Page 1 of 1

AI scripts for jap CV TFs

Posted: Sun Aug 16, 2009 4:34 pm
by PavelA
warning: May contains spoilers...

Hi all,

I play a scenario #2 as allies with AI difficulty "very hard". Till the end of the year 41, japanese CVs returned from Pearl Harbor attack and they began with an aggressive strategy in DEA. It means: AI divide Kido Butai into several TF and set sail to the Makassar strait and Java Sea. To my surprise, AI is not able to set up appropriate escort for these CVs nor it is able to set an surface group with/ahead them.

My surface TFs encounter these TFs in night attacks in the Makassar strait, and Java sea (near Sorebaja); it's 31st of December 1941, and japanese navy mourns for two heavy CV, CVL (and several other type of ships), which were pummelled by a group of dutch DDs and remnants of Force Z.

Here I post a composition of particular japanese CV task forces, you can judge why AI set TFs like this.

1) 1xCVL + 1x DD
2) 2x CV + 2xCA + 3xDD
3) 1xCV + 2xDD

I have to remark that these TFs have never been attacked by allied forces before - they were fresh and new and intact. I know that in WITP it was common to send jap CVs to Java sea because they effectively cut off all naval movement of allies. However in AE, this strategy is not good - these TFs can be attacked during theirs movement in the straits by allied surface attack TFs. It is in contrast with WITP, where you have to be a lucky man to left your surface TF at the hex where the CV TF end the resolve phase.

Resume: I suggest to have a look at the composition of CV task forces produced by AI and than re-value AI scripts based on the non-AE (=WITP) strategy.





RE: AI scripts for jap CV TFs

Posted: Sun Aug 16, 2009 5:27 pm
by herwin
ORIGINAL: PavelA

warning: May contains spoilers...

Hi all,

I play a scenario #2 as allies with AI difficulty "very hard". Till the end of the year 41, japanese CVs returned from Pearl Harbor attack and they began with an aggressive strategy in DEA. It means: AI divide Kido Butai into several TF and set sail to the Makassar strait and Java Sea. To my surprise, AI is not able to set up appropriate escort for these CVs nor it is able to set an surface group with/ahead them.

My surface TFs encounter these TFs in night attacks in the Makassar strait, and Java sea (near Sorebaja); it's 31st of December 1941, and japanese navy mourns for two heavy CV, CVL (and several other type of ships), which were pummelled by a group of dutch DDs and remnants of Force Z.

Here I post a composition of particular japanese CV task forces, you can judge why AI set TFs like this.

1) 1xCVL + 1x DD
2) 2x CV + 2xCA + 3xDD
3) 1xCV + 2xDD

I have to remark that these TFs have never been attacked by allied forces before - they were fresh and new and intact. I know that in WITP it was common to send jap CVs to Java sea because they effectively cut off all naval movement of allies. However in AE, this strategy is not good - these TFs can be attacked during theirs movement in the straits by allied surface attack TFs. It is in contrast with WITP, where you have to be a lucky man to left your surface TF at the hex where the CV TF end the resolve phase.

Resume: I suggest to have a look at the composition of CV task forces produced by AI and than re-value AI scripts based on the non-AE (=WITP) strategy.

If you do a google search on ijn cruiser and battleship doctrine there's a whole lot of interesting articles relevant to these issues.

RE: AI scripts for jap CV TFs

Posted: Sun Aug 16, 2009 5:35 pm
by Andy Mac
its all interesting but you need to make the AI form TF's the AI IS NOT CAPABLE of prioritising CV TF's over other forms so if I specifiy a large CVTF of max size then it will only for the TF if the correct escorts are present which basically means it will never form
 
There are limits to what I can do guys as it si if the AI loses to many CA's it probably wont form KB
 
As I said there are limites we work with what we have

RE: AI scripts for jap CV TFs

Posted: Sun Aug 16, 2009 6:03 pm
by PavelA
ORIGINAL: Andy Mac

its all interesting but you need to make the AI form TF's the AI IS NOT CAPABLE of prioritising CV TF's over other forms so if I specifiy a large CVTF of max size then it will only for the TF if the correct escorts are present which basically means it will never form

There are limits to what I can do guys as it si if the AI loses to many CA's it probably wont form KB

As I said there are limites we work with what we have
Forgive me guys, it was not intended in negative way (if it was written so, it is due to my English - I'm not a native speaker). I'm really happy to play the witp and ae and I really don't want to mean the text above bad.
Frankly, I was surprised with the weak escort of CVs, I tried to find the issue in this subforum without result, thus I wrote the report as I take it as a possible bug.

You have done an impossible thing with releasing AE - such a complex game. May be I took it as if the AI can be unbeatable.


RE: AI scripts for jap CV TFs

Posted: Sun Aug 16, 2009 7:12 pm
by Andy Mac
Thats ok but there is a reall issue
 
The Ai has a number of settings when I tell it to form TF's
 
Large/Normal/Small
Min/Normal/Max
 
Whithin these parameters the Ai will try to form TF's if I specify Large Max TF it needs a load of escorts and will NOT form if lessa re available - this is a BAD thing if you need a Carrier TF to form so Max setting only get used where you genuinelly dont want a TF to form if it cannot be well escorted.
 
More typically are Normal/Normal or Large/Minimum or Normal/Minimum settings where the AI will try to create a TF using as many escorts as are available as long as they are above the  minimum sometimes if all the DD's and CA's are off doing other things e.g. the opening where there are a lot of TF's to cover the AI will have smaller than normal TF's
 
Later as the initial invasion phase is over more fleet assets return to various ports and become available so TF's get stronger.
 
(This is where Scens like the ironman and scen 2 help because there are more DD's and other appropriate escorts floating about to strengthen these screens)