Events outcome predetermined?

[Louella]

did the console thing to see the wights and ofc is 600 vs 35 like Louella indicate. If I play with DOD_Romania.txt file and try give a lower factor to the puppet path i still get 600 vs 35. Maybe I don't understand the way factors work, but they look pretty easy.

Huh. Sounds like the game's not reading your changed file properly, and it's still reading the original.

time to doublecheck you have your mod enabled in the launcher, and check for any duplicate or .bak files in your mods /events/ folder

 

[walliart]

I have only the DLCs I have bought but no other mods. Therefore I got no mod folder in Hearts of Iron IV root folder. In the events folder I modify directly in the DOD_Romania.txt and I keep a copy file of original (named DOD_Romania - Copy.txt but being with other name pratically i don't see why game should try read it). Atm looks like I modify in vain that file. Like the event is stored somewhere else and I don't understand why I get these results.
I will do test on the laptop to see if I get same results.


LE on laptop works as intended giving other values. (it is true that i put my copy of original file on desktop now, not leaving in same directory).

 

[SchwarzKatze]

I have only the DLCs I have bought but no other mods. Therefore I got no mod folder in Hearts of Iron IV root folder. In the events folder I modify directly in the DOD_Romania.txt and I keep a copy file of original (named DOD_Romania - Copy.txt but being with other name pratically i don't see why game should try read it). Atm looks like I modify in vain that file. Like the event is stored somewhere else and I don't understand why I get these results.
I will do test on the laptop to see if I get same results.

That's what you're doing wrong. Never directly edit files in the game folder, and the game doesn't care about file names, just the folder the file is in, so it's still reading the old one.

 

[eastcoastceojam]

To add extra info to the original post on this thread, I just started a new game as Nationalist China. I progressed up to the "Subjugate the Warlords" focus, then made a backup copy of my save 3 days before the focus finished.

When I re-loaded the save, 4 of 5 warlords agreed to be subjugated, which is fairly unlikely. I then quit the game, deleted the current save, then copied my old save back to the main directory.

When I re-loaded the save, I got the same 4 out of 5 subjugation result. I tried this 4 more times, but kept getting the same 4 warlords to agree to become my vassal.

Like bitmode reported on the 1st page of this thread, the focus results are no longer random.

 

[TropikThunder]

To add extra info to the original post on this thread, I just started a new game as Nationalist China. I progressed up to the "Subjugate the Warlords" focus, then made a backup copy of my save 3 days before the focus finished.

When I re-loaded the save, 4 of 5 warlords agreed to be subjugated, which is fairly unlikely. I then quit the game, deleted the current save, then copied my old save back to the main directory.

When I re-loaded the save, I got the same 4 out of 5 subjugation result. I tried this 4 more times, but kept getting the same 4 warlords to agree to become my vassal.

Like bitmode reported on the 1st page of this thread, the focus results are no longer random.

Necro alert! I found this thread while searching for info on how AI selects event responses, specifically for the Puppet Bulgaria event while playing DoD Romania.

Just to maybe add a nugget of info to the conversation: I was playing around with changing the event probabilities to get Bulgaria to defy (I don't like puppets) so I changed the "sure" odds to 0 and the "no" odds to 100 and it worked as desired. One thing I noticed is that the factors multiply, which I didn't know (I thought they were one or the other). For example, the unadulterated odds for Bulgaria to accept puppeting are 600 to 35 as stated above. In the Event files, as shown above:

option = { #sure
name = DOD_romania.40.a
ai_chance = {
factor = 30
modifier = {
strength_ratio = { tag = ROM ratio < 0.8 }
factor = 2
}
modifier = {
strength_ratio = { tag = ROM ratio < 0.5 }
factor = 10
}
modifier = {
is_in_faction = yes
factor = 0.5
}
}
ROM = { country_event = DOD_romania.41 }
}

The base for "Sure" is 30, and Romania always has >2x Bulgaria's forces (you need 400k to start the focus). So the odds for "Sure" are 30x2x10 = 600 since they stack. News to me, not new to most of you I'm sure.

The other thing I noticed is that if you re-load a game from within the game (savescum), the event seems to have already been decided, i.e., reloading doesn't change anything. However, if you exit the game and re-start, the game creates a new checksum and then when you load the save, the event decision is now re-rolled. Anyone else experience this, or am I wrong?

 

[kettyo]

One thing I noticed is that the factors multiply

Yes, factors always multiply, the "add" will add or subtract an amount.

 

[bitmode]

A modifier block combines its components like an AND trigger. The only exceptions are factor and add. Both are scoped variables, and if the trigger components evaluate to true, the current MTTH value is multiplied by factor before adding add. Can appear multiple times and all applicable modifiers are applied in the order they were defined in. The calculation stops early if the intermediate value is zero after applying a modifier. This means it is not possible to increase the value again using add after a previous modifier set the value to zero.

(linked from https://hoi4.paradoxwikis.com/AI_modding#AI_chance)

The other thing I noticed is that if you re-load a game from within the game (savescum), the event seems to have already been decided, i.e., reloading doesn't change anything. However, if you exit the game and re-start, the game creates a new checksum and then when you load the save, the event decision is now re-rolled. Anyone else experience this, or am I wrong?

Are you sure you didn't edit the files in the mean time? Restarting should have no impact as far as I can tell:

In the old patch the AI's selection randomization was seeded with the current real world time in nanoseconds, so it was completely unpredictable across reloads.
In 1.9.1 it uses predictable numbers as input:

• unique game seed (contained in the save file, created once when starting the game)
• in-game time when the event triggers
• country ID

All of these values will be the same after a reload (unless you can somehow delay the completion of your NF), which is why the result is the same too.