Pros and cons of killrate modding alternatives

[Louis de la Ferte Ste Colombe]

Many people like RTW 1.2 and feels its potential... and at the same time are somehow frustrated that vanilla is not fulfilling this potential.

One aspect of the game that all mods are trying to change is the killrate: everyone seem to lower it, and in my opinion, although most mods are actually successfull in lowering kill rate, the aftermath is somehow worse...

What have been tried so far....

- change HP
- change defence stat
- delay between attack
- change lethality
- change mental stat (although it does not change kill rate, it keeps unit fighting longer)

Either mods have gone too far changing those factors, and by tweaking them differently a better compromise can be achieved, or something is missing on that list....

What are the pros and cons of those methods for you? How would you do it? Which one is your favorite?

Louis,

PS: please, no "I did the XXXXX mod, and it rules, you got to try it" post: just explain what you did, why you think it's good, and what are the downside to it (like... "but I also had to mod that to make it work overall"...). Be sure I have tried lot of mods around, and I got some answer ready for "mod XXXXXX rulez" fanboys

[professorspatula]

I've made two changes affecting killspeeds on your list: lethality and mental stats.

I gave every unit either a 3,4,5, or 6 point mental boost. Basic untrained units got a +3, trained or those better than militia got a +4 bonus, elite units got a +5 and the super elite and advanced generals got +6. This had led to battle lines that don't fold instantly. I'm considering adding another 1 to all units because even now a cavalry charge can reduce a unit's morale to breaking point too fast. However I don't want to add too much morale: units should break if they're losing badly. It's just a shame when they rout they die so easily. I think that's where much of the battle length problems lie.

Regarding lethality, this is a good one to change. Something I will note first, even before changing lethality, the mental boosts can make battles last long. I had a 1 Auxiliary unit versus 1 Auxiliary unit battle before making changes to lethality. The battle lasted six and a half minutes. I refought the battle and the next one lasted over 10 minutes! That's 242 men in total (121 on each side inc. captain) fighting for over ten minutes! I had to use the 'Rally Troops' command twice to keep the men fighting for so long, and in the end I had only 4 men left before they routed. Reducing their lethality would in theory lead to silly lengths of battles.

However, it had to be done, especially when considering often several units are engaging each other at once - it's not always spearmen versus spearmen. I reduced most units lethality by around 0.20-0.27 and this makes the fights in the game last a bit longer. With the mental stat boost, I even see the AI will sometimes pull out it's wounded men, then charge them in again, whereas before they would probably have routed before getting the opportunity.

Still, cavalry and missiles are still overpowered and can turn potentially long battles into nothing more than short skirmishes. My next step is to reduce the number of cavalry in a unit and try and figure a good way to reduce the lethality of missiles.

I've considered upping the hitpoints as well, but I don't want unnecessary long and tedious slug-fests where nothing much happens for a while. Plus I'm not sure if the battles are slowed down by too much whether the AI would benefit more, or myself. I don't need any more advantages over the AI.

[Epistolary Richard]

What have been tried so far....
- change HP
- change defence stat
- delay between attack
- change lethality
- change mental stat (although it does not change kill rate, it keeps unit fighting longer)

Is 'lethality' the unmarked variable after the delay between attack?

HP - I don't like so much, it leads to soldiers being cut down then standing up again, dusting themselves off and fighting normally. Fine for generals and officers and certain elites but not for line troops.

Delay between attack - again, not so much, it's already a little weird how long it takes in between each swing.

Mental - a little tweaking can be good but too much and you end up with too many units virtually fighting to the death which isn't right. IMO the only problem with mental is that routers should be able to run faster than pursuers, this would give them a chance to get a little distance and rally. Plus you should lose control of pursuers for a short time as well (a la Hastings etc.) this would stop a single unit being able to roll up your opponent's line.

So defence/armour/shield would be my preferred options. The battles look the same as vanilla, units still rout, but the action is a little more extended and involved.

You should also consider the effect of movement speeds, while they aren't directly related to combat they obviously havean impact on the duration of the battle.

[Red Harvest]

Lethality seems to work best. It is linear based on my testing. Going from 1 to 0.5 doubles the combat time in my tests. It doesn't affect archery--projectile stats are independent. It should blunt cavalry charges somewhat. I tested it briefly in campaign and it did ok. I haven't had any interest in RTW campaigns for the past few weeks. If I ever start playing again it will be with the reduced lethality.

The next best would be applying a couple of points of armour to everyone. This would have the positive impact of reducing archer effectiveness. Note that defensive skill is limited to particular quadrants (like shields) according to Jerome, so it is less effective as a tweak.

Multi-hit points are just bad, in my opinion. Units don't "act" right with multiple hit points.

Mental state tweaks are not good in my opinion. The units are already too eager to fight to the death.

[lt1956]

Units too eager to fight to the death in the vanilla? lol what are you smoking! j/k

[hrvojej]

I have changed several things.

I have decreased frequency of attacks, while leaving it higher for phalanx pikes and chariot launching attacks (which makes them more effective than if everything was balanced, and I feel they need to be). This has no effect on charges, but it does in prolonging the ensuing melee. And so far I haven't found the way to decrease the frequency of ranged attacks (stat_fire_delay doesn't work as far as I can tell).

I have decreased lethality for everybody but phalanx pikes. This dulls down the effect of charges a bit, as well as prolonging the ensuing melee. Pikemen however are way more effective than they are if everything was equal. Again, it has no effect on missiles.

I have upped the morale by 3-5 (mostly 4-5, only missiles and placeholders get +3). This gives units more staying power, allows for a bit more tactical thought, and generally feels better. Units don't fight to death more often, only when they're surrounded as usual. It however means that the enemy won't rout as soon as they touch my line, or even before that. And the differential addition means that e.g. archers will still break from cavalry charge, yet they won't do so from a breeze that cavalry causes when they are a mile away. Despite the pitfalls of possibly losing the balance and ratios of morale, I just cannot stand those 15+:1 casualty ratio victories any more, and watch all those beautiful enemy stacks run away before I even managed to get a good look at them.

I have also decreased the movement speeds, and added no horse jumping skeleton file.

All this results in a relatively enjoyable brawls for me on medium/hard. It still isn't quite perfect, but it sure as hell feels much better than before. Maybe dabbling with the defense/armor would make it even better but I don't think I'll be trying because 1) I could again lose the interesting ratios etc. between units, and this time even more seriously than in the case of morale, 2) I don't think it will really change much, and 3) I can't be bothered to edit all of that and test it to satisfaction as well. The only thing I can think of that could really make it better would be having some knowledge about combat calculations (so that I know what I am doing exactly), and even better yet being able to mod all morale modifiers and combat parameters per se.

Unfortunatelly, the end result of it all is that I did all this work only to realize that the loadgame bug makes all my effort quite pointless. Yipee.

[Simetrical]

HP - I don't like so much, it leads to soldiers being cut down then standing up again, dusting themselves off and fighting normally.

This happens to all troops. I don't know if the frequency is greater for more HPs. Regardless, wounding is important for realism; more men should die on the second volley than on the first, and this is what increased HP accomplishes. The main problem is that it nerfs elephants, what with the 15-HP cap.

-Simetrical

[lt1956]

I thought the cap was 29-20 on HP?

Lt

[Louis de la Ferte Ste Colombe]

I guess I ought to have made it clearer; I'd like to reduce kill rate without affecting too much balance.

It is obvious changing killrate will somehow affect balance; the question is really, what kill rate reduction solution has the least impact on balance?

Red harvest, my experience with lethality is it's not really linear when compared to lenght of fight. Reducing lethality by 20% does not change fight duration by 20%. It looks like it works per plateau, and one plateau might be around 40% reduction....

I guess I got to try a few more things

Thanks for your feedback, does anyone feel like one stat has been overlooked that could possibly affect killrate? Or do we got them all, and just need to find a good one, or a combination of good one?

Louis,

[Kraxis]

I'm certainly for lethality rather than any of the others.

Even if what Louis is right it is better than the others. Red pretty much said it all. Especially better than Morale and Def ability.
Morale because it seems the penalties for large stacks are significant, but if we get down to individual units duking it out there will be a lot of fights where the loser will fight on until about 95% losses. Not cool.
Def ability of course only works forwards and to the right. So tweaking that will create odd battles where certain units will get crushed while other equal units win out easily.

Armour is my second choice. But since we are not certain how that impacts fatigue (if at all) it can't become my favourite.

[Red Harvest]

I thought the cap was 29-20 on HP?

Lt

I've see Jerome state it was 15 before (I can't remember where.)

[Red Harvest]

I haven't tested lethality heavily enough to determine if it is truly linear. But I did test 1 vs 1 melee with 0.5 and 1. I measured how long it took to produce X number of casualties (about 1/4 losses if I recall correctly.) It took almost exactly twice as long with 0.5 as it did with 1. The reason for only using a fractional loss was to minimize the interference of shrinking sample size. When I've gone down to 0.2 battles take much longer and both sides are exhausted.

One other positive aspect of reduced lethality: fatigue. Units begin to tire before they kill. And when they tire, they kill more slowly. At the same time they become much more vulnerable to a fresh unit. In vanilla RTW fatigue isn't a big factor, except in the chase (or on uphill marches.)

[Crandaeolon]

What about randomness? Does decreasing lethality cause end results to be less random, like increasing defense (and armour) does?

For example, when two units of unmodded Iberian inf clash, the results may differ quite a lot. But when you add armour the results tend to be less lopsided, i.e. the units will stay within a few kills of each other and fight for a long time.

Edit: Gave it a quick test. All methods seem to produce a similar end result; melees last longer and are less random. I guess the side effects determine which method is preferable. Did only a few tests, so these "results" are more like "feelings" than actual usable data.

Decreasing lethality has the side effect of making charges less powerful. Red Harvest stated that lethality has no effect on ranged weapons, so "precursor" weapons (such as pila) and archery should retain their effectiveness.

Increasing armour has the side effect of affecting ranged weapon effectiveness. Armour is also omni-directional; adding lots of armour would probably devalue flank / rear attack bonuses somewhat.

Edit: More precise testing indicates that the influence of armour to fatigue is actually very small or non-existent. That would make it a quite good candidate for lowering kill rates.

Increasing defense skill produces a bit more random results than increasing armour or decreasing lethality. This would seem to suggest that DS is indeed directional at least to some degree. However, in a typical head-on melee, both sides have roughly equal flanking opportunities, which does seem to result in a nice, slow, and fairly predictable result. Flanking charges would be an entirely different story...

[professorspatula]

The tricky thing with trying to get a good kill rate balance in RTW is that there are two kinds of 'clashes' in any given battle. There's the small skirmish: maybe a couple of infantry units going at it alone; then there's the big melees with several units piling in from all angles.

If you reduce kill rate by too much, those small skirmishes tend to become drawn out tedious battles of attrition. The fact that I had 2 auxiliary units fight for 10 minutes before altering anything but morale highlights this, and adding hitpoints or lowering lethality would make the event comical. But then in the larger clashes where units are fighting on many fronts, without reducing the killrate, entire units can be wiped out in seconds.

Even with lowering lethality and increasing morale, a cavalry charge into the rear or flank of a unit will destroy it fast, as will a bunch of irate axemen or whatever. In RTW it is incredibly easy to flank the enemy, so you end up wiping out your foe pretty quickly. So do you compensate for this and end up with upsetting the balance of the smaller fights or what?

I really think one of the biggest problems with kill rates is something that is difficult to resolve. That is the silly rate that routing units die at. As soon as you rout a unit in close combat, you barely have to do more than just touch them to kill them. Tough elite units crumble like mere peasants and sending in several units to make sure you kill them off becomes a familiar sight. Because of this you try to over-compensate by making sure units don't rout too quick in the first place, even if it upsets the balance. Really CA ought to make killing routing units far more difficult so you are less willing to chase after them, which would also give them a greater chance of reforming and getting back in the fray.

When all is said and done though, I definitely think it's worthwhile tweaking morale and lethality and even movement speed by a bit, just to give a slight improvement over the vanilla battles if nothing else. Hitpoints and extra defensive bonuses are a possibility but are more likely to upset the general balance of things unless more effort is put into tweaking them.

[Red Harvest]

Yes, single unit clashes can be long affairs. Multi-unit clashes are usually short. It is like the combat engine gives defensive penalties for facing multiple foes across a front, but no substantial defense boost for having supporting units on ones flanks.

[hrvojej]

Multi-unit clashes are usually short.

And hence the thread I started a while ago about morale penalties for being outclassed being too great. You kill one unit really quickly due to the high killrate, and this in turn makes the whole army snowball and panic in seconds. The real treat would be to be able to work on such individual parameters to avoid this snowballing.

I agree that one-on-ones can be a bit drawn out when you do all the changes listed above, but in full-stack clashes it feels much better. I rarely get one-on-ones in the game except in explicit testing anyway.

[WesW]

Many people like RTW 1.2 and feels its potential... and at the same time are somehow frustrated that vanilla is not fulfilling this potential.

One aspect of the game that all mods are trying to change is the killrate: everyone seem to lower it, and in my opinion, although most mods are actually successfull in lowering kill rate, the aftermath is somehow worse...

What have been tried so far....

- change HP
- change defence stat
- delay between attack
- change lethality
- change mental stat (although it does not change kill rate, it keeps unit fighting longer)

Either mods have gone too far changing those factors, and by tweaking them differently a better compromise can be achieved, or something is missing on that list....

What are the pros and cons of those methods for you? How would you do it? Which one is your favorite?

Hi guys. This is my first post regarding the modding of Rome, so forgive me if I am a little rusty regarding some of the stats.
From what I have read about the various unit bonuses and the complexity of the overall combat system, as well as my experience modding Medieval, I would generally discourage altering the base attack or defence stats. You have a lot of factors that enter into the overall formula for a battle between given units, and anything other than a wholistic approach is going to skew the stew, so to speak. An alteration that improves things between Hastati and Phalanx might screw things up between Hastati and Warbands, for example.
Typically, the various aspects of the battle system are developed separately, and only combined right before release. This means that the individual components commonly make sense when taken separately, but can become skewed when thrown into the pot with all the other ingredients that make up combat "stew".
Therefore, try and find factors that have the least interaction with other components, and which pertain to those fundamental issues which you noticed when you were just beginning to understand the game's mechanics.
One trait I have come to highly value when modding is naivete in regards to game flaws. By that, I mean the things that bothered you when you were still fairly new to the game, and had not developed any preconceived notions about what things "should" be like. I just started playing Rome a couple of weeks ago, so hopefully I am at a pretty good point as far as noticing the basic things that bug me about battles, before I get good enough at the game where I began to see all the little things that affect the outcome. I have found that if you can identify and correct those fundamental settings, it will also correct many of the secondary issues, at least well enough to get by.

Using these principles, the first issue I would address is unit speed. Even using Large unit sizes, galloping cavalry units can go from one end of the battle line to the other in about 3 seconds, which is barely manageable even if your "twitch" reflex is to hit the Pause button every few seconds of realtime. Time and again I have looked up to see a cavalry unit hitting me from the rear which wasn't even near the battlefield only a few seconds earlier.
I have't studied the Rome stats yet, but in Medieval the stats for one of the basic spear units were MARCH_SPEED( 6), RUN_SPEED( 9), CHARGE_SPEED( 10), while those for the cavalry units were basically MARCH_SPEED( 9), RUN_SPEED( 21), CHARGE_SPEED( 23).
If I had to guess, I'd say the Rome settings for cavalry are a comparative 30 Run and 36 Charge, which is ridiculous even given those weird circular movements during mop-up. (Whenever I see something that appears so obviously skewed, I try and figure out what happened, and I think that the cavalry speed was set for the pursuit manouver, rather than the run and charge modes.)
Speed is also relatively unrelated to other combat factors, so you should have a minimum of unforseen complications.

On a related issue, I also think that the developers decided to half the size of unit sprites, as well as unit size, in order to facilitate urban combat.
One of the first things I noticed about the new maps was how huge individual trees were. Then I noticed how the terrain seemed to swallow up the armies, especially with the comparatively bland maps compared to Medieval. Then you have the movement speed issue, and even the inordinate space between units in the default formations.

If you think about it, increasing the size of the sprites would address all of these problems, though I think cavalry speed would still need to be adjusted. I have this theory that units were supposed to form up in column formation during urban assaults, ala those manning the siege engines, but that CA decided to go away from this for some reason, which has resulted in havoc in other parts of the game.
I have heard so much about how CA has nerfed up battles, which are the heart of the game, and which they had such a good handle on with Medieval, so I have to figure that it's due to a problem with integrating one of the new aspects of the game, and street battles are the most complex and tricky addition to the combat system.

Changing lethality sounds like a good option, though slowing down casualities is going to exacerbate any speed issues.

Slowing down the attack rate may not be good if it's not uniform between all unit types. This may also affect the delay in reacting to orders.

From what I have seen, units don't rout unless they get in appropriately bad situations, which is what Red alluded to; the problem is that either the AI sends them into awful situations much too often due to poor scripting, or that they get into such situations all too often due to the speed issues (both yours and the AI units). I would wait until you have tried adjusting the other factors, and see if improving them doesn't render the morale issue mute.

Nice to see you back on the forums, Red. (Btw, as a point of advice for you other guys, Red generally knows what he's talking about, and scoffing at his observations can often end up making you look like an ass.)

[Red Harvest]

The unit speed is not directly moddable in a simple fashion. It is actually tied to the animation (although it does slow for fatigue.) There are some terrain modifiers that can be used to manipulate it. Kraxis has experimented with this some, as have others.

Scale issues are the real problem with respect to speed. Even if the unit speeds were 100% realistic, armies/units are of ~1/10th scale. Formation depths are usually no worse that 1/2 scale. This means that frontage is typically about 1/5th of what it should be. So lateral movements and such take far less time than they should (1/5th as much if the speed were accurate to begin with.) Honestly, lower than "realistic" movement speeds would be needed to simulate this properly. A true battle line could stretch well over a mile. You couldn't go running that length in 30 seconds. So to simulate the true "flanking potential" you need slower than actual movement speeds.

[Kekvit Irae]

I agree. Slowing down the killrate in a RTW battle is like replacing Formula One race cars with Hondas. I dont want my men to be fighting any longer than they have to, even if it means I'll be at a disadvantage when trying to manuever my cavalry behind the fight.
It's the same reason I hate slowing down the movement. I play this game to have fun. When I have to open up a book to wait for my men to march across the battlefield, it's no fun. Unless I'm specifically delegating something, I keep my battles at top speed.

[Kraxis]

Modding the terrain is nigh imposible to make work. Sadly.
Suddenly the units seems to move too slow when marching and you end up running everywhere. Meanwhile the AI still marches along under fire from archers (who are now mightily buffed).
I initially modded the speeds to be relatively slow (half speed in swamps and so on) but that turned out to ruin the experience, even after I nerfed the archers quite a lot. And to add it to it the animations suddenly looked rather bad.
Now I'm running a much less defined mod of speeds which I like.
Oh and remember that the speed of mud is applied to all mud roads (doubletracks) and roads in simple barbarian settlements. That can mess up a lot when you have to cross such a road in a nice plain, or have your phalanx march along the road to get to the enemy (will result in one or two units being broken up by men lagging behind).

I think I have come to a personal point of experience... hehe, nice little comment there. Anyway, I think a small moraleboost might be enough to help out in the large battles while not inflicting too much damage on individual fights, such as 2 moralepoints. Add that to lowering certain weapon (those of lethality 0.73) to 0.5 and other to 0.7 I think we will get more static battles that aren't going to be too static.

[hrvojej]

Movement modifiers of 0.8 - 0.75 for most terrains are reasonable. Units don't look like they're moonwalking and it doesn't take them forever to get across the map. At the same time they do not seem to run on high octane fuel either. IMO anything slower than that is too slow, and 1.0 is really too fast. Also, this potentially lowers the plowing effect of charges as well.

I also think that units having different lethality stats as they do in the vanilla is not so great. The reason I have left phalanx pikes at 1 lethality while I decreased it for everybody else is due to the my own perception of ineffectiveness of a formed phalanx formation. However, I don't really see the point of having a lower lethality for, say, triarii or berserkers than what is set for iberian infantry or greek cavalry.

[Duke John]

Posted by Red Harvest:

A true battle line could stretch well over a mile. You couldn't go running that length in 30 seconds. So to simulate the true "flanking potential" you need slower than actual movement speeds.

Lateral movement may now be 5 times too quicky, but forward movement is not (with some edits in animations). If you make movement slower to get a better lateral movement you are messing up the forward movement. I think edited animations so that the running/charging speeds are on par with M:TW will fix most issues.

WesW
Good to see you back! But "sprites"? Boy, do you need to learn some things about R:TW

[hoom]

Really CA ought to make killing routing units far more difficult so you are less willing to chase after them

S& M gave routers their charge bonus.
That meant that they could potentially desperately cut their way through an enemy blocking their exit. (a trapped foe is at his most dangerous)

It also tended to make me more careful about who/what I went chasing with.
However I seem to recall this was 'fixed' in the VI patch?

I'm pretty sure that RTW has routers unable to attack at all or with highly reduced stats to the point of irrelevancy.

Interesting idea about the soldiers having been scaled down.
I'd come to the conclusion that various objects had been scaled up to be more 'epic' but scaling down the soldiers kind of makes more sense with what we see.