Retinues in 2.5.2?
- Thread starter Kardo
- Start date
-
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.
Did another quick test and first impression is that the troop type leadership traits work like this:
new damage = ( old damage ) ^ [1+bonus]
So with a normal leadership trait (20% bonus), this becomes: new damage = old damage ^ 1.2
new damage = ( old damage ) ^ [1+bonus]
So with a normal leadership trait (20% bonus), this becomes: new damage = old damage ^ 1.2
- 1
Does it only impact damage?
Q2: does other traits also work this way? I.e. winter warrior improves damage by defined % to all unit types when in winter?
Q2: does other traits also work this way? I.e. winter warrior improves damage by defined % to all unit types when in winter?
Also do they stack? In winter vs religious enemy if I am winter soldier/ holy warrior would I get 30+50% = 80% bonus to damage?
Further testing showed my hypothesis was wrong. A +200% bonus to all troop types elevated damage from about 16-17 to bout 1500-1600. According to the hypothesis however the damage should've been around 4000-5000.
What I am sure of however, is that troop type bonus only affects damage and that it's not a linear increase.
I don't know if or how it stacks with anything else. That would still need to be tested, but it's not very high on my priority list. I neither know if winter soldier works in the same way, but it might.
What I am sure of however, is that troop type bonus only affects damage and that it's not a linear increase.
I don't know if or how it stacks with anything else. That would still need to be tested, but it's not very high on my priority list. I neither know if winter soldier works in the same way, but it might.
- 1
Then what the hell does morale damage do and why do bonuses to it exist? So confused...
Keep up the good work Dragatus
Probably adds to coefficient how much troop loss translates into moral loss.
- 1
Bonuses, ok, sure, that makes sense. But morale damage itself? One of the tooltips in a combat window for a flank will popup something like Damage 67 Morale Damage 87. That's what's made me think that there might be 2 morale loss terms, one for kills and one for damage.
No, it doesn't. Morale loss values were too high for that.
I think morale damage as displayed in the combat window is simply regular damage multiplied by the morale damage bonus of the flank leader (for example if the leader is Cruel). And either this morale damage is used to calculate hypothetical kills and these hypothetical kills are then inserted in the morale loss formula. Or the game calculates morale loss from real kills and the at the end multiplies it by the morale damage bonus. Either way morale damage becomes "effective damage for the purposes of calculating morale loss".
I'll have to double check, but I thought the savegame files stored a single, variable morale number for the morale of each unit. If morale loss comes only from kills, why bother with this? You could just multiply the number of survivors of each type by the base morale of that unit type and sum it up. It seems like morale damage should reduce that number above and beyond kills. But I'm not the one with the combat test mod and actual experimental data, so I'll defer.
Your first explanation is pretty much what I was thinking yesterday and worked out formulas for. Real kills based on damage, morale loss based on real kills, and "kills" due to panic based on morale damage.
Lots of hypotheses, insufficient data. Care to share your testing mod?
Your first explanation is pretty much what I was thinking yesterday and worked out formulas for. Real kills based on damage, morale loss based on real kills, and "kills" due to panic based on morale damage.
Lots of hypotheses, insufficient data. Care to share your testing mod?
I appreciate all the effort people do in threads like these so thanks! 
- 1
Well, I don't know what information the savegames store and how to check that, so I guess that can be your contribution to the thread.
And here's the mod: https://www.dropbox.com/s/hkbhja4yrp2ixoh/ck2_combattest.zip?dl=0
You'll need to unzip the archive into your mod folder, but that should be it. Also, I realized that I had previously forgotten to remove the morale bonus provided by training ground buildings and military organization technology, which might explain the 12-20% extra morale. That has now been fixed.
I appreciate all the effort people do in threads like these so thanks!
I enjoy figuring it out, so you're welcome.
Well, sometime in the past few years savegames all became encoded / compressed, except autosaves. Autosaves (with the correct options selected) are stored in plain text like all saves used to be. So we can still learn things, but it's important to have our test battles around 1 Jan and 1 July.
Each unit in an army(retinue unit, single holding levy, merc unit, etc.) has its own entry in a save, which includes the number of troops of each type in that unit (stored to 3 decimal places) and a single number for the morale of that unit. I suspect the 3 decimal places of troop count account for some of the roundoff error noise we see.
So I suspect that morale damage is totaled for a flank, then proportionally distributed to each unit in that flank. When a unit's morale hits zero it flees, regardless of combat phase.
Each unit in an army(retinue unit, single holding levy, merc unit, etc.) has its own entry in a save, which includes the number of troops of each type in that unit (stored to 3 decimal places) and a single number for the morale of that unit. I suspect the 3 decimal places of troop count account for some of the roundoff error noise we see.
So I suspect that morale damage is totaled for a flank, then proportionally distributed to each unit in that flank. When a unit's morale hits zero it flees, regardless of combat phase.
Last edited:
The stored morale value could be current morale, while maximum morale is calculated from the troops that form the unit.
Almost definitely. The interesting test will be if, in the test mod where all values are 1, the stored morale always changes as a constant linear function of kills / troops remaining, a non-linear function of kills, or a function of kills and morale damage as separate factors. And someday I'll have both the time and tools to test that.
Just remembered the in game date is stored at the top of the save file. So we can manually adjust the date in each autosave to create a day by day record of test battles, giving us precise daily data on troop numbers and unit morale. Set up a test battle to start on July 2, record the prebattle data out of the July 1 autosave, edit the date of that save back to Jun 30, load it up, and get a new autosave from day 1 of the battle, repeat as needed.
Just remembered the in game date is stored at the top of the save file. So we can manually adjust the date in each autosave to create a day by day record of test battles, giving us precise daily data on troop numbers and unit morale. Set up a test battle to start on July 2, record the prebattle data out of the July 1 autosave, edit the date of that save back to Jun 30, load it up, and get a new autosave from day 1 of the battle, repeat as needed.
Last edited:
- 1
You can actually save at any time in plain text so long as you're saving locally. There's a little box next to where you name the save file that you can untick. It's good for checking whether your events are firing.
Thanks for the great work everyone I'm very interested to see what you guys come up with! Personally I want to find out whether there's any point building camel cavalry in my Socotra game
Considering Camel Cavalry had all the defensive values halved in the 2.5 patch and that a pure Camel Cavalry flank is going to be triggering Disorganized Harass, I'd say the answer is probably a "no".
Thanks, never noticed that checkbox and spent quite a bit of time looking for that function.You can actually save at any time in plain text so long as you're saving locally. There's a little box next to where you name the save file that you can untick. It's good for checking whether your events are firing.
OTOH, at least according to my kill ratio math (that we've already discussed the problems with, but...), they needed those nerfs since they are still an order of magnitude more effective than heavy infantry post-nerf.
Much depends on how morale works. Need some data to curve fit. My guess ATM is that daily morale loss on a flank is proportional to 3 * (morale damage multiplier / morale defense multiplier) * (daily kills) ^ some power and then the result is allocated per capita among the units on that flank.
Last edited:
Looked thru Dragatus' data and some of my own and concur with his assessment. Morale loss is linear with kills.
Reworking my spreadsheet in terms of the ratio of survivors each day and compounding that for 18 days of skirmish and 18 of melee, I get some different and interesting results.
Without any tactics (like during assaults), the top ten and their cumulative kill ratio after 36 days in battle with 1000 all 1 stat punching bags
Berber LC 6.48
Schiltron 6.41
Italian Pike 5.41
Andalus LC 4.98
Caballero LC 3.19
Knight 1.75
Baltic HI 1.62
Cataphract 1.55
Defense 1.49
Hussar 1.47
With best tactics (except for disorganized harrass for the pure light cav units)
Berber 738
Knight 572
Cataphract 452
Scottish 386
Italian Pike 378
Lombard 188
Andalusian 144
Gusar 111
Caballero 108
Baltic 106
In the latter case, even the best skirmishers don't settle things during skirmish. Melee is by FAR the decisive phase. Defense is the best generic retinue at 20.6.
Reworking my spreadsheet in terms of the ratio of survivors each day and compounding that for 18 days of skirmish and 18 of melee, I get some different and interesting results.
Without any tactics (like during assaults), the top ten and their cumulative kill ratio after 36 days in battle with 1000 all 1 stat punching bags
Berber LC 6.48
Schiltron 6.41
Italian Pike 5.41
Andalus LC 4.98
Caballero LC 3.19
Knight 1.75
Baltic HI 1.62
Cataphract 1.55
Defense 1.49
Hussar 1.47
With best tactics (except for disorganized harrass for the pure light cav units)
Berber 738
Knight 572
Cataphract 452
Scottish 386
Italian Pike 378
Lombard 188
Andalusian 144
Gusar 111
Caballero 108
Baltic 106
In the latter case, even the best skirmishers don't settle things during skirmish. Melee is by FAR the decisive phase. Defense is the best generic retinue at 20.6.