become that retreating crowd moving out of the field), I guess it's 5%. You can try to find it and increase.