Troops are killed instead of wounded when participating

Currently viewing this thread:

kab

Recruit
In my vanilla 1.5.8 campaign I've noticed that "participating" in battles is turning nearly all casualties into ***deaths***. This causes Medicine to level extremely slowly because there are few wounded, it also causes many lord deaths if you have that enabled. By contrast, using auto-resolve ("Send troops...") generates more casualties but fewer ***deaths***. This seems to happen across the board regardless of Medicine or Tactics.

As an example with 45 Medicine, 85 Tactics with a battle I tested repeatedly (roughly 100 vs 100):

1. Click "Send troops..." get ~10-15 dead, ~25-40 wounded, 40-50 prisoners.
2. Click "Attack" and hit F6, get ~20-25 dead, < 10 wounded, 20-30 prisoners.

It's a consistent pattern. I get loads of prisoners and Medicine actually levels pretty well when I get 20+ wounded out of every fight using send troops.
 

AnandaShanti

Sergeant Knight at Arms
Yeah it's pretty good for getting a little medicine to just take all recruits and just ram(auto calc) them into looters of similar size.
Anything else you want to aim for ZERO units getting defeated in live battle, because even with a 120 medic it's not that much better, though it's more effective on higher tier troops.
I would never use f6 or f1-f3 unless the enemy is broken.
 

kab

Recruit
My point is the ratio of dead to wounded is wildly off in live battles. A number of complaints including too many dead lords in live battle vs simulation, Medicine leveling too slowly, etc. are at the very least impacted by this and perhaps even caused by it.

The algorithm/implementation of wounding in live battle likely needs a quick balance pass or has a bug.

Edit: To put this in perspective, we're talking about it swinging from a somewhat realistic 1:3 ratio with 10% dead and 40% total casualties in simulations to a 3:1 ratio with 20%+ dead and 30% total casualties in live battles. That's leaving the AI on it's own to do it's thing. That's a gigantic discrepancy and all I can do is reduce the number of casualties, it's still all fatal with very few or no wounded. This is why you end up targeting "zero" which is unrealistic and silly, because anyone who goes down frickin dies unless you simulate.
 
Last edited:

ImperialDane

Recruit
Well i suppose the counter argument there is that in the live battle you have more influence to actually negate casualties via direct tactics and strategy. So it's there to balance out the live battles one might surmise.

Additionally the AI is a bit of a dunce and letting it handle everything will also result in at times very pointless casualties. For example one thing the AI likes to do is present your troops as target practice to enemy archers.
 

kab

Recruit
Well i suppose the counter argument there is that in the live battle you have more influence to actually negate casualties via direct tactics and strategy. So it's there to balance out the live battles one might surmise.
That's not a counter argument though. I don't need to do anything to have fewer casualties, it simply happens by clicking the button. These things do not balance each other, they rest on the same side of the scale and make it more unbalanced.

Additionally the AI is a bit of a dunce and letting it handle everything will also result in at times very pointless casualties. For example one thing the AI likes to do is present your troops as target practice to enemy archers.

Who cares if the AI is a dunce? It doesn't matter if I let it handle things or not. You take fewer casualties simply by being in a live battle instead of a simulation. Yet somehow mortality rate jumps from 20% to 75% at the same time. So the AI is a dunce, sure, yet somehow manages to take fewer casualties while simultaneously turning most of them fatal? This is silly and unbalanced.

We need to see more casualties in live battle, with far more of them wounded.
 

Bannerman Man

C# Sleuth
Knight
@kab I think what you're noticing is the difference between how damage types are handled in live battles vs auto calc. In auto calc, all units have a 30% chance to deal blunt damage with each blow (and if the final blow deals blunt damage it will always result in being wounded). In live battles, it's based on the actual weapon's damage type(s) and the way in which it is wielded (i.e. thrusted or swung), along with some other physics based elements like horse charges or striking with the hilt/haft, meaning it is highly dependent on the combat situation. Depending on the types of equipment your opponent's troops have, the proportion of blunt damage dealt in live battles can be greater or lesser than that of an auto calc battle, but in most cases I think it will probably be < 30% of the overall damage dealt.

Other than that though, auto calc and live battles share the exact same survival formula to determine if a troop is killed or wounded, so things like medicine skill shouldn't cause any differences in casualty ratios.

FYI, you will still earn medicine skill for each fallen troop, regardless of whether they were killed or wounded. In live battles, it's 10 raw medicine xp per wounded troop and 10 per killed troop, but in auto calc it is 10 per wounded troop and 5 per killed troop for whatever reason.
 

AnandaShanti

Sergeant Knight at Arms
My point is the ratio of dead to wounded is wildly off in live battles. A number of complaints including too many dead lords in live battle vs simulation, Medicine leveling too slowly, etc. are at the very least impacted by this and perhaps even caused by it.

The algorithm/implementation of wounding in live battle likely needs a quick balance pass or has a bug.

Edit: To put this in perspective, we're talking about it swinging from a somewhat realistic 1:3 ratio with 10% dead and 40% total casualties in simulations to a 3:1 ratio with 20%+ dead and 30% total casualties in live battles. That's leaving the AI on it's own to do it's thing. That's a gigantic discrepancy and all I can do is reduce the number of casualties, it's still all fatal with very few or no wounded. This is why you end up targeting "zero" which is unrealistic and silly, because anyone who goes down frickin dies unless you simulate.
That's not a counter argument though. I don't need to do anything to have fewer casualties, it simply happens by clicking the button. These things do not balance each other, they rest on the same side of the scale and make it more unbalanced.



Who cares if the AI is a dunce? It doesn't matter if I let it handle things or not. You take fewer casualties simply by being in a live battle instead of a simulation. Yet somehow mortality rate jumps from 20% to 75% at the same time. So the AI is a dunce, sure, yet somehow manages to take fewer casualties while simultaneously turning most of them fatal? This is silly and unbalanced.

We need to see more casualties in live battle, with far more of them wounded.
I'm all for enchantments to the effect of medicine in live battle and/or the survivability of troops, especially high tier troops, however I think if TW would change it they would actually make auto-calc harsher instead of live battle more forgiving and everyone would be more annoyed :smile:

In auto calc, all units have a 30% chance to deal blunt damage with each blow
I bet this was just a time saver to not worry about what troops actually have for weapons in auto-calc :smile:

FYI, you will still earn medicine skill for each fallen troop, regardless of whether they were killed or wounded. In live battles, it's 10 raw medicine xp per wounded troop and 10 per killed troop, but in auto calc it is 10 per wounded troop and 5 per killed troop for whatever reason.
That's great to know! I had Thought you got for kills too but I read other posters saying it was wounded only and wasn't sure anymore.
 

kab

Recruit
In auto calc, all units have a 30% chance to deal blunt damage with each blow (and if the final blow deals blunt damage it will always result in being wounded). In live battles, it's based on the actual weapon's damage type(s) and the way in which it is wielded (i.e. thrusted or swung), along with some other physics based elements like horse charges or striking with the hilt/haft, meaning it is highly dependent on the combat situation.

That would certainly do it. I think the live battle algorithm should produce a lot more casualties but at a much lower mortality rate.

I am a bit surprised it's that simplified for simulations. I'd have thought it at least considers troop composition.

Other than that though, auto calc and live battles share the exact same survival formula to determine if a troop is killed or wounded, so things like medicine skill shouldn't cause any differences in casualty ratios.

FYI, you will still earn medicine skill for each fallen troop, regardless of whether they were killed or wounded. In live battles, it's 10 raw medicine xp per wounded troop and 10 per killed troop, but in auto calc it is 10 per wounded troop and 5 per killed troop for whatever reason.

That's interesting, but a lot of the Medicine experience comes post battle while they recover which is lost entirely.

The core problem is how it affects game play. If I play live battles a huge number of the casualties will be fatal, therefore my only recourse is to eliminate the casualty entirely. That in turn means even less Medicine gain from no wounded, no map speed loss from wounded and so on. I'll see what it's like at 100+ Medicine too.
 

kab

Recruit
I'm all for enchantments to the effect of medicine in live battle and/or the survivability of troops, especially high tier troops, however I think if TW would change it they would actually make auto-calc harsher instead of live battle more forgiving and everyone would be more annoyed

Yeah no doubt.

It would be more forgiving in that there's a lower mortality rate and therefore fewer permanent losses. Less forgiving in that more casualties with far more wounded will slow the player down on the map, lower the prisoner limit, can result in a herd penalty, etc. so might actually make it harder to win successive fights.
 

Bannerman Man

C# Sleuth
Knight
I am a bit surprised it's that simplified for simulations. I'd have thought it at least considers troop composition.
I bet this was just a time saver to not worry about what troops actually have for weapons in auto-calc
When the game was first released, they actually did base a troop's damage type in auto calc on their primary weapon, but I think they changed it because it caused certain troops with blunt weapons to always knock out their enemies in auto calc.

I think part of the reason they've left auto calc so abstract is to give them more freedom to balance troops' loadouts around live battles, since those are the central focus of the game. If the gear they gave troops affected both versions of battle, then any changes they made might impact each version differently. I don't think they would ever be able to bring the balance of the two modes into congruence with each other, since live battles are just too chaotic. If, for instance, they wanted to give a certain faction's recruits a mace for live battle purposes, it may end up altering the faction's campaign map balance in an undesirable way, and so that limits their options.

It also gives modders the opportunity to rework loadouts without having to consider how it impacts the balance of the overall campaign as well.

That's all a discussion I don't really want to get into though.

That's interesting, but a lot of the Medicine experience comes post battle while they recover which is lost entirely.
That's a good point. One of the devs posted today that they are looking into alternative methods to train medicine, so this may get better with time.

The good thing is that medicine's boost to the survival rate is front loaded (it's not a linear increase, there are diminishing returns), so the biggest benefit comes within the first 50-60 levels, which isn't too hard to get.
 

kab

Recruit
If the gear they gave troops affected both versions of battle, then any changes they made might impact each version differently.

You don't have to do that. The value assigned to the troop in simulations can be an entirely separate table from their live battle equipment composition. This actually makes it easier to mold simulation to the live battle results. You don't have to touch the equipment at all, you just tweak the table to make them more or less fatal. Deriving it from equipment would be a way bigger PITA and prone to balance problems.
 
Top Bottom