Home Trending Latest threads New posts Developer diaries

Land Efficiency FAQ - Demo version

Jump to latest Follow Reply
  • 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.
anoldman said:
Felix,
I assume you have updated this from the actual game now?

This was a REALLY helpful post for me, I hope it is still accurate. ;)

Thank you! :)
--An Old Man
Click to expand...

Its Fiendix ;).

At first glance it seems correct, but I havent had time to look everything.

F
 
Last edited:
I agree that the defense is too much favored in HOI2.

for example, in the Desert Fox scenario as the Germans, I destroyed most of the British army, but did not support Syria, so I had no units across the Suez Canal. When I reached it, although the British had no air force, I was unable to even make a dent in their defense, because there is only one province to attack from. all the British have to do is stack up on the other side of the canal, and no amount of bombing or attacking with any number of units can hurt them.

Not only did my attacks fail to cross, but they failed to diminish the org or defense strength of the British, although I was bombing them with 3 different stacks and attacking with every unit I had.

I had a similar but less critical problem with the Germans against the soviets (see my AAR in the AAR section) when I found myself in a two-front war with France and the soviets in 1939. The Soviets kept moving in units from adjacent provinces when I attacked, and my Panzers were nable to make any hole in the soviet line despite superior Panzers (Pz3 vs BT5) and concentration of force with Rommel and Manstein bonuses.

With human players, in practice, when one attacks a province, he will in effect be fighting against the enemy units in 3 provinces, assuming that there are no reserves behind the province - in which case he will find himself fighting them too!

This means that breakthroughs will be possible only when one has a big advantage in numbers or when the enemy is overstretched.

Henri
 
To calculate the efficiency we always start from the base of 100. We then multiply by
all the modifiers if the following way:
In the above example we have (displayed as 85% attacker and 113% defender)
100% base attack efficiency:

*1.1 - offensive leader
*1.0005 - exp
*1.04 - ESE
*0.9 - forest
*0.75 - frozen
*1.1 - leadership

= 84.9844 %

100% base defence efficiency
*1.1 - offensive leader
*1.0005 - exp
*1.04 - ESE
*0.9 - forest
*1.1 - leadership

= 113.82 %

Although your numbers do calculate correctly as you offered your formula.... they are incorrect and the compluter is calculating them incorrectly.

Let me explain using your sample data.
*1.1 - offensive leader
*1.0005 - exp
*1.04 - ESE
*0.9 - forest
*0.75 - frozen
*1.1 - leadership

In this case, offensive leader and forest have directly opposite effects (+ and - 10%). This being so they cancel each other...

now we are left with the following:
+0.05% - exp
+4% - ESE
-25% - frozen
+10% - leadership

Now merely adding/sutracting these numbers should provide the total modifier to the original 100 base. If we do this (ignoring the 0.05 number as it is so small it is virtually irrelevent) we get +14% and -25%.... leaving us with -11%. Why then does the cpu and your calculation display 85%?


Here is why. You cannot calculate these using multiplication. Easy example to prove it. Lets say 2 modifers of +10% and -10%. Now order there are calced is irrelevent (or better be because if one is more important I would sure like to know why). A 10% gain and a 10% loss calced = a net gain of zero. However using the formula you used above we get this:

100 *1.1 * 0.9 = 99% ALSO 100 *0.9 *1.1 = 99%

order is irrelevent. Your formula is wrong because you turned a % into a factor which is based on the original base number.

The correct formula should be:
100 + (10-10)

or in the above example:
100 + (+10-0.05+4-10-25+10)


These formulas should be corrected if they have not already been by Paradox. As it is your thread I will let others post contradictory data to mine to dispute it should I be incorrect, but let you let them know of the mistake.

Cop115
aka - The DR.
 
cop115 said:
Here is why. You cannot calculate these using multiplication. Easy example to prove it. Lets say 2 modifers of +10% and -10%. Now order there are calced is irrelevent (or better be because if one is more important I would sure like to know why). A 10% gain and a 10% loss calced = a net gain of zero. However using the formula you used above we get this:

100 *1.1 * 0.9 = 99% ALSO 100 *0.9 *1.1 = 99%

order is irrelevent. Your formula is wrong because you turned a % into a factor which is based on the original base number.

The correct formula should be:
100 + (10-10)

or in the above example:
100 + (+10-0.05+4-10-25+10)
Click to expand...

Hmmm... yes and no. I don't think it's wrong, exactly, it just represents a different interpretation of bonuses and maluses, with the effect of making the maluses a little larger than they might appear.

The original post did point out the big significant change of going to multiplicative modifiers rather than additive modifiers. The effect you noticed is just a consequence of that change.

I'll offer a counter-suggestion. At the heart of the issue is that percentage maluses are a tad ambiguous. When we say "a 10% increase", we unambiguously mean "multiply by 1.1". However, when we say a 10% decrease, there can be some question as to whether we mean "multiply by 0.9" or "divide by 1.1". When a store posts a sign that says 10% Off, they mean the former. E.g. sales taxes in Ontario, Canada sum to 15%, but I might observe that if a product is marked down by 15% the total cost ends up being less than the original price -- it's exactly the same phenomenon.

But if you want a 10% decrease to be the inverse operation of a 10% increase, then the answer is the make the decrease a divisive operation.

So your example now gets:
100 x 1.1 x 1.0005 x 1.04 / 1.1 / 1.25 x 1.1 = 91.6%

-----

Let's compare the three systems. Consider a situation with 2 10% bonuses and 1 15% malus.

First, the additive (HoI1) system, which you seem to be advocating a return to:
Eff = 100 + 10 + 10 - 15 = 105%

Second, the existing (HoI2) system:
Eff = 100 x 1.1 x 1.1 x 0.85 = 102.8%

Third, my proposal:
Eff = 100 x 1.1 x 1.1 / 1.15 = 105.2%


The advantages of my model is that you get the nice effect of modifiers being cumulative, and it also makes bonuses and maluses symmetric.
 
Pro (et all)

They can use whatever they want :) It's their game. Only don't say i get a 10% increase of I only get 8.9%, that's all I'm saying. Don't see what the problem was of having it non-cumalitive.
The current formula is just wrong if all the bonuses are to be applied equally. Should there be a weighting to these? What does this number mean because if it is not a sum total. Asolutely nada. Why not just increase the % for a certain variable if your loking for a weight. This would allow everyone to see what is actually being represented in the numbers, not just APPROXIMATE.
The correct formula of 100 + 1(1+.05) - 1(1-.06) weights all factors equally.
The existing formula of 10 + 10 -15 = 2.8 just seems kinda silly to me.
 
wanted to follow up my last post with a quick one...

The current formula makes a +10% and a -10% modifier = a -1% penalty for the player. Why is this good? How can this possibley mean ANYTHING?
 
Yea I noticed it when I origianally calculated this. I used the forumula you suggested at first, but since I (to my suprise) came up with different answers I had to go into the more "illogical" formula. I presented the way to get the displayed eff corectly.

I dont know why P did it in the way they did. However with P one never knows and I doubt they would want to fiddle with this even if the whole world would tell them its a BUG, so they probably will claim it to be a WAD. I suggest you post in the enchancement forum to see if they want to change it. I havent bothered since there is no bug forum in which this IMO belongs.

F
 
Last edited:
Sent it in as a bug and offered the solution. Anyway just tryin to help. Great post full of usefull information and appreciate you posting it.

Cop115
aka - the DR.

edited for relevence
 
Last edited:
cop115 said:
Sent it in as a bug and offered the solution. I guess the important thing to note is that you said that Leader and XP are the most important. Everyone should make note of that if that is indeed true (did not test at all personally) as they would be more significant than terrain modifiers. etc. Anyway just tryin to help. Great post full of usefull information and appreciate you posting it.

Cop115
aka - the DR.
Click to expand...

scrap that - too little coffee in the morning.. just make sure you have as little negative modifiers as possible...

If you see anything else that you think is wrong just let me know I dont mind. I definately dont have everything figured out. The rates of exp is something that really gets me.

F

PS i edited that crap so nobody understands it wrong
 
Inbrainsane said:
The only Advantage of BIG defending stacks would be at a bottleneck. But where are such good bottlenecks? I dont think, there are any at all.
Click to expand...

i am in one right now. i'm turkey in axis, and the british are in iraq directly across from me. both persia and vichy france are neutral, so the border is only one province wide. no one has dared to attack in 2 years. i'm vastly outnumbered (35 to 100), but no british attack. i guess it's working in my favour
 
cop115 said:
The current formula makes a +10% and a -10% modifier = a -1% penalty for the player. Why is this good? How can this possibley mean ANYTHING?
Click to expand...

I am 99% (100% x 1.1 x 0.9 :p) sure it is WAD and not a bug, and Paradox if they can be bothered will tell you so.

I would weigh in on the opposite side, though. I really like that modifiers are now cumulative, so when you start piling on bonuses your efficiency goes through the roof. It always kind of bothered me in previous paradox games that modifiers were labelled as percentages but were really additive increases.

So I would rather keep the current system than go back to the old one, but my suggestion of balancing penalties with bonuses stands: you just have to make penalties a division operation. Then your +10%, -10% complaint will resolve itself.
 
Fiendix said:
Also if anybody has the time can you please check this:

Combined Arms
when units number <= CPC(not double),
and (less than 100% and 30% or more) is Hard Units(arm mech...)...
attacker: Eff+5%
defender: Eff+15%

F
Click to expand...
Checked in demo. Works out just fine.
All units not Pz or mech even HQ will make the CA come into effect. Most effective just after an attack, when the enemy counterattacks you.
But i'm a bit puzzled. My instict was that a combo of Mech and armour should give a CA bonus. Mech was introduced to keep up with the Pz DIVs. A mech unit would mostly be a MOT with 1 regiment in halftracks, the 1 or 2 other regiments in trucks+ some xtra SP AT+ some more assorted vehicles. What use are Mech if they cant support tanks?
 
Nice on the 99% reference btw.

I don't want to beat a dead horse (wife, boss, coworker, etc.) here but I still think the application is wrong and will post one final summation to show why.

When we call something a modifier we mean it affects some base number. We do not label this a cumulitive % modifier nor do we call this "the mystery factor". As a modifier, and a displayed modifer to boot, these numbers represent % increases and % decreases for our unit's combat value(100). These +'s and -'s are done at the same time (they affect our unit at a specific hour and change hourly) and therefore SHOULD BE CALCED AT THE SAME TIME. Time is the relevent part of this equation becuase we are CALCULATING A SPECIFIC VALUE AT A SPECIFIC TIME.
Example:
A unit in the forest gets a penalty. Ok fine. But should the penalty for a unit BEING in the forest out of supply be different from the penalty for a unit just being in the forest. There is a specific penalty already for a unit being out of supply. This will be applied, but the current formula adds ANOTHER penalty other than the 2 listed above. When we do this we call it COMPOUNDING. We are adding a 3rd penalty in the above equation. The penalty is applied when the player has more than 1 penalty.

As I have shown previously the formula used currently is not appropriate. The only application in the real world I can think of that this formula could be used for is compounding (daily, annual, whatever).
Example: if you had $100 in the stock market... lost 10% the first year and gained 10% in the second year. You would not have $100 in this scenario.
This is exactly what the game is doing. But we don't have a reason to compound for an application of a modifier BECAUSE THE MODIFIER ITSELF REPRESENTS THE CHANGE TO THE ORIGINAL BASE NUMBER and is calculated at a specific time. Compounding these numbers is innapropriate becuase it assumes there is a relevance for other modifiers, when in fact they should all be applied independantly. Instead the MODIFIERS THEMSELVES should be adjustable if the programmers feel a weighting must be applied.

Example: as we have used this one many times I will continue.
100 +10 -10 =99%

instead we can offer this:
100 +9.3-10.3 = 99% WHERE 9.3 = x+y+z-g/9*4.5r9i239320 (etc)

Calculating modifiers with whatever formula seems appropiate is fine. But DISPLAYING them must be done as an additive/subractive value to the base number BECAUSE TIME/COMPOUNDING IS IRRELIVENT and does not truly display a modifier but a cumulative total (which is meaniless in this application. So if you want to make penalties/rewards display correctly (and be appropraite) you must use the formula above. Using the current formula we get a number displayed which is meaningless to the player.
 
I'm sure it is "WAD". This is a change (and a very important one) that we lobbied for from HoI 1. Having the penalties "additive" (with the values that Paradox provides) meant it was far too easy for things like weather, night and mountains to combine to make it "impossible" to attack (in fact disasterous). The new system is a lot less prone to disasterous mistakes.
 
Mithel said:
I'm sure it is "WAD". This is a change (and a very important one) that we lobbied for from HoI 1. Having the penalties "additive" (with the values that Paradox provides) meant it was far too easy for things like weather, night and mountains to combine to make it "impossible" to attack (in fact disasterous). The new system is a lot less prone to disasterous mistakes.
Click to expand...

Can you go into more detail as to how this fixes this issue? Couldn't this have been fixed by changing the modifiers to make them more realistic if they were giving unrealistic results?

I ask this because if this is the reason why then it doesn't appear to be working as INTENDED. In the example the original poster displayed the player is actually penalized an additional 4% above the stated values. I fail to see how this corrects a player having too many penalties making it impossible to attack.
 
Last edited:
This is all pretty subjective (which is why Paradox will tell you it's not a bug), but I'll just offer a bit of rationalization for the view I offered above.

cop115 said:
Example:
A unit in the forest gets a penalty. Ok fine. But should the penalty for a unit BEING in the forest out of supply be different from the penalty for a unit just being in the forest. There is a specific penalty already for a unit being out of supply. This will be applied, but the current formula adds ANOTHER penalty other than the 2 listed above. When we do this we call it COMPOUNDING. We are adding a 3rd penalty in the above equation. The penalty is applied when the player has more than 1 penalty.
Click to expand...

Compounding indeed. Offering nothing as justication other than my own feeling based on life experiences, I feel the penalty for being in a forest out of supply should be worse than the sum of the individual forest and out of supply penalties. It seems to me that, in life, when a lot of things go wrong at the same time, each thing going wrong tends to increase the severity of the other things.

Conversely, things going well does the same thing. There is definitely a notion that if you have multiple good things that can work together, the net effect is greater than the sum of the good things separately -- in business this idea is called synergy.

cop115 said:
Example: if you had $100 in the stock market... lost 10% the first year and gained 10% in the second year. You would not have $100 in this scenario.
Click to expand...

I prefer the simile I offered earlier. If I see a product in a store that costs $100, and it has a sign above it that says 15% off, and sales tax is 15%, when I go to the cash and pay for it, it will cost me $100 x 1.15 x 0.85 = $97.75.

It's not about compounding over time, it's just that the modifiers are not considered independent of other modifiers.

And I that is indeed what this argument comes down to: should modifiers be considered isolated factors, or should we consider there to be a relationship between them where their interaction can be considered a modifer in itself.

I vote for the latter, you for the former; I don't think we can get any further than that, but it doesn't matter since this isn't a democracy anyway :).
 
Mithel said:
I'm sure it is "WAD". This is a change (and a very important one) that we lobbied for from HoI 1. Having the penalties "additive" (with the values that Paradox provides) meant it was far too easy for things like weather, night and mountains to combine to make it "impossible" to attack (in fact disasterous). The new system is a lot less prone to disasterous mistakes.
Click to expand...

Uh, gotta side with Cop115 on this one.

Multiplying the penalties together actually makes them worse.


-------

Quick change of subject (but not really) ....

What is the penalty for tanks fighting in cities?
 
Proaxiom said:
I feel the penalty for being in a forest out of supply should be worse than the sum of the individual forest and out of supply penalties. It seems to me that, in life, when a lot of things go wrong at the same time, each thing going wrong tends to increase the severity of the other things.

Conversely, things going well does the same thing. There is definitely a notion that if you have multiple good things that can work together, the net effect is greater than the sum of the good things separately -- in business this idea is called synergy.



I prefer the simile I offered earlier. If I see a product in a store that costs $100, and it has a sign above it that says 15% off, and sales tax is 15%, when I go to the cash and pay for it, it will cost me $100 x 1.15 x 0.85 = $97.75.

It's not about compounding over time, it's just that the modifiers are not considered independent of other modifiers.

And I that is indeed what this argument comes down to: should modifiers be considered isolated factors, or should we consider there to be a relationship between them where their interaction can be considered a modifer in itself.

I vote for the latter, you for the former; I don't think we can get any further than that, but it doesn't matter since this isn't a democracy anyway :).
Click to expand...

Just for reference the application of the seller paying the additional sales tax must be a Canadian thing... so I can't really relate as in the US our sellers do indeed just take off 15% and then apply sales tax. Just FYI :)

As for your reasoning listed above you offer a fair argument as to why it should be compounded. Stated in such a way it could be viewed that way. Could we not then at least call it compounded modifier.
Your last statement at least opens my eyes to the "Why?". Don't know if that really solves the problem as the poster listed above, but at least defines the logic behind using it. I suppose using this logic if you found that the results still gave you "disastrous results" you could modify the values while keeping the logic of "each effects the other".

In summary. I see why now at least, as posted by Pro, it does at least appear a reasonable reason for using a compoun formula. Anyway, thanks pro for finally pullin up my eyelids. Was missing the part that they indeed could be viewed as compounding. Had not considered it and now that I do I guess I can see why some would think this way.
 
You must log in or register to reply here.
Top Bottom
Back