Anybody who does any in depth reading of this forum, or who plays the Japanese side in a game, knows how the Japanese economy works. Oil is used to make fuel; fuel and resources feed the heavy industry; fuel allows ships to move; resources also feed light industry; and supplies make the world go round and so forth. Ships and ground transportation are used to move these goodies around the map to where they are needed.
I would like to inquire if there would be any interest amongst the AI players to examine the capability of the computer in it's ability to manage it's economy? My proposal is to make it very easy for the AI. We will play stock scenario one as Allied on the "hard" difficulty level. Subs will be controlled by the human player and we will simply let them park themselves wherever they desire. We do not care about issuing any orders to the Allied forces. They will only carry out the initial preset orders and then sit and defend. We have no interest in reinforcements. We are simply not interested in fighting the Japanese empire. We are simply interested in allowing the Japanese economy to operate in the most optimal form manner.
Punching the "end turn" button is our only goal so that we can then check and see if the oil and fuel and resources and supplies and being moved and production is occurring as it should. We want to see if the economy is operating in the manner in which it should. If it doesn't, then we should examine why. Testing I have done indicates that these "goodies" are not being moved by the computer as one would expect. We will use Tracker to see what the economy is doing. No rocket science here - just plain, good old fashioned testing. I am confident that this was tested to death during the initial development of the game and even more so during the AE upgrade, but over time errors can creep back in.
Anybody interested? Any comments or suggestions as to what you would like to see examined? My thinking at the moment is to run X (to be decided) turns using the method described above, run those X turns thru Tracker, and the post the results from the Industry view/WitP Chart and Global views along with each Xth save file for independent examination by anyone who desires to do so. We should also take a peek at selected bases to see it their goodies are simply accumulating or are being transported to or from. We can then delve into whatever is felt requires further delving into. The testing process itself will take little time, so we should be able to test as far into the war as is felt necessary. Discoveries made will certainly assist any AI player who desires to make changes to the Japanese economy and know if these changes have any actual effect.
A Modest Proposal to Examine the Japanese Economy
Moderators: wdolson, MOD_War-in-the-Pacific-Admirals-Edition
A Modest Proposal to Examine the Japanese Economy
USMC: 1970-1977. A United States Marine.
We don't take kindly to idjits.
We don't take kindly to idjits.
RE: A Modest Proposal to Examine the Japanese Economy
I do not think this is necessary to be honest. I am playing Japan both in a PBEM and against the Allied AI. In both games, I use Auto Convoy's a LOT which is the same system the AI uses although with ALL the ships versus just the ones I let it use [:)].
Since the beta p8 patch, the AC system is working MUCH better in supplying everything to everywhere. Before p8, the AC was totally broken which caused the Japanese AI's economey to collaspe sometime in late 42/early 43. To me this was the major reason that Japan was so weak against the Allies. After all, as you stated above, if Japan is not getting fuel and resources inbound and supplies out bound, then nothing will help them.
I currently have over 40 AC TFs running around and 90% of my bases are set to AC on. It has been this way for almost a year in game time (also RL too lol). So I can attest that with the AC system now working as designed. Japan's economy will be in better shape than ever [:)] with it under AI control.
Of course as I AM controling the production system [:D], I cannot answer how the AI woud handle production. I assume that would be part of the AI scripts. If so then you would need to somehow control which AI script Japan's AI would use. Otherwise, everyone would be getting different results because differnt AI scripts would be used.
Also even with the Allies doing nothing each turn, this process would still take a long time as you would still need to run the the turn execution every turn. Even with the reports turned off, it could still take a while just to run each and every turn. To get a really good idea of what the AI was actually doing, you would need to run this for a year of game time to truely see the long term effects the AI would be making on their production. It is not enough to just track the resources/supplies. You would need to see how the ship construction panned out, AC R&D, factory expansions (when does it happen, where does it happen, what gets expanded, and to see if there is any trigger as to why the AI did it this turn versus the next one or the turn before), etc.
So to me a lot of time involved, versus very little gain, especially since what I believe the root casue of Japan's economy's demise was the AC system being broken.
Since the beta p8 patch, the AC system is working MUCH better in supplying everything to everywhere. Before p8, the AC was totally broken which caused the Japanese AI's economey to collaspe sometime in late 42/early 43. To me this was the major reason that Japan was so weak against the Allies. After all, as you stated above, if Japan is not getting fuel and resources inbound and supplies out bound, then nothing will help them.
I currently have over 40 AC TFs running around and 90% of my bases are set to AC on. It has been this way for almost a year in game time (also RL too lol). So I can attest that with the AC system now working as designed. Japan's economy will be in better shape than ever [:)] with it under AI control.
Of course as I AM controling the production system [:D], I cannot answer how the AI woud handle production. I assume that would be part of the AI scripts. If so then you would need to somehow control which AI script Japan's AI would use. Otherwise, everyone would be getting different results because differnt AI scripts would be used.
Also even with the Allies doing nothing each turn, this process would still take a long time as you would still need to run the the turn execution every turn. Even with the reports turned off, it could still take a while just to run each and every turn. To get a really good idea of what the AI was actually doing, you would need to run this for a year of game time to truely see the long term effects the AI would be making on their production. It is not enough to just track the resources/supplies. You would need to see how the ship construction panned out, AC R&D, factory expansions (when does it happen, where does it happen, what gets expanded, and to see if there is any trigger as to why the AI did it this turn versus the next one or the turn before), etc.
So to me a lot of time involved, versus very little gain, especially since what I believe the root casue of Japan's economy's demise was the AC system being broken.
RE: A Modest Proposal to Examine the Japanese Economy
The AI scripts do not control the convoy system. The scripts are used for moving the LCU's, air groups and ships around. I am now running with the q5 beta.
I decided to go with this, however I changed it to 4 day turns after running the 2nd day. In less than two hours I am now at 7 Feb 42, so time to do this is minimal as far as processing the turns. The "do nothing" defense actually works quite well. Singapore and Manila are still holding out. The Adelaide is still parked at Port Morsby, and the Houston still sits at IIlilo. It hurts to see my little electronic soldiers starving, but they are doing it for a good cause (in my most humble opinion). I haven't run tracker yet. Maybe do three months and then toss the data into tracker. For the first few months the goodies in Japan should be in a state of decline until the goodie centers are captured, so after that is when the serious checking will begin.
What else have I to do when the economy is in the process of dying in the game I was playing?
If you have any interest, I would welcome your comments on how the AI is handling it's economy.
I decided to go with this, however I changed it to 4 day turns after running the 2nd day. In less than two hours I am now at 7 Feb 42, so time to do this is minimal as far as processing the turns. The "do nothing" defense actually works quite well. Singapore and Manila are still holding out. The Adelaide is still parked at Port Morsby, and the Houston still sits at IIlilo. It hurts to see my little electronic soldiers starving, but they are doing it for a good cause (in my most humble opinion). I haven't run tracker yet. Maybe do three months and then toss the data into tracker. For the first few months the goodies in Japan should be in a state of decline until the goodie centers are captured, so after that is when the serious checking will begin.
What else have I to do when the economy is in the process of dying in the game I was playing?
If you have any interest, I would welcome your comments on how the AI is handling it's economy.
USMC: 1970-1977. A United States Marine.
We don't take kindly to idjits.
We don't take kindly to idjits.
RE: A Modest Proposal to Examine the Japanese Economy
Now at 7 Mar 42. Manila, Bataan and Singapore have fallen in the last month. Houston and Adalaide still guarding their start locations. Time for some Tracker revelations. This won't be quite as thrilling as taking one of the Vettes to Roebling Road for a 130 mph workout, but I am sure it is going to be an eye opener.
USMC: 1970-1977. A United States Marine.
We don't take kindly to idjits.
We don't take kindly to idjits.
RE: A Modest Proposal to Examine the Japanese Economy
Why do this and unveil the AI seems a little pointless no one has denied the AI gets help not sure I get the logic in removing the mask
Its your time so fair enough but please somewhere in your first post put some sort of spoiler alert because most AI players dont want to know this stuff....
Its your time so fair enough but please somewhere in your first post put some sort of spoiler alert because most AI players dont want to know this stuff....
RE: A Modest Proposal to Examine the Japanese Economy
I can certainly put in a "spoiler alert". However, I am not going to be posting anything other than general observations and tracking the movement, or non movement, of various resources. If someone requested specific information I would be willing to converse with them via personal messages. Like I said, this is not rocket science. In fact, it is so easy even an AI player can do it.
My reason for doing this is as I have stated: I do not think that the computer Japanese player is managing the transport of it's various resource types in an efficient manner, and I feel that this is the most obvious reason that the economy would fail. As no one will respond to a request for tech support in answering this (including yourself) then the only other viable solution is for me to investigate it further. I have put in literally thousands of hours playing this game in the last seven plus years, and at some point within the next few years I would like to be able to complete a game. Far too many other games were a complete waste of time due to the collapse of the Japanese economy. At least, that is the goal I have set for myself. If, through researching this, I find that the lack of resource transport is the issue, then someone else may want to see if it can be corrected. I would certainly never presume to attempt to get an AI problem reported and subsequently fixed, but it should be within the scope of any player who would like to see improvement of the game to report deficiencies.
My reason for doing this is as I have stated: I do not think that the computer Japanese player is managing the transport of it's various resource types in an efficient manner, and I feel that this is the most obvious reason that the economy would fail. As no one will respond to a request for tech support in answering this (including yourself) then the only other viable solution is for me to investigate it further. I have put in literally thousands of hours playing this game in the last seven plus years, and at some point within the next few years I would like to be able to complete a game. Far too many other games were a complete waste of time due to the collapse of the Japanese economy. At least, that is the goal I have set for myself. If, through researching this, I find that the lack of resource transport is the issue, then someone else may want to see if it can be corrected. I would certainly never presume to attempt to get an AI problem reported and subsequently fixed, but it should be within the scope of any player who would like to see improvement of the game to report deficiencies.
USMC: 1970-1977. A United States Marine.
We don't take kindly to idjits.
We don't take kindly to idjits.
RE: A Modest Proposal to Examine the Japanese Economy
My main reason for not reacting is that despite what tracker says the Ai does still produce stuff - so does the economy actually 'fail' or is it only because you can track it that it appears to fail.
Remember tracker did not exist when we wrote the AI so a lot of what was done was done after months of testign AI v AI to get the game out to 45 and then analysed via spreadsheet.
Basically what I have seen from your posts in the other thread is that you would rather the AI economy worked and didnt cheat a bit. Fair enough but there aint enough years left in my life to fix that one....
ps the development of this game ended about 2 years ago
Michael is doing what he can on tech support to fix bugs and make minor tweaks as the support team lead but there are no devs any more (other than one miracle working Michael) all of the original team has disbanded and while we are still about Joe made it clear we are not developing the game any more - is it perfect no but its as good as we could make it.
Andy
Remember tracker did not exist when we wrote the AI so a lot of what was done was done after months of testign AI v AI to get the game out to 45 and then analysed via spreadsheet.
Basically what I have seen from your posts in the other thread is that you would rather the AI economy worked and didnt cheat a bit. Fair enough but there aint enough years left in my life to fix that one....
ps the development of this game ended about 2 years ago
Michael is doing what he can on tech support to fix bugs and make minor tweaks as the support team lead but there are no devs any more (other than one miracle working Michael) all of the original team has disbanded and while we are still about Joe made it clear we are not developing the game any more - is it perfect no but its as good as we could make it.
Andy
-
Dan Nichols
- Posts: 863
- Joined: Mon Aug 29, 2011 11:32 pm
RE: A Modest Proposal to Examine the Japanese Economy
And I must say that the team did a darn fine job. I really do not want to know what the AI is doing, I just want it to put up a good fight and not do too many stupid things. If it has to "cheat" to do it, so be it, just don't tell me about it.
I think that the two obligations you have are to be good at what you do and then to pass on your knowledge to a younger person
RE: A Modest Proposal to Examine the Japanese Economy
When will the team rejoin for WITP 2 ? Love AE and need new fix
RE: A Modest Proposal to Examine the Japanese Economy
The economy failed many times in the WitP days. Played a few times into 43 and a couple into 44 and find that basically nothing was happening on the J side, so being a curious type of guy I would go in, change to head to head, and look to see what was going on, and it was enough to convince me that nothing was going on. Naturally there were countless restarts as well due to fixing data errors or scenarios being updated.
For AE I began playing it when it came out, but - in my opinion - found various bugs which caused me to remain in a wait mode for awhile. With 1.06 - in my opinion - the stability was there to make the long go of it again. Scenario updates and my own mods caused numerous restarts. Since I had learned from WitP that it is better to check and verify what was happening to the other sides economy, I continued doing so. When I saw the economy doing what I perceived as in the stages of failing, my past experience caused me to stop and question. I have rarely posted bug reports, probably the three I commented on recently in this forum far surpasses anything I did previously. No one would answer my question "should I be concerned about this"? Therefore, I was.
If the economy does work I would be most happy to quit checking it and enjoy the game as it should be enjoyed. I do have a couple questions that I would like to ask you in a PM if that is permissible? You have my word here in this open forum that if any answer you give me is not for public consumption then I will honor any such request.
Went off for other interests for a bit but returned when it occurred to me what I had actually said in my last post - about thousands of hours spent, countless restarts and never completing a game. That does fit the definition of utter stupidly, does it not?
I will praise you again for your work on the AI. It is a vast improvement from WitP and puts up the good fight, which is want the AI player wants from the beginning to the end of the game.
For AE I began playing it when it came out, but - in my opinion - found various bugs which caused me to remain in a wait mode for awhile. With 1.06 - in my opinion - the stability was there to make the long go of it again. Scenario updates and my own mods caused numerous restarts. Since I had learned from WitP that it is better to check and verify what was happening to the other sides economy, I continued doing so. When I saw the economy doing what I perceived as in the stages of failing, my past experience caused me to stop and question. I have rarely posted bug reports, probably the three I commented on recently in this forum far surpasses anything I did previously. No one would answer my question "should I be concerned about this"? Therefore, I was.
If the economy does work I would be most happy to quit checking it and enjoy the game as it should be enjoyed. I do have a couple questions that I would like to ask you in a PM if that is permissible? You have my word here in this open forum that if any answer you give me is not for public consumption then I will honor any such request.
Went off for other interests for a bit but returned when it occurred to me what I had actually said in my last post - about thousands of hours spent, countless restarts and never completing a game. That does fit the definition of utter stupidly, does it not?
I will praise you again for your work on the AI. It is a vast improvement from WitP and puts up the good fight, which is want the AI player wants from the beginning to the end of the game.
USMC: 1970-1977. A United States Marine.
We don't take kindly to idjits.
We don't take kindly to idjits.
RE: A Modest Proposal to Examine the Japanese Economy
Amen.ORIGINAL: Andy Mac
My main reason for not reacting is that despite what tracker says the Ai does still produce stuff - so does the economy actually 'fail' or is it only because you can track it that it appears to fail.
Remember tracker did not exist when we wrote the AI so a lot of what was done was done after months of testign AI v AI to get the game out to 45 and then analysed via spreadsheet.
Basically what I have seen from your posts in the other thread is that you would rather the AI economy worked and didnt cheat a bit. Fair enough but there aint enough years left in my life to fix that one....
ps the development of this game ended about 2 years ago
Michael is doing what he can on tech support to fix bugs and make minor tweaks as the support team lead but there are no devs any more (other than one miracle working Michael) all of the original team has disbanded and while we are still about Joe made it clear we are not developing the game any more - is it perfect no but its as good as we could make it.
Andy
I've gotten +500 turns against the AI and had to re-start as it was kicking my butt in China. Looking forward to testing the "new and improved" Ironman.
Thanks for all your efforts. They are great and much appreciated by many of us, and definitely me.
[&o][&o][&o]
Pax
RE: A Modest Proposal to Examine the Japanese Economy
I ne3ver played v the AI in orig WITP (not my bag) but I did consult extensively on the problems AI players found with the AI - the AI economy should not be collapsing in AE unless you are physically sinking most of the AI merchant fleet or hitting the key industry sites in Japan PM fine
RE: A Modest Proposal to Examine the Japanese Economy
What I saw with the Japanese economy is written on the first post of my "faltering Japanese economy" thread. The home islands of Japan were running way short of the fuel required to feed heavy industry, and running short of the resources which are required mainly by the light industry, which leads to drastic reductions of HI points and supply. The various "goodies" were simply not being transported by the AI back to Japan. There was a great deal of fuel sitting in Palembang but very little of it being transported. There were 295 ships, the vast majority to be cargo transports, sitting in Singapore and not performing any transport duties. Again, the first post of that thread contains more info that I want to rewrite here.
In my view, this is indicative of an economy that was failing. Which lead to my question "Should I be concerned about this"? You indicate that the economy is viable regardless of what I see. Since that must be so, then I will accept it. I don't mean to argumentative, just stating what I saw in Japan when I examined it and what tracker tells me is occurring there and expressing a concern regarding it.
I will stop questioning it here in these forums.
In my view, this is indicative of an economy that was failing. Which lead to my question "Should I be concerned about this"? You indicate that the economy is viable regardless of what I see. Since that must be so, then I will accept it. I don't mean to argumentative, just stating what I saw in Japan when I examined it and what tracker tells me is occurring there and expressing a concern regarding it.
I will stop questioning it here in these forums.
USMC: 1970-1977. A United States Marine.
We don't take kindly to idjits.
We don't take kindly to idjits.
RE: A Modest Proposal to Examine the Japanese Economy
I dont mind questions and as I said its your time and is it perfect no of course not I just dont see much scope for re design or improvement at this stage
RE: A Modest Proposal to Examine the Japanese Economy
I've never had a game against the AI collapse due to lack of production. AI will occasionally do some things that make you scratch your head, but that's part of the surprises that were built into it no doubt. (What IJN player expects a carrier to swoop through the South East Fleet area of the Empire in January 1942?) [;)]
Distant Worlds Fan
'When in doubt...attack!'
'When in doubt...attack!'
RE: A Modest Proposal to Examine the Japanese Economy
ORIGINAL: Andy Mac
I ne3ver played v the AI in orig WITP (not my bag) but I did consult extensively on the problems AI players found with the AI - the AI economy should not be collapsing in AE unless you are physically sinking most of the AI merchant fleet or hitting the key industry sites in Japan PM fine
The issue was that Japan's economy WAS collapsing in the mid game without the Allies having to do anything. Again as I pointed out above, this was due to the AC system not working. To be specific, AC TFs would get stuck at ports waiting for resources/oil that would never show up. Over the course of time a significent number of Japan's cargo fleet would be tied up in this manner. The AC system no longer does this in beta p8 and higher.
Of course if the AI behind the scenes was mitigating this somehow that is not visible to the players and/or Tracker, then I am fine with that and the AC system not working may not have had as big of an impact as I have thought. So if the AI has 'special' abilities and now the AC system is working much better, the Allied players should have an even better game experience than before [:)] even on historical.
Of course as an unofficial JFB at the moment, I will have to wait until I finish my current games as Japan in order to find out for myself [:D]
RE: A Modest Proposal to Examine the Japanese Economy
AI on hard the economy will not collapse
RE: A Modest Proposal to Examine the Japanese Economy
ORIGINAL: Andy Mac
AI on hard the economy will not collapse
I always play against the AI on very hard (I think that is highest difficulty) and the AI doesn't collapse at all. It is definately a challenge.
Distant Worlds Fan
'When in doubt...attack!'
'When in doubt...attack!'
RE: A Modest Proposal to Examine the Japanese Economy
ORIGINAL: Andy Mac
AI on hard the economy will not collapse
Might it collapse on "Historical"? Or is it just the supply rules that are different from "Hard"?
RE: A Modest Proposal to Examine the Japanese Economy
Not sure it didnt when I last tested it during the gold release version haven't tested it since as all my testing is done on hard now after the cvhanges made.