Printing Press - Heavily biased towards Protestantism?

[Rider_of_Doom]

I had nothing to do, so I decided to create a savegame right before the arrival of the Printing Press to play a little bit with statistics, because the religion map is very interesting and in a good state of balance

There are 81 Protestant provinces, 44 Reformed, and 50 Catholic provinces within the German regions. I reloaded the same save 50 times now (wanted to do it 100 times, but I think this isn't necessary anymore).

One time the PP fired in a Catholic province, three times in provinces, which are Reformed and 46 times in Protestant provinces. That seems odd, if I consult the description of the PP in the wiki:

"The province must be either protestant, reformed, or in one of the German regions."

Certain provinces seems to get it very often too (Slesvig 5 times, Danzig and Bahuslehn 4 times). The PP also never fired in a Reformed province outside of the German regions. The description does not mention any limitations, so I thought, you could get the institution even outside of the German regions, if you go Reformed, because it's possible with Protestant. Is it possible with Reformed?


Apart from that, the spawning seems off. It's correct for the historical context of the press, but in my opinion it's wrong for the gameplay purpose. What do you think?

Edit: Forgot the screenshot of the religion map:

http://images.akamai.steamuserconte...387/18580989190283A9EFF2262E9FD3AA8177F4A750/

 

[bbqftw]

Might be a concern that some of your experiments are using identical seeds (e.g. if you save a campaign on day 1 you will always roll the same general via estate interaction - same pips / same name). I'll do some digging into game files.

edit:

can_start = {
is_year = 1550
is_in_capital_area = yes
is_island = no
OR = {
religion = protestant
religion = reformed
region = north_german_region
region = south_german_region
}

}
start_chance = 5 #in 100 for when a province can_start

doesn't look like there are any biasing factors, except that it must spawn in a capital area (assuming Bahuslan is in same area as Akershus, and Danzig tag is alive this would at least partially explain the biasing). Perhaps @Trin Tragula can shed more light on this.

 

[Rider_of_Doom]

Might be a concern that some of your experiments are using identical seeds (e.g. if you save a campaign on day 1 you will always roll the same general via estate interaction - same pips / same name). I'll do some digging into game files.

I thought that too, but I'm no expert, so I haven't digged in the game files. I saw the same thing with Colonialism, which spawned every time in the English region in one save game. If this would be the case, the institutions aren't that random.

 

[alexti]

I wonder how the "start" check is performed. I can imagine an algorithm that traverses provinces in numerical order and if the province is eligible performs 5% start check and if not successful moves to the next province. The provinces you've listed would get a fairly high chance with this logic

 

[bbqftw]

If the provinces follow the tag numbering system (Scandinavian region first) than that makes complete sense...

 

[Rider_of_Doom]

Might be a concern that some of your experiments are using identical seeds (e.g. if you save a campaign on day 1 you will always roll the same general via estate interaction - same pips / same name). I'll do some digging into game files.

edit:


doesn't look like there are any biasing factors, except that it must spawn in a capital area (assuming Bahuslan is in same area as Akershus, and Danzig tag is alive this would at least partially explain the biasing). Perhaps @Trin Tragula can shed more light on this.

Thanks for the digging.

Bahuslehn is in the same area as Akershus, but Danzig is a core of Pomerania with Stettin as capital. Slesvig belongs to Denmark (Sjaelland still capital). Tried it 10 times more now. Had it 1 time outside of the German regions in Reformed Desmond (Ireland), so it's possible. On the other hand, the 9 other tries were in Protestant provinces again. 3 times Danzig...

 

[alexti]

If the provinces follow the tag numbering system (Scandinavian region first) than that makes complete sense...

The province numbers start in Scandinavia (from Stockholm, of course) and then generally expand around the Baltic sea

 

[Råttan]

I thought that too, but I'm no expert, so I haven't digged in the game files. I saw the same thing with Colonialism, which spawned every time in the English region in one save game. If this would be the case, the institutions aren't that random.

I have always seen it spawn in Spain/Portugal (and one time in France, just at the border to Spain.)

RNGs aren´t perfect, and it shows, but its still as random as computer gets...

 

[Tweakee]

To really test this, you'd need to either use multiple games, or have a way to reliably reset the random seed.

Otherwise, I'd just trust that the code works until more data comes in.

For my one data point, in my last game I turned Aragon+Castille+Naples Protestant, and the Press spawned in Protestant Switzerland.

 

[tre3qwerty]

doesn't look like there are any biasing factors, except that it must spawn in a capital area

Just a quick note: is_in_capital_area actually means "Is connected to the capital" (meaning that there must be a string of adjacent provinces connecting it to the country's capital), it has nothing to do with areas

 

[steveh11]

Johan literally *just* posted this on Twitter:-

In 1.19 for @E_Universalis you will be able to see what influences where an Institution will spawn!

 

[TorquemadaStark]

Germany? Ha, they only got enlightenment in my latest Byzantium game. And yes, manufactories and global trade spawned both just 1 province away from the rennaisance spawn. Fun times.

 

[MaXimillion]

Johan literally *just* posted this on Twitter:-

We can already see that information by looking at the game files, doesn't change anything as far as this thread is concerned.

 

[Rider_of_Doom]

To really test this, you'd need to either use multiple games, or have a way to reliably reset the random seed.

Otherwise, I'd just trust that the code works until more data comes in.

For my one data point, in my last game I turned Aragon+Castille+Naples Protestant, and the Press spawned in Protestant Switzerland.

It works perfectly for the Protestant side. The problem is, that Catholic (within German regions) and Reformed provinces don't get the institution the same amount too.

Can you reset the random seed for a game?

 

[SolSys]

I had nothing to do, so I decided to create a savegame right before the arrival of the Printing Press to play a little bit with statistics ...

This is your main problem.
While the institution has a chance to appear in any province the fits the criteria, by using a save file with short time frame until the event there is not much time for additional provinces to qualify [as opposed to starting a new game].

You also have to take into account that the mechanic, in general**, favors richer provinces.


**as opposed to individual requirements.

 

[Rider_of_Doom]

This is your main problem.
While the institution has a chance to appear in any province the fits the criteria, by using a save file with short time frame until the event there is not much time for additional provinces to qualify [as opposed to starting a new game].

You also have to take into account that the mechanic, in general**, favors richer provinces.


**as opposed to individual requirements.

I saved on the 1st December and let it go from there. After that I closed the game, restarted and loaded the save again. And what would change it, if I would do a run with more time to pass? Shouldn't the provinces, which were selected before, still in the seed?

Can't really confirm this from this. Only 5 out of 25 provinces are yellow or higher (4 yellow, 1 green). Currently at 66 and Danzig is the leader with 8 times with 16 development.

 

[SolSys]

I saved on the 1st December and let it go from there. After that I closed the game, restarted and loaded the save again. And what would change it, if I would do a run with more time to pass? Shouldn't the provinces, which were selected before, still in the seed?

Can't really confirm this from this. Only 5 out of 25 provinces are yellow or higher (4 yellow, 1 green). Currently at 66 and Danzig is the leader with 8 times with 16 development.

Could you cross-reference the origin provinces with development?
It would be helpful to know where provinces that got the institution the most times place on the development chart against other less fortunate provinces.

 

[Rider_of_Doom]

Could you cross-reference the origin provinces with development?
It would be helpful to know where provinces that got the institution the most times place on the development chart against other less fortunate provinces.

I have noted every province, so that's no problem. I will complete the 100 (currently at 68) and post the list here shortly.

 

[Issac1709]

I have noted every province, so that's no problem. I will complete the 100 (currently at 68) and post the list here shortly.

You mind using another save file (AKA new start) and then doing this another 30 or 50 times and see if the results are largely different?

 

[Rider_of_Doom]

You mind using another save file (AKA new start) and then doing this another 30 or 50 times and see if the results are largely different?

I can do this until tomorrow. No problem at all.

Current list will take a little bit longer (RL that bastard).