• We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.
Ears up folks, it's time for another Legacy of Rome and patch 1.07 dev diary! This time, I'll reveal what we've done to rebalance combat. All these things are actually in patch 1.07, so you'll get them whether you buy the DLC or not.

To give you some background, we were never entirely happy with the combat system in the game; the outcome was too dependent on numerical advantage and there was little you could do to affect it beside throwing more troops into the fray. Part of the problem was simply the lack of proper random elements. This was fixed in patch 1.06 with more varied, decisive and longer lasting combat tactics. Another issue was that the composition of the Holding levies was largely beyond your control; all you could do was try to focus on Cavalry buildings in your own Holdings if you wanted to try having more Cavalry on the flanks in battle. We have addressed this in part with more specific unit type buildings and cultural versions of buildings. The major fix though, is the Retinue system in Legacy of Rome. Lastly, we realized that the most important tactical consideration for players, and something they can influence, is the choice of commander and the effect he has on the battle. In patch 1.06, we added a new type of trait called "Leadership Traits", and now we're adding even more of them:

  • Flat Terrain Expert
  • Rough Terrain Expert
  • Mountain Expert
  • Desert Expert
  • Holy Warrior
  • Unyielding

LoR_03_Holy_Warrior.jpg

The choice of flank leader for the right role is now an important tactical choice with a real effect on the outcome of the battle.

We've also tweaked the combat tactics and unit types a bit more, and made sure that pure archer flanks are not imbalanced, knights are slightly less overpowered, etc. Oh, and all cultures now have a cultural building, but some cultures have the same as others.

Lastly, though not exactly related, I should probably mention that the military AI has been improved a lot, to focus on what's important and avoid attrition.

Hmm, that was a bit short, but it's all I have for you today. Next time: Orthodoxy.
 
Last edited by a moderator:
Would it be more generalised that slecting the commander for the province you wish to invade.
I would imagine if the land traits were more abundant in areas of that type of province then the choice would be, once I have invated say Tunis and got some local lords go native then if I need to war against my neightbours then I may pick the local lords as commanders over the ones from Europe.
that said I never have the CA to appoint leaders
 
you're forgetting that not ALL leaders will have terrain specific traits.
just like not all martial-education leaders even HAVE traits, the terrain-specific traits will be just as randomly assigned, so it really shouldn't be a problem (unless you have close to a hundred vassals with martial traits, then that's a whole different issue anyway and is less of a problem than an advantage depending on your situation)
 
Many Frenchmen at Agincourt, Crecy, and Poitiers would disagree with this.

Tell that to the Persians at Thermopylae. Sometimes strategic considerations override tactical.
You seriously need to watch less TV :)

Thermopylae was a hugely unusual battle which is why people still talk about it almost 3000 years later and make one movie after another about it. In any other circumstances a small force of 3000 men without horse would NOT give battle to a force that is ten-fold larger. They would do hit and run, or, if they chose to defend a strongpoint the way Leonidas did, they would try to slip away during the night once it's clear the enemy is going to force his way through.

My remark about archer fire was a reply to a fellow poster who said suggested that a small force would stand no chance to even get into melee combat against a larger force, because if they got into a battle the "doomstack of archers" (whatever that is supposed to be in reality??) would just hide between a shieldwal or something and destroy the smaller force before they even got close.

This may be how RTS games work but it's definitely not realistic. You don't win medieval battles without melee combat. And in meelee combat there's a limit to how many men you throw bring into battle against a given length of frontage. The Romans lost Cannae in a massive way despite being 3x superior in numbers, because after the Carthaginians turned the Roman flanks the Roman infantry got so constricted on the battlefield that the ranks could no longer rotate and got utterly disorganized. The Carthaginian infantry slaughtered them, one row at a time, one after another.
 
you're forgetting that not ALL leaders will have terrain specific traits.
just like not all martial-education leaders even HAVE traits, the terrain-specific traits will be just as randomly assigned, so it really shouldn't be a problem (unless you have close to a hundred vassals with martial traits, then that's a whole different issue anyway and is less of a problem than an advantage depending on your situation)
Terrain traits are still shit. They're the exact opposite of "tactics" and "random". Their effect is utterly predictable when in fact combat should have a random element to it, that you can influence via tactics.

Your role is to choose tactics that suit the combination of:
- What your force is and what state it is in;
- What the enemy force is, what state it is in and what the personalities of its commanders are;
- what the battlefield MIGHT be (note that you as an overal commander SHOULD NOT KNOW in advance what the battlefield will be, local actions and local commanders decide that)

And so on.
 
Would you rather go back to "the army with most units wins battle"?

I think it will matter very little if you battle desert with a forest commander, less than a trivial matter.
So then what's the point in having that trait in the first place? Also the Dev Diary markets this new system as something that will make battles more fun. So they want it to play a significant role. However I think it sounds like a nightmarish system of micromanagement, that has nothing to do with "tactics".

He's just hoping that the devs will devise a more intuitive system. Being able to save army schemes might help.
Absolutely.

It seems like the new hassle for battle is:

1. Hover mouse over target province to determine terrain type.
2. For each army involved:
a. Select army
b. Click little arrow button that secretly means "change flank leaders" (because that's clearly what an arrow indicates)
c. Click each flank leader, to open the terrible Select a Leader AI.
d. Scroll through the randomly generated characters who you have no influence over, and see if any of them happen to have the appropriate attribute
e. Since there's no way to sort or filter on attributes (but you can totally sort on your leader's age, stewardship skill and first name!), try to remember who you saw with the desired attribute.
f. Compare the other combat modifiers of each candidate leader, and evaluate what effect they'll have on the opaque, convoluted combat system.


How can this possibly be considered fun?
I agree fully with your analysis of what the Dev Diary's described system would lead to. It sounds like an absolutely nightmarshly bad system, something like what they implemented in the initial HoI3 release.

I see several aspects that will make it frustrating, non-immersive and non-rewarding as a game element:

  • In any decent sized army, you have dozens of potential characters for flank leaders. Your selection among these is in no way restricted by their rank, their expertise, their culture, their religion, their prestige or ANY of the many fun things that set characters apart so nicely in the other facets of CK2.
  • With so many existing and new leader traits, there is an endless combination of possible effects that a leader may have on the battle. It used to be that MIL score was the most important thing, which is prominently displayed in the leader selection menu to aid the player. BUT this is no longer the case, many other things now matter. The interface is in no way suited to aid the player with this!!!
  • Once a battle has started, you see the flank leaders and you see the tactics they choose. This is nice but you never actually get feedback on WHY you win a particular battle, unless you pay really a lot of attention. This kills player motivation - it's immensely frustrating that you have to spend so much effort in picking leaders the best way (you think) possible, and then there's nothing that tells you whether this really paid off or whether you won the battle simply due to the enemy having a center wing leader with MIL score 0.

How could this be remedied for Legacy of Rome?

Firstly: Restrict the choice of leaders!! When you have an army full of dukes, princes and counts, you should not be allowed to have a 75 year old 21-MIL mayor lead the left wing, a homosexual heretic MIL-19 count the right wing and your nameless random wrong culture courtier the center. Regardless of crown law, you should be restricted to maybe the top 8-10 guys for flank leaders, picked by a mix of rank, prestige and MIL score but weighted heavily towards rank and prestige. This would make it less of a chore to choose leaders, AND makes more sense from a gameplay standpoint. When you raise an army you would already have to pay attention to who the leaders are, because if you put count X who is a god of war into an army with ten dukes you know he's not going to be available as flank leader.

Secondly: Leave the terrain traits in if must be, but make it so that the terrain of the battle is picked at random (like in EU3).

Thirdly: Make the personality of your ENEMY matter!! Let the player spy who is leading the enemy army wings, BEFORE a battle. Make it so that when you know two of his three leaders are hotheads, you pick experienced commanders for the best effect. When the enemy has a god of war in command, make it so that it's SENSIBLE to pick the most cautious commanders you can get. When the enemy has a cautious guy, pick bold guys or religious enemies who can deal morale damage to his entrenched warriors.

Fourth: Add random battle events! Examples:
- "left wing under Duke X charges enemy without orders -> center and right wings follow". Effect: all three flanks pick a charge tactic.
- "Count X taunts the Emir Y" -> enemy picks charge tactic against shieldwall before his archers have a chance to work your guys over.
- "Battle is interrupted for single combat between champions" -> neither side takes casualties but the side whose champions are killed suffers big morale damage.
- character deaths during battle: Have them be battle events with HUGE influence on morale!! Yeah this is going to be a heavily random element, but so be it. Don't pick rash leaders if you want to avoid this, and don't make your monarch flank leader if you fear his death.
Of course a suitable area for the display of such events has to be added to the battle interface. See my next point.

Lastly: Give the player feedback when a battle is going on!! Spell out what is happening. Add a little area on the side of the battle screen where, like in a chat window, you can read about the events of that battle. So you see, what did Duke X do before he got his flank routed, why is Grandmaster Y's flank still in such good shape; when did they lose my precious knights; how they manage to get my son and heir killed??

Solution: expand the battle window, add a message list on the right side where you can read about significant events that happened in the battle:

attachment.php


battle.jpg (click to enlarge)

It's not like the battle window is too large at present. It's really tiny and takes up almost no space on a widescreen monitor. Use that space and make battles and commanders be FUN!
 
Great Idea Leviathan would love to see that implemented, and is the next dev diary the last one before the expansion is released.
 
While I do think having more unique commanders is a good idea (because frankly all I do these days is try to find my highest ability commander with the "flanker" trait and then put whoever the heck is the highest non-flanker in the center), as others have said, it does look rather like too much micromanagement. While I don't agree with all of Leviathan's ideas, I think you're on the right track, particularly with the battle message system. I mean, right now in the game, all I see is the morale bar going down and the casualty numbers going up and I have no idea what's going on. There has to be a way of making the decisions seem more final and binding, because being able to change commanders every other day is what will turn this into micromanagement - i.e., "Oh, I'm going into a desert in Muslim territory, better find my desert specialist and holy warrior... where are they?" If, somehow, being able to choose generals was given a bit more weight and something that you couldn't change so easily, then maybe these new traits will have much more importance.


Also, it'd be nice if I'm viewing a battle screen once the battle is over the game would pause. I guess some dont need this, but since I tend to run the game at fastest speed it gets kind of annoying.
 
I'm not so keen on your first, second and fourth ideas, Leviathan. The first one seems dreadful to me as a modder-- the scenarios I enjoy working on include periods where some successful and prominent generals were not the most powerful landed nobility, and to hardcode in something that limits general selection to just dukes, courtiers of the exact same culture, etc. and so forth detracts from the scenarios I enjoy working on. Why shouldn't my Roman army be able to use a Frankish courtier as a general? Why can't the lowly patrician of Thessalonica lead the left flank against an Ostrogothic army?

If you really want to emulate the dominance of nobility over the skill of a general, a way that enables more flexibility (so I could still field a Gothic courtier as a general in a Roman army) while making it more appealing for a player or the AI to choose a noble with crap MIL skills would to simply give generalships a relationship score. A Duke who becomes a general could maybe a +5 relation with his liege and a -10 malus (pulling numbers out of the air here) if a count or courtier is chosen over him. That way upper nobility could dominate military leadership at the expense of actual skill, a player who wishes to ignore this and field skillful generals could do so for a penalty, and the choice of military leaders is tied back into the broader themes of gameplay in Crusader Kings II, what with vassal relations now being quite important to leadership selection.

Your second idea could lead to some very awkward moments. It would be weird to fight amongst randomized mountains when the armies are fighting in the heart of Mesopotamia, or to be stuck fighting in the midst of a forest near Medina. There should be limits of terrain randomization, at the very least. What's the point in selecting a general with certain terrain traits if there's no guarantee that said terrain will show up for the battlefield, even if it makes probable sense (like a desert in Nefoud)?

Random battle events could detract from pre-battle micromanaging. After the first few times where my careful pre-battle planning is thrown into chaos because an event causes one of my generals to drown in a river or rushes into battle because of a taunt, I'd probably give up planning altogether and just throw whoever into battle the next time around.
 
To be fair to the developers, I usually don't put much thought into organizing my army and its leaders unless there's a crucial battle coming.
Then, I'll pick my best generals and mold the flanks to their specialties. Alternately, I'll pick generals according to my own specialties (like, heavy infantry for Ireland, or heavy cavalry for Toulouse). It would be easier if I could get tooltips from the selection roster, though.
Only when I'm unsure that I can win a battle or series of battles will I worry about all that. So maybe the micromanagement won't be that bad.

Also, I'd like to repeat my request that the aggressive and defender traits be mutually exclusive. Really sucks to get them both.
 
I'm not so keen on your first, second and fourth ideas, Leviathan. The first one seems dreadful to me as a modder-- the scenarios I enjoy working on include periods where some successful and prominent generals were not the most powerful landed nobility, and to hardcode in something that limits general selection to just dukes, courtiers of the exact same culture, etc. and so forth detracts from the scenarios I enjoy working on. Why shouldn't my Roman army be able to use a Frankish courtier as a general? Why can't the lowly patrician of Thessalonica lead the left flank against an Ostrogothic army?

If you really want to emulate the dominance of nobility over the skill of a general, a way that enables more flexibility (so I could still field a Gothic courtier as a general in a Roman army) while making it more appealing for a player or the AI to choose a noble with crap MIL skills would to simply give generalships a relationship score. A Duke who becomes a general could maybe a +5 relation with his liege and a -10 malus (pulling numbers out of the air here) if a count or courtier is chosen over him. That way upper nobility could dominate military leadership at the expense of actual skill, a player who wishes to ignore this and field skillful generals could do so for a penalty, and the choice of military leaders is tied back into the broader themes of gameplay in Crusader Kings II, what with vassal relations now being quite important to leadership selection.
Even easier solution: 5-6 honorary titles that you can award to any adult male in your realm - "Grand General", "Commander of the Horse", "Magister Militum" and so on - that qualify a character for command.

All titles in CK2 already have a "dignity" value and when you sort a character list by rank the game ranks characters according to that value. So a queen consort comes before a duke, because in the game files the "queen consort" automatic title has a dignity value higher than that of a duke. The same system could be applied (or duplicated) to give you an easy hierarchy of characters. It's not a complicated system at all.

A hypothetical Roman empire would have a ton of such honorary titles so you can basically pick and choose who your generals are. A feudal realm has maybe one such title or none at all so you're not able to "push" people higher in the ladder than they ought to be. Easy system, unlimited moddability, endless possibilities.

Your second idea could lead to some very awkward moments. It would be weird to fight amongst randomized mountains when the armies are fighting in the heart of Mesopotamia, or to be stuck fighting in the midst of a forest near Medina. There should be limits of terrain randomization, at the very least. What's the point in selecting a general with certain terrain traits if there's no guarantee that said terrain will show up for the battlefield, even if it makes probable sense (like a desert in Nefoud)?
Randomized terrain does not mean you get swamps in Libya or mountains in the Netherlands. :) The way it worked in EU3 was that a province had, say, 65% forest, 20% plains and 15% mountains and the dice roll would just determine in which part of the province the battle happens. The % numbers came straight from the map files where these are the "colors" in which the EU3 map terrain file is painted. Basically you look at a province on the terrain map and what you see there within the province's borders is what's possible.

Random battle events could detract from pre-battle micromanaging. After the first few times where my careful pre-battle planning is thrown into chaos because an event causes one of my generals to drown in a river or rushes into battle because of a taunt, I'd probably give up planning altogether and just throw whoever into battle the next time around.
Micromanagement is EXACTLY what I would want to avoid!! Giving the player the possibility to micro each and every battle, and then making it so hard to do it that the player won't bother after the first few tries, is bad game design. It frustrates the player because while he KNOWS he could do even better, he gives up because it's tiresome. Much better to not allow such micro in the first place.

Limiting commander choice to the top 5-10 available people and penalizing frequent commander swaps will put a stop to micromanagement while still giving the player great influence. Do you appoint the experienced old general with mediocre MIL? Or the hot young but inexperienced prodigy? Better check who the enemy is putting in charge and then make your choice. THAT is fun. Not having 1000 possible commanders available and calling that "tactics".
 
I think the choice of tactic every few days by both sides is the "random" element already in the game. Its not as simple as "archer fire has disrupted someone's shield wall", but more of "Count X does a heroic countercharge against the timid advance of Emir Y".

Your role is to choose tactics that suit the combination of:
- What your force is and what state it is in;
- What the enemy force is, what state it is in and what the personalities of its commanders are;
- what the battlefield MIGHT be (note that you as an overal commander SHOULD NOT KNOW in advance what the battlefield will be, local actions and local commanders decide that)
The current tactics system is chosen randomly and is influenced by your army composition and your MA/culture/traits. IMHO it needs to also be influenced by your opponents army composition, and explained a little better.

The current problem is that those tactics give you linear advantages, but numbers scale with their square. Fix that and give tactics an explanation of why they were selected (eg. due to the heavy infantry and brave leader) and I think the combat system would be good.
 
I think the choice of tactic every few days by both sides is the "random" element already in the game. Its not as simple as "archer fire has disrupted someone's shield wall", but more of "Count X does a heroic countercharge against the timid advance of Emir Y".


The current tactics system is chosen randomly and is influenced by your army composition and your MA/culture/traits. IMHO it needs to also be influenced by your opponents army composition, and explained a little better.

The current problem is that those tactics give you linear advantages, but numbers scale with their square. Fix that and give tactics an explanation of why they were selected (eg. due to the heavy infantry and brave leader) and I think the combat system would be good.
Sorry I didn't express myself clearly... with "tactics" I did not mean the actions of the flanks during the battle. Here I meant the "tactical" actions that a player can make before battles - i.e., how does he form up his armies, how does he maneuver the armies on the map, how does he assign leaders to them. "Leader choice tactics" or "Army maneuver tactics" so to say. More like strategy actually, but in the dev diary they talked about "tactical decisions" that the player would make so I used that word.

My main problem with what Doomdark outlined in the dev diary is that there actually isn't any tactical component to the leader choice. For any given situation, there would always exist an "optimal" choice of leaders that will give you the best chances at battle performance, and you could find it by comparing the bonuses of the available leaders and then picking the one that maximized offensive and defensive bonuses. There is no downside to picking the best available leaders anew each time before a battle so it really is not a game, it is just min-max micromanagement.

I think it would be much better if you were restricted in your leader choice, if you could not reassign them all the time, and if there were several possible and equally promising but differently weighted leader choices. Such as, between offensive and defensive leaders; between young prodigies and old experienced leaders; and between leaders with high authority (i.e. high ranked nobles) but mediocre stats and leaders with low authority (foreign courtiers at your court) but high stats. With the present system (and the one proposed in the dev diary) there is no trade-off involved, you don't have to think about any of that so you can always stick with the same strategy (pick whatever random dude has the highest bonuses) and that will work best.
 
Last edited:
Sorry I didn't express myself clearly... with "tactics" I did not mean the actions of the flanks during the battle. Here I meant the "tactical" actions that a player can make before battles - i.e., how does he form up his armies, how does he maneuver the armies on the map, how does he assign leaders to them. "Leader choice tactics" or "Army maneuver tactics" so to say. More like strategy actually, but in the dev diary they talked about "tactical decisions" that the player would make so I used that word.

My main problem with what Doomdark outlined in the dev diary is that there actually isn't any tactical component to the leader choice. For any given situation, there would always exist an "optimal" choice of leaders that will give you the best chances at battle performance, and you could find it by comparing the bonuses of the available leaders and then picking the one that maximized offensive and defensive bonuses. There is no downside to picking the best available leaders anew each time before a battle so it really is not a game, it is just min-max micromanagement.

I think it would be much better if you were restricted in your leader choice, if you could not reassign them all the time, and if there were several possible and equally promising but differently weighted leader choices. Such as, between offensive and defensive leaders; between young prodigies and old experienced leaders; and between leaders with high authority (i.e. high ranked nobles) but mediocre stats and leaders with low authority (foreign courtiers at your court) but high stats. With the present system (and the one proposed in the dev diary) there is no trade-off involved, you don't have to think about any of that so you can always stick with the same strategy (pick whatever random dude has the highest bonuses) and that will work best.

Well, there is one additional tactical possibility: you can decide to use terrain specialists, then position yourself such that the battle will be on the correct terrain. It's opportunity cost is that it forces you to be either more passive (not attacking a sieging enemy since he is on the wrong terrain) or more aggressive (attack the enemy even if he is on the other side of a river since he is on the right terrain). Not that exciting, but here nonetheless.
 
Well, there is one additional tactical possibility: you can decide to use terrain specialists, then position yourself such that the battle will be on the correct terrain. It's opportunity cost is that it forces you to be either more passive (not attacking a sieging enemy since he is on the wrong terrain) or more aggressive (attack the enemy even if he is on the other side of a river since he is on the right terrain). Not that exciting, but here nonetheless.
If daily swapping commanders is disabled or penalized, then yeah, that would make combat more interesting. However as long as you can swap commanders in and out at random it's not really making anything more interesting. If you follow the troop movements closely (slow speed and frequent pausing) you know where and when they'll meet in battle and commander choice can be optimized every time. You might say, a player should just choose not do it, but there is the AI too and there really is no reason not to code them to do this commander-optimization all the time.
 
Seems that most of your propositions are moddable leviathan
A true diversification of commander choices - yeah. But the combat events aren't all that moddable. Any event that happens during combat is a regular pop-up for the player and you cannot have too many of those before the player gets seriously annoyed and turns them off in disgust.

Also I think you can't at present make combat events where individual actions matter that much - one flank charging prematurely cannot make the other flanks charge as well although that would be a fun thing. You also don't really get to see what is causing these events, which commander is really doing well and who is being a total dunce. (See point above.)

Lastly the most important thing, restricting the choice of a commander (with the possibility of making it moddable e.g. through honorary "general" titles) cannot be done right now. Neither can you as a modder add a cooldown to commander swaps or penalize swaps by opinion maluses. (EU:Rome has that system, commanders may even refuse being removed or moved from their command.)