Infrastructure building queue for AI in AoD 1.09

[Pang Bingxun]

There has been some controversy about the infrastructure builds i added to the building queue of AI in the 1936 grand campaign. Some think i made AI build too much infrastructure and as it spends great amounts of its available ic into Infrastructure this critique has some substance. I would like to lay down the reasoning for the Infra builds in 109 Beta 11(set33) and later versions of 1.09 and have that as a starting point for discussion.

The most important reason for Infrastructure is increased provincial efficiency that increases the output of factories and resources. In most cases factories are more important as ic is short in the short run. In the long run the output of resources becomes more and more important. A collapse of the world economy by AI building factories instead of Infra must be avoided at all costs.

When AI builds Infra it gets a 30 percentage point gearing increment per finished unit. It has to pay no malus for hard terrain, everything is threated as plains(details may apply). Both change the balance in favour of building much infra, especially for USSR and China.

AI adding Infrastructure to the building queue is a rare instance. It ignores factories and only seems to care about resources even when factories are much more important. It only addes one unit infrastructure und thus makes no use of gearing bonus. It does even ignore the maximum of 200% Infra in a province, so it wastes a lot of icd for no use at all. The inital building queue tries to compensate for that.

Of some importance for AI building Infra is what AI would build else. AI would else build military, that reduces yields from unused manpower, costs salaries, costs supplies and will cost substancial upgrading because models are built that are going to be outdated long before they are going to be used. So building military as a huge negative rate of interest which the low positive rate of interest on Infra needs to be compared with. Also changing sliders and advancing in tech usually changes the relative building efficiency in favour of building military as late as possible.

The effect infra has on reorg of land units and ese is often of secondary importance, but it is important nonetheless. It might also be important for allied forces. Finland is the best example for this as it may need other things more urgently, but its german ally utilizing increased ese is a big advantage for its alliance in total.

Edit: There is something else to consider:

It might be worth to note that the USA, the UK and France are 3 special cases as their peacetime penalties had been reduced in 1.09. The purpose was to increase resource supply and to allow the human player to fund research properly. As AI has no need to fund its research it has been made to spend the gained icd into some additional Infra, so that the goal, to keep those nations sufficiently unprepared for war at their historic dates of war entry, is not harmed.

 

[MagooNZ]

Sounds reasonable. I didn't know about the 30% AI infrastructure gearing bonus. I often delay releasing a puppet in order to improve the infrastructure in several provinces, because once released, I know the AI is never going to build infrastructure. It would be over to players to make their own customised starting production queues for the nations that they play, so this only has to be done once for all new games. The human player would have more factory and fewer infrastructure builds than the ai. There is also the grouping and location of divisions (OOB?) at the game start that takes a lot of time to re-organise. I should be producing my own allocations and replacing the existing files. This would go a long way to making the start of a new game much easier to setup.

 

[Commander666]

There has been some controversy about the infrastructure builds i added to the building queue of AI in the 1936 grand campaign.... ... for the Infra builds in 109 Beta 11(set33) ....

I also agree what you write sounds reasonable. While not having seen v1.09 yet, the AI's bonus of 30% gearing increment and no terrain malus is a huge advantage that should let AI get the total project mostly finished early enough to build an army. The exception might be building a 200-infra line from capital to outlet (nearest port with level 10 naval base) which might not be completed until later. In any case, certain AI countries should complete that special project because a maximum outlet to realize maximum world-wide ESE is so huge a strategic benefit it can't really be termed "secondary importance." For the USA and Japan, a 200-infra capital (which also serves as those country's outlets) is strategically more important than any increased resources in Washington or Tokyo; or increased concentration bonus for the factories there.

Some other countries also should build a 200-infra outlet, but they need some extra provinces in the line from capital to naval port also built to 200-infra - even if there is not a single factory or any resource there - because they majorly need best world-wide ESE. The obvious choice would be UK with the AI building a 200-infra line to Dover or Norwich (or Liverpool is pretty good too). Italy might also be an AI candidate to validate its longer needed 200-infra line from Rome to Venice, Genoa or even Taranto. But when it comes to discussing Germany or SU, I don't think that the cost of a 200-infra line from capital to outlet is worth it because these AIs really should not be pre-building infra in 1936 to mostly serve for some improbable major overseas invasion.

That AIs mostly choose to not build infra (and when they do then not serially) is a problem. It is aggravating to delay liberating a country because you know the puppet will never construct the much needed infra improvements you have started. Can this not get changed so the infra serials laid down before liberation continue?

I note player can place any air or naval base they construct in any other country they are allied with if building was not done thru province. However infra only constructs "thru province". The discrepancy is absurd. If I can build an air base or naval base out-of-province and "drop" that, once constructed, in any friendly country I chose, how much more logical to "drop infra improvements in any country" as - in reality - most of those improvements really were Bailey bridges manufactured elsewhere, transported, and dropped where needed. The airbases were made serviceable with steel matting brought in from elsewhere; and I suppose the naval bases would be Mulberries or similar towed from overseas. Whatever the comparisons to RL are, player getting his infra builds cancelled in the country he liberates is just very wrong.

The best solution would be some clever code – just like player inherits nuke or rocket test site upon conquering that province – so the puppet inherits your relevant infra builds from your construction queue when you liberate them.

 

[Pang Bingxun]

That AIs mostly choose to not build infra (and when they do then not serially) is a problem. It is aggravating to delay liberating a country because you know the puppet will never construct the much needed infra improvements you have started. Can this not get changed so the infra serials laid down before liberation continue?

That would require hardcoded changes.

The best solution would be some clever code – just like player inherits nuke or rocket test site upon conquering that province – so the puppet inherits your relevant infra builds from your construction queue when you liberate them.

That would indeed be clever. For myself i use a similar solution. I add long serials of Infra to freshly liberated puppets, but that requires savegame edit.

While at peace you could save, load as the puppet, add the infra serials, save and load as your regular nation. That works fine for the most part while you are at peace, albeit you risk involuntary promotions of leader you would rather prefer to stay at lower ranks for a while longer. During war it is much worse as AI will take command of your country while you take care of the puppet. Once you will be back you will find quite a mess.

 

[Commander666]

That would indeed be clever. For myself i use a similar solution. I add long serials of Infra to freshly liberated puppets, but that requires savegame edit.

While at peace you could save, load as the puppet, add the infra serials, save and load as your regular nation. That works fine for the most part while you are at peace, albeit you risk involuntary promotions of leader you would rather prefer to stay at lower ranks for a while longer. During war it is much worse as AI will take command of your country while you take care of the puppet. Once you will be back you will find quite a mess.

Please clarify if your second paragraph is an example of savegame edit. I am not really sure what the term means. What I have done many times is what you describe in 2nd paragraph - save game, load as other country and do some change with that country, save and load as your regular country. Is that "savegame edit"? Importantly, the game must not run - even just 1 hour.

Whatever called, yes that is usually a mess.... and I think we can assume player would be at war. Your convoy display will be disrupted and many stacks broken down with those units now missioned to run everywhere, and perhaps more. I reserve this for really important things like loading an ally to give to me some needed province so I can liberate some new nation once playing my regular country again.

 

[Pang Bingxun]

Please clarify if your second paragraph is an example of savegame edit. I am not really sure what the term means. What I have done many times is what you describe in 2nd paragraph - save game, load as other country and do some change with that country, save and load as your regular country. Is that "savegame edit"? Importantly, the game must not run - even just 1 hour.

Sure, in game time must not pass. But as the save is "edited" only by the game itself this does not constitute a savegame edit. That term desribes editing the savegame with a different program than the game itself, usually notepad++ is used.

 

[Commander666]

Thanks Pang.

 

[Mr_B0narpte]

I did not know the AI gets 30% gearing bonus per unit, and that its infra production is terrain blind. That certainly reduces my opposition to such substantial infrastructure builds.

However that still doesn't persuade me of the use in developing the infrastructure of 1-3IC provinces - perhaps even 4/5 IC provinces. I haven't done the exact maths, but surely that only pays off in the late 1940s, if not early 50s? IMO anything that does not pay off until 1943 is pointless - in fact for the AI (barring perhaps an AI USA) I would say anything that does not pay off until 1940/41 is of little use to the AI if faced with a human major. This is because most players, quite rightly, reach their 'peak' by late-39 or 1940.

On a slightly different note:

The USSR AI developing infra in Ukraine and surrounding areas primarily benefits Germany, not the USSR.

The same for France spending much IC on infra development across its northern territories (for southern territories, it primarily benefits Vichy).

For Dutch AI builds, it benefits Germany & Japan.

For Belgium AI builds, it benefits Germany.

For the UK AI, its infra builds in India are just an invitation for any Japan player, if not the Japan AI. I think we all know the UK AI stands little chance of holding India, if only because it can't handle defending such an expansive empire.

I guess for Czechoslovakia it matters little what it builds, but the infra builds there pretty much just make the nation an extension of Germany from 1 January 1936 - the same can be said for Luxembourg too, in both regards.

For the USA, I wonder the need of so many infra builds. Ultimately it is not low in either resources or IC, but manpower. I know an infra serial costs only 0.1 manpower, but it all adds up. It looks as if there are around 80 full infra serials for the USA - totaling around 192 manpower. Assuming we cut out 50% of those infra serials, another 10 arm-harm should be much more useful to the USA then more infra in low IC/resource provinces.

-----------------

However, Norway's and Sweden's builds actually for me are the examples to follow. Infra production need not continue until 200%, especially with war looming in September 1939 for many nations.

 

[Pang Bingxun]

I haven't done the exact maths, but surely that only pays off in the late 1940s, if not early 50s?

Compared to building money instead of infra in provinces with 2 factories infra will give you(more precisely AI which will not stockpile money) a bit more than 10% interests per year. So you need about 10 years for the interests to amount to the principle. If you do the math with highly negative return on investment for alternative military investements infra will look a lot brighter.

For the UK AI, its infra builds in India are just an invitation for any Japan player, if not the Japan AI. I think we all know the UK AI stands little chance of holding India, if only because it can't handle defending such an expansive empire.

That is probably true, too. UK AI can defend India only when it is not attacked by a serios force. However, it is possibly to take India back later. Then the investment pays off again. The same principle applies to soviet union, there it is much more important.

For the USA, I wonder the need of so many infra builds. Ultimately it is not low in either resources or IC, but manpower. I know an infra serial costs only 0.1 manpower, but it all adds up. It looks as if there are around 80 full infra serials for the USA - totaling around 192 manpower. Assuming we cut out 50% of those infra serials, another 10 arm-harm should be much more useful to the USA then more infra in low IC/resource provinces.

The manpower is an issue for the USA. I have seen it run low on manpower because it builds too many land division and because it loses those when transporting them. USA losing 1.2 million soldier till early 1944 i bit much. That is worth more than 3 years of recruiting.

in fact for the AI (barring perhaps an AI USA) I would say anything that does not pay off until 1940/41 is of little use to the AI if faced with a human major.

Probably true.

This is because most players, quite rightly, reach their 'peak' by late-39 or 1940.

And this a major mistake and it ruins the game. Those human players use a much bigger portion of their economic potential for military investments than IRL. Some degree of restrainment is needed in the early years in order to enjoy the game. Earlier patches had the opposite problem. The AI used a much higher portion of the economic potential for military investments and thus those AIs could be made to fall back by simply delaying war till 1943+. There needs to some kind of minimal consensus about when the war is to start to which degree. For 1.09 i chose to make nations spend not too much more into military than IRL. And one can still very well argue it is too much.

 

[Mr_B0narpte]

Compared to building money instead of infra in provinces with 2 factories infra will give you(more precisely AI which will not stockpile money) a bit more than 10% interests per year. So you need about 10 years for the interests to amount to the principle. If you do the math with highly negative return on investment for alternative military investements infra will look a lot brighter.

You look at AoD economic gains; I look at AoD effective gains. Having a strong military (or, in other words, the strongest military the nation can build) by late-1939 is the optimum - at least for all nations bar USA, USSR & perhaps UK. China is obviously another exception as there's little way it can use all of its' 4k manpower by that stage.

That is probably true, too. UK AI can defend India only when it is not attacked by a serios force. However, it is possibly to take India back later. Then the investment pays off again. The same principle applies to soviet union, there it is much more important.

However I get the feeling your AI Germany v. AI USSR tests typically end in Bitter Peace? If that's true, then the investment does not pay off. IIRC the highest IC province in India is 4 or so? I hardly see that as a worthwhile investment for an IC restricted UK - especially considering it loses India after the war.

The manpower is an issue for the USA. I have seen it run low on manpower because it builds too many land division and because it loses those when transporting them. USA losing 1.2 million soldier till early 1944 i bit much. That is worth more than 3 years of recruiting.

You & I are well aware of this hardcoded 'feature' of AoD whereby the AI likes to send its TPs undefended into hostile seas. AFAIK there's nothing that can be done about that, but the infra serials are obviously within our purview.

And this a major mistake and it ruins the game.

I can see how it ruins the game, but fail to see how it's a major mistake. The aim of any game is to win - and at whatever date we set war in AoD to start, players will focus their strategy around that date.

Those human players use a much bigger portion of their economic potential for military investments than IRL.

I know I'm pointing out the obvious, but AoD is a war game. The ultimate aim is not to have the biggest economy but to annex all your foes. Of course, players will focus their strategies on this aim.

Some degree of restrainment is needed in the early years in order to enjoy the game.

I guess by enjoyment you mean to make it less of an intense arms race? I'm not really sure that's possible considering my points made above.

Earlier patches had the opposite problem. The AI used a much higher portion of the economic potential for military investments and thus those AIs could be made to fall back by simply delaying war till 1943+.

I'm not exactly sure how that works. IMO an AI that has maxed out its manpower usage for military purposes - without strangling it's TC to any significant degree - has had it's production maximised to its full potential. Anything less is inefficient.

There needs to some kind of minimal consensus about when the war is to start to which degree. For 1.09 i chose to make nations spend not too much more into military than IRL. And one can still very well argue it is too much.

Another problem with this assumption is that it also assumes AoD handles military production very accurately. As I'm sure you're aware, it is certainly not 100% accurate. Not all nations had 10,000 men infantry divisions. Not all panzer divisions consisted of ~300 tanks. Aircraft carriers did not take 1.5 years to build.

 

[Pang Bingxun]

However I get the feeling your AI Germany v. AI USSR tests typically end in Bitter Peace?

There would seem to be that bias.

If that's true, then the investment does not pay off. IIRC the highest IC province in India is 4 or so? I hardly see that as a worthwhile investment for an IC restricted UK - especially considering it loses India after the war.

Those are details i chose to ignore. India contributes to the economy of the UK, so it is only fair it receives an adequate portion out of it. Also the UK can evade the loss by becoming a dictatorship. Well, i suppose AI cannot.

I guess by enjoyment you mean to make it less of an intense arms race?

No. Also i am not sure what precisely less intense means in this context. But having a similar timetable as IRL is something that i assume to be desired. If all nations simply mobilize their manpower 2 year too early they will also run out of manpower way too early and the arms race will end before it reached its historic peaks.

I'm not exactly sure how that works.

The AI cripples its economy by building military instead of something useful while you build up something useful and then build up a real army that will be superior to what AI will have then.

You look at AoD economic gains; I look at AoD effective gains. Having a strong military (or, in other words, the strongest military the nation can build) by late-1939 is the optimum - at least for all nations bar USA, USSR & perhaps UK.

If we postpone the critical date to mid 1941 or 1942 things look different. And if one looks at 1943 instead it will be much more biased towards infra. A reasonable choice of criteria could be some sort of average over the historic time of WW2. For europa half time of the war has passed in July 1942. If players were to optimize their plans to such dates it would make for a better game imo.

I'm not exactly sure how that works. IMO an AI that has maxed out its manpower usage for military purposes - without strangling it's TC to any significant degree - has had it's production maximised to its full potential. Anything less is inefficient.

A nation that mobilizes its manpower fast mobilizes its manpower in an ineffcient manner because the quality of the troops will suffer. If constructing military is delayed till 1938 tools are researched, till 1940 assemly line experimentation is reached, till the different 1940 assembly line techs are researched and sliders are maxed out, than a much greater military can be built. The total gotten base icd(7x95 for Infantry etc.) would be much higher then. Setting mid 1942 as a target to have strong military to utilize an adequatly high tc and still have sufficient manpower reserves by then is a proper goal for both AI and the human player.

 

[Mr_B0narpte]

Those are details i chose to ignore. India contributes to the economy of the UK, so it is only fair it receives an adequate portion out of it. Also the UK can evade the loss by becoming a dictatorship. Well, i suppose AI cannot.

This just proves my point. The infra issue is really only relevant to the AI as a human can obviously change the production as they wish. If you want a historical argument, I'm sure there is much evidence the UK did little to develop the Indian sub continent in this time period. The Bengal famine is the extreme example of such apathy towards their colonial 'subjects'.

No. Also i am not sure what precisely less intense means in this context. But having a similar timetable as IRL is something that i assume to be desired. If all nations simply mobilize their manpower 2 year too early they will also run out of manpower way too early and the arms race will end before it reached its historic peaks.

The Axis players do not want a re-run of historical WWII, they want to win. Time is against them, those two years can be critical.

The AI cripples its economy by building military instead of something useful while you build up something useful and then build up a real army that will be superior to what AI will have then.

The only thing useful to an AI is a sizeable land army, if it does not build such a thing by 1940 then it has wasted precious time.

If we postpone the critical date to mid 1941 or 1942 things look different. And if one looks at 1943 instead it will be much more biased towards infra. A reasonable choice of criteria could be some sort of average over the historic time of WW2. For europa half time of the war has passed in July 1942. If players were to optimize their plans to such dates it would make for a better game imo.

Such a move just benefits the Allies at the cost of history - as I presume you mean 'Danzig or war' as the critical date.

A nation that mobilizes its manpower fast mobilizes its manpower in an ineffcient manner because the quality of the troops will suffer. If constructing military is delayed till 1938 tools are researched, till 1940 assemly line experimentation is reached, till the different 1940 assembly line techs are researched and sliders are maxed out, than a much greater military can be built. The total gotten base icd(7x95 for Infantry etc.) would be much higher then. Setting mid 1942 as a target to have strong military to utilize an adequatly high tc and still have sufficient manpower reserves by then is a proper goal for both AI and the human player.

A player who maximises his army size by late-39/mid-40 stands a much better chance of winning compared to one aiming for mid-1942.

 

[Commander666]

For the USA, I wonder the need of so many infra builds. Ultimately it is not low in either resources or IC, but manpower. I know an infra serial costs only 0.1 manpower, but it all adds up. It looks as if there are around 80 full infra serials for the USA - totaling around 192 manpower. Assuming we cut out 50% of those infra serials, another 10 arm-harm should be much more useful to the USA then more infra in low IC/resource provinces.

Well, I should have downloaded beta v1.09 (or whatever is the current latest) before I gave my support for the changes in it. Of course, MogooNZ and I were only responding to Pang’s original post, and what Pang wrote in Post #1 does “sound reasonable.” However, Mr_BOnarpte’s revelation that the version has so changed AI infrabuilds that now USA has allegedly got 80 serials of them is a humongous change that is just “wacko”. I withdraw my earlier stated support totally.

Pang, instead of having written a well polished post trying to substantiate your immense changes, you should have focused on the obvious of why you needed to justify yourself in the first place. You start your thread with:

There has been some controversy about the infrastructure builds i added to the building queue of AI in the 1936 grand campaign. Some think i made AI build too much infrastructure and as it spends great amounts of its available ic into Infrastructure this critique has some substance..

… and then continue it to explain your reasons – so comprising a post that sounded “reasonable.”

There is nothing reasonable in the changes you made. On nearly every point in your discussion with Mr_BOnarpte above, he is correct and you are not.

And if you don’t know what the “critical date” is you don’t understand what players want out of this game. To be succinct, they want AIs that are militarily a challenge to beat provided player abides by the major historic events and does not start wars earlier.

But you justify UK building up infra in India – instead of using that IC for greater air force and navy? And the USA does not need a single infra build anywhere except Washington increased to 200%. USA is already flooded with resources. Unfortunately, your over abundant infra builds do nothing to make the AIs play better at war, but they certainly must diminish what they have left to play with.

The controversy regarding your v1.09 infrabuilds is probably not as small as you state (“some controversy”) but more important to me is how a single DEV can apparently ram down our throats what he thinks is best for us.

I am not interested in discussing your v1.09 changes in detail. Others can do that better than me. However I am most interested in the future of this game… and am wondering why you earlier stated something like “1.09 will probably be the final version.” Would you mind explaining that honestly? To me it seems to imply that “only Pang Bingxun can do version changes… and he’s not planning on doing any more.” Is that basically it?

While I might personally welcome you not changing AoD anymore, I simply cannot understand the politics behind why future AoD patches should end if you aren’t involved anymore. Sorry if you find my inquiry harsh or brutally frank, but you really over did yourself this time, IMO. So, I stand with Mr_BOnartpte’s views in the many points above because he seems to understand much better what players want from AoD vanilla, and also the ramifications of your major changes.

 

[Commander666]

Those are details i chose to ignore. India contributes to the economy of the UK, so it is only fair it receives an adequate portion out of it. Also the UK can evade the loss by becoming a dictatorship. Well, i suppose AI cannot.

I think you ignored much more than details. My basic belief is that ever since your massive changes done with China in earlier version, you have basically further changed AoD vanilla as to "your vision." Sadly, your vision has never included being an actual multi-player to so better learn what players want from AoD to get a falr and good game challenge.

There is nothing "fair" about India's portion of total IC spent on infra construction. There is only historic comparison (which you way over shot) and what is best for UK player. In more instances than I care to detail, your over abundance of infra constructions amounts to nothing but "gifts for German player". Your elsewhere stated view that SU and UK can recapture the provinces where they built high infra to eventually recover the investment totally fails to understand what really happens once German player has conquered that. That leads to end of game... and not some mis-lead notion that a beaten UK or SU will recover.

Finally, your statement of, "Also the UK can evade the loss by becoming a dictatorship" is about the silliest thingI have read to defend really poor massive game changes. And was "SU becoming a democracy" not another of your suggestions to defend earlier major game changes? Like the game's major powers should reverse basic ideology so your game changes then work?

 

[Pang Bingxun]

The Axis players do not want a re-run of historical WWII, they want to win. Time is against them, those two years can be critical.

Since when does AI want anything? And by building Infra they make time to be more in their favour.

The only thing useful to an AI is a sizeable land army, if it does not build such a thing by 1940 then it has wasted precious time.

But it will have built sizeable landarmies by mid 1940. The first thing that happens when Italy joins axis that it gets crushed in africa by the UK. The UK you claim to be short on ic makes any AI attempt of a sealion a failure.

A player who maximises his army size by late-39/mid-40 stands a much better chance of winning compared to one aiming for mid-1942.

Well, the precise dates varies from nation to nation and one needs to examine what maximizing means concretely. It means that the maximum speed of mustering new divisions reaches its peak long before then. AI Germany has about 100 mobile divisions for its historical Barbarossa "albeit" it spams out Infra like nuts. Or should that albeit be a because? The same is true for soviet union.

Such a move just benefits the Allies at the cost of history - as I presume you mean 'Danzig or war' as the critical date.

Both logical deduction and empirical evidence suggest otherwise. AI is rather strong at building up an army, but it is rather poor at managing it.

 

[Pang Bingxun]

However I am most interested in the future of this game… and am wondering why you earlier stated something like “1.09 will probably be the final version.” Would you mind explaining that honestly?

Did i say such a thing?

I could probably answer to other parts of what you wrote, but as you wrote it before analyzing the actual situation this seems to make little sense. This thread refers to the concrete situation of AI Infra in set33(Beta 11) and the soon to be published set34. There the results of AI Infra seem to be rather fine.

 

[Mr_B0narpte]

Since when does AI want anything? And by building Infra they make time to be more in their favour.

Of course the AI does not want anything, but the relevant thing is what we want the AI to be. I want it to be a challenging as possible within a historical context (mainly regarding dates, i.e war beginning 1 Sep 1939 in Europe).

But it will have built sizeable landarmies by mid 1940. The first thing that happens when Italy joins axis that it gets crushed in africa by the UK. The UK you claim to be short on ic makes any AI attempt of a sealion a failure.

As I'm sure you're aware, the amphibious invasion AI is appalling - even USA & UK AI struggle to land against the German AI. The UK has the biggest fleet in the world so obviously it should make any AI attempt of a Sealion a failure. The overseas AI is only slightly better; and with the Italy AI there's the never-ending problem of finding a balance between it defending its mainland and it sending units abroad - I believe that having 1/2 divisions locked on each Italian beach would help resolve this. They can be unlocked if either Italy is human or if the Axis AIs have taken Gibraltar and Suez.

Anyway, as I'm also sure you're aware, Italy performed awful IRL from start to finish - particularly in 1940. 36,000 Commonwealth troops took 140,000 Italians prisoner by February 1941.

Well, the precise dates varies from nation to nation and one needs to examine what maximizing means concretely. It means that the maximum speed of mustering new divisions reaches its peak long before then. AI Germany has about 100 mobile divisions for its historical Barbarossa "albeit" it spams out Infra like nuts. Or should that albeit be a because? The same is true for soviet union.

I doubt Germany is stronger by 1941 for building infra in Magdeburg, Kiel, Hamburg etc. The same for the USSR building infra in Kiev, Vinnitsa, Rostov etc. You do know those provinces lose their IC via the movement of industry events, right?

Both logical deduction and empirical evidence suggest otherwise. AI is rather strong at building up an army, but it is rather poor at managing it.

So you don't believe delaying Danzig benefits the Allies? Common sense suggests otherwise. The IC, manpower and resources of the UK, USA, France, USSR, Poland and all the Allied minors are much greater then Germany, Italy, Japan and the Axis minors.

 

[Mr_B0narpte]

Did i say such a thing?

I could probably answer to other parts of what you wrote, but as you wrote it before analyzing the actual situation this seems to make little sense. This thread refers to the concrete situation of AI Infra in set33(Beta 11) and the soon to be published set34. There the results of AI Infra seem to be rather fine.

I imagine you said you fear 1.09 being the last update due to a lack of popular interest in its further development.

I am grateful for Commander666 supporting my views on this matter regarding the AI infrastructure builds, I hope other people express their views on the topic.

 

[Commander666]

Did i say such a thing?

Yes, you did. I clearly remember it. But without also remembering the exact phrasing you used, it's impossible to search for your comment as searching for "1.09" gives you over 100 pages of posts listed. Other searches including "last" or "final" still leaves far too much to read thru for making a simple point: Sometime ago you wrote that 1.09 would probably be the last version, or wrote something there abouts.

... but as you wrote it before analyzing the actual situation this seems to make little sense.

I did analyze it - on the basis of it being reported that USA has about 80 infra serials. If true, that is NUTS! If not true, you should correct Mr_BOnarpte's statement regarding the number of infra serials with USA. You know, not everybody needs to fly on a new airplane to judge the experience. It is quite enough to analyze it based only on the reports of others who have flown on the new plane - provided one trusts their reports. Further, I did read your and Mr_BOnarpte's comments to each other very thoroughly.

There the results of AI Infra seem to be rather fine.

Yes, in your opinion. Obviously, your opinion must not fit numerous other opinions since you went out of the way to explain your changes. I'm glad others are taking the time to go out of their way also to state honestly what they think about the changes. And I'll be glad to discuss one or two things with you in detail - like new infra for USA and India. Unfortunately, I'm afraid that you will probably only prove my point that you are changing the game into more of an economics game than a war game. Certainly just reading some particular comments of yours in the above discussion with Mr_BOnarpte sounds like pure economics and little else. I'll stand by my statement that an AI USA does not need any more infra except in Washington (and India none) to make this a better war game.

Further huge changes usually do not result in improving games... they just create worse versions. AoD does need improving, but it needs many small things tweaked and other things fixed. You, however, had your vision of major infra changes scheduled for 1.09 very long ago (just discovered that in the searching I did do). Clearly, you have a vision for what you think AoD should be...clearly others don't agree... and clearly I don't like my favorite game being majorly changed as to what I see as mostly "one person's opinion". Just how that's happening I'm still trying to figure out.

 

[Commander666]

I imagine you said you fear 1.09 being the last update due to a lack of popular interest in its further development.

No, that is not what I remember. There was no "fear" mentioned, nor was the statement as long as explaining why. "popular interest" was also not what I remember. Rather it was just a short thing like "Aod will probably be the last (or maybe final) version. Anyway, I searched quite a while to find it, but didn't. Seems Pang would rather remain "mystery man" since he didn't just simply clarify.... preferring instead to ask me if I was sure. The answer is "YES"