Home Trending Latest threads New posts Developer diaries

Ai

Jump to latest Follow Reply
  • 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.
Ok right, having written my last exam for this year soon, I should be able to address some of the grave issues. What especially worries me are the constant reports about Austria-Hungary (which I can verify), and I will look into that issue as it could be considered a possible game breaker.
 
Thanks everyone :D

TurnerBenton: Sure, thanks. I downloaded the 19th December fix and will look into what values you exactly changed so I can incorporate them into my future AIs.
 
Good luck in your AI work. :)

Keep us updated on what files you are working on, then we can be able to provide our opinions on what deployment and construction piorities these files should have.
 
Xennik said:
My computer will be out of repair one of these next days. This means that I will be able to restart work. :)
Click to expand...

Good news. :) I am very glad that you are keeping up updated.
 
I am sure you remember that we intended to solve the problem of the AI's erratic ship movements by writing AI events that lock the naval units in place until the outbreak of war.

I have just written events for the German AI with the intention of preventing Königsberg and the Ostasiengeschwader journeying to the Jade in January 1914.

It appears that the 'lock_division' causes the game to crash. (perhaps someone could test this command on their version of the game?)

So I propose a different course of action: AI events that delete the relevant units on the outbreak of war and automatically replace them with identical units in their correct locations.
 
Additionally, if anyone has an idea or two about how to prevent the French AI from immediately moving large numbers of troops into Belgium upon the outbreak of war, I would be delighted to read them.

This habit of the AI is ahistorical - French divisions take the principal role in defending Belgium long before General Lanrezac's Fifth Army actually did.
 
Allenby said:
Additionally, if anyone has an idea or two about how to prevent the French AI from immediately moving large numbers of troops into Belgium upon the outbreak of war, I would be delighted to read them.

This habit of the AI is ahistorical - French divisions take the principal role in defending Belgium long before General Lanrezac's Fifth Army actually did.
Click to expand...

Speaking of long absences...

I believe it was a fault of mine - I think the code for WW2 still exists that forces troops into Belgium. A believe a delay (passivity) for France into Belgium for 20-25 days would be sufficient.

It would also be worthwhile to delay the movements of German troops across the Belgian border for 15 days to represent the "concentration" portion of the opening days of the war, and to broadly represent the difficulty Germany had in capturing Liege prior to the arrival of heavy artillery.
 
under 'front = ' it is the 'enemy_handicap =' and it works handily I generally use 15-18 days.

However, with ships I have recently come to understand three factors to control ship movement in the case mentioned above.

Under the ai command, 'invasion = ' have 'ignore' set to 'yes' and then under ai 'admiral = ' list by name the sea zones with the 'ignore =' command there.

Lothos just said in another thread main offense to this command is the ai attempting to set up or support an amphibious assault in any province lying along or adjacent to one of those ignored sea zones.

With this it would seem if you set the other 'invasion' modifiers for Germany such that it would not consider any amphibious invasions?

One idea from my test would be to list under the ai 'invasion' command, under 'target = ' here list provinces throughout Europe proper or north with say 9 = -1000 etc. This should make it very clear to the German invasion ai to stay clear of trying to set up any amphibious invasion relating thereto.
 
Yukala said:
under 'front = ' it is the 'enemy_handicap =' and it works handily I generally use 15-18 days.

However, with ships I have recently come to understand three factors to control ship movement in the case mentioned above.

Under the ai command, 'invasion = ' have 'ignore' set to 'yes' and then under ai 'admiral = ' list by name the sea zones with the 'ignore =' command there.

Lothos just said in another thread main offense to this command is the ai attempting to set up or support an amphibious assault in any province lying along or adjacent to one of those ignored sea zones.

With this it would seem if you set the other 'invasion' modifiers for Germany such that it would not consider any amphibious invasions?

One idea from my test would be to list under the ai 'invasion' command, under 'target = ' here list provinces throughout Europe proper or north with say 9 = -1000 etc. This should make it very clear to the German invasion ai to stay clear of trying to set up any amphibious invasion relating thereto.
Click to expand...

Thanks for posting - I think this will have to be tried in order to see if it can effectively stop the AI from sending large numbers of naval units into the North Sea too often.

This is important with respect to the Royal Navy in particular. The Grand Fleet ought to be confined to port much more often than the High Seas Fleet given how keen the Germans were to attack the eastern coast of England and coax a portion of the Grand Fleet out to battle.
 
TurnerBenton said:
Speaking of long absences...

I believe it was a fault of mine - I think the code for WW2 still exists that forces troops into Belgium. A believe a delay (passivity) for France into Belgium for 20-25 days would be sufficient.

It would also be worthwhile to delay the movements of German troops across the Belgian border for 15 days to represent the "concentration" portion of the opening days of the war, and to broadly represent the difficulty Germany had in capturing Liege prior to the arrival of heavy artillery.
Click to expand...

At the moment, the values in the French war file are as follows:

Code: 
      enemy_handicap = { GER = 15 } # Number of days before AI takes any action against a named enemy

Code: 
	passivity = { # % chance of taking no action against that enemy
			ITA = 50
			U11 = 80
			GER = 50
			 }

I think that the French ought to act well before 15 days against the Germans - they should be attacking Alsace-Lorraine immediately but without sending anything much into Belgium. That is the tricky part.

I wonder whether these passivity figures ought to be altered? Surely it should read 'GER = 5' or 'GER = 0'? :confused:

The most problematic aspect of the German AI is not that it cannot concentrate enough divisions, but that it leaves too many of them in Cologne when Belgium is attacked.
 
You must log in or register to reply here.
Top Bottom
Back