Once the Japanese AI captures the SRA and the Solomons/Port Moresby does it do anything else???

[Tanaka]

I have a CHS AI vs AI going where the Japanese AI is doing very well everywhere except Burma where it let a couple of divisions get trapped. It has captured everything the Japanese captured historically except in the Burma area and unhistorically has captured Port Moresby. But since all that has happened....nothing. And the game goes on and on and on....and nothing...

Does the AI do anything after capturing what the Japanese did historically???

Will it ever attack Midway???

The Aleutians???

Australia???

Noumea???

Hawaii???

Or does it just stop and die???

[wild_Willie2]

what do you expect of the AI, be fortunate that it took the SRA at all.
Never hapened in my game, let alone letting it grab port mor.

[Banquet]

Unlike the 'unofficially proposed' WitP II, WitP has an AI so I imagine he wants it to do something.. hardly an unreasonable expectation..

Tanaka, there were some reported problems with the AI going to sleep on the CHS mod.. I wonder if this is what you're experiencing?

[kaiser73]

ORIGINAL: Banquet

Unlike the 'unofficially proposed' WitP II, WitP has an AI so I imagine he wants it to do something.. hardly an unreasonable expectation..

Tanaka, there were some reported problems with the AI going to sleep on the CHS mod.. I wonder if this is what you're experiencing?

More the game is complex, and harder is to get a good AI. There is no strategic game around which has a good AI (and AI which can give an average human player some challenge). Giving AI unhistorical advantages is what they do in every game.

AI conquered all SRA as Japan did historically. I guess that is good for the AI. if you want smart moves and unexpected plans...play PBEM. it may sound harsh, but that is how it is in every game out there.

[Banquet]

ORIGINAL: kaiser73

ORIGINAL: Banquet

Unlike the 'unofficially proposed' WitP II, WitP has an AI so I imagine he wants it to do something.. hardly an unreasonable expectation..

Tanaka, there were some reported problems with the AI going to sleep on the CHS mod.. I wonder if this is what you're experiencing?

More the game is complex, and harder is to get a good AI. There is no strategic game around which has a good AI (and AI which can give an average human player some challenge). Giving AI unhistorical advantages is what they do in every game.

AI conquered all SRA as Japan did historically. I guess that is good for the AI. if you want smart moves and unexpected plans...play PBEM. it may sound harsh, but that is how it is in every game out there.

I certainly don't disagree. Obviously for the best challenge (especially after learning how an AI plays) PBEM is the way to go.

I don't agree that all stragegic level AI's are completely lacking in challenge. I find HoI2's AI gives me plenty to think about as long as I don't go too far off the rails.

However, my point was that the AI should be doing something. (I dont' recall ever seeing a post where someone demanded an AI that was on a par with playing a human)

This game did ship with an AI and I don't see why, on the odd occasion it's ever mentioned on this forum, the majority of responses seem to be either 'well, it's the AI, what did you expect' or 'play PBEM' Fair enough if people are complaining about the AI's lack of a genius plan of operations.. but hardly relevant when Tanaka is asking why the AI is doing 'nothing'

[hithere]

one thing though is, what can the AI really do once everything historically captured (and plus port moresby) is captured. even the original japanese plan did not call for anything else to be captured, if i recall correctly. japan does not have the resources to capture anything else save maybe more of china.

also, is maybe both AI's waiting for the other to do something to trigger a response?

[Banquet]

ORIGINAL: hithere

one thing though is, what can the AI really do once everything historically captured (and plus port moresby) is captured. even the original japanese plan did not call for anything else to be captured, if i recall correctly. japan does not have the resources to capture anything else save maybe more of china.

also, is maybe both AI's waiting for the other to do something to trigger a response?

That might well be the case (although I'm a bit surprised by the lack of Midway)

[kaiser73]

ORIGINAL: Banquet

ORIGINAL: kaiser73

ORIGINAL: Banquet

Unlike the 'unofficially proposed' WitP II, WitP has an AI so I imagine he wants it to do something.. hardly an unreasonable expectation..

Tanaka, there were some reported problems with the AI going to sleep on the CHS mod.. I wonder if this is what you're experiencing?

More the game is complex, and harder is to get a good AI. There is no strategic game around which has a good AI (and AI which can give an average human player some challenge). Giving AI unhistorical advantages is what they do in every game.

AI conquered all SRA as Japan did historically. I guess that is good for the AI. if you want smart moves and unexpected plans...play PBEM. it may sound harsh, but that is how it is in every game out there.

I certainly don't disagree. Obviously for the best challenge (especially after learning how an AI plays) PBEM is the way to go.

I don't agree that all stragegic level AI's are completely lacking in challenge. I find HoI2's AI gives me plenty to think about as long as I don't go too far off the rails.

However, my point was that the AI should be doing something. (I dont' recall ever seeing a post where someone demanded an AI that was on a par with playing a human)

This game did ship with an AI and I don't see why, on the odd occasion it's ever mentioned on this forum, the majority of responses seem to be either 'well, it's the AI, what did you expect' or 'play PBEM' Fair enough if people are complaining about the AI's lack of a genius plan of operations.. but hardly relevant when Tanaka is asking why the AI is doing 'nothing'

You get that responce by disillusioned gamers i guess. I myself played HoI, HoI2, EU2 (just to mention the Paradox games). In all of them i won easily since first game. In HoI2 i had to edit config files to give AI 3 times the industrial/research and still they were no brainer.
And i don't think i am Rommel. Actually, i am no more than an average wergamer.

So yes, when i get a wargame, i do know it's to play PBEM/online. if there isn't that option i don't even take it in consideration. i never found a good AI in any strategy game made in last 3-4 years. now i don't even hope i get.

An AI which is able to get all Japan got historically and then set on defensive is good.

[Sardaukar]

i've got one Campaign into July 1944 from start vs. IJ AI. After certain areas have been captured, AI seems to get quite cautious. I don't think it falls asleep, but it doesn't want to extend it's defensive perimeter too much. I did find out that it did react to me capturing Tinian etc., resulting to carrier battle which didn't go too well to IJN..it's 1944 anyway. Also, IJA started ground attacks in China against closest B-29-capable airfield where I was flying night attacls to mainland Japan. That is quite historical.

So, I don't think AI falls asleep, it just becomes "reactive" after it has captured all areas it wants. AI doesn't seem to play to win in early 1943, but is seeking tie in 1945-46. It's not very agressive after easy victories at start.

Cheers,

M.S.

[Speedysteve]

One thing I did notice from the 'other side' when I played Japanese against Allied AI in 43 campaign is that over 3 months he didn't invade 1 base and sent numerous small carrier TF's into harms way (1 Cv + CVL past Rabaul near Truk etc. Foolish. I gave up after a while since I thought he should have invaded a base by October. May just have been my game though.

Steven

[Erik Rutins]

If the AI expands further, it's just going to lose more in the long run. Human Japanese players will often not expand beyond what they feel is a good defensive perimeter for the long-term. Why should the AI do differently? I can't speak to whether the CHS mod has anything to do with this, but I've noticed that the AI does tend to switch from aggressive to conservative at some point in stock games as well. It's waiting for you to make the next move. Generally, if you start counterattacking it will make an effort to smack you.

Regards,

- Erik

[Banquet]

ORIGINAL: Erik Rutins

If the AI expands further, it's just going to lose more in the long run. Human Japanese players will often not expand beyond what they feel is a good defensive perimeter for the long-term. Why should the AI do differently? I can't speak to whether the CHS mod has anything to do with this, but I've noticed that the AI does tend to switch from aggressive to conservative at some point in stock games as well. It's waiting for you to make the next move. Generally, if you start counterattacking it will make an effort to smack you.

Regards,

- Erik

Can't argue with that.

I've no particular complaints about the AI in WitP. I certainly understand that with the size of the map, number of units, and amount of logistics involved it must have been a nightmare to program.. and overall it seems to do ok.

Bearing in mind the sequel (if done) won't have AI at all, I'm now quite grateful that there's any form of AI in this version and much more liable to overlook it's 'quirks'

[Speedysteve]

Hi Erik,

How about Allied AI in the liberation period (07/43 onwards)? In my 3 months of playing it did no advancing.

Regards,

Steven

[Tristanjohn]

ORIGINAL: Banquet

ORIGINAL: kaiser73

ORIGINAL: Banquet

Unlike the 'unofficially proposed' WitP II, WitP has an AI so I imagine he wants it to do something.. hardly an unreasonable expectation..

Tanaka, there were some reported problems with the AI going to sleep on the CHS mod.. I wonder if this is what you're experiencing?

More the game is complex, and harder is to get a good AI. There is no strategic game around which has a good AI (and AI which can give an average human player some challenge). Giving AI unhistorical advantages is what they do in every game.

AI conquered all SRA as Japan did historically. I guess that is good for the AI. if you want smart moves and unexpected plans...play PBEM. it may sound harsh, but that is how it is in every game out there.

I certainly don't disagree. Obviously for the best challenge (especially after learning how an AI plays) PBEM is the way to go.

I don't agree that all stragegic level AI's are completely lacking in challenge. I find HoI2's AI gives me plenty to think about as long as I don't go too far off the rails.

However, my point was that the AI should be doing something. (I dont' recall ever seeing a post where someone demanded an AI that was on a par with playing a human)

This game did ship with an AI and I don't see why, on the odd occasion it's ever mentioned on this forum, the majority of responses seem to be either 'well, it's the AI, what did you expect' or 'play PBEM' Fair enough if people are complaining about the AI's lack of a genius plan of operations.. but hardly relevant when Tanaka is asking why the AI is doing 'nothing'

It's possible the AI has decided it doesn't have enough strength to reasonably occupy more ground. If you were playing that game as the Japanese would you invest yet more troops to occupy more ground? Assuming the AI has established for itself a decent perimeter in the Central Pacific, and has defended the home islands, where else might it logically be expected to move? For Japanese players in AARs where a big push has been made on Ceylon and India, or deep into the South Pacific regions of the board, what sort of reserve do these players have on hand? Any reserve at all? Do they simply strip their rear areas of troops to keep going? Maybe the AI does not.

I don't know, mind you. I'm only posing questions.

[Bradley7735]

Hi all,

Here's what I've noticed about the japanese AI. I think it's programmed to take certain bases. (all the bases on the rail line in Burma, Port Moresby, Lunga, Gilberts, Wake, Guam, and all spots in between. If you (as allies) decide to oppose the AI at any spot they want, you'll get large battles. If you don't, then the AI will stop progressing after it's taken the bases on it's list.

For example, if you fight at Mandalay, it will fight you back. It'll try real hard to take it from you. However, if you retreat to Imphal (I think that's the base I'm thinking of), it will only take Mandalay, Lashio and Mykitnia. It won't advance or even try to advance to Imphal.

The same goes for Lunga. if you don't try to keep it, it won't go further. IE, it won't go for Noumea or Luganville.

To make the game interesting, you should try a bit to stop the AI as it moves. However, you can try real hard to keep certain places (Wake, Mandalay, others?) and cause the AI to destroy itself. I've held Wake for months as the AI just loses transports after transports. you can literally sink dozens of ships by putting one group of SBD's and some more troops on Wake. If you want a good game, don't reinforce one area too much. Pull back like the allies did until May 42. Then hold them at Port Moresby and Lunga. The fun part comes when you advance from mid 42 to mid 43 (after then, the AI will probably have lost all it's combat ships and good pilots)

Against a human, its' a different matter.

[Banquet]

ORIGINAL: Tristanjohn

It's possible the AI has decided it doesn't have enough strength to reasonably occupy more ground. If you were playing that game as the Japanese would you invest yet more troops to occupy more ground? Assuming the AI has established for itself a decent perimeter in the Central Pacific, and has defended the home islands, where else might it logically be expected to move? For Japanese players in AARs where a big push has been made on Ceylon and India, or deep into the South Pacific regions of the board, what sort of reserve do these players have on hand? Any reserve at all? Do they simply strip their rear areas of troops to keep going? Maybe the AI does not.

I don't know, mind you. I'm only posing questions.

That's possible and I'd be rather impressed if that were the case. After all, better that than leave gaping holes in their defenses for the player to exploit.

[Banquet]

ORIGINAL: Bradley7735

Hi all,

Here's what I've noticed about the japanese AI. I think it's programmed to take certain bases. (all the bases on the rail line in Burma, Port Moresby, Lunga, Gilberts, Wake, Guam, and all spots in between. If you (as allies) decide to oppose the AI at any spot they want, you'll get large battles. If you don't, then the AI will stop progressing after it's taken the bases on it's list.

For example, if you fight at Mandalay, it will fight you back. It'll try real hard to take it from you. However, if you retreat to Imphal (I think that's the base I'm thinking of), it will only take Mandalay, Lashio and Mykitnia. It won't advance or even try to advance to Imphal.

The same goes for Lunga. if you don't try to keep it, it won't go further. IE, it won't go for Noumea or Luganville.

To make the game interesting, you should try a bit to stop the AI as it moves. However, you can try real hard to keep certain places (Wake, Mandalay, others?) and cause the AI to destroy itself. I've held Wake for months as the AI just loses transports after transports. you can literally sink dozens of ships by putting one group of SBD's and some more troops on Wake. If you want a good game, don't reinforce one area too much. Pull back like the allies did until May 42. Then hold them at Port Moresby and Lunga. The fun part comes when you advance from mid 42 to mid 43 (after then, the AI will probably have lost all it's combat ships and good pilots)

Against a human, its' a different matter.

I'm sure someone from the dev or beta team (possibly Pry?) indicated that the size of the ports and/or airfields had an influence on what the AI decides to attack. This would be good as it infers a degree of replayability (you build up a base to a large size that you didn't build up last game - and the AI responds) However this may mainly apply to US/Allied AI.

The Japanese AI certainly seems to be scripted. The question is, is it completely scripted (for offensive operations at least) or does it have some freedom to react.

[rogueusmc]

Sooooo...what we have concluded is that the AI has a plan from the beginning and sticks to it?!?!?! Sounds pretty intelligent to me...[:D]

[ADavidB]

It's possible the AI has decided it doesn't have enough strength to reasonably occupy more ground. If you were playing that game as the Japanese would you invest yet more troops to occupy more ground? Assuming the AI has established for itself a decent perimeter in the Central Pacific, and has defended the home islands, where else might it logically be expected to move? For Japanese players in AARs where a big push has been made on Ceylon and India, or deep into the South Pacific regions of the board, what sort of reserve do these players have on hand? Any reserve at all? Do they simply strip their rear areas of troops to keep going? Maybe the AI does not.

One thing that limits the AI is that it does a terrible job of handling its naval assets. Other than trying to keep the KB around where it might be able to surprise the human player, the AI tends to send TFs willy-nilly everywhere. And more times than not it will let the KB run out of fuel in the middle of nowhere. I played a couple of games against the AI where it stopped doing anything after about 6 months game time. I switched sides and found that it had hundreds of small TFs going nowhere useful - either empty or taking fuel/supplies away from front line bases.

Dave Baranyi