German build up: A simple comparison between a CIC build and pure MIC.

[Black_Shade]

But that seems to me a pretty valuable result: that by foregoing CIC entirely, you boost your military production by a mere 8%, explicitly discounting trade effects. Or in other words, building CIC means exchanging 8% of raw output, again without calculating the effects of trade, for the benefits said CIC give you.

I do not understand how one would properly include trade into the calculation. Is trade static? Is it not subject to considerably price fluctuations? A worthwhile study must be reproducible, with trade, can you have that or will we end up discussing whether or not some big trade deal was particularly good or bad, distorting the entire experiment?

It's not real production output. You'd never actually be able to get that 8% output.

So what did you actually learn? That in some instance that would literally never, ever happen, you might be 8% ahead. But that's not actually valuable information, because the things you need to produce to actually win don't conform to the experiment!

I don't understand how not including trade in the calculation makes a useful result, you've learned nothing of value. Trade has to be included for the test to actually be useful, since you lack the rubber and tungsten to produce tanks and planes as Germany and that obviously impacts your production capabilities on things you actually need to produce. Also, I'm not sure what you mean by price fluctuations. Its a civ for 8 resources no matter what.

I don't think it would be that hard to do exactly the same test, but produce only fighters and make sure that you were sitting at near peak efficiency for resource use. Parameters for trade (most important for the MIC I guess)- once you get to -4 rubber start another trade for another batch of rubber. The other requirement is 15 synths by Aug 1939. Obviously trade is not static because the number of mils coming online dictates how much you're trading. But the impact of said trading is what you want to test, so including variable trade in the experiment is desirable, not a drawback. Just build mils only in one case, with a switch to synths so you hit the 15 in Aug '39 (i'm guessing that means you stop building mils in novemberish of '38 since you're going to be civ starved). For civs, do that same path that SM did in his first test, but just build fighters instead of ACs. Similarly, keep your resources balanced so that you never drop below -4 rubber. Again, you still have to hit the 15 synths by 1939, but you have many more civs, so that means start building these in febuaryish of '39. After that, just keep building mils on both builds with a synth dropped in every 6 mils to take care of the rubber requirements. Take identical focus paths.

Compare the number of fighters in Aug '39, June '41.


edit: actually, after thinking about it the best thing to do is actually probably make Medium Tanks 1 non stop. The problem with fighters is that 15 synths is enough to cover ~60 mils worth of fighter production, which is likely what you'll be putting on them, but producing more fighters means you have to keep building synths, as you can't trade for rubber... but tungsten doesn't have this problem as Portugal and the Soviets have plenty up until 1941. The requirement to keep building more synths would obviously restrict the MIC build more than CIC, so would skew the test. However, medium tanks 1 and medium tanks 2 don't suffer from the trade problem, but also have the same tungsten cost. I usually hold off building MT until MT2, but since they cost the same tungsten it doesn't matter for the test, so the best thing to do is the following:

Cheat and research 1939 medium right off the bat for both builds. Put your mils on tank production, keeping tungsten balance relatively ok (add in a new trade once you get to -4 tungsten). You still have to build 15 synths (even though you're not using the rubber output) because you'd build at least that many in a normal game anyways.

Compare medium tank 1 numbers in August '39 and June '41.

This would give you a realistic look at the production outputs including trade for a MIC only vs. a CIC only build that at least kind of matches what a German player might get in a real game.

 

[Simon_9732495]

If you did the test and produced only fighters, you'd find the opposite result: the CIC build up would get you more net fighter output by 1941 than the MIC build,

I doubt that. Have you done a test on this? Any calculations to prove your point?

It seems you forget that in 1939 with the MIC build you have a lot more Fighters AND MILs. Yes trade trade slows you down your MIC building more at that point because you have less CIC, but that isn't slowing down your fighter production.
The trade also slows down the CIC build. And the CIC build has to overtake in MIC AND in production efficiency before it can start retaking the lost ground in number of fighters.

I DID a test and it felt I had more fighter with the MIC build than with my normal 2 years CIC Germany run. (But it can also be because I put more factories on fighters. It wasnt a dedicated test, so it's only a feeling.)

 

[Black_Shade]

I doubt that. Have you done a test on this? Any calculations to prove your point?

It seems you forget that in 1939 with the MIC build you have a lot more Fighters AND MILs. Yes trade trade slows you down your MIC building more at that point because you have less CIC, but that isn't slowing down your fighter production.
The trade also slows down the CIC build. And the CIC build has to overtake in MIC AND in production efficiency before it can start retaking the lost ground in number of fighters.

I DID a test and it felt I had more fighter with the MIC build than with my normal 2 years CIC Germany run. (But it can also be because I put more factories on fighters. It wasnt a dedicated test, so it's only a feeling.)

In SMs test, he had only an 8% boost with MIC over CIC in 1941. Thats with literally the entire test being stacked in MICs favor using a ridiculous production output that only used steel... and it barely won. If you did a realistic test, there's no way that CIC doesn't make up that ground and then some.

I also just realized that he used concentrated industry over dispersed, which I doubt many players do, which also heavily biases the test in favor of MIC. Not only does dispersed let you start at a higher production efficiency (lowering the lag time for CIC to catch up there), it also gives a big boost to the initial output since SM never changed the production line and was making AC1 the entire time.

So basically, the test needs to be redone as I outlined in post #81 with MT1, and you should also use dispersed industry instead of concentrated.

 

[BeauNiddle]

In SMs test, he had only an 8% boost with MIC over CIC in 1941. Thats with literally the entire test being stacked in MICs favor using a ridiculous production output that only used steel... and it barely won. If you did a realistic test, there's no way that CIC doesn't make up that ground and then some.

But it won by a lot in Jan 1940. SM's point is that you have to tailor your approach to your desires. If you have already captured the Soviet factories by 1941 the 8% difference is irrelevant.

Some people on this forum throw the meta around as holy writ - you must build 2 years of civ, you must use 10/0 infantry, you must X/Y/Z. The game is a lot more flexible than people think and often there are choices made that you are overlooking. In this case which year will you be kicking off against soviets. If you do a 2 year civ build then the Soviets know they have time for a 3 year civ build. If the soviets prep for a 3 year civ build and the Germans go for a MIC build then 1940 is going to be deadly.

Against the AI (especially on historical focuses) a lot of this is irrelevant but it's always interesting to read about high end play from those that actually stress the system.

Paradox have once again written an interesting and flexible system, then hidden it from the normal players

 

[Zauberelefant]

In SMs test, he had only an 8% boost with MIC over CIC in 1941. Thats with literally the entire test being stacked in MICs favor using a ridiculous production output that only used steel... and it barely won. If you did a realistic test, there's no way that CIC doesn't make up that ground and then some.

I also just realized that he used concentrated industry over dispersed, which I doubt many players do, which also heavily biases the test in favor of MIC. Not only does dispersed let you start at a higher production efficiency (lowering the lag time for CIC to catch up there), it also gives a big boost to the initial output since SM never changed the production line and was making AC1 the entire time.

So basically, the test needs to be redone as I outlined in post #81 with MT1, and you should also use dispersed industry instead of concentrated.

But building MARM1 from the get go produces the same issue as building AC1.
I also argue that fighters are not priority 1 in 1936, as you need manpower in the field for the focuses.

 

[Fulmen]

I also argue that fighters are not priority 1 in 1936, as you need manpower in the field for the focuses.

You only need about 3 factories on guns, a few on artillery (I like to do 4, so I don't have to import tungsten) if you're using line artillery (as opposed to using infantry as sponges and making more ARM & CAS instead) and 3-4 on support equipment. Then you just build the starting templates and later upgrade them to 40 widths with the XP you're getting from attachés, lend-lease, etc. No need to rush-deploy green divisions, nor exercise trained ones to have the necessary manpower in the field for a historical timetable. Actually following this method you can go for Anschluss already in mid-1937.

 

[Black_Shade]

But building MARM1 from the get go produces the same issue as building AC1.
I also argue that fighters are not priority 1 in 1936, as you need manpower in the field for the focuses.

Why does Marm1 produce the same issue as building AC? It requires 2 tungsten, which you have to trade for, giving you a more realistic production outcome.

Fighters are absolutely priority #1 in 1936 as Germany. You should put the bare minimum number of factories on guns, 3-4. What you do is build cav divisions and transform them to normal infantry. They have a lot less equipment requirements so they train faster.

I usually put 3-4 factories on CAS, 3-4 on guns, a few on support, and everything else fighters (15-20).



As an aside, I already tried to do this test using MIC going MARM, because the early game through the fall of france doesn't take very long. It turns out that you don't have enough Civs to actually do the trades required to get the tungsten until the fall of poland/france, you'll literally have 0 civs to build by the time 1939 rolls around. Once you take France/Poland though, it turns out there isn't enough tungsten in the world to actually supply your factories. At the time of the war outbreak with Poland I had 27k MT1. I also did build up the spy agency and ran 2 collaboration governments on Poland and France. Still have to run the test with the civs but its obviously going to be a whole lot less.

Because of the tungsten supply issue I guess to do a real test you actually will have to split production between planes and tanks. Probably focus on planes until 60 mils, then have the new mils that come online to tanks after that. This is also probably pretty close to what a german player would do in MP anyways (assuming no human france and not building tanks until MT2).

 

[Fulmen]

What you do is build cav divisions and transform them to normal infantry.

You're wasting equipment. You can easily achieve the required manpower figures without green divisions or excessive on-map training and its related attrition, whilst also keeping a bare minimum of factories dedicated to infantry divisions. See my post above.

 

[Black_Shade]

You're wasting equipment. You can easily achieve the required manpower figures without green divisions or excessive on-map training and losing of equipment from attrition, whilst also keeping a bare minimum of factories dedicated to infantry divisions. See my post above.

In terms of min/maxing I like to run Anschluss as the 3rd focus, because it lets you switch to war economy right away.

I do rhineland, Four Year Plan (rushing construction 3 with the tech bonus here), followed by Anschluss, which gives you a WS boost to switch to War economy right away.

 

[Fulmen]

In terms of min/maxing I like to run Anschluss as the 3rd focus, because it lets you switch to war economy right away.

Early Anschluss is not optimal, and you can get War Economy through attaché to Spain anyway.

 

[Black_Shade]

Early Anschluss is not optimal, and you can get War Economy through attaché to Spain anyway.

Why? You also get 9 additional civs+3 military factories right off the bat, which in terms of civ output is huge. 9 Civs in particular at the start it makes Anschluss the biggest production boost in the game.

Only thing I could think would be a WT bump. UK will get to run Shadow Scheme but they can do that because of the SCW if they are paying attention, but they wont be able to do General Rearmament either way. Plus if you do it early it will actually decay a couple % by the time Japan attacks China, which might prevent other WT gated focuses from being able to trigger.

 

[Simon_9732495]

Why? You also get 9 additional civs+3 military factories right off the bat, which in terms of civ output is huge. 9 Civs in particular at the start it makes Anschluss the biggest production boost in the game.

Your feeling on this is wrong.

See this thread:

1.9 German Economy Benchmark Thread

There were some changes to economy in Husky and I'd like to optimize my Germany Buildup 1936-1939 again. To see where I stand I would like to compare my builds with others. Therefore I'd suggest the following benchmark: Date: 1.6.1939 (Short...

forum.paradoxplaza.com

Several tests show that early Anschluss is not "the biggest production boost in the game".
Reasons: The other early focuses of Germany are also very strong. Early war economy doesnt help that much because you are not building MILs. With later Anschluss Austria has more Industry.

If you want to make factual statements like "A is better than B" you should do more research and tests or have reliable sources. Otherwise you should stick to "i think A is better than B" or "A might be better than B".


That is also and especially valid in this thread. You state:

If you did a realistic test, there's no way that CIC doesn't make up that ground and then some.

But you bring nothing to back it up.
It could easily be that in a "realistic test" (whatever that might be, please provide a clean definition) the MIC build is still better in 1941.
You have a feeling that it's that way, but you have no evidence. That is what makes @Secret Master s test valuable. It provides evidence for 1 scenario.

 

[Zauberelefant]

Why does Marm1 produce the same issue as building AC? It requires 2 tungsten, which you have to trade for, giving you a more realistic production outcome.

Fighters are absolutely priority #1 in 1936 as Germany. You should put the bare minimum number of factories on guns, 3-4. What you do is build cav divisions and transform them to normal infantry. They have a lot less equipment requirements so they train faster.

I usually put 3-4 factories on CAS, 3-4 on guns, a few on support, and everything else fighters (15-20).



As an aside, I already tried to do this test using MIC going MARM, because the early game through the fall of france doesn't take very long. It turns out that you don't have enough Civs to actually do the trades required to get the tungsten until the fall of poland/france, you'll literally have 0 civs to build by the time 1939 rolls around. Once you take France/Poland though, it turns out there isn't enough tungsten in the world to actually supply your factories. At the time of the war outbreak with Poland I had 27k MT1. I also did build up the spy agency and ran 2 collaboration governments on Poland and France. Still have to run the test with the civs but its obviously going to be a whole lot less.

Because of the tungsten supply issue I guess to do a real test you actually will have to split production between planes and tanks. Probably focus on planes until 60 mils, then have the new mils that come online to tanks after that. This is also probably pretty close to what a german player would do in MP anyways (assuming no human france and not building tanks until MT2).

MT 1 from 36 produces the same issue as AC1 from 36 - maxed efficiency.
And fighters are more valuable when upgraded, so you are better off saving CIC, getting Air XP and then mass produce the 109 B from 37 onwards. Upgrading obsolete planes is wasteful.

You can debate what's better for min maxing, but I hold that unless you skewer the comparison into one direction, any playthrough will produce meaningful results.

 

[Black_Shade]

Your feeling on this is wrong.

See this thread:

1.9 German Economy Benchmark Thread

There were some changes to economy in Husky and I'd like to optimize my Germany Buildup 1936-1939 again. To see where I stand I would like to compare my builds with others. Therefore I'd suggest the following benchmark: Date: 1.6.1939 (Short...

forum.paradoxplaza.com

Several tests show that early Anschluss is not "the biggest production boost in the game".
Reasons: The other early focuses of Germany are also very strong. Early war economy doesnt help that much because you are not building MILs. With later Anschluss Austria has more Industry.

If you want to make factual statements like "A is better than B" you should do more research and tests or have reliable sources. Otherwise you should stick to "i think A is better than B" or "A might be better than B".


That is also and especially valid in this thread. You state:


But you bring nothing to back it up.
It could easily be that in a "realistic test" (whatever that might be, please provide a clean definition) the MIC build is still better in 1941.
You have a feeling that it's that way, but you have no evidence. That is what makes @Secret Master s test valuable. It provides evidence for 1 scenario.

How does Austria have more industry if you wait? They will get through 3 focuses themselves, if they start on the industrial path that means they finish 2 factories, and you miss out on 3 that they could have gotten via focus. The ones they build themselves is with lower tech and slower than what you would have done if you built them yourself (due to a combination of tech and snowballing), and the civs you get from them are far better than Goring and KdF. Plus, you're going to take those immediatly after Anschluss anyways, what you are doing is changing the order to get 9 civs 140 days sooner than 6 civs.

Also, you realize I posted in the thead you linked several times, and have some of the highest scores in there going Anschluss third, right?

If you could point to specific posts that show its clearly worse please do, because I'm not seeing many (if any) scores higher than what I had already and that I had posted in the thread. I'm not sure how me having some of the highest scores means that it's clearly worse.

 

[Simon_9732495]

If you could point to specific posts that show its clearly worse please do

Please be precise. I never said it's clearly worse.
I said:

Several tests show that early Anschluss is not "the biggest production boost in the game".

This posting shows the best result in the thread. And it's without early Anschluss:
https://forum.paradoxplaza.com/foru...benchmark-thread.1387471/page-4#post-26551698

If early Anschluss was "the biggest production boost in the game" it should give a lot better results in the economic benchmark than other runs.

 

[Black_Shade]

Please be precise. I never said it's clearly worse.
I said:


This posting shows the best result in the thread. And it's without early Anschluss:
https://forum.paradoxplaza.com/foru...benchmark-thread.1387471/page-4#post-26551698

If early Anschluss was "the biggest production boost in the game" it should give a lot better results in the economic benchmark than other runs.

So that's worse than the one I posted, because I had 2.35 mil but also had 4 collaborations done at the same time, so you'd get nearly full factories from Poland+France as well once they were down.

 

[Nuklearius]

#

 

[Simon_9732495]

So that's worse than the one I posted

Please Link.

because I had 2.35 mil

What do mean by that?

What's the point of this entire experiment? No one builds equipment that uses steel only.

Was explained several times in this thread. Everybody does a build different. To find a common ground a simple, granular item was chosen for production.

The Test can easly be redone for other production. Just sit down some hours and do it.

 

[Black_Shade]

Please Link.


What do mean by that?



Was explained several times in this thread. Everybody does a build different. To find a common ground a simple, granular item was chosen for production.

The Test can easly be redone for other production. Just sit down some hours and do it.

Post 52 in the thread you linked, I had 2.35 mil according to your scoring system, but also had already paid for the civ costs for 4 collaboration government events, which is worth far more than the extra few factories that got that 2.5 million score.

 

[Simon_9732495]

Post 52 in the thread you linked, I had 2.35 mil according to your scoring system, but also had already paid for the civ costs for 4 collaboration government events, which is worth far more than the extra few factories that got that 2.5 million score.

You refer to that one:
https://forum.paradoxplaza.com/foru...conomy-benchmark-thread.1387471/post-26547764

You had a score of 2.337.900 but you ran the test 1 month longer. 1 month is about 5 MIls so your comparable score would be around 2.300.000.
That is 225.000 less than the top benchmark I linked above.
That equals 225.000/7200 =~ 31 MILs less.

The 4 collaboration are worth a lot but the don't cost much. I think it was 5 CIVs, 5 MIL if I remember correctly. Certainly not 200.000 benchmark points.

Your result is very good, but it's by far not the top result in the thread. The top result was without early Anschluss.
If early Anschluss was the "the biggest production boost in the game" it would be a must for good results in that benchmark. It is not.