Weird AI behaviour (both sides) in Bulge to the Rhine scenario (Full Version)

All Forums >> [New Releases from Matrix Games] >> Gary Grigsby's War in the West



Message


rocketman71 -> Weird AI behaviour (both sides) in Bulge to the Rhine scenario (7/17/2021 2:30:46 PM)

I first started playing this as Axis and had a first turn breakthrough in the Ardennes (as expected) and to the south as well. To my surprise the AI then retreated completely from Aachen and hexes nearby, Nijmegen and several hexes west of the Maas. Seemed uncalled for. I then started the scenario as Allies and the AI advanced through the Ardennes (as expected). I plugged the line best I could with available forces, but then on turn 2 the AI retreated from all conquered hexes and withdrew from the Ardennes completely, despite a successful advance. I started the scenario again as Allies and the same thing happened again. It seems something is wonky with the AI IMO.




Naughteous Maximus -> RE: Weird AI behaviour (both sides) in Bulge to the Rhine scenario (7/18/2021 1:53:00 AM)

Sometimes you never know what the AI is going to do. It's best to play a human opponent but if you don't have that option, I usually will play myself. Just play your best for both sides and you won't be disappointed. If you have a dual personality, works even better! [sm=00000016.gif]




rocketman71 -> RE: Weird AI behaviour (both sides) in Bulge to the Rhine scenario (7/18/2021 1:36:23 PM)

I want to play more before taking on a human opponent. But a scenario dealing with "the Bulge" can't have German AI retreat after the initial breakthrough and not even doing a single attack on turn 2. Makes no sense.




loki100 -> RE: Weird AI behaviour (both sides) in Bulge to the Rhine scenario (7/18/2021 2:04:22 PM)

few bits.

I genuinely doubt there will be any more work on the WiTW AI as such. From the development of WiTE2 I'm aware how hard it is to first get the AI to become flank aware and then to actually stand if it feels its flanks are exposed. The WiTE2 Red God of War scenario has a similar problem to the Bulge situation described and took a lot of work to get the German AI to behave.

The attacking AI may well be scripted to bail out after one turn, or the above applies. it will only have a formal script for T1.

Generally giving it bonuses will help address the first point. If you have any grasp of the basic game systems I'd say the AI needs to be on 110 for morale if defending and 120 if you want it to sustain an attack. In WiTW those values don't trigger the secondary AI behaviour that it does in WiTE2 so its just a linear boost to CV.

One reason for doing this, is the AI is less than stellar in constructing its HQ chains so will lose a lot of the bonus that a player gains that way, in effect +110 is probably a net +105.

in WiTW, the Axis AI handles the 1943 GC pretty well, esp at 110. It gets the big things 'perfectly ok' and that masks the occasional localised mistakes. In a small scenario, esp one that opens with a pretty silly offensive, its not a surprise the AI tries to resolve the situation in its own way.




rocketman71 -> RE: Weird AI behaviour (both sides) in Bulge to the Rhine scenario (7/18/2021 5:49:04 PM)

I get that AI is hard to perfect, but in this case it might be better to bill the scenario as "best suited for PBEM play or Axis vs Allies AI". I was surprised because I felt that the AI did good in the NA and Italy scenarios I've played.




Page: [1]

Valid CSS!




Forum Software © ASPPlayground.NET Advanced Edition 2.4.5 ANSI
2.734375E-02