User avatar
Crimguy
Lieutenant
Posts: 144
Joined: Sun Jun 17, 2007 4:49 am

AI's decision making . .

Sat Jun 30, 2007 2:37 pm

I've had a few instances in the first year of the war where the AI has been behaving in a suicidal manner. Just bringing it to your attention in hopes that it might be addressable. These examples have been, in my experience, the exception rather than the rule, so it's not as if I can march my way straight to Savannah. . .

First, Ft. Pickens. Braxton Bragg has been throwing his command at that Ft. just about every turn since July '61. His losses have been staggering (1-2000 casualties PER BATTLE versus my <100 per battle). He must have thrown 10,000+ soldiers away at that fort.

Second, the ANV. They seem to have split up their forces and sent leaders off to conduct attacks of very limited value or chance of success. An example is TJ Jackson, who now commands a cavalry regiment, and went down to Ft. Monroe to attack the fort. By himself. His whole unit (994 soldiers) were killed, to my 6 losses.

Solution?: I'm have no idea how the AI does it's calculations, but it seems that it needs to be a little more mindful when calculating whether to attack - perhaps just a bit more conservative. Additionally, it might stand to benefit from knowing that it needs to create a "Death Star" army in the East to apply pressure (or the threat of it) to any opposing force.

User avatar
Winfield S. Hancock
Captain
Posts: 176
Joined: Tue Apr 17, 2007 10:14 pm
Location: Lovettsville, VA, USA

Sat Jun 30, 2007 2:47 pm

I have seen some of these behaviors as well.

I had to stop my last game (1.05a) and restart because the AI sent the Army of the Potomac on a suicide attack across the high mountains of West Virginia towards Pittsburgh. The attack started in late September. By the time they were stopped short of Pittsburgh by my river fleets and figured out they needed to turn around, it was winter and the return march was a frozen hell through blizzards and frozen wastelands. By the time they got back, a very large bulk of the main Reb force in the East was finished. I easily took Richmond, and sensing that the game had lost its 'fun' factor without a force in being to oppose me in the East, decided to restart.

My current game under 1.05d is going much better. I am thinking that this may be due to at least two things -- 1) Pocus's fix to tone down raiding forces, since I have seen a lot less of them other than the CSA cav regiment now sitting in Sandusky OH and 2) variability in what plan the AI picks to use.

I am playing on the Hard difficulty level with normal AI aggressiveness. I am beginning to wonder that for games with the North against the South if it wouldnt be better to set the AI to the 'less aggressive' setting, since it should be on the strategic defense for virtually the whole game.
"Wars are not all evil; they are part of the grand machinery by which this world is governed, thunderstorms which purify the political atmosphere, test the manhood of a people, and prove whether they are worthy to take rank with others engaged in the same task by different methods" -- William T. Sherman addressing the Grand Army of the Republic in 1883

Second in War, Second in Peace, First in the Hearts of His Countrymen -- General Winfield Scott Hancock, USA

User avatar
Pocus
Posts: 25673
Joined: Wed Oct 19, 2005 7:37 am
Location: Lyon (France)

Sun Jul 01, 2007 2:36 pm

Strange, the AI is rather shy most of the time if not always. I'm anyway rechecking several algorithms for 1.06, so there is still light at the end of the tunnel for Athena :)
Image


Hofstadter's Law: "It always takes longer than you expect, even when you take into account Hofstadter's law."

Return to “AGEod's American Civil War”

Who is online

Users browsing this forum: No registered users and 18 guests