• We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.

unmerged(19545)

Second Lieutenant
Sep 11, 2003
185
0
Visit site
Someone posted a while back in Math Guy's HSP announcement thread something like:

We're testing what's happening inside the black box, by tugging on the bits of string hanging out, and seeing what happens.

Well, inspired by that, I've starting tugging on every bit of string I could lay hold of. Here are the first few hour's results. All assertions are tested rigorously, according to proper scientific method. Assumptions and areas of uncertainty should be clearly labeled as such. Anyone who has tested other things that seem silly or overly obsessive are cordially invited to add them to this thread. Most particularly if they can fill in a blank I identify, or point out an assumption I have blithely made without proper testing.

1: Unit build costs can only be integers, and the engine rounds down, not up. I tried setting infantry build cost at 0.219, and putting 10 in my queue. If the engine read past the decimal point, cost would have been 2.19, or 2.2, or even 2 depending on rounding. But it was 0. To ensure this wasn't a disparity between actual ic's used, and what the display said, I cheated some manpower, built 100, and set IC to 0. All 100 were built at full speed, which is something considering I was using Luxemburg. The ability to use build costs with at least one decimal place would have such enormous value to modders that I am considering pestering the designers to add it to 1.06. It would be an easy change.

2: Unit supply and oil costs can be set to 0, and they will consume 0. However, armour units seem to be hard coded that oil is required. The little tank picture on the formation display did not have the blue oil drop on it, but the corp it formed did. And when there was no oil, the unit itself was not considered out of oil, but the corp it was alone in was. Whether this would affect movement rates I haven't tested. :eek:o This might make units able to move and survive in low-infrastructure provinces, like rangers, or beduins, or tough local militia.

3: Unit fuel and supply costs can be set to negative values. I'm not sure how this would happen, but the string was hangin' out, so I yanked on it. Negative supply units are read as 0, both in screen display, and in actual fact. When Luxemburg's only unit was a tank that needed -10 supply a day, supply requirement stayed at 0, and no supply accumulated. I don't know if 0 supply requirements would make a unit immune to being cut out of supply, or able to survive in low-infrastucture provinces.
Interestingly, units that require negative fuel, seem to break the conversion cycle. The oil does not appear anywhere in the ledger, but it is converted to rubber. Unfortunately for Luxemburg, they needed to convert 20 oil a day to have enough rubber, and the tank only created 5 oil. The game-engine converted the oil from the tank, but did not convert any more than that, so the economy fell to 2 IC/day, while great quantities of coal accumulated in the stockpile.:eek:o D'oh! I didn't check what would have happened if there already had been enough rubber. Perhaps the created oil might have accumulated, either in the province the unit was in, or in the national stockpile.

4: Decimal combat values - not yet tested I've read in a few places that non-integer values are possible. These would be very handy for tech modders. There are far too many applications on the trees that don't do anything other than enable another application, which chances are, won't do anything either. Everything should do something. Adding 0.25 to Hard_Attack might not seem like much, but it could make a very big difference to a unit, both because other techs would be adding fractions too, and because the game-engine, after calculating efficiency, multiplies the relevant value by that, and rounds down. For example 1 HA at 95% efficiency = 0, but 1.25 HA at 95% efficiency = 1. The first unit is doomed, even if a stack of 24 attacked by 1 lousy MG tank. The second would win.
I'll test this soon, unless someone beats me to it.
 

unmerged(12680)

I'm Lazy.
Dec 12, 2002
2.818
1
Visit site
Sounds very good! Let me comment upon some specifics...

1. Agree 100%, it wouldn't be a computer-resource hog, and would open up several more possibilities for modders. I'm putting my trust in Paradox's tradition of good modding support.

2. Sounds resonable, except for that last bit, even if units consumed 0 supplies they would still die during a Saharan Death March, wouldn't they?

3. No supply-creating units then, I'm a bit sorry as it would have been a rather cool option, but I'm not devastated. The fuel thing sounds odd and screams for further investigation.

4. As with #1, it should be there for modders, and I can't imagine it being all to hard to add, really...

Once again, outstanding work!
 

unmerged(19545)

Second Lieutenant
Sep 11, 2003
185
0
Visit site
Then test!

Hi Anti-Strunt.

You too can join the whacky team of testers. If the odd negative fuel thing cries out for further testing, then please test it. Seriously. And post it in this thread. And anything else weird you can think of. Or if you lack the skill to test something, mention it here, and someone (hopefully) with a good grasp of the scientific method will work it over.

Just because you can't think of a use for it, doesn't mean some brilliant modder like Mithel, or Soapy Frog, or one of the Core Geniuses (you know who you are) won't use it to back-door or workaround some fiddly little issue.

And when we get a good database of oddities, we or I will summarize it, and post it to the FAQ forum, where it will live in infamy forever.;)
 

unmerged(15893)

Forest of Metaphors
Mar 31, 2003
1.649
0
DC, you bring up some good points, and i've been following the work of Mithel/MathGuy/Yourself for quite some time.

As for a lead modder on the MDS team, I'd like to see the ability to use non-integral values added, as it may be a help for weapons like the RPG and stuff which do only a small amount of damage.

Thanks guys, and keep up the great work.
 

unmerged(19545)

Second Lieutenant
Sep 11, 2003
185
0
Visit site
Hi Marty Wolf

I should repeat: I believe it IS possible to use decimal combat values, but not decimal costs. I've read it several times in the HSP thread that you can do it. However, it would be very useful if someone(you for example;)) would actually test this. Anyway, thanks for the interest.

Has anyone else out there tested weird things, just to see how they worked, or confirm everyone's assumptions?

Remember a few months back, when Math Guy et al were blithely assuming that Ground Defense had a small chance of blocking a hit, and so if you had enough, you would be immune to damage? All the debate was about it was 66%, or 10% or somewhere in between, and it turned out the basic assumption was wrong.

If you have tested anything, even something banal, tell the world. Don't be shy. I led a cavalry charge on a dachsund for god's sake. :rolleyes:
 

Gwalcmai

©
8 Badges
Mar 14, 2003
5.341
22
Visit site
  • Crusader Kings II
  • Europa Universalis III Complete
  • Europa Universalis IV
  • Europa Universalis III Complete
  • Europa Universalis III Complete
  • Victoria 2
  • 500k Club
  • Pride of Nations
What does GD do, then? Sorry for leaving the thread subject, but I must have missed that discussion.
 

Mithel

General
8 Badges
Nov 29, 2002
1.725
0
www.mnstarfire.com
  • Hearts of Iron III
  • 500k Club
  • Hearts of Iron IV: Cadet
  • Hearts of Iron IV: Together for Victory
  • Hearts of Iron IV: Death or Dishonor
  • Hearts of Iron IV: Expansion Pass
  • Hearts of Iron IV: Expansion Pass
  • Hearts of Iron IV: La Resistance
I think Mathguy & Mikel are the experts on GD, but since Mathguy's busy and not around much, I'll make a comment on how I've observed it to work.

Basically one point of GD will possibly stop one attack. The odds of it doing this is dependent upon the ground defense efficiency (which I believe to be 80% initially). Extra GD points do nothing!

So say you have an attack of 10 points against 15 GD points. Each of the 10 points has only a 20% chance to actually do damage. Thus on average those 10 attack points do 2 attacks per hour. Raise this up to 15 attacking points against the same 15 defense and you now have an average of 3 attacks per hour hitting. A 50% higher casualty rate!

Now take two 10 attack divisions hitting a 15 GD division and you've got 15 attacks being blocked for an effective 3 attacks per hour plus five more attacks (unblocked) for a total of 8 attacks per hour hitting. By overwhelming the enemy like this you will quickly destroy them.

Thus what tripped a few of us up was thinking that we could increase attack strengths and compensate by increasing GD values too. Wrong! That leads to very accelerated (bloody) battles. Mathguy guy mentioned this is a reason the later war period combat becomes less realistic.

The trick or fix is that while you increase your attack values you must ALSO increase ground defense efficiency.

- Mithel
 

unmerged(14603)

Captain
Feb 10, 2003
464
0
Visit site
Mithel your explanation is right on target. Other things we concluded were:

1. Changes to the ground_def_eff variable affect your opponent's defense not your own. So changing Germany's gound_def_eff by +10 actually increases it's opponents defense.

2. The ground_def_eff variable can not exceed 99% nor be lower than 1%.

3. Strength loss averages .15 per successful attack (except when a hard unit attacks a soft unit - then the value is .1125). This number is not dependent on the strength level of the attacking unit. This means a 10% strength unit hits just as hard as a 100% strength unit.

4. Mechanized units always defend as hard targets - they do not choose the better of being a hard or soft target as stated in the manual.
 

jdrou

Field Marshal
74 Badges
Jun 10, 2002
24.161
461
  • Stellaris: Federations
  • Europa Universalis IV
  • Europa Universalis IV: Common Sense
  • Europa Universalis IV: Rights of Man
  • Hearts of Iron IV: Cadet
  • Stellaris: Galaxy Edition
  • Europa Universalis IV: Mare Nostrum
  • Europa Universalis IV: Cossacks
  • Paradox Order
  • Stellaris: Digital Anniversary Edition
  • Humble Paradox Bundle
  • Europa Universalis IV: Pre-order
  • Europa Universalis IV: El Dorado
  • Europa Universalis III: Collection
  • Europa Universalis III: Chronicles
  • 500k Club
  • 200k Club
  • Europa Universalis IV: Dharma Pre-order
  • Stellaris: Lithoids
  • Stellaris: Ancient Relics
  • Imperator: Rome Deluxe Edition
  • Europa Universalis IV: Golden Century
  • Stellaris: Megacorp
  • Europa Universalis IV: Dharma
  • Stellaris: Distant Stars
  • Stellaris: Apocalypse
  • Stellaris: Humanoids Species Pack
  • Europa Universalis IV: Cradle of Civilization
  • Stellaris: Synthetic Dawn
  • Europa Universalis IV: Third Rome
  • Europa Universalis IV: Mandate of Heaven
  • Stellaris - Path to Destruction bundle
  • Stellaris: Leviathans Story Pack
  • Deus Vult
  • Darkest Hour
  • Crusader Kings II
  • Hearts of Iron II: Armageddon
  • Arsenal of Democracy
  • Hearts of Iron Anthology
  • Victoria 2
  • Rome Gold
  • Victoria: Revolutions
  • Europa Universalis IV: Res Publica
  • Hearts of Iron III
  • For The Glory
  • Europa Universalis IV: Call to arms event
  • Europa Universalis IV: Wealth of Nations
  • Europa Universalis IV: Conquest of Paradise
  • Europa Universalis IV: Art of War
  • Stellaris: Galaxy Edition
Originally posted by Dog Cavalry
1: Unit build costs can only be integers, and the engine rounds down, not up.
. . .
The ability to use build costs with at least one decimal place would have such enormous value to modders that I am considering pestering the designers to add it to 1.06. It would be an easy change.
Actually build costs can be fractional until they are actually in the build queue. If you have a unit with a base cost of 2.9 and a minister that gives a 10% discount the cost in the queue will be 2 since the modifier is applied before truncating. It would be useful to have the build queue hold decimal values though.
 

unmerged(19545)

Second Lieutenant
Sep 11, 2003
185
0
Visit site
EMPIRICALLY DETERMINED:

Reinforcing takes 10% of build cost, in supplies, prorated to state of damage, for infantry types; 30% for armor types, 20% for all others. With supplies costing 3 IC’s each, full recovery is 30% of cost for infantry, up to 90% of cost for armor. This seems easily rationalized, considering that until an infantry division collapses, it’s just infantry men being killed, and they are the cheapest part of the division. Much higher armor costs reflect the fact that it is tanks being killed from the start, and they are the most costly part of the formation. Actually, I like these values. When Math Guy’s mod is implemented, this will result in reinforcement costs of:
34.5% for infantry types, 69% for most, and 103.5% for armor units. Best Supply efficiency improvements bring this down to 32.3%, 64.5%, and 96.8%. Very satisfying.

I wonder though jdrou, if the game's calculation of unit cost for reinforcing purposes take in to account these decimal values, even if they aren't factored in to actual construction. Do you have time to test for this?
I think for many units, particularly Soviet or German ones, that reinforcement cost might become more important than initial build costs, and therefore it might be worth modding the techs to include decimal cost increases for improved equipment.

Actually it certainly would. Enough different improvements, either in electronic, artillery, armor or infantry techs would add up to another integer, and the high performance teched up divisions would cost more to build, reinforce, and supply, than a no-tech division.
 
Last edited:

jdrou

Field Marshal
74 Badges
Jun 10, 2002
24.161
461
  • Stellaris: Federations
  • Europa Universalis IV
  • Europa Universalis IV: Common Sense
  • Europa Universalis IV: Rights of Man
  • Hearts of Iron IV: Cadet
  • Stellaris: Galaxy Edition
  • Europa Universalis IV: Mare Nostrum
  • Europa Universalis IV: Cossacks
  • Paradox Order
  • Stellaris: Digital Anniversary Edition
  • Humble Paradox Bundle
  • Europa Universalis IV: Pre-order
  • Europa Universalis IV: El Dorado
  • Europa Universalis III: Collection
  • Europa Universalis III: Chronicles
  • 500k Club
  • 200k Club
  • Europa Universalis IV: Dharma Pre-order
  • Stellaris: Lithoids
  • Stellaris: Ancient Relics
  • Imperator: Rome Deluxe Edition
  • Europa Universalis IV: Golden Century
  • Stellaris: Megacorp
  • Europa Universalis IV: Dharma
  • Stellaris: Distant Stars
  • Stellaris: Apocalypse
  • Stellaris: Humanoids Species Pack
  • Europa Universalis IV: Cradle of Civilization
  • Stellaris: Synthetic Dawn
  • Europa Universalis IV: Third Rome
  • Europa Universalis IV: Mandate of Heaven
  • Stellaris - Path to Destruction bundle
  • Stellaris: Leviathans Story Pack
  • Deus Vult
  • Darkest Hour
  • Crusader Kings II
  • Hearts of Iron II: Armageddon
  • Arsenal of Democracy
  • Hearts of Iron Anthology
  • Victoria 2
  • Rome Gold
  • Victoria: Revolutions
  • Europa Universalis IV: Res Publica
  • Hearts of Iron III
  • For The Glory
  • Europa Universalis IV: Call to arms event
  • Europa Universalis IV: Wealth of Nations
  • Europa Universalis IV: Conquest of Paradise
  • Europa Universalis IV: Art of War
  • Stellaris: Galaxy Edition
Re: EMPIRICALLY DETERMINED:

Originally posted by Dog Cavalry
I wonder though jdrou, if the game's calculation of unit cost for reinforcing purposes take in to account these decimal values, even if they aren't factored in to actual construction. I think for many units, particularly Soviet or German ones, that reinforcement cost might become more important the initial build cost
Should be tested but I would expect that the reinforce cost would use the decimal value since it seems to only be truncated because the build queue only holds integers.
 

unmerged(19545)

Second Lieutenant
Sep 11, 2003
185
0
Visit site
Most Excellent!!

Okay, some more obscure testing here. I have a game well advanced, with many damaged units. 1st Corp under FM Crerar is a little beat up, and will require 68 supply to reinforce. So I exited the game, opened the db file, and changed the infantry.txt file so that infantry costs 6.9 ic for 95 days, rather than 6 ic for 95 days, then re-started the game. Now 1st Corp requires 73.6 supply. 1st Corp consists of 2 inf-A divisions, and 11.9/11 * 68 is indeed 73.6. Reinforcement does check the cost to build after the decimal place, even if the build queue for some reason doesn't.

I think this is great. I hope to do a DC mod, where every application does something, even if only raising HA by 0.25, and has a tiny increased build cost, on the order of 0.1. This demonstrates that to rebuild a division with lots of expensive equipment can be made to cost more than rebuilding one with nothing. The game engine will factor in the additional costs.

This is more reason not to blunt the effect of the tech trees. If every improvement also carries an increase in reinforcement cost, then allowing players to choose between quality and quantity of divisions is more viable than ever.

The DC mod would include tiny supply cost increases for ammo greedy weapons like big arty, and most particulary rocket launchers. It will also include increases to build times for most doctrines that increase org or any other value. After all, how long does it take to train a fellow to 40 org, and how much longer would it take to train the same guy to 100 org? It's not the same training. If it was, conscript militia would be exactly as effective, man for man, as the SAS, or US Navy Seals. Clearly they aren't. Nor are the training times the same.

If you know, please post some representative training times for different military units.
 

Mithel

General
8 Badges
Nov 29, 2002
1.725
0
www.mnstarfire.com
  • Hearts of Iron III
  • 500k Club
  • Hearts of Iron IV: Cadet
  • Hearts of Iron IV: Together for Victory
  • Hearts of Iron IV: Death or Dishonor
  • Hearts of Iron IV: Expansion Pass
  • Hearts of Iron IV: Expansion Pass
  • Hearts of Iron IV: La Resistance
DC, excellent discovery!

Per Siegfried Knappe's biography German infantrymen (pre-war) received twelve months of infantry training (while artillery men only received six weeks of infantry training). Siegfried Knappe was in the artillery and it looks like he had about seven months of basic training before starting officer training (roughly one year of officer training).

From what I've heard the training times were dramatically different from country to country.
 

unmerged(19545)

Second Lieutenant
Sep 11, 2003
185
0
Visit site
Kudos from Mithel!

Yeah, I did the happy dance when the math confirmed this. I remember there used to be a guy named Math Guy who was interested in this sort of thing. Oh well, we carry torch now. I just hope we hear from him some day. If he doesn't post soon, I'm going to phone him, just to make sure he's okay.

Thanks too for the training times. I bet later on in the war, German infantrymen didn't get any year of training. I bet their division's org fell, rather than rising too.

Perkele - no, org isn't number of days training, it's troop skill and quality, which comes almost entirely from good training and good leadership. Leaders cover the leadership, and unit build time is the only way to model training time. If not build time, then what? Why should Germany be able to build a teched-out division with 100 org, in the same time and for the same cost, as Ethiopia builds a basic inf div, with default attack, defense, and org? The answer is obviously, it shouldn't.
 

Mithel

General
8 Badges
Nov 29, 2002
1.725
0
www.mnstarfire.com
  • Hearts of Iron III
  • 500k Club
  • Hearts of Iron IV: Cadet
  • Hearts of Iron IV: Together for Victory
  • Hearts of Iron IV: Death or Dishonor
  • Hearts of Iron IV: Expansion Pass
  • Hearts of Iron IV: Expansion Pass
  • Hearts of Iron IV: La Resistance
DC, yes I think it would be reasonable for German divisions org to fall later in the war as the average soldier received less training once the pressures of war were upon them.

However I suspect the very experienced and well trained German officers are largely what kept the German divisions from falling apart at the end of the war. They did fight very effectively until the very end.

I think of org as not just training but also moral. In the Soviet case, the org rose as the Germans approached Leningrad, Moscow and Stalingrad probably mostly from patriotic desparation.

I was just reading that Stalin was trying to negotiate a surrender even after they won at Stalingrad. (The Russian's were shocked at how quickly and effectively the Germans counter attacked after suffering a major defeat)

If the Germans hadn't been so foolish and actually worked out an agreement with the Russians to surrender then they probably would have won WWII and the world would be a very different place today!

Opps... I've drifted off target. Anyway, DC, I intend to use your new information in the Starfire Historical mod to help increase the difference in cost between building vs reinforcement vs supplying divisions.

Hopefully Mathguy will return eventually. The last few e-mails I exchanged with him he sounded rather busy.

Great work! Thanks!

- Mithel
 

unmerged(19545)

Second Lieutenant
Sep 11, 2003
185
0
Visit site
How does Org loss work during combat?

Lots of talk in other threads have once again pointed out that we really don't know what is going on here. When we were all talking about cost of reinforcement, naturally Strength loss seemed most significant.
What we overlooked is that for units in combat, they will keep fighting, and keep losing strength, until they hit 0 org and lose.
It seems likely that Org loss happens during hits in combat. Certainly I don't recall units losing org against enemies that couldn't hit. Therefore, Org loss must, like strength loss, be distributed in some range of possible range of values, and averaged, be some ratio of strength loss. Or maybe not -- we don't really know.

Or maybe someone out there does ? If so, please throw us a bone.

I've done some very preliminary testing, just tracking Str and Org loss for an infantry div getting the boots put to it by other inf. Str/Org

2/1 2/3 1/0 4/5 1/1 0/1 2/2 8/8 3/4 2/2 1/1 1/1 2/1 2/2 4/6 4/5 5/6 5/5 5/6 9/11 Totals: 63/70.

I don't yet know if it makes a difference what types of units are int the combats, but I do notice that the ratio of org to str loss seems pretty constant.

Over to you, world.
 

unmerged(14603)

Captain
Feb 10, 2003
464
0
Visit site
Unfortunately when we were testing the combat system last summer we did not finish the effect on org (IIRC we entered a plea of mental exhaustion).

The best formula (which is far from exact) that I can give you is:

Org Loss = (Hits - 3) / 5. Unlike strength, Org tends to be lost in whole numbers. There is a rounding or grouping impact that I was never able to figure out.

Hope this helps.
 

Soapy Frog

Lt. General
60 Badges
May 1, 2001
1.324
1
Visit site
  • Stellaris: Leviathans Story Pack
  • War of the Roses
  • 500k Club
  • Cities: Skylines
  • Cities: Skylines Deluxe Edition
  • Europa Universalis IV: El Dorado
  • Hearts of Iron II: Beta
  • Pride of Nations
  • Europa Universalis IV: Common Sense
  • Stellaris
  • Hearts of Iron IV Sign-up
  • Hearts of Iron IV: Cadet
  • Stellaris: Digital Anniversary Edition
  • Victoria 2: Heart of Darkness
  • Hearts of Iron IV: Together for Victory
  • Stellaris - Path to Destruction bundle
  • Steel Division: Normandy 44
  • BATTLETECH
  • Stellaris: Synthetic Dawn
  • Hearts of Iron IV: Expansion Pass
  • Stellaris: Humanoids Species Pack
  • Stellaris: Apocalypse
  • BATTLETECH - Digital Deluxe Edition
  • Cities: Skylines - Parklife Pre-Order
  • Cities: Skylines - Parklife
  • Stellaris: Necroids
  • Lead and Gold
  • Europa Universalis III
  • Europa Universalis III Complete
  • Divine Wind
  • Europa Universalis IV
  • Europa Universalis IV: Art of War
  • Europa Universalis IV: Conquest of Paradise
  • Europa Universalis IV: Wealth of Nations
  • For the Motherland
  • Hearts of Iron III
  • Hearts of Iron III: Their Finest Hour
  • Heir to the Throne
  • Europa Universalis III Complete
  • Crusader Kings II
  • March of the Eagles
  • Europa Universalis III Complete
  • Naval War: Arctic Circle
  • Europa Universalis IV: Res Publica
  • Victoria: Revolutions
  • Semper Fi
  • Sengoku
  • Sword of the Stars
  • Sword of the Stars II
  • Victoria 2
Well has anyone seen this little gem?:

manual_01.gif


Would seem to indicate that there is not a constant relationship between strength and org loss per hit, but rather a variable one.

In short, in an infantry vs infantry assault, each hit will inflict 1-3 points of strength and 1-3 points of org.