Did anybody actually lose the GC?

[karonagames]

The only way the Axis AI can get close to matching the 1941 historical casualties is if the SU player handicaps himself by adjusting the settings to increase the chance of "shatter" results occurring, as these are equivalent to "surrender" results. A setting somewhere between challenging and hard should be a reasonable starting point. Another approach could be to play on hard until the AI gets to an historical benchmark, then switch to challenging until you get ahead of an historical benchmark, then switch to hard again and so on.

You can adjust the AI settings every turn if you want to. I still can't face a blizzard on challenging, and I see no shame in switching to normal settings for 12 turns.

[amatteucci]

ORIGINAL: Singleton Mosby

ORIGINAL: redmarkus4

- Finally, we need to see the AI make some random (and stupid) offensives, to add the human error that is an essential part of military conflict. At the moment the AI plays it too safe strategically, IMO.

That's a splendid suggestion.

Well, in my recent game the AI did make some unexpected (and in some respect stupid) offensives I don't think there's a necessity to implement this.

What would be interesting (and I guess this is also the actual idea behind redmarkus4 proposal) is to have the AI select some script for renewed offensive when things go static. As I said the "problem" with the game is that the AI didn't show strategic thinking after the eastward rush was contained. It defended competently and counterattacked smartly but if you're only going to play safe like this you'll eventually lose. So, it would be great to randomly select a target like Leningrad-Moscow-Stalingrad-younameit and pull an unexpected grand strategic offensive... but not to make things easier for the player but to make things more difficult!

To do such a thing it would be necessary to script such operations... the problem is that it won't be that easy to implement such a thing since the possible starting situations would be countess (sort of...).
For example, if in WitP you want the Axis AI to assault the Midways starting from the Home Island you can make a script that is appliable in 1941, 1942, 1943, 1944 etc. The only difference will be in the actual units available and the efficacy of the enemy defence. If I want to make a script for the Axis AI in WitE, to take the oilfields near Baku, I cannot assume I'll start from a given place and march along given routes in 1941, 1942, 1943 etc. well, I presume you got the picture! [:)]

[Flaviusx]

There's a big difference between scripting offensive operations to break up a static situation and simply programming an AI to do stupid and random offensives, just because.

Scripting has its own problems: after repeated plays, the human will know exactly what the script is and optimize his response to deal with it. Only way to get around that is to have a selection of scripts chosen randomly given a particular situation. This is a huge project we are talking about here.

[amatteucci]

ORIGINAL: Flaviusx

There's a big difference between scripting offensive operations to break up a static situation and simply programming an AI to do stupid and random offensives, just because.

Yes, and this is why I said that I don't think it's a good thing to implement such a suggestion taken at face value.

Scripting has its own problems: after repeated plays, the human will know exactly what the script is and optimize his response to deal with it. Only way to get around that is to have a selection of scripts chosen randomly given a particular situation. This is a huge project we are talking about here.

Well, it seems that AI scripting didn't create a big replayability problem for WitP/AE. Anyway, I don't know the inner workings of WitE so I cannot comment further on this issue.

[Grotius]

You can adjust the AI settings every turn if you want to.

Can I switch from "Normal" to a higher setting in mid-game? Or do I have to change the Logistics/Fort/etc settings instead?

Assuming it's the latter, what settings would you recommend for playing against the Axis AI? It's now late October 1941, and I'm doing reasonably well (stopped the AI short of Leningrad/Moscow), so I'm considering increasing the difficulty level.

I do think this is the best wargame AI I've ever seen, but even so, it needs a bit of a boost.

[karonagames]

On the games option screen, every number in the the "levels" boxes can be adjusted each turn. Morale has the biggest impact on combat performance; the others are pretty obvious.

I think a level somewhere between Challenging and Hard would be a starting point. I have an an AAR of Road to Moscow in the developers forum played at 115/85 in favour of the Axis, which was a lot of fun. You have to adjust to having less than half of your units in a combat ready state.

[Zovs]

Assuming it's the latter, what settings would you recommend for playing against the Axis AI? It's now late October 1941, and I'm doing reasonably well (stopped the AI short of Leningrad/Moscow), so I'm considering increasing the difficulty level.

Challenging

I do think this is the best wargame AI I've ever seen, but even so, it needs a bit of a boost.

Try it on Hard, if that does not cut the bacon then switch to Impossible.

I only know of two people that have tried Impossible, myself and Andy. I tried it in the 1943 Campaign game with the Germans, the Soviet GAIA created such huge gaps by turn 5 and 6 that I knew I could not fix the leaking so by turn 9 I threw in the towel and said forget it. It was almost as bad (but not quite) as being beat by Andy.

[Grotius]

I just set everything to 120/100 to favor the Axis, except for Axis Admin, which I set at 125. I didn't think of decreasing my own levels, but maybe I should try that?

[karonagames]

In 41, if you want to increase the chance of shattering, you have to reduce Soviet morale by as much as you feel comfortable with.

[Grotius]

Thanks. One thing I noticed with setting Axis (AI) logistics to 120 instead of 100: the German front-line units were almost all in "green" supply, even though it's a snow turn after mud (Turn 22). I might tone that one down, and decrease my own levels a bit instead. I want to help the AI, but I also want to see the historical effect of supply shortages.

[jomni]

ORIGINAL: redmarkus4
- Next we need to see the AI holding onto cities more effectively - the manual says that they are used to calculate VPs, right? Should be important to capture and hold them then.

You've been requesting a fix for this in a lot of your posts. But do you have a suggestion on how to actually fix this?
If they stay in the city, that makes the units easily encircled and eliminated... unless you change the game mechanic that units stuck in a city will last for years under seige.

[amatteucci]

ORIGINAL: jomni

ORIGINAL: redmarkus4
- Next we need to see the AI holding onto cities more effectively - the manual says that they are used to calculate VPs, right? Should be important to capture and hold them then.

You've been requesting a fix for this in a lot of your posts. But do you have a suggestion on how to actually fix this?
If they stay in the city, that makes the units easily encircled and eliminated... unless you change the game mechanic that units stuck in a city will last for years under seige.

I see (in the commander's report) that the amount of supply that the game tracks the amount of supply in each single city. Have units in a city the possibility to draw supplies from that stockpile?

[jomni]

Good idea.  Just like WITP. Wonder if it's feasible.