The CK2 To EU3 Converter Project

[Pode]

@ the boss: yay armies!

@ tech groups: to sum up, we have 3 proposals on the table right now. I've withdrawn the first Pode method from contention, it makes too many people western. The second Pode method is based on hard standards for how many provinces "should" have maxed out schools. Right now a standard of 2 in 7 maxed is what's been tested & gives the 2/1 west east split. The 2 in 7 parameter especially is up for adjustment if the results aren't to our liking. The key thing is the fixed standard relative to the fixed ROW.

Idhrendur prefers groups based on % of max learning score, i.e. most developed schools. I went through this to get to the fixed standards, so the 52%/20%/5%/23% thresholds I came up with for this approach have the same thought process behind them as my fixed standard approach. This will always give a reasonable distribution of tech within the ck2 area, but may not accurately reflect tech relative to ROW.

The third option is a culture/geographical mapping. I can see this applying for Greek traditionalism and Muslim use of Arabic preventing use of the printing press, otherwise I just don't find this sort of determinism credible. So if you want this system, you get to define it.

I'd like to throw out a 4th contender, the johnjohansson method, since he just gave me the idea. CK2 "ending" techs spec out a tech level historically achieved by endgame for various cultural/religious groups. We could map those groups to EU3 groups, compare achieved tech to this historical standard, and westernize or easternize realms that achieved significantly more or less tech development than their historical level. This would be more culturally /geographically driven, encompass more factors of tech development than just schools, and still be a fixed historical standard. I'm going to work on fleshing this approach out.

 

[Idhrendur]

I seem to recall a variety of people have experienced a crash bug at 'merging titles'. I ran across it today. Fortunately, it was an easy fix.

 

[Idhrendur]

just ran Version 0.6D, still getting the odd no_goverment, no tech issue. Tried with a old saved and created a new save just to make sure there wasn't a issue with the old one.

Sorry, I had missed your post before. Are there any modifications to your EU3 game? The save itself *should* come out fine right now, and that's a sign of something wrong in either the save or the game itself.

 

[Idhrendur]

I've got the army regiments going properly now, using the technique in this post: http://forum.paradoxplaza.com/forum...er-Project&p=15090508&viewfull=1#post15090508

I did a first pass on manpower (the summation of unraised levies throughout the country, modified by the technique in the post above), but the numbers come out far too high. Arbitrarily dividing the manpower by four seems to bring it back into a reasonable range, but it'd be nice to use a number with something behind it, not just something plucked out of the air. Any ideas?

Also, how about naming the regiments? If someone can cook me up a not too complicated way of getting a name, it'd be nice to have them not blank.

 

[johnjohansson]

Also, how about naming the regiments? If someone can cook me up a not too complicated way of getting a name, it'd be nice to have them not blank.

In eu3, regiments are typically named "<origin province>'s <n>th Regiment", e.g. "Stockholm's 1st Regiment".

I'm not sure what the code looks like (and even if I did, it wouldn't help), but we should be able to get the name of the capital province at least? Possibly through going back to the ck2 data and check the location of the main holding for the ruler of the eu3 country (if there's not an easier way)? Might be a bit strange for all regiments to hail from the capital, but it's better than no name at all.

It would be really great if it was possible to get the main holdings for vassals absorbed into the eu3 country to use for regiment names too, or maybe just non-main holdings of the ruler. I have no idea how difficult that would be, though. I suppose you'd have to know how many regiment names are needed, to know how many vassals to check, so I guess it also depends on in what order things are done.

The numeral should be pretty easy to generate I hope (I'm thinking "some_variable = some_variable + 1", but my coding skills are less than stellar...).

If all else fails, just name them "<country name>'s <n>th Regiment". Better than nothing.

 

[Grifen88]

Just a normal "First/Second/Third Army" for each regiment? EU3 defaults to those once you run out of pre-defined names, so it isn't out of place.

(Unless there's some difficulty I'm not aware of here. The alternative is using provincal naming. Army of <province> which you spawn these regiments.)

 

[dtremenak]

I'm not sure what the code looks like (and even if I did, it wouldn't help), but we should be able to get the name of the capital province at least?

We have the home province, so it shouldn't be a problem to generate real names. EU3ToVic2 does something similar, but V2's naming convention is more complicated (and we have to select specific pops instead of just home provinces).

 

[Pode]

I've got the army regiments going properly now, using the technique in this post: http://forum.paradoxplaza.com/forum...er-Project&p=15090508&viewfull=1#post15090508

I did a first pass on manpower (the summation of unraised levies throughout the country, modified by the technique in the post above), but the numbers come out far too high. Arbitrarily dividing the manpower by four seems to bring it back into a reasonable range, but it'd be nice to use a number with something behind it, not just something plucked out of the air. Any ideas?

My subconscious had apparently run the math and come to the same conclusion, as for about a week the idea of using pike as the infantry standard instead of heavy infantry has been nagging at me. With that change alone, infantry would convert at:
Pike = 1
HI = 0.610
Arch = 0.384
LI = 0.305

So a levy of 1000 of each type would convert to 2299 pike equivalents instead of 3770 men under the HI standard setup. So 57.5% of CK2 numbers instead of 94%. That's basically half of your factor of 4 adjustment, totally justified.

For the other factor of 2, how about a "distant liege" penalty tied to the top liege's tier? If my local baron calls me to fight, the problem is local, my family is in danger, and I'll turn out. The Emperor wants me to go fight in east bumfuckistan, not so much, thanks, there's plenty of other jobless sots in the empire he can ask, I'll be right here farming / bandit hunting. Emperors get their manpower cut in half, Kings get 3/4 of theirs, Dukes 7/8, Counts 15/16, and Barons get it all. This'll cut blob manpower down to something reasonable without making OPMs into snacks. I want the period equivalent of the Netherlands to stand a chance against period equivalent Hapsburg Spain.

Edit: rethinking the tier penalties, the stacking powers of 2 aren't hard enough on the lower tiers. Thinking emperors lose half, kings a third, dukes a 4th, counts a 5th, barons none. So

if tier(topliege) > 0 then manpower == manpower * (1- 1/(6-tier))

 

[DasGuntLord01]

If they're going to be combined in to a stack, call the stack "Army of [Nation]". Then you can easily call them "1st Regiment", "2nd Regiment", or if you like to be a bit more colourful: "1st Infantry Regiment", "2nd Cavalry Regiment" and so on.

 

[Idhrendur]

Thanks, everyone. Army names were easy because there was a 1-1 correspondence between CK2 armies and EU3 ones.

But with the many-many relationship between CK2 subunits and EU3 regiments, I was struggling with how to assign a home province to get the name. Just after I posted last night, I realized I could just use the probalistic algorithm from the EU3->V2 converter, and then everything was easy.

One shortcut: I'm using the province's capital's name instead of the province's name because the converter only know the former.


Now for morale: as I see it, there's two choices.
1) Get the morale of each subunit, determine how much that applies to the army's morale, aggregate those contributions, determine the max morale for the EU3 nation, translate the CK2 morale to EU3 morale, apply this morale to all regiments in the army.
2) Give every regiment a decent default morale.

Now that I've a game to convert and am impatiently waiting for the convert to be more advanced, I'm inclined to go with 2.


Edit: Also, generals and admirals. Seeing as leaders are really easy to assign in CK2, and generals aren't terribly common in EU3, I say we don't bother with them at all. But what is everyone else's thoughts?

 

[Dustz]

My suggestion: convert the Marshal as a leader, and if the ruler has military education convert it too.

 

[Pode]

Shifting to EU3 armies means standardizing on a unit type, so a standard morale makes sense. Roll in the morale bonuses from state martial and realm average morale tech level if you want to get fancy/accurate.

Generals: you have a system for making vassals & courtiers into advisors of various star ratings. Could you run similar code to test for characters with martial above say 12? Make a guy with 12 a zero star leader, one with 15, 1, 18, 2 etc. l'm just making up numbers, pick ones that make sense. Add a star for each leader trait. Cav and heavy leaders get their first star in shock, 2nd (if they rate 2) in fire, 3rd in siege (specialist in tactics for melee). Light leaders go fire, siege, shock (bow users). Nonspecialists get siege, shock, fire (overall focus of medieval tactics).

Set thresholds such that there are a reasonable number of excellent generals (i.e. far enough above average to be an eu3 general) but only a handful of multistar legends. Most will be siege specialists with maybe one or 2 tactical geniuses bringing multiple stars to shock or fire. And no admirals, because no naval combat.

 

[Idhrendur]

I've started work on navies. We're getting close to next release.

Except we haven't done anything for ROTW nations. Can some of you analyst types try and figure out either how EU3 knows what armies and navies to place in a new game or a rule that's close enough? Manpower too.

 

[Idhrendur]

Also, I've uploaded testresults-ArmyStuff.zip to the sourceforge page if the analytical types want to see how well armies and manpower are converting.

 

[johnjohansson]

Except we haven't done anything for ROTW nations. Can some of you analyst types try and figure out either how EU3 knows what armies and navies to place in a new game or a rule that's close enough? Manpower too.

Doesn't seem to be an easy answer to this one, as it's calculated at the beginning of a game based on an unknown formula.

This little thread indicates that starting armies are based on economy, force limits, ongoing wars etc. Troop costs(/maintenance?) in particular (i.e. what you can afford) seems to be a big driver.

Not much help , but I haven't had time for more than that, sorry.

 

[Pode]

I'd like to throw out a 4th contender, the johnjohansson method, since he just gave me the idea. CK2 "ending" techs spec out a tech level historically achieved by endgame for various cultural/religious groups. We could map those groups to EU3 groups, compare achieved tech to this historical standard, and westernize or easternize realms that achieved significantly more or less tech development than their historical level. This would be more culturally /geographically driven, encompass more factors of tech development than just schools, and still be a fixed historical standard. I'm going to work on fleshing this approach out.

Figured if I'm going to flesh this out I better do so before the boss gets much closer to a new release.

CK2 tech growth assumes the following 1453 tech levels in each culture group:
Catholic 3.5 Muslim 3.0 Greek 3.5 Other (Pagan, non-Greek Orthodox) 2.5 Renaissance 5.0

So my proposed mapping to EU3 groups:
Catholic & dejure HRE westward = western if tech average >= 3.5, else eastern
Greek = western if tech average >= 4.5, else eastern if tech >= 3.0, else ottoman
Orthodox, Pagan, Catholic east of HRE = western for tech >=4, else eastern for tech >=2.5, else ottoman
Muslim = western for tech >= 4.5, else eastern for tech >= 4, else ottoman for tech >=3.5, else muslim

Basically, if you're a full tech level ahead of historical tech, you get bumped up an EU3 tech group, if you're behind you get downgraded one group. Exceptions are Polish/Hungarian Catholics who'll find it a bit easier to westernize, and Muslims who'll find it hard but possible to fully westernize. If we sort realms into these groups by what de jure their capital is in, we give players some control over how they convert via moving their capital.

 

[Idhrendur]

Doesn't seem to be an easy answer to this one, as it's calculated at the beginning of a game based on an unknown formula.

This little thread indicates that starting armies are based on economy, force limits, ongoing wars etc. Troop costs(/maintenance?) in particular (i.e. what you can afford) seems to be a big driver.

Not much help , but I haven't had time for more than that, sorry.

Yeah, it's one of those unknowns that needs to be reverse engineered. As usual with this projects we're going to have to be the ones to do it. There's a few days before I have any time to work on things, and a few navy-related features to do first of all, but then this will be the roadblock holding us back from a release.

It should be straightforward enough to investigate, though. Just mod simple factors in the game, then start a new game and see how that affected the starting army. I'm guessing base tax (probably just summed across all provinces) is a major factor. Likewise various slider values. Sliders probably account for the army composition.

Army starting position is a weird one, though. That might take some cleverness to figure out.

 

[Idhrendur]

Figured if I'm going to flesh this out I better do so before the boss gets much closer to a new release.

CK2 tech growth assumes the following 1453 tech levels in each culture group:
Catholic 3.5 Muslim 3.0 Greek 3.5 Other (Pagan, non-Greek Orthodox) 2.5 Renaissance 5.0

So my proposed mapping to EU3 groups:
Catholic & dejure HRE westward = western if tech average >= 3.5, else eastern
Greek = western if tech average >= 4.5, else eastern if tech >= 3.0, else ottoman
Orthodox, Pagan, Catholic east of HRE = western for tech >=4, else eastern for tech >=2.5, else ottoman
Muslim = western for tech >= 4.5, else eastern for tech >= 4, else ottoman for tech >=3.5, else muslim

Basically, if you're a full tech level ahead of historical tech, you get bumped up an EU3 tech group, if you're behind you get downgraded one group. Exceptions are Polish/Hungarian Catholics who'll find it a bit easier to westernize, and Muslims who'll find it hard but possible to fully westernize. If we sort realms into these groups by what de jure their capital is in, we give players some control over how they convert via moving their capital.

Can this be extrapolated to seamlessly handle saves from 1399-1453? I don't want to limit users to only converting in 1453. If you can write that up for me, I'm try implementing it so we can see how well it works.

 

[Pode]

Forgive the possibly stupid question, never played around much in EU3 saves, but do we actually need to figure out how to handle ROW? If we're creating a save file, couldn't we just copy all the ROW bits from an unplayed EU3 start? Build an ROW database from unplayed saves starting every Jan 1 between 1399 and 1453, and load the info from the most recent one into the converted save?

 

[Idhrendur]

Forgive the possibly stupid question, never played around much in EU3 saves, but do we actually need to figure out how to handle ROW? If we're creating a save file, couldn't we just copy all the ROW bits from an unplayed EU3 start? Build an ROW database from unplayed saves starting every Jan 1 between 1399 and 1453, and load the info from the most recent one into the converted save?

We could, but that would require redoing a lot of previous work of just reading the history files to figure it out. Which worked perfectly well until we hit an area where the data we need isn't in the history files.