Matrix Games Forums

I have played both SF,WF, and now both versions of WIR, so I believe I have a good frame of reference. Now, the first thing that I noticed was a more coherent german attack in 41. I think the actions the AI takes pretty well follows history, though not quite as successful. BTW, I have only played Russia into 42 so far this time. Then I noticed the bad stuff.

1)The game can be very unstable. This was never the case in previous versions.
2)The germans are sending way too many units on suicide runs behind my lines. Seems more pronounced in this version.
3)What up with the readiness jumps? Ill have a unit 1 turn have a readiness of 60, and the next turn it will be 5. The units dont perform any action and there was no change in the supply situation or weather. Its crazy! The readiness calculations seem very inconsistent and buggy.

Combining 2 and 3, that is just killing my fun. Ill have to undig my original WIR Im afraid.


Bob L

Originally posted by swatter555
1)The game can be very unstable. This was never the case in previous versions.
2)The germans are sending way too many units on suicide runs behind my lines. Seems more pronounced in this version.
3)What up with the readiness jumps? Ill have a unit 1 turn have a readiness of 60, and the next turn it will be 5. The units dont perform any action and there was no change in the supply situation or weather. Its crazy! The readiness calculations seem very inconsistent and buggy.

Combining 2 and 3, that is just killing my fun. Ill have to undig my original WIR Im afraid.


Bob L

Bob,

Thanks for posting this. Number 1 may, or may not, be a memory issue. Compared to the initial release, the program uses a little more memory, which for me sometimes results in the program locking up, but it is rare. The longer I play, the more likely it is to happen. Could this explain what you are seeing?

#2. There has been no AI changes according to Arnaud, the programmer on this. However, I can see other changes resulting in changes in AI thought, as units may fit whatever conditions the AI wants before sending units into an attack. Not sure what can be done about it, as Arnaud has consistently avoided touching the AI because of the complexity of the code.

#3. The only readiness problem I have seen is that combat doesn't appear to have any real impact on it anymore, unless a unit is forced to retreat. If you have a save game before and after that shows this, could you send it to me at rbancroft@att.net so that I can see what may be causing the problem and get the info to Arnaud? Without a save file, it will be hard to determine what may be causing it since we haven't seen it ourselves.

Thanks,

Thanks for the thoughtful reply. Its not just my imagination, the AI is sending more units on suicide runs. Not only that, there is a nasty bug where the AI is using the march command to "teleport" units behind my solid line.

It wont be hard to catch the readiness problem, Ill send you a before and after when I find one.

The march teleport issue, at least in bad weather is known, but Arnaud has not found the cause yet. The funny thing about that one is I never saw it in the earlier versions, but it can happen, only against the AI now. I have felt the same way as you about the AI play in penetrating, but haven't played it enough to really tell myself, so I can believe it. I have seen it more in AI vs. AI tests though.

Arnaud did fix one AI issue though, where it could stack multiple armies in one hex making your attacks harder to carry out. I always thought it cheated on the spot and sent the extra army in when the first one retreated, but found out in our testing it was placing the second (or even more in some cases) during its turn, so at least it wasn't cheating during your turn:rolleyes: .