Rough Analysis of IC Improvement Returns to Investment

[Young Hickory]

I'd love to make the "definitive IC improvement" post, unfortunately that isn't going to happen right now, but given my observations of the topic's discussion here I think I can at least take it to a more analytic level and dispel some of the misconceptions I've seen.

There are almost certainly some mistakes, so read critically

I. Relevant Factors

1. IC improvement build time

Build time starts at 360 days (a paradox year), but is reduced by the sliders:
a. Hawk lobby
b. Free Market

The lower the build time, the more quickly the investment will pay off.
These sliders will affect build time, but not build cost (Cost will still be 5 IC, but the time will be reduced by the relevant percentage).

Note that serial builds will not reduce build time, and slider effects will not be applied to improvements already underway.

Example:
With full free market (-20%) and a -5% build time from hawk lobby an IC improvement will take 360 * .75 = 270 days.

2. Base IC increase vs. Actual IC input

The 5 IC cost of improvement is taken out of the modified IC total (like all other production), but the IC increase is applied to the unmodified base IC. This means that the input to output ratio will be better then 5:1.

IC is modified by:
a. Central planning
b. Ministers (Armaments, Intelligence, heads of state/government AFAIK)
c. Technology (The three industrial techs in the upper left corner)
d. Difficulty level

For example, if you have:
+25% (full central planning) + 20% (ministers) +15% (technology) = +60%

So for every IC increase you will get 1.0 * 1.6 = 1.6 IC increase, for a input output ratio of 5:1.6 (I hope I'm not being annoyingly simple, I just figure that if I'm trying to be informative I might as well lay it all out)

Generally the bigger your IC modifier, the quicker IC improvements will pay off.

3. Modified IC to Consumer Goods (CG) ratio (from MadViking)

Not all of the additional IC will go towards increased productive capacity; some will go towards consumer goods. Exactly how the consumer goods ratio is determined from the various modifiers is eluding me, but it's not necessary to have a formula for this analysis (I assume it is calculated from a percentage of base IC with the modifiers applied in some counterintuitive order of operations).

IC:CG ratio is affected by:
1. Central Planning/Free Market slider
2. Interventionism slider
3. Minister effects
4. At war or not

The fewer ICs that need to go to CG the more IC investment will pay off.

For example:
If 20% of your IC goes to CG (and the above situation) then your productive IC from one increase will be 1.6 *0.80 = 1.28

4. Opportunity costs

In addition to the basic opportunity cost of present v. future IC (a largely qualitative judgment beyond the scope of this discussion), there are objective changes in the costs of other production.

This will be affected by many factors including:
a. Slider changes (Hawk Lobby, Free Market)
b. Technology (Mass production techs)
c. Ministers that reduce build time
d. Gearing Bonuses (and thus drafted army)

For Example:
In 1936 it will cost Germany 7*95 = 665 IC*Days (ICD) to build an infantry unit, while in 1940 with full Hawk lobby and small arms mass production the cost will be 5.6 * 61 = 341.6 ICD, 51.4% of the 1936 cost.
(From MadViking: 61 days = 95 (base) *0.8 (hawk lobby) * 0.8 (small arms assembly); not 57= 95*.8 (hawk) - 19 (SAA) as one migh think).

II. Rough analysis

Given the above information, and considering that when changes in these factors takes place is important, it would appear to be difficult to make a meaningful formula to determine when IC investment will pay off. I'll try to run through an analysis of when 1936 IC investment pays off for Germany given reasonable assumptions.

Assumptions:
1. The first slider move is towards central planning (no immediate IC effect), and this is continued until central planning is complete
2. Interventionism & hawk lobby are allowed to progress normally
2. IC builds are started on January 3, 1936 right after the re-occupation event (giving a 5% time bonus)
3. Ministers and tech are chosen to max IC (20% minister bonus, and 15% tech bonus is the best AFAIK, done as quickly as possible)
4. Difficulty level is normal

Total IC bonuses will be assumed to be (IMHO good estimates, but not perfect):
+40% IC (15 slider, 15 minister, 10 tech), 23.2% to CG for 12/15/36 to 12/30/37
+50% IC (20 slider, 15 minister, 15 tech), 19.5% to CG from 1/1/38 to 12/30/38
+60% IC (25 slider, 20 minister, 15 tech), 15.4% to CG from 1/1/39 to 10/1/39
+60% IC (25 slider, 20 minister, 15 tech), 7.7% to CG from 10/2/39 on

Percentage of IC used on CG

Build time in 1/3/36 is 360 * .95 = 342 (done 12/15/36, for a total cost of 5 * 342 = 1710 ICD

Payoff from 12/15/36 to 12/30/37: 1.4 * 375 = 535 ICD * 76.8% = 411 ICD
1/1/38 to 12/30/39: 1.5 * 360 = 540 ICD * 80.5% = 435 ICD (total 411 + 435 = 846 ICD)
1/1/39 to 10/1/39: 1.6 * 300 = 480 ICD * 84.6% = 406 ICD (total 846 + 406 = 1252 ICD)
10/2/39 to 8/10/40: 1.6 * 310 = 496 ICD * 92.3% = 458 ICD, for a total 1252 + 485 = 1737!!!

So by early August of 1940 you have broken even in productive IC. Given that a fully geared infantry division will cost about 224 ICD at that point, you can get about 2.4 additional infantry divisions per year for every IC improvement you did in 1936.


III. Other considerations

1. TC

For every IC improvement you will be getting: 1 * IC modifier * TC modifier additional TC (this will be 1.6 * 1.8 = 2.88 additional TC given the above IC modifiers and all TC improving tech but no TC improving ministers).

2. Additional tech teams (from Malch)

A nation gets 1 tech team for every 20 IC rounded up, so the intervals 21, 41, 61, and 81 are very important (each will grant an additional tech team). Without going into a deep analysis it should be apparent that should be a serious consideration, particularly if you are close to one of the cut off points.

3. Increased energy to oil conversion (from MadViking)

When energy reserves are more than twice oil reserves then energy is converted to oil such that you get 10% of you IC in oil every day (the conversion rate is determined by tech). Naturally as you increase IC you will also get more oil from conversion.

For example if you have a +60% IC modifier then you will get 1.6 IC * 0.1 = .16 additional oil per day from conversion for every IC increase you invest in.

4. "You go to war with the army you have, not the army you wish you had, or the army you hope to have at some future time" - Rummy (paraphrased from memory)

I assume I don't need to expand

 

[unmerged(14035)]

with full Hawk lobby and small arms mass production the cost will be 5.6 * 61 = 341.6 ICD, 51.4% of the 1936 cost (the data is from a save, my calculations would lead me to think the time would be 57 days, 95 *.8 - 19 = 57; but 61 it is, not sure why)

The reason it says 61 instead of 57 is this: the way small arms assembly tech (and the other similar assembly techs) works is that it reduces the cost by a fix percent of the base time (check the files for this, it's 20% for almost all units except marines/para/mountain/engineers which are only 10%). Any bonus from sliders is not added, instead they are mutiplied, so in your example its 95 x 0.8 x 0.8 = 60.8, which is rounded upwards (seems it always rounds up) to 61.

Also, in your analysis you are forgetting one very important thing: CG demand. All the extra IC you get from building a factory will NOT be used to buy new troops etc, some will be turned into CG. This must definately be taken into consideration, as this is quite a considerable amount.

Another benefit of having more IC (besides the TC) is that you can convert more energy into oil.

 

[Young Hickory]

The reason it says 61 instead of 57 is this: the way small arms assembly tech (and the other similar assembly techs) works is that it reduces the cost by a fix percent of the base time (check the files for this, it's 20% for almost all units except marines/para/mountain/engineers which are only 10%). Any bonus from sliders is not added, instead they are mutiplied, so in your example its 95 x 0.8 x 0.8 = 60.8, which is rounded upwards (seems it always rounds up) to 61.

Ahh, that makes sense, thanks.

It could also be, (95-19) *0.8 = 60.8, which would be in line with the stated effects of small arms assembly, but given's Paradox's inclination towards multiplication the .8*.8 is probably more likely.

I should have though of that

Also, in your analysis you are forgetting one very important thing: CG demand. All the extra IC you get from building a factory will NOT be used to buy new troops etc, some will be turned into CG. This must definately be taken into consideration, as this is quite a considerable amount.

Good point, that's certainly going to be a non-trivial amount. If there is sufficient interest I'll work that into the analysis tonight.

Another benefit of having more IC (besides the TC) is that you can convert more energy into oil.

Somewhat minor considering that the increased IC from investment will be a rather small percentage of your total IC by the point in the game where you might be running into oil shortages, but an important point nonetheless.

 

[unmerged(18911)]

excellent analysis, thanks for the info!

 

[Malch]

Don't forget the techteam factor. For some minors its better to build IC's to get that extra tech team

 

[Young Hickory]

excellent analysis, thanks for the info!

Thanks, I'm glad it was helpful

Don't forget the techteam factor. For some minors its better to build IC's to get that extra tech team

Good point, added to "other considerations."


I also updated the original post to incorporate the CG considerations raised by MadViking.

 

[unmerged(14035)]

Somewhat minor considering that the increased IC from investment will be a rather small percentage of your total IC by the point in the game where you might be running into oil shortages, but an important point nonetheless.

Well, as Germany (on normal/normal) I regularly build about 50 more IC during 36/37, which is about +35% of the starting base IC. Even after you take the usual nations (Poland, Benelux, France etc) this will be around 20-25% more IC, which would translate into 20-25% more oil from conversion. The 50 base IC with good sliders/tech/ministers equals 80 IC more or 8 oil per day, wich is about 3000 more oil per year. Probably won't win the game for you, but during a 3-year period this is around 10k oil. Every drop counts if you want to keep the panzers rolling

 

[Young Hickory]

Well, as Germany (on normal/normal) I regularly build about 50 more IC during 36/37, which is about +35% of the starting base IC. Even after you take the usual nations (Poland, Benelux, France etc) this will be around 20-25% more IC, which would translate into 20-25% more oil from conversion. The 50 base IC with good sliders/tech/ministers equals 80 IC more or 8 oil per day, wich is about 3000 more oil per year. Probably won't win the game for you, but during a 3-year period this is around 10k oil. Every drop counts if you want to keep the panzers rolling

That's certainly some serious IC investment you're talking about. Is that all you do until the second round is finished?

I suppose at that level it really does make a difference in oil conversion, a significant and useful side effect of intensive IC investment. Of course if you're main concern is oil trading supplies will probably be your best bet.

 

[unmerged(14035)]

That's certainly some serious IC investment you're talking about. Is that all you do until the second round is finished?

If I plan on playing at least semi-historically (starting WWII via the Danzig event) then I build 20-23 IC in 1936 and 30-32 IC in 1937. Usually I make 52-53 total, which is quite a bit considering that you start with 142. You also get 19 more from the Anschluss, and up to this event I don't build or upgrade a single unit (I might finish the ships in the build que, but not necessarily). All I do is build factories and supplies, and research. And lots of trading

Once Anschluss happens I will be at 214 base IC and a total of +50% from sliders/ministers/tech for a total of 321 IC. In other words, it's time for some serious rearmament of das Vaterland

The extra oil conversion you get for the 53 IC might not win the game for you, but it also gives you 160 extra TC when at full tech. And 160 TC might actually win the game for you. Yhat, and some panzers of course

 

[Graymane]

I know this is a late entry and possibly as a result of 1.1, I'm not sure. The Infrastructure of the province affects the Factory build time. 100% Infra provinces will build factories something like 1 month faster or so than a 60% Infra province.

 

[TheLand]

I've linked to this thread from the HOI2 Wiki:

http://hoi2.nsen.ch/wiki/index.php/Do_I_build_Factories?

 

[unmerged(54503)]

I've linked to this thread from the HOI2 Wiki:

http://hoi2.nsen.ch/wiki/index.php/Do_I_build_Factories?

That is how I found it thanks. This is a great thread to read for newbies like myself.

There is another major advantage to improving IC in the beginning, but its kind of abstract.

It will allow you far more versatility in the future should you experience an unexpected disaster that might have well dragged your war effort to a grinding halt.

It furthermore buffers you from the chance of finding yourself in an economic crunch sown the road.

How can you quantify the advantage of preventing something that might have happened.

Also bulding IC can also be considered a defensive strategy moreso than an aggressive one.

I think it is a better stratagy for less experienced players than it could ever be for an expert who can plan on the certain addition of IP from conquests.

 

[Mork]

Now that this has been dug from the deeps of oblivion, I just has to say that the CG needed by extra IC also adds money, money that can be used to trade, or in times of war, get the nessesary money for tech teams (I quite often find that I have to run at higher than minimum at war, extra IC lessen that).
Point is that the % of IC used to build CG is not wasted as might be surgested by this thread.

 

[blue emu]

One other observation that might be worth making...

The percentage of IC's required by Consumer Goods also depends largely on your form of government... Democracy, Communism or Fascism. That is... it depends to a great extent on your Authoritarian slider and your Open/Closed Society slider.

For certain countries... Democratic nations in particular, such as the USA... you will get a much better profit from using the Laissez-Faire Capitalist (-20% to Consumer Goods) Industry Minister than you would obtain from any Industry Minister who boosts IC production, since it allows more of your existing IC's to be actually USED productively.

As a side-benefit... that Minister does not increase your consumption of raw materials, unlike the IC-boosting Industry Ministers. In effect, he obtains his results by making more efficient use of your existing resources, rather than by increasing your consumption of resources.

 

[unmerged(13991)]

democratic nations , USA ... I wonder why some words can not fit together??? For proof the people who will complain I used my free speech rights (guarantee by the first amendment I think).

Anyway, building IC will also increase something that can be considered as the 'cash flow' concept in economy : How much of something of value you can have at one time?

Let say we paly germany, there is 2 ways to build my army :

1) it will take 3 years and will gradually appear. Good, like that we will pay for uselss soldiers in 1936, 1937 and 1938.

2) create it in 4 month.

For the later, you must have a huge IC ready to use, yes you have to build it before. The interrest is you will not pay (supply) for the army before you need it. It will also push you to think a bit about what you will need and stop the building of army with 200 + divisions in 1939 as I have read in some post.Also, it will allow you to get directly up to date units (you will not pay for the upgrade).

I just did a game with france with following those concepts:

a) build IC up to septembre 39 (but fort 10 all along the border)
b) start to build a small army from sep 39 up to may 40.
c) crash axis forces by the 1st of november 40 (germany surrended, Italy surrended, slovaquia annexed, hungry annexed, romania annexed).

For the record , I did not build any infantery France have enought with the starting unit, but 6 armored corps (2 arm + 1 cav or mots). I got some wings of CAS + the 2 TAC already here. Around 450 IC in nov 1940 after I released as pupet Italia, thechoslovaquia, austria, hungria and Romania.The base IC in the national provinces must be arount 250 I think.

Huge IC allow to build many queues on the same time. I runned 9 queue of CAS and 3 of ESC. 4 queues of ARM....

 

[Long Lance]

Just a sidenote: For me factories are too expensive. I changed the build time to 240 days.

 

[unmerged(3221)]

Even in mid game, you might as well build some more factories. The more factories that you have, the more supplies that you can produce. The more supplies that you can stockpile, the easier it is for you to do enormous upgrade or naval/aircraft builds.

After you reach very high stockpiles, if you are not ready to build units at that point because they are not needed yet or the research is not done yet, build more factories. I built 60 factories pre war in a normal Germany game and 36 more in 1942/43 because my supply stockpile was over 370K.

This means that each additional factory gives me 1.5 IC (minus a small at war CG need) and slightly more than 3 TC. So my 96 factory build gives me 150 more available IC and 461 more TC (which is about 30% of my TC).

I would argue that the IC math is irrelevant. As long as you have enough resources, and that's never been a problem for me yet in 1.3, you always will have the ability to build more units, upgrade more units, or to build up large supply stockpiles for the times when you need to build and upgrade. You get much more TC. You can always idle factories if resources ever get real low, and that doesn't affect TC at all.

Everyone more or less ends up conquering the same foreign lands anyways as Germany, so I regard any argument that mentions that as irrelevant. Just build the factories instead and you are much better off in the long run and much better off when you need to balance the cycles in the game (build cycles, upgrade cycles, normal day to day cycles, and factory build cycles).