Dynamic Campaign Idea

[Lincolns Mullet]

Just a fun design theory on how a dynamic campaign might be implemented using the existing game. Maybe it could work, maybe not but I find this kind of stuff fun to do and what harm is there starting a discussion on "what if"? Would love to hear any and all additional ideas! A dynamic campaign could exist alongside the current campaign structure the devs are working on, offering two distinctly different experiences.

Vid: 

 

[Don't Escrow Taxes]

Key location and national morale concepts are very smart, also I like the resting vs. movement options per turn and their corresponding impact on national morale.  Bonuses for the key location battles could stands of arms, money, manpower, a free unit of veteran militia joins you, etc.  This would ensure each campaign is different every time.  Also having a random direction of attack / defense in each battle would add to replay-ability and ensure it's different every time.  The capitol idea is interesting because in the real war, Lincoln and Davis were constantly concerned about the defense of their respective capitols. 

Fun video, I enjoyed it.  Good resume, too .

 

To be the biggest thing I would want to see here are the random starting points for the battles, depending on what direction the armies are approaching from on the campaign movement map.  Hopefully there would not be an impossible amount of programming needed, in which case I can understand why these ideas cannot be implemented.  Also if you choose to make your army idle for a turn while national morale might drop you might see more recruits filter into camp or casualties from previous battles heal up.  

 

[Butch]

Outstanding ideas man I can tell you put a lot of time and thought into making this video. I loved every single proposal particularly the random deployment zones in each battle and the unpredictability of when your reinforcements arrive. This was the essence of UG Gettysburg in the sense that reinforcements could come early or late and it changed the way you played the battle every single time. I loved the idea of having different 'variables' affecting when a corps arrived into battle; I'm gonna make a new thread with some ideas for an officer trait system that could actually go hand in hand with this proposal. I really hope some of this stuff makes it into the final version! 

[Lincolns Mullet]

DET:  Thanks, glad you enjoyed it!  Good ideas here too.  As far as directional attack/defense on the strategic map, I considered this as well.  In order to do that, each playable map on the grand map would need to be in the shape of an octagon and all of them connected to each other.  So if my army resides in Antietam, and the enemy army attacks me directly on my left, then the battle map would load exactly that way (attackers on left, defenders on right).  Certainly doable but would add a bit more complexity to the design of the grand map.  While randomization may not make as much sense (I'm being attacked from the north, but our starting positions are the exact opposite on the battle map), it simplifies how the maps are connected to each other on the grand map.  And in a way, you could justify the random setup by imagining that prior to the battle starting there was some maneuvering as the attacker converged on the defender.  But, a more predictable system would be making all the maps octagons on the grand map so that the direction of an attack translates into where each army is positioned on the battle map.  That would be cool too!

And yes, "resting" would grant some sort of pro and con to doing so (national morale drop, but also an increase in gold and men, or healed casualties returning as you mention, etc).  The existing skill tree already has all the elements to support these new designs as well.  If you rest your army, for example, your medicine skill automatically adds 2% or 6% or whatever your level is back into your existing brigades (but only if they aren't maxed out yet, which they probably wouldn't be if you're trying to recover from a major battle).

Butch:  My mind is always churning, and it helps that I love UGG and UGCW and think Game Labs has a bright future in wargaming ahead of them! Would love to see your thread on the officer traits. I think the basic idea I had would be that each officer has 1 pro and 1 con.  If you have Pope leading a Corp, for example, his con may be insubordination (you never know when his corp will arrive or how many brigades he'll release) and his pro is Command (+20 across all units in his Corp).  So you have to take the good with the bad with every officer who leads a Corp. In Pope's case maybe he is "late" one too many times and you sack him, even though his command bonus is a huge help once he does take the field.

The variables on the battlefield are what makes it so excited, just as long as both sides are subjected to the same kinds of variables to make it fair.  Sometimes luck goes your way, other times it doesn't!

[guidon101]

@LincolnsMullet

First of all, great ideas. I can tell you put a lot of thought and effort, and love in the game and improving on it. Also grateful to your resume, as I enjoyed many of those titles you've worked on.

I like thinking and discussing ideas like yours. I am an aspiring game developer, and in this case I had some questions, comments and ideas to offer: 

1) Enemy Army continuity and scaling -- I watched your vid and tried to read the white paper, but I feel like I didn't get a good sense of what the enemy army would look like and how they would evolve from battle to battle? e.g. Let's say the initial enemy army has 10k, you inflict 8k loss in one battle...will you face a 2k army the next battle, or will the enemy also be replenished in a some sort of scaled way relative to some variables (national morale, etc.)?

Will enemy armies see veterancy, equipment, and commander upgrades the way the player army does?

This touches on a bigger overall design question of: How would the difficulty scale as the game progresses? In the current design, in my opinion, it seems the game difficulty curves upward, with large spikes at the Grand Battles... later battles more or less are generally more difficult than earlier battles, usually due to scale (managing 1 division at first, then later multiple Corps)

If we examine how a Total War campaign plays, there is usually a point in the campaign where the player reaches "critical mass" about half-way through a campaign; from that point onward, Victory is pretty much assured, and it's only a matter of time. If anything, campaigns at that point become tedious clean-up of enemy forces off the map. The difficulty is in the early stages of a campaign, and then the difficulty peters out after mid-campaign. Arguably, in most "dynamic" campaigns or battles in games, this power curve phenomenon can be observed; the games that avoid that kind of power curve are less dynamic and more "story-like" "on-rails" type experience or maintain the difficulty ramp in contrived ways, like in this game.

I saw how you really thought about the balance of both sides in terms of reputation and post-battle reward calculations, etc. So I wonder how you see the enemy power curve play out relative to the player. I fear it might play out like Total War campaigns, where a campaign is pretty much decided well before the war is officially over, which makes the last half of a campaign usually not very fun due to the lack of challenge or tension.

2) Really love the "dynamic" battle scenarios idea with more or less unpredictable deployment, orders of battle, and reinforcements, etc. The only thing I might add is to add some random scripted events or deployment configurations. For example: (a) Similar to the Shiloh battle Union-side, your troops are "caught with their pants down" in encampments, while the enemy deploys a full-scale assault, or vice versa; (b) some of your troops got lost and entered on the "wrong side" of the battle field; (c) enemy forces execute a flanking maneuver outside of the current battle map, so some of their forces enter on "your side" of the field

3) While I would certainly love to see a more dynamic campaign, I'm not sure how well your proposed Campaign map mechanics would scale with more battle locations. I like the "key location" idea, but I wonder if that would be enough to encourage exploration and variety in battle locations.

I can't speak for how others play, but if there is an arbitrary turn-limit on a campaign game (personally, I don't like arbitrary imposed time-limits, but alas that is a reasonable and game mechanic), I tend to go for the most effective/efficient route... So someone like me, if I'm playing to win, would probably end up always playing on a few battle sites in the beginning areas, and the only reason I would go to the other areas would be to "draw out the game longer". I would always play for the favorable grounds, draw the enemy army there to a meat-grinder session, then just play with the enemy army afterwards like a cat playing with its mouse dinner. This is how I see your current Campaign map mechanics play out in my head because there is only one enemy army.

This is what happens in Total War campaigns in 1v1 army maneuvers -- it usually only takes 1 decisive battle to seriously cripple an army, which makes sense. Which begs into question...

4) Multiple armies on the campaign map -- While I would also much prefer managing only one personal army on the field, having a Campaign map that you are proposing with 2 mobile opposing armies begs into question: Why is there not another enemy army? Or, why wouldn't I be able to split my army?

Having mobile Armies on a campaign map might be a step too far from the current abstraction of the campaign map because seeing two armies move around the map seems a bit counter-immersive and counter-intuitive-- why is there only one army per side, you know? Unless maybe you justify it by limiting operations to Theatres, or something, which contradicts the purpose of having the wide breadth of the country available for battle sites.

 

I offer some possible alternatives or food for thought for points #1 and #3/4:

Alternative 1: I propose a dynamic campaign design with "rogue-like" concepts in mind... keep all the concepts you propose except for the mobile Armies. Instead, have a quasi-procedurally generated campaign with multiple branched and linked quasi-procedurally generated random scenarios. Enemy force presence is defined as parameters in each scenario, based on whatever procedural algorithm for the campaign generation. The idea is to abstract Army maneuvers while giving sufficiently randomized/unique battle scenarios, and maintaining the ramp of campaign difficulty by having the scenario-generation procedure increase weighted difficulty parameters to avoid the "player critical mass effect" as the campaign progresses.

Let's say as Union, your goal is to start with Battle A, and to win your campaign, you must win Battle Z. But to get to Battle Z, you may have different quasi-randomized battle sequences available on the same campaign. You could do:

A -> B -> C -> Z

or

A -> C -> D -> B -> Z

etc.

In the two possible campaign paths above, if you fight Battle B early in the campaign, it will be "easy" in terms of scaled difficulty. If you fight the "same" Battle B (some but not all initialization parameters will be the same) later in the campaign, Battle B will be "harder" in terms of scaled difficulty.

The first sequence of battles will lead to an earlier victory, compared to the second sequence of battles.You could have a short campaign or a long campaign depending on the sequence of battles you choose, but the campaign won't be cut short or made easier by "destroying the enemy army" (this can be a positive or a negative, depending on player preference). In this case, you never really "destroy the enemy army" -- you just win the war by winning the final Battle Z, but the battles you fight to get there would be different in each campaign.

Pros/Cons:

- remove mobile Armies on a campaign map

- generated scenarios can be weighted to increase in difficulty as the campaign progresses (maintain or increase challenge throughout the game)

- enemy forces are pre-determined which takes away from the player experience of "destroying the enemy army"

- more or less unpredictable sequence of battles and battle configurations

 

Alternative 2: Keep all the concepts you propose, including mobile Armies. Add the concept of having "multiple armies" on the enemy side, like "Army lives", so to speak. After an enemy army has suffered enough casualties, let's say they are reduced to a threshold of 30% of original troop strength, that Army is "destroyed" or disbanded. But another, more powerful and more numerous Army will emerge and face you on the field. Let's say we have up to 3 enemy armies of increasing power in each campaign (and add a difficulty option to modify this value). To win the full war, you have to fight and "destroy" all three armies, one at a time.

Not a true history buff here, so hope I don't offend with the difficulty scaling I propose below, but for example, let's say I'm playing CSA campaign, I'd go up against the following Union Armies:

1st Union Army: McClellan: 25,00 Infantry, 1000 Cavalry, 100 guns

2nd Union Army: Meade: 50,000 infantry, 10000 cav, 200 guns

3rd Union Army: Grant: 100,000 infantry, 20,000 cav, 500 guns

After I "destroy" one army, the next one appears. This basically just extends the lifetime of the enemy (which needs to be balanced relative to the campaign length) but I think it adds the feeling of "I finally destroyed McClellan the pushover, but now I have to face Grant..." Goosebumps, hehe.

It helps immersion knowing the damage you inflict on the enemy persists, but then you also have to be prepared for the next stronger Armies that the enemy nation will send. This also somewhat alleviates the question, "Why only 1 army in the war?"

This might also keep the difficulty curve going up as the campaign progresses, as opposed to going down after whittling down only1 army. The downside is that there will be severe difficulty spikes when a new stronger army enters the campaign, but can you imagine the anticipation of preparing to fight Grant in this context? Sort of like the way in Early Access we anticipate fighting at Antietam and prepare for that "final battle" from the very beginning of the campaign  

Sorry for the long wall of text, but I do enjoy discussing this in detail, and I hope some of these ideas are adopted if not by this game, at least by others. The game is great as it stands, and it would take it a whole other level to add dynamic campaigns and scenarios for practically limitless replayability.

[fallendown]

This is pretty cool, but wasn't this how the campaign in CWG2 worked?

I had envisioned something much more dynamic. As in the TW series, I'd have the historical battles completely separate from the campaign. You can create a generic height map of say the eastern US and import it into Unity3D fairly easily. Populate that map with foliage, roads, towns, rivers, etc and you have a nice scalable campaign map. You can also create grid files and cut those into multiple map pieces of higher resolution and detail and load them dynamically depending on the location. This would be used for the actual battles which wouldn't be conducted on the campaign map, but on the smaller maps drawn from grids. This whole idea would allow you to play the campaign free of historical battles and allows you to conduct the campaign the way you see fit, not unlike TW. This is actually the game concept I've been working on, but I got busy doing other things. 

[Grognard_JC]

Having played the dynamic campaigns of Matrix Games Pike and Shots, I must testimony that it looks very good on the paper, but it gets very boring once you've played some campaign against the AI, because the AI usually makes it predictable. Therefore, going full dynamic like this, or going free dynamic (like in Total War games) may look appealing at first but they're, in truth, less appealing in the long run, than what we already have in the early access.

Therefore, I put much hope in Lincoln's Mullet proposals. We could have this semi-dynamic campaign too.

The devs must be thinkig we are mad...

Edited November 29, 2016 by Grognard_JC

[guidon101]

12 minutes ago, fallendown said:

This is pretty cool, but wasn't this how the campaign in CWG2 worked?

I certainly had that game in mind as inspiration. Used to play it a lot a few years ago and modded the campaigns/scenarios for replayability. CWG2 campaign had branching and linked battles, yes, but the branches weren't "randomized" per se and neither were the scenario setups, if I recall correctly. You could win the war in just several battles if you kept winning decisive victories, etc, but the branching paths are predefined in the vanilla game (you can edit it through modding though)

Correct me if I'm wrong, but I also remember the CWG2 scenarios being somewhat more static in design (predefined Orders of Battle, objectives, timer, etc.). If I remember correctly, there was a lot of discontinuity in both the enemy forces and your own... in the scenario editor, you can see which battles are linked to other battles based on Victory/Draw/Defeat, and you can also see which units are flagged to "carry over" -- but I found the "carry over" mechanic inconsistent or broken. Might have just been me and not knowing what I was doing/seeing at the time.

Otherwise, yes, I think some of the "dynamic" campaign concepts from CWG2 would be great to see here, although I wouldn't like the idea that I won't get to see all the battle locations if I do too well with Decisive Victories, and I'd have to try to "find" specific battles that I want to play based on the branching campaign choices, as opposed to being exposed to optionally fight each scenario available throughout the course of a campaign.

 

28 minutes ago, fallendown said:

I had envisioned something much more dynamic. As in the TW series, I'd have the historical battles completely separate from the campaign. You can create a generic height map of say the eastern US and import it into Unity3D fairly easily. Populate that map with foliage, roads, towns, rivers, etc and you have a nice scalable campaign map. You can also create grid files and cut those into multiple map pieces of higher resolution and detail and load them dynamically depending on the location. This would be used for the actual battles which wouldn't be conducted on the campaign map, but on the smaller maps drawn from grids. This whole idea would allow you to play the campaign free of historical battles and allows you to conduct the campaign the way you see fit, not unlike TW. This is actually the game concept I've been working on, but I got busy doing other things. 

I know what you're talking about working with Unity and using sectors, etc. At that scale though, you really are making a Total-War-esque game, with a dynamic campaign map and separate tactical battles. However, at that point, one might consider the alternative: modding or using existing Civil War Era-like mods in Total War games, albeit you'd have the Total War battle engine and BAI, which is ironically what Darth was known for making not-suck.

Whatever you were working on sounds awesome, and I hope it sees the light of day (or picked up for Ultimate General 2 :).

[fallendown]

The tactical battles on TW are so small. I mean whats the max scale of those maps? Maybe 5 sq miles? I feel so confined when I play Napolean TW or Empire TW. You have very few  options for maneuver warfare. I have to admit though I was extremely mad when TW went full retard with Warhammer, so many of us were expecting a Civil War TW game.

 

[Caesar15]

5 hours ago, fallendown said:

This is pretty cool, but wasn't this how the campaign in CWG2 worked?

I had envisioned something much more dynamic. As in the TW series, I'd have the historical battles completely separate from the campaign. You can create a generic height map of say the eastern US and import it into Unity3D fairly easily. Populate that map with foliage, roads, towns, rivers, etc and you have a nice scalable campaign map. You can also create grid files and cut those into multiple map pieces of higher resolution and detail and load them dynamically depending on the location. This would be used for the actual battles which wouldn't be conducted on the campaign map, but on the smaller maps drawn from grids. This whole idea would allow you to play the campaign free of historical battles and allows you to conduct the campaign the way you see fit, not unlike TW. This is actually the game concept I've been working on, but I got busy doing other things. 

Issue I see with this is that what we have now, very detailed maps for very specific battles, will all be lost. All the maps are very precisely drawn, and if there was some sort of grid based system based on terrain as a whole, I feel that the end result would be a lot more free range for the player, but at a loss of the maps being a lot less detailed and more generic. Also that sounds quite a bit of work for the dev team, but you seem to know more than me so I won't argue that point.

[Lincolns Mullet]

Great discussion! I love hearing different ideas and proposals.  Game design is a fantastic mental exercise, and it's like playing the game you'd love to see but only have it in your head (and yes, the devs probably think we've gone mad lol).  Admittedly I never played CWG2. The only two civil war games I've played at any length was North vs South (Some cheapy PC game from 1988 or something) and Sid Meier's Gettysburg/Antietam. I'm a huge fan of Sid in general.  He had that special knack for knowing how to make a game "fun". You could tell he loves what he does for a living and is a true gamer at heart.

Guidon: I'll reply as best I can to your nicely detailed post.

1.  How would the enemy AI army appear on the battlefield?  This part I didn't touch on in the video because I didn't fully detailed it out yet.  There's at least a few ways to do it, depending on how detailed and programming intensive you want it to be.

One way would be to have the enemy AI army mirror exactly how you manage your own army.  They create corp/divisions/brigades which are "saved" and carry over from battle to battle. They receive rep/gold/men just like you do through battles and/or resting.  There could be added differences specific to the Union and Confederate sides, so you're not fighting a symmetrical war where the uniforms are the only things that are different (Union gets more resources and men, but Confederates rookies have more base experience than their Union counterparts, etc).  However, creating an Army Camp for the AI would probably be fairly programming intensive.  But, the "information" about the enemy AI army would basically be an excel file organizing the army composition by Corp/Division/Brigade, the info of which is updated and saved from battle to battle.  When the AI army deploys into battle, it draws the relevant info from this file for the Corp names/structure, unit info, weapon type, total size, etc and then created as a unit on the battlemap.

IE: 1st Division: Cooke (1500 men, springfield rifles, etc), Kemper (1288 men, farmers rifle, etc). Each unit would have its corresponding stats tracked in a master database and referenced whenever the game needs to create the unit as a physical object on the battle map.  Think of how AI reinforcements appear during a battle now. Each unit has specific stats stored somewhere so when they appear, they have specific rifles, numbers, officers, etc. and all organized into a division.

After a battle, the AI receives gold/men to then replenish losses and/or build additional brigades.  The AI programming to control that aspect would probably be the hardest part, but I'm no programmer, I only knew one a long time ago. lol  In your example quoted below:

Quote

e.g. Let's say the initial enemy army has 10k, you inflict 8k loss in one battle...will you face a 2k army the next battle, or will the enemy also be replenished in a some sort of scaled way relative to some variables (national morale, etc.)?

Both sides will receive gold/men/rep after the above battle, and are allowed to rebuild their army prior to the next engagement.  If it was the enemy army that attacked the last battle, it would now be my turn to either attack, move, or rest.  We can still use the existing system in UGCW to award resources and men, but with the added ability to rest as a safe way to rebuild your army (but not with huge amounts...successful, large battles will still be the best way to get more men/gold).

In terms of the power curve, I think that's inherent to any domination type game.  For the purpose of my idea, it may be more important to emphasize morale while lowering casualty rates.  Nothing too dramatic, but only to prevent one massive battle from single-handedly destroying the fighting capability of your opponent.  Inflicting 15k casualties on your enemy in a single day, for example, might be considered extremely high under my proposal.  This way you can't wipe out 50k enemy troops in a single battle and eliminate their ability to fight any subsequent battles.

Then gameplay becomes less about annihilating your enemy (ala Total War), but rather achieving enough significant victories to force your opponent to surrender/sign peace treaty.  By reducing mass-casualty battles, and focusing more on morale, this will allow a defeated army to defend a follow-up attack and have a chance to win. New map, new conditions, new reinforcements, etc.  Your ability to fight will remain intact long enough that you'll likely "lose" the game because you've lost too many battles, not because your army was outright destroyed.

2. Yes, I like that idea too. Elements of unpredictability are key!  They also have to be handled in a way that's not a constant, ruinous occurrence for either side.  The focus should still be on management of your armies, but with some of these unpredictable events/circumstances forcing you to think on your feet from time to time.  I like the idea of a "pool" of scripted events that could occur at the start of the battle, depending whether you are on attack or defense like the examples you gave of being surprised in an attack, etc.  Used sparingly, of course!

3.  Excellent point, and my "key locations" idea was a simple remedy to prevent fighting over the same maps over and over.  There's a more graceful way to spread out the playing field, so to speak, but I haven't thought on that too much yet. If you attack and defeat an enemy, for example, it would push them "back" a map.  An incentive or benefit needs to be dangled out there for each army to maneuver around the map as much as possible. Maybe controlling specific map locations increases national morale, increases cash/gold benefits, etc. so that slowly losing those locations weakens you long term.

4. I only envisioned 1 army to try and stay in line with the original vision of the devs that you are solely responsible for one army.  My proposal would focus on a specific theater you operate within, and while it simplifies the actual conflict in that area, it provides for a dynamic campaign within the boundaries of the current engine.  Or maybe a weee bit past the current boundaries. lol.  By creating additional armies, you may as well open up the entire theater of operations of the Civil War. Which, I suppose if they can make 1 moveable army on a map, then they can make multiple armies.  They'd have to create a separate army camp screen for each army.

Phew, that's it for now!  Need some sleep lol

[guidon101]

12 hours ago, Lincolns Mullet said:

the devs probably think we've gone mad lol

THIS IS SPARTA, Illinois. Some of us have always been mad lol. The good news is Darth/Dartis responded on a Steam thread-- I can't find it now... something about addressing the issue of enemy force continuity or long-term campaign impacts from player actions "ASAP"! Not quite all-out dynamic campaign, but still some promise in a good direction, as this is a concern very commonly voiced in the playerbase. So these discussions are not for naught.

 

14 hours ago, Caesar15 said:

Issue I see with this is that what we have now, very detailed maps for very specific battles, will all be lost. All the maps are very precisely drawn, and if there was some sort of grid based system based on terrain as a whole, I feel that the end result would be a lot more free range for the player, but at a loss of the maps being a lot less detailed and more generic. Also that sounds quite a bit of work for the dev team, but you seem to know more than me so I won't argue that point.

Definitely would like to keep the current campaign and scenarios! I like the current "historical campaign" as it is, but for more replayability and variation I'm with Lincoln Mullet's proposal: "A dynamic campaign could exist alongside the current campaign structure the devs are working on, offering two distinctly different experiences. " Historical Campaign (as-is) and a Fictional Campaign (dynamic). Either one is epic enough, both would just blow my mind.

 

12 hours ago, Lincolns Mullet said:

In terms of the power curve, I think that's inherent to any domination type game.  For the purpose of my idea, it may be more important to emphasize morale while lowering casualty rates.  Nothing too dramatic, but only to prevent one massive battle from single-handedly destroying the fighting capability of your opponent.  Inflicting 15k casualties on your enemy in a single day, for example, might be considered extremely high under my proposal.  This way you can't wipe out 50k enemy troops in a single battle and eliminate their ability to fight any subsequent battles.

Then gameplay becomes less about annihilating your enemy (ala Total War), but rather achieving enough significant victories to force your opponent to surrender/sign peace treaty.  By reducing mass-casualty battles, and focusing more on morale, this will allow a defeated army to defend a follow-up attack and have a chance to win. New map, new conditions, new reinforcements, etc.  Your ability to fight will remain intact long enough that you'll likely "lose" the game because you've lost too many battles, not because your army was outright destroyed.

Wow, I think that's it! I've been struggling about the idea and asking other people on Steam how they see balancing the enemy force to maintain difficulty if the enemy has to maintain a "persistent army", and I think this is the best idea I've heard, and it coincides with the points brought up in another thread about how Casualty Rates are too high currently, and how some combination of morale and combat effectiveness need to play more of a role than Casualty Rates do now.

 

13 hours ago, Lincolns Mullet said:

One way would be to have the enemy AI army mirror exactly how you manage your own army.  They create corp/divisions/brigades which are "saved" and carry over from battle to battle. They receive rep/gold/men just like you do through battles and/or resting.  There could be added differences specific to the Union and Confederate sides, so you're not fighting a symmetrical war where the uniforms are the only things that are different (Union gets more resources and men, but Confederates rookies have more base experience than their Union counterparts, etc).  However, creating an Army Camp for the AI would probably be fairly programming intensive.  But, the "information" about the enemy AI army would basically be an excel file organizing the army composition by Corp/Division/Brigade, the info of which is updated and saved from battle to battle.  When the AI army deploys into battle, it draws the relevant info from this file for the Corp names/structure, unit info, weapon type, total size, etc and then created as a unit on the battlemap.

It is not a coincidence that I used to have a personal program that did something exactly like that by using elements of two different games: the campaign elements from AGEOD Civil War 2 with the real-time tactical battles of Take Command: Manassas. Basically I would play out a campaign in AGEOD CW2, and when a battle occurs, I extract the army composition (Order of Battle) and convert it into an Order of Battle file that feeds into Take Command: Manassas, where I would play out the battle in real-time, and then extract the resulting casualties, etc. to re-import back into the AGEOD CW2 campaign. So the "persistent army data" was maintained in AGEOD CW2, basically a csv file that can be treated like an Excel spreadsheet, as you mentioned. (but that was a terribly hacky way to do it...I suspect in this game built from Unity, Darth created special object classes, so there is a hierarchy of classes and attribute variables stored much more neatly in the game files/memory, which they could expose in XML files or other openly readable format, if they wanted).

In this case, Rough Order of Magnitude for what you suggest, it would require creating a centralized "persistent army database", so the "programming" of that in itself isn't hard (since they already have all the data class structures to work with). They would just need to basically change the "persistent data location/reference" of enemy army composition to be associated with a Campaign, instead of stored disparately in each Scenario, and have each Scenario reference the Campaign persistent data on demand-- so likely the toughest part about that actually is designing a good data structure and addressing the gameplay balance ramifications that comes from that persistent army data, which implies an Army Camp AI or a "Campaign AI", which I agree would be the toughest part from a programming and AI design perspective.

But an alternative to an AI Army Camp is just a plain scaling formula for troop replenishments and experience/equipment upgrades based on some factors (e.g. National Morale, difficulty level, National Manpower, etc.) For example, a linear model for AI "power curve growth", like let's say on Normal difficulty, all else equal, AI forces grow/replenish at a rate of 10% or let's say 1-2 new full brigades per turn/battle depending on "national factors", just as an example. Enemy brigades that survive a battle get a notch of XP, and maybe equipment upgrades on a weighted scale, but new brigades start off green with basic equipment.

So in the context of the above scaling, if the player doesn't inflict more than 10% casualties or destroy 1-2 brigades per battle, then they would see the AI army grow relatively in total size. Something simple like that I think might work to abstractly substitute from a proper Army Camp AI, and would be relatively easier to implement.

You could also get creative with making the "simple scaling" more dynamic depending on Campaign decisions, events, and player progress (e.g. Emancipation Proclamation event causes Union ranks to swell with green troops; Sherman's raids causing new CSA militia units to form, etc.; CSA winning Gettysburg gives their soldiers better shoes, etc.  

Food for thought

 

[Barry Maynard]

 You don't have to have a campaign showing the entire war. Many people do not realize that it was a war of maneuver. You could have mini-campaigns focused around certain battles. Check out Great Campaigns of the American Civil War by Multimanpublishing. Divisional level units with Corp and Army leaders. Cavalry really shines at that scale.

[Lincolns Mullet]

9 hours ago, guidon101 said:

THIS IS SPARTA, Illinois. Some of us have always been mad lol. The good news is Darth/Dartis responded on a Steam thread-- I can't find it now... something about addressing the issue of enemy force continuity or long-term campaign impacts from player actions "ASAP"! Not quite all-out dynamic campaign, but still some promise in a good direction, as this is a concern very commonly voiced in the playerbase. So these discussions are not for naught.

Yes, and don't forget people like myself, Koro, CSA Watkins, VegasOZ, etc are beta testers and have the ear of the devs.  We can make very strong cases for certain changes, additions, etc. It's still up to Nick and the devs of course whether to make a change or addition, but we like to think we can help bring the "voice of the people" to them in a concise manner.  Nick of course prowls the forums and likely sees the same complaints all of us see too, with or without the beta testers chiming in. But I imagine he spends more time working on the game than reading about it lol.  In regards to that above matter, it's been discussed before and now it will be interesting to see how/when it is implemented and how that'll affect overall gameplay. Personally, I think it'll help a lot in immersing the player in the overall campaign.

9 hours ago, guidon101 said:

Wow, I think that's it! I've been struggling about the idea and asking other people on Steam how they see balancing the enemy force to maintain difficulty if the enemy has to maintain a "persistent army", and I think this is the best idea I've heard, and it coincides with the points brought up in another thread about how Casualty Rates are too high currently, and how some combination of morale and combat effectiveness need to play more of a role than Casualty Rates do now.

The only possible issue with this is that the game, as it stands currently, doesn't have sufficient feedback for morale on individual brigades. Take Sid Meier's Gettysburg, for example. Each regiment had morale bars that would stack higher and higher depending on the "bonuses" they received.  These bonuses were very specific: friendly regiment on right flank, friendly regiment on left flank, nearby officer, veterancy, etc.  Each unit also had a flag above it that, when at 100% morale, stood fully unfurled.  As morale dropped, the flag would sink and look sad and pathetic.  This provided an easy, visceral way to see the morale of your entire army and specific regiments in a single glance without having to click on every single brigade.  Since UGCW doesn't focus on morale as much as it does casualty rates, this added morale feedback isn't entirely necessary because physical manpower loss is the biggest determiner of success.  And this you can currently see very easily as the battle rages.

I'm not exactly sure how morale is handled in UGCW other than an individuals morale stat, nearby officer, and how many casualties they take per volley.  In other words...I don't know if having nearby friendly units adds morale bonuses.  From game experience I feel morale importance goes in this order: individual brigade morale (1-100), rate of casualties, nearby officer.  If morale became much more important to the success of a battle, we'd need better feedback of the morale of your entire army ala Sid Meier's Gettysburg so you'd know which brigades to pull back and which ones to plug the gaps, or when to begin withdrawing altogether if your entire line is wavering.  But yes, the general theory would be: "Greater morale sensitivity = less overall casualties".

The alternative, if casualty rates remained the same, was to provide greater men/gold to each side when they win/lose a battle, or rest.  The historical numbers on how many people died might be greatly exaggerated, but the game would need to be balanced towards the idea you could lose 40k men in a single battle.  The player or AI would need the ability to replenish at least half those losses between battles in order to stay competitive.  Realism = greater emphasis on morale or Gameplay = greater emphasis on casualties.  Either way can work.

9 hours ago, guidon101 said:

It is not a coincidence that I used to have a personal program that did something exactly like that by using elements of two different games: the campaign elements from AGEOD Civil War 2 with the real-time tactical battles of Take Command: Manassas. Basically I would play out a campaign in AGEOD CW2, and when a battle occurs, I extract the army composition (Order of Battle) and convert it into an Order of Battle file that feeds into Take Command: Manassas, where I would play out the battle in real-time, and then extract the resulting casualties, etc. to re-import back into the AGEOD CW2 campaign. So the "persistent army data" was maintained in AGEOD CW2, basically a csv file that can be treated like an Excel spreadsheet, as you mentioned. (but that was a terribly hacky way to do it...I suspect in this game built from Unity, Darth created special object classes, so there is a hierarchy of classes and attribute variables stored much more neatly in the game files/memory, which they could expose in XML files or other openly readable format, if they wanted).

In this case, Rough Order of Magnitude for what you suggest, it would require creating a centralized "persistent army database", so the "programming" of that in itself isn't hard (since they already have all the data class structures to work with). They would just need to basically change the "persistent data location/reference" of enemy army composition to be associated with a Campaign, instead of stored disparately in each Scenario, and have each Scenario reference the Campaign persistent data on demand-- so likely the toughest part about that actually is designing a good data structure and addressing the gameplay balance ramifications that comes from that persistent army data, which implies an Army Camp AI or a "Campaign AI", which I agree would be the toughest part from a programming and AI design perspective.

But an alternative to an AI Army Camp is just a plain scaling formula for troop replenishments and experience/equipment upgrades based on some factors (e.g. National Morale, difficulty level, National Manpower, etc.) For example, a linear model for AI "power curve growth", like let's say on Normal difficulty, all else equal, AI forces grow/replenish at a rate of 10% or let's say 1-2 new full brigades per turn/battle depending on "national factors", just as an example. Enemy brigades that survive a battle get a notch of XP, and maybe equipment upgrades on a weighted scale, but new brigades start off green with basic equipment.

So in the context of the above scaling, if the player doesn't inflict more than 10% casualties or destroy 1-2 brigades per battle, then they would see the AI army grow relatively in total size. Something simple like that I think might work to abstractly substitute from a proper Army Camp AI, and would be relatively easier to implement.

You could also get creative with making the "simple scaling" more dynamic depending on Campaign decisions, events, and player progress (e.g. Emancipation Proclamation event causes Union ranks to swell with green troops; Sherman's raids causing new CSA militia units to form, etc.; CSA winning Gettysburg gives their soldiers better shoes, etc.  

Food for thought

 

Very cool idea!  You essentially bridged two games into a single experience.  Well done, and creative!  To put it simply...if you can do this, then it can be programmed to do it for you as part of the gameplay experience.

Again, I'm no programmer (I only did if/then/else programming, while referencing object classes that were "what you see is what you get") but what you're explaining sounds about right.  Info storage already exists, they'd just need to make it accessible as a global reference for the AI from battle to battle.  And yes, creating an AI that determines which brigades to create, which ones to reinforce with vets/rookies, weapons, etc would be a whole new ballgame.  If I were to try a simple version, I would create ratios the AI referenced against available gold/manpower.  A ratio could be referenced for army composition (1 veteran infantry brigade per 8 infantry brigades : 1 artillery per 4 infantry brigades : 1 cav per 8 infantry brigades : 1 skirmisher per corp), etc.  AI would do a check of available gold/men against current force composition and then expand their army based on above ratios.  In other words, army size would come before army strength (veterancy, weapon type).  Once army size was fulfilled under current Army Org restrictions, they'd turn their gold/men inward by upgrading units by veterancy and newer weapons.  Maybe a variable could be added that any brigade under 900 strength should be upgraded to 1500 first, prior to expanding total army size.  But the goal would be for the AI to have a general rule-set for their Order of Battle and attempt to fulfill that OOB with their available gold/men.

Then, once their army is created, the AI would need to know your general force composition before deciding on attacking or resting.

To expand on your idea of a general "scaling" of AI army growth, that could work too based on a pre-set OOB for the AI.  They can only expand to a certain size (growing larger as time goes on), so your success against them will minimize/mitigate their growth.  Having an OOB based on Army Org level would also prevent the AI from expanding into a huge monstrosity.  They'd have a cap for how many new brigades they get until more time passes, or as you say, historical events pop up.  The scaling would definitely be a nice shortcut in programming terms, anyway.  The game would just have to check what their max OOB/army org is and see if they can add any additional brigades to it (after replenishing their existing brigades).  The Union AI maybe grows at a faster rate than the Confederates, putting the burden on the CSA to win as many early, decisive victories as they can to force peace.  While the Confederate AI grows at a slower pace, so the burden on the Union is to hold on long enough before the CSA wins too many battles.

Good ideas!