Half the game is unplayable
- Thread starter PossibleTango
- 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.
Really? Shouldn't the game simulate the pops in countries you don't control as well? Or do they get less detailed when there is no player involved?
I think paradox has finally developed a game that needs proper comparison benchmarks at like 1900 mark so people can actually get an idea what kind of computer Victoria needs. There is alot of confusion what is actually causing their game to slow down. Benchmark functionality could really help player diagnose their own issue as well as improve paradox understanding of their own game performance.
My guess is that people are having an issue with their GPU not their CPU, why do you think the performance improves so much while so zoomed out or maxed zoomed in? Being zoomed out would not have any effect on pop calculations they are always working unless the game is paused. However the game creates a graphical asset for each X pop amount and each building, so the GPU resources required to run the game in 1836 and 1900 is going to be vastly different. I noticed I would get much bettter performance in the max zoom , where the game doesn't have to render so many buildings as compared to that mid level zoom when the game switches to rendering buildings. Also playing a game as japan where it's mostly water on the screen is much smoother than to move the camera mid level zoom over central europe, watch your FPS go to crawl in those dense building areas.
My guess is that people are having an issue with their GPU not their CPU, why do you think the performance improves so much while so zoomed out or maxed zoomed in? Being zoomed out would not have any effect on pop calculations they are always working unless the game is paused. However the game creates a graphical asset for each X pop amount and each building, so the GPU resources required to run the game in 1836 and 1900 is going to be vastly different. I noticed I would get much bettter performance in the max zoom , where the game doesn't have to render so many buildings as compared to that mid level zoom when the game switches to rendering buildings. Also playing a game as japan where it's mostly water on the screen is much smoother than to move the camera mid level zoom over central europe, watch your FPS go to crawl in those dense building areas.
- 3
- 2
As far as I can tell merging mechanics don't work like that.
Every POP is defined by 4 major factors - culture, religion, job and location, as per the description. Plus there are attributes - SOL, literacy, citizenship status, radical/loyalist, IGs, political strength and current political movements. wages, consumption and dependents.
As long as there is a single person of a particular culture, job and religion - there would be a POP for him. When you start the game - POPs rarely move around, and even when they do, they do that inside your country. No colonies, no sprawling empires (the British is the only one), restrictive citizenship/migration laws and SOL on the lower end means all these varied nations of the world stay at home.
Give the game 30 years - all of these factors change. Now there are colonies everywhere, SOL is all over the place, people migrate. I think in vanilla game in observer mode it would be better - since AI is braindead, their laws and SOL won't go up much. But if a player controls a decently sized nation - or even starts and minor and goes up in size and rank, their country inevitably would skyrocket at some point. And would start attracting people.
Here an example - the standard game of Russia for me always ends up taking Persia for sugar, Afghanistan for opium and Sindh for dyes, Congo for rubber plus Mali for Mosque of Djenne wonder (20% education access is a must for low literacy countries) I don't want to have these lands (for roleplaying reasons mainly), but since I can't invest and AI NEVER builds up needed stuff - I have to take them. All of that already makes internal migration tricky. But add to that inevitable SOL increase by the midgame to 16-18 and voila, now I have Irish farmers in Tver and Belgian clerks in Bryansk. Each and every one of these POPs have their own calculations, for all of their attributes.
The only way to fix performance right now is to engage full genocide mode - set assimilation rate to something absurd, like 200k a week. So freshly migrated POPs would be almost immediately assimilated into the borg upon arrival. That's not the worst idea for monocultiral nations, but as soon as we look at Austria, Russia, Ottomans and whole lot of others that becomes an issue - for the 'fix' to work, these countries need to finish off minor cultures in their initial borders in the first 20 years.
The issue itself is that it appears the game can't properly group POPs for ease of calculation. At the same time you can't fix it by instakilling tiny POPs (<100) - since they might be just the first ones in the upcoming wave of refuges, migrants or colonizers. And POP mechanics is amongst the core mechanics of the game. To be honest - kinda moot, even big multicultural repressive empires don't have an issue with discriminated POPs - the only issue is potential revolt, nothing in the internal politics that would be interesting. So right know setting cultural assimilation into genocide mode practically won't change anything in terms of game mechanics but might fix the performance.
- 1
I dont understand what that has to do with what i posted. I tried modding the game to increase pop merging and assimilation but its obviously not working, these tiny cultures are not being insta merged/assimilated in my country.
If i set min assimilation to 10k, that should mean that a min of 10k pops of a culture will assimilate each month, but that is not happening.
I was talking about merging POPs initally.
As for the assimilation - the wording in the commentary leads me to believe min_assimilation calculated after the overall assimilation and hence can't be higher:
ASSIMILATION_RATE = 0.02 # Percentage of pops who will Assimilate into a different Culture each month, if eligible
MIN_ASSIMILATION = 100 # If there is any assmiliation at least this many people will assimilate
But the be sure you need to check overall assimilation speed before and after changes to both rates
Also game don't specify how it chooses the target for assimilation, it might be on a single state level or country-wide. Probably hardcoded
BTW I set mine to 0.02 and 100, and played mid-game for 5 more years. The effect, as in so far as I can tell, is minimal of even a placebo
I would love to be able to compare the length of time a week takes to simulate for someone who's game is running "well" to have a more definable metric here. It's painfully slow for me in the end game now -- I left and read a book for 2 hours to get my end game achievements, on a rig built to run VR games on an Index (32 gigs of ram, 2070 super, ryzen 5 3600). My GPU is not being maxed out, my CPU isn't, and despite the game using 14 gigs of RAM I still have plenty left.
Would anyone be willing to post a video of post 1910 to show how long a week takes them to process? Someone who says the game runs well on their setup. It feels like we just have different standards here -- there are so many people with top of the line specs reporting drastic performance issues late game that I feel like we need a basis of comparison.
Would anyone be willing to post a video of post 1910 to show how long a week takes them to process? Someone who says the game runs well on their setup. It feels like we just have different standards here -- there are so many people with top of the line specs reporting drastic performance issues late game that I feel like we need a basis of comparison.
- 6
The last time I rand a test was with 1.05 so before the performance fix of merging pops. Then I had 4~6s per week in 1916 (https://forum.paradoxplaza.com/forum/threads/1-06-when.1556073/page-4#post-28612076)
Now with 1.1.2 it is again/still around 4s (stopwatch says 4.27s). This is in 1901 with France with 1.9billion GDP and decent amount of auto build active. Didn't focus on economy that much, was mostly a game for the Frech Comune achivement.
Now with 1.1.2 it is again/still around 4s (stopwatch says 4.27s). This is in 1901 with France with 1.9billion GDP and decent amount of auto build active. Didn't focus on economy that much, was mostly a game for the Frech Comune achivement.
I didnt see changes in that setting help. But i tried to broaden the pop fragmentation fix settings pdx introduced. Its set to pop sizes smaller then 10 and minimum 4 different cultures per workplace. After increasing minimum pop sizes and reducing minimum pop numbers to 2 did it really kick in.
I tried it in my first run where i tried to find out "how much can the world support in GDP". It immedeatly started to increase my mainculture by millions per week. So it does work. Question is simply on how high one wants to set it and what consequences to demographics one can accept for the sake of performence.
The Oberoni Fallacy, though originally formulated for tabletop RPGs, is fully functional for cases like this as well. "The end user can fix this issue on their own" doesn't mean the issue doesn't exist, nor that it is the end user's duty to fix it.
- 14
- 6
I kinda figured it out. At least I think I did - POPs can only be assimilated when living inside incorporated states
You can set assimilation rate to basically instantly assimilate anything - you need to have multiculturalism, since you can only assimilate non-discriminated cultures. Having it around even 50 is enough. But you have to make sure the state is incorporated.
I checked in my Russia game, for the sake of testing I released conquered Persia as a subject - it took 3 weeks to clean up all micro-POPs
Since I don't see any mentions of assimilation apart from the base speed and some multipliers, I have to conclude that you can't enable assimilation in non-incorporated states as of right now. You can always incorporate to avoid POP mess, but you can't Force AI to do the same and the scramble for Africa inevitably would make a mess.
Maybe If you disable both multiculturalism and open borders, so POPs from discriminated cultures can never move in...
You can set assimilation rate to basically instantly assimilate anything - you need to have multiculturalism, since you can only assimilate non-discriminated cultures. Having it around even 50 is enough. But you have to make sure the state is incorporated.
I checked in my Russia game, for the sake of testing I released conquered Persia as a subject - it took 3 weeks to clean up all micro-POPs
Since I don't see any mentions of assimilation apart from the base speed and some multipliers, I have to conclude that you can't enable assimilation in non-incorporated states as of right now. You can always incorporate to avoid POP mess, but you can't Force AI to do the same and the scramble for Africa inevitably would make a mess.
Maybe If you disable both multiculturalism and open borders, so POPs from discriminated cultures can never move in...
- 3
What I dont get is why POP_MERGE_MAX_WORKFORCE doesnt seem to work, at least not for high values. Im not seeing all these tiny pops getting merged if i set it to have a workforce of 100 or less. Unless theres some kind of hidden max and going above that breaks pop merging...
I set POP_MERGE_MIN_NUM_POPS_SAME_PROFESSION = 2 and POP_MERGE_MAX_WORKFORCE = 100 but i still see lots of small pops everywhere...
I set POP_MERGE_MIN_NUM_POPS_SAME_PROFESSION = 2 and POP_MERGE_MAX_WORKFORCE = 100 but i still see lots of small pops everywhere...
As far as I can tell the game can't force merge POPs of different culture and religion. But maybe something is broken, it is supposed to do exactly that but can't
I suggest to install the "save your CPU" mod. It simplifies the graphics and has a function to merge pops. Worked well for me and allowed to play the game to 1870 on my surface book 1.
- 1
I tried POP_MERGE_MIN_NUM_POPS_SAME_PROFESSION = 2 POP_MERGE_MAX_WORKFORCE = 10000000 for my World GDP run. It immedatly started to merge millions into Brititsh.
Could it be that it doesnt merge them in your case because the moment you have a lvl 10 building the workforce working there probably has a minority employment far greater then 100?
While it seems that it is not solely graphical issues, I do have to beat one of my personal dead horses:
Strategy games don’t need pretty graphics, and they *especially* don’t need dynamic graphics. No waves, no clouds, no birds, no trains, etc. If it moves, it better be because the player told it to. Same for meticulously rendered topography.
Strategy games don’t need pretty graphics, and they *especially* don’t need dynamic graphics. No waves, no clouds, no birds, no trains, etc. If it moves, it better be because the player told it to. Same for meticulously rendered topography.
- 14
- 5
- 2
But the whole point of force merging is to merge pops of different culture and religion....otherwise there is nothing to merge them with...
Hmm not what im seeing in the screenshot here : https://forum.paradoxplaza.com/forum/threads/half-the-game-is-unplayable.1563014/post-28692278
I will try a larger number like 1000 or 10k just in case.
Sure, but the map looks gorgeous. My problem is that the rendering while in state view / mid zoom is too much. I don't want to take those performance hits because I'm trying to build something, thanks.
Agreed on this. I don't really get who's asking Paradox for all the graphics work? I don't think I've ever seen a thread on thr Paradox forums or subreddits asking for fancy dynamic maps. They just showed up one day and we were stuck with them.
So it's a little galling to me that long-standing issues like performance that have never been truly solved are being exacerbated in their name. I wouldn't really care so much if it had no net impact elsewhere (though an option to have just a paper map view would still be nice...) but it's costing everyone heavily in exchange for something extremely few people if anyone actually wants or cares about.
- 14
- 3
- 3
I don’t want a gorgeous map. Take, for example, EU4’s simple terrain mode. Or Civ 5’s and 6’s strategic view. I’m not a fan of those two games, but the strategic view was a masterstroke of game design right there.
Paradox really should take that design philosophy and run with it: design the entire game, every single interface and map mode to work like Civ’s strategic view. Then, after that, take the time to make a more high-res mod with over-the-top animated doodads.
- 6
- 3
- 1