Attrition is basically irrelevant now

Reman · Dec 18, 2020

Taw did some tests and made a writeup on Reddit. Here's his post. I know some people on the forums might not use Reddit, so I'll repost it here for posterity.

If you were wondering why you aren't running low on manpower in 1.30, here's why:

Fresh game for testing, 1444 Muscovy, no leaders, just put whole army and mercs in Yuriev:

Tooltip says: Attrition from Supply Limit 268.50%. Weight: 39k, Supply: 14.52k. => 2.6%

By wiki formula: (39-14)*(10/14)=17.8% (which would get capped to 5%).

Deleted some:

Tooltip says: Attrition from Supply Limit 206.60%. Weight: 30k, Supply: 14.52k. => 2.0%

By wiki formula: (30-14)*(10/14)=11.4% (which would get capped to 5%).

Deleted some more:

Tooltip says: Attrition from Supply Limit 179.00%. Weight: 26k, Supply: 14.52k. => 1.7%

By wiki formula: (26-14)*(10/14)=8.5% (which would get capped to 5%).

Deleted even more:

Tooltip says: Attrition from Supply Limit 137.70%. Weight: 20k, Supply: 14.52k. => 1.3%

By wiki formula: (20-14)*(10/14)=4.3%

Added general with 3 pips, 20k troops:

Tooltip says: Attrition from Supply Limit 117.00%. Weight: 17k, Supply: 14.52k. => 1.1%

By wiki formula: (17-14)*(10/14)=2.1%

Accidentally waited enough to get out of winter, so adjusted troops just over limit:

Tooltip says: Attrition from Supply Limit 101.70%. Weight: 16k, Supply: 15.72k. => 1.0%

By wiki formula: (16-15)*(10/15)=0.6%

So new formula looks like:

weight still equals (soldier/1000 - leaderpips), just as wiki says

supply doesn't get rounded at any point, fractional values are used in calculations

new formula for from supply: attrition = (weight / supply) * 1%, except you only get it if it's >1%

(I'm assuming to this you'd add attrition from sieges, province modifiers etc.)

so you cannot ever get below 1% attrition, anything less is 0%

This is enormous reduction in attrition. Hitting 5% cap used to be really common.

Assuming new formula is correct, here's how many troops you'd need to get 5% attrition by old and new formulas:

for 10 supply limit; 15k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 50k

for 11 supply limit; 17k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 55k

for 12 supply limit; 18k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 60k

for 13 supply limit; 20k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 65k

for 14 supply limit; 21k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 70k

for 15 supply limit; 23k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 75k

for 16 supply limit; 24k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 80k

for 17 supply limit; 26k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 85k

for 18 supply limit; 27k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 90k

for 19 supply limit; 29k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 95k

for 20 supply limit; 30k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 100k

for 21 supply limit; 32k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 105k

for 22 supply limit; 33k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 110k

for 23 supply limit; 35k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 115k

for 24 supply limit; 36k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 120k

for 25 supply limit; 38k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 125k

for 26 supply limit; 39k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 130k

for 27 supply limit; 41k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 135k

for 28 supply limit; 42k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 140k

for 29 supply limit; 44k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 145k

for 30 supply limit; 45k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 150k

for 31 supply limit; 47k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 155k

for 32 supply limit; 48k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 160k

for 33 supply limit; 50k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 165k

for 34 supply limit; 51k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 170k

for 35 supply limit; 53k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 175k

for 36 supply limit; 54k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 180k

for 37 supply limit; 56k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 185k

for 38 supply limit; 57k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 190k

for 39 supply limit; 59k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 195k

for 40 supply limit; 60k would be 5% attrition due to cap; now 1.5%. To get 5% attrition you now need 200k

The "attrition cap" is so high it's basically irrelevant, back with old formula you'd hit it in no time.

holyvigil · Dec 18, 2020

Tell that to my 200k stack.

Jarvin · Dec 18, 2020

I think, if anything, this really goes to show just how poorly designed attrition is in general.

Anyhow, I checked some random saves from Skanderbeg's database, and IMHO the claim about attrition being irrelevant is at least partly disproven by attrition losses in wars still often being the main cause of casualties on either side:

Top wars from some random SP saves from 1.30:

(note that you sort of can only look at the bigger attrition casualties number between the two sides because of how stupidly they are counted)

Some random MP one:

From my experience just solely observing the data there seems to be no massive difference between how relevant attrition was in 1.28 and now. Circumstancial evidence suggests that attrition is still extremely significant as far as MP goes. It's still often the main cause of troops' loss in war.
As far as SP goes can't really tell beside what's in my databases because I don't exactly play SP

Curious as to why that is tbf

Battlex · Dec 18, 2020

Jarvin said:
I think, if anything, this really goes to show just how poorly designed attrition is in general.

Anyhow, I checked some random saves from Skanderbeg's database, and IMHO the claim about attrition being irrelevant is at least partly disproven by attrition losses in wars still often being the main cause of casualties on either side:

Top wars from some random SP saves from 1.30:
View attachment 663165

View attachment 663166

(note that you sort of can only look at the bigger attrition casualties number between the two sides because of how stupidly they are counted)

Some random MP one:
View attachment 663168

From my experience just solely observing the data there seems to be no massive difference between how relevant attrition was in 1.28 and now. Circumstancial evidence suggests that attrition is still extremely significant as far as MP goes. It's still often the main cause of troops' loss in war.
As far as SP goes can't really tell beside what's in my databases because I don't exactly play SP

Curious as to why that is tbf

Could there be confirmation bias in that the meta tells players not to assault forts but to just siege them out, and once big enough that's your only losses in war

Reman · Dec 18, 2020

"Irrelevant" might indeed be too strong a word, as there's still the 1% minimum on sieges. However, I just finished a 1740s WC with the Mughals without taking a single military idea or merc, and never even remotely ran into manpower issues. This never would have happened in previous patches. Then again, part of this might be the shift in the meta away from TCs to full states.

Also, I believe the attrition casualty counter takes casualties from ALL wars currently running, which can inflate things a bit.

Deliberus · Dec 18, 2020

Reman said:
"Irrelevant" might indeed be too strong a word, as there's still the 1% minimum on sieges. However, I just finished a 1740s WC with the Mughals without taking a single military idea or merc, and never even remotely ran into manpower issues.

Also, I believe the attrition casualty counter takes casualties from ALL wars currently running, which can inflate things a bit.

Then again you're playing as Mughals

Jarvin · Dec 18, 2020

Reman said:
"Irrelevant" might indeed be too strong a word, as there's still the 1% minimum on sieges. However, I just finished a 1740s WC with the Mughals without taking a single military idea or merc, and never even remotely ran into manpower issues. This never would have happened in previous patches. Then again, part of this might be the shift in the meta away from TCs to full states.

Also, I believe the attrition casualty counter takes casualties from ALL wars currently running, which can inflate things a bit.

I mean yea but still

Top row is a player, the rest is AIs

From that MP game

this one should count everything properly actually

Manwe_Sulimo · Dec 18, 2020

The problem is that you still have a lot of factors that give you fixed +attrition irrespective of supply limit when not in friendly territory (terrain types such as arid, tropical, ..., monsoon, harsh winters, ...), in those provinces you can easily get to 5. Also you have the massive attrition when transporting troops between continents.
Its a bit easier to move around larger armies now, but in attrition heavy areas the relief is quite small. As others said sieges have not changed either.

EarlKonrad · Dec 18, 2020

Wouldn't the best way of confirming or denying this be comparing two exact attrition ticks, one running a 1.29 version and another a 1.30? If you can fix all of the variables and let a month row, wouldn't that be enough evidence that something was changed or not under the radar?

Manwe_Sulimo · Dec 19, 2020

EarlKonrad said:
Wouldn't the best way of confirming or denying this be comparing two exact attrition ticks, one running a 1.29 version and another a 1.30? If you can fix all of the variables and let a month row, wouldn't that be enough evidence that something was changed or not under the radar?

It was in the dev diaries - they stated that they reduce attrition by supply limit.

EUIV - Development Diary - 25th of February 2020

Hello everyone! So we have worked on this update for quite a long while and already covered a lot of features in previous diaries. However just because we’ve already covered something doesn’t mean we’ve stopped working on those features but we...

forum.paradoxplaza.com

EarlKonrad · Dec 19, 2020

Manwe_Sulimo said:
It was in the dev diaries - they stated that they reduce attrition by supply limit.

EUIV - Development Diary - 25th of February 2020

Hello everyone! So we have worked on this update for quite a long while and already covered a lot of features in previous diaries. However just because we’ve already covered something doesn’t mean we’ve stopped working on those features but we...

forum.paradoxplaza.com

Well, nevermind then.

Search

Attrition is basically irrelevant now

Reman

Field Marshal

holyvigil

General

Jarvin

Colonel

Battlex

Field Marshal

Reman

Field Marshal

Deliberus

Banned

Jarvin

Colonel

Manwe_Sulimo

Major

EarlKonrad

Field Marshal

Manwe_Sulimo

Major

EUIV - Development Diary - 25th of February 2020

EarlKonrad

Field Marshal

EUIV - Development Diary - 25th of February 2020