• Blitz Shadow Player
  • Caius
  • redboot
  • Rules
  • Chain of Command
  • Members
  • Supported Ladders & Games
  • Downloads


Replacement question
12-24-2021, 06:10 AM,
#11
RE: Replacement question
(12-24-2021, 04:46 AM)ComradeP Wrote: True, for recovery the difference is not particularly significant, but recovery tends to be the icing on the cake unless units are very small or local supply is very poor.

Let's use PzC for the example, as (updated) FWWC titles use fractional numbers in some OOB's that make it difficult to compare conditions between titles. Unlike PzC, Recovery might also be 0 (to prohibit loss recovery for artillery units).

Some PzC titles, such as Budapest '45, feature small battalions compared to other titles. In most cases, infantry battalions will have a strength of 400-600 men or so depending on whether the heavy weapon company is abstracted and added to the regular infantry companies a battalion consists of.

Let's assume the scenario is long enough for results to move towards the average result, removing variability from the examples.

First example: a battalion with a full strength size of 400 Men, and a strength of 300 Men in this situation. A loss of 100 Men compared to full strength.

Recovery 2% at C quality for 2% results in a recovery of 2 Men: 100x0.02=2.


Replacements at 1% (no OOB modification) result in 4 Men being replaced at 50 local supply or higher: 400 x 0.01=4 Men.

The local supply value would need to be 34 or lower in order to be below Men regained by Recovery in this example.
Local supply modifier: (34-20) / 30= 0.4666.
400 x 0.01=4.
4 x 0.04666=1.8664 Men

Bigger battalions regain more strength through Replacements than smaller units regardless of losses suffered, whereas all units of the same quality benefit equally from Recovery. That's also why cutting maximum strength in a Moscow '42 update and decreasing Axis replacements in Kharkov '43 dramatically reduced the Wehrmacht quickly ballooning in size like in the stock versions, for example.

Second example: four non-combinable 20 vehicle at 100% strength Soviet tank companies/"battalions" compared to four combinable 20 vehicle at 100% strength Panzer companies.

Let's assume all companies have 15 runners and let's assume all units are C quality.

The game turns all guns/vehicles into "Men" to determine percentage chances. 1 Gun/Vehicle=10 Men.

Recovery 2%: each company/"battalion" has a 10% chance to recover a Vehicle each turn:
5 Vehicles= 50 Men.
50 x 0.02 = 1 Man or 0.1 Vehicle.

Replacements are also identical for both sides, with a 1% Replacement rate with no local supply modification resulting in a 20% chance to recover a vehicle:
20 Vehicles=200 Men.
200 x 0.01 = 2 Men or 0.2 Vehicle.

In this case, the result is the same, but only because all units are below maximum strength and other conditions are equal.

Third example: Now let's look at a situation where two of the Soviet tank companies/"battalions" have lost 10 tanks, and the others have lost none.

Recovery 2% gives a 20% chance to recover a Vehicle:
10 Vehicles = 100 Men
100 x 0.02 = 2 Men or 0.2 Vehicle

Replacements are as above, with a 20% chance to recover a Vehicle.

Combining and breaking down the German units results in a loss of 5 Vehicles for each company. The results are the same as in the previous example.

The Vehicles regained from Recovery will decrease with each Vehicle that is recovered, whereas the Vehicles recovered through Replacements remain stable as long as the local supply value doesn't change to a value of 49 or lower.

If the two Soviet units that have lost Vehicles move, they can't recover Vehicles. If three out of four German companies move, the fourth can still recover a Vehicle.

The ability to distribute losses between component units and being able to give some units rest whilst others fight on gives a significant advantage over time.

ComradeP,

I may be misunderstanding what you are saying, but replacements are not applicable unless there is an OOB modification. If there isn't the PDT value is irrelevant. And such OOB modifications are extremely rare in PzC, so it is not normally a consideration.

Also, recovery does not convert tanks to men as there is no need. The losses are in tanks and the recovered units are in tanks. The recovery rate is the same but the number of tank losses tend to be low compared to men so the number recovered is also comparatively low. But losses of only 3 men would take the same time to recover as losses of only 3 tanks, on average. So if the recovery rate is 2%, then it is 2% for everything. In your vehicle calculations you have multiplied by 10 and then divided by 10, which produces the right result but may cause confusion.

John
Quote this message in a reply
12-24-2021, 06:51 AM, (This post was last modified: 12-24-2021, 06:56 AM by ComradeP.)
#12
RE: Replacement question
I meant: no OOB modification for the base value of 1 in the parameter data, so the OOB value is also 1.


Quote:Also, recovery does not convert tanks to men as there is no need. The losses are in tanks and the recovered units are in tanks. The recovery rate is the same but the number of tank losses tend to be low compared to men so the number recovered is also comparatively low. But losses of only 3 men would take the same time to recover as losses of only 3 tanks, on average. So if the recovery rate is 2%, then it is 2% for everything. In your vehicle calculations you have multiplied by 10 and then divided by 10, which produces the right result but may cause confusion.


For other calculations in the game, such as stacking limit and loss calculation, strength in Guns/Vehicles is given as 1 Gun/Vehicle=10 Men. I assumed the same would apply for recovery/replacements, in the sense that the reverse situation of loss calculation would apply. As you say, mathematically the end result is the same.

If recovery/replacements don't turn Guns/Vehicles into Men for calculation purposes, that might be the beginning of an explanation to some of the results I saw when I ran a few tests before typing this. I observed results which, mathematically speaking, shouldn't be possible with my current understanding of how Recovery/Replacements work.

An A quality battalion with 88 out of 96 Vehicles recovered 5 Vehicles using a parameter data Recovery value of 2% and a Replacement value of 1%. Even if results are anywhere between 0 and twice the value in the PDT file, it shouldn't be possible to recover 5 Vehicles when I calculated Vehicles as being the equivalent of 10 Men.

Assuming a high end result of "results are doubled":

Recovery:

8 Vehicles = 80 Men, 80 x 0.04 is 3.2 Men or 0.32 Vehicle.

Using twice the value: 6.4 Men or 0.64 Vehicle.

Even if there are two separate checks in the sense that the result of the first check is doubled and that result is "the unit regains a Vehicle", the maximum gain is 2 Vehicles. If the result of the first check isn't doubled, but the Recovery value is doubled (or rather: quadrupled from the base value, as the quality modifier already doubles it) the result is less than 1 Vehicle so it can never recover 2.

Replacements:

96 Vehicles = 960 Men, 960 x 0.01 is 96 or 0.96 Vehicle.

Again, the result is <1 so even if the gain in the first check is multiplied by 2 the results can never be higher than 2 Vehicles.

Yet the unit recovered/replaced 5.
Quote this message in a reply
12-27-2021, 06:51 AM,
#13
RE: Replacement question
There may be a bug but I also suspect that there is an undocumented element in the replacement calculation. It probably handles its % calc in the same way as the recovery % in that it can range from zero to twice the nominal rate. So a replacement rate of 1% would correspond to 1% on average. I am guessing but the testing I have done always gives values within this range. Apart from this your calculations are the same as mine although I would express them without the conversion to and from Men:

Recovery:

8 vehicles = 8 x .04 x 2 = .64

This gives a maximum of 1 vehicle.

Replacements:

96 vehicles = 96 x .02 = 1.92 

This gives a maximum of 2 vehicles.

So, 3 in total should be the highest number possible. I have tried repeatedly to replicate your result but only ever get 0, 1 or 2 vehicles on every attempt. I am using a quality A vehicle unit with 8 losses and a full strength of 96. I have used the J46 OOB and PDT. While I never got a value of 3 vehicles on a single turn, this is not surprising as the probability for this is low. 

To get a value of 5 vehicles, as you did, suggests that unless you happened to hit the maximum possible value, then even higher values are potentially possible. This is not consistent with my tests so there must be some important difference that is being missed. Any ideas?

John
Quote this message in a reply
12-28-2021, 12:05 AM,
#14
RE: Replacement question
You could revamp it to have so many replacements in men, vehicles and air assets each turn to be allocated to which ever units you wish. But, they would also be split by country. All replacements by the U.S. could only be allocated to U.S. units. Like replacing some losses in the Campaign Series. But, that would make a lot of work to get implemented.
Quote this message in a reply
12-28-2021, 03:38 AM, (This post was last modified: 12-28-2021, 03:40 AM by ComradeP.)
#15
RE: Replacement question
(12-27-2021, 06:51 AM)Green Wrote: There may be a bug but I also suspect that there is an undocumented element in the replacement calculation. It probably handles its % calc in the same way as the recovery % in that it can range from zero to twice the nominal rate. So a replacement rate of 1% would correspond to 1% on average. I am guessing but the testing I have done always gives values within this range. Apart from this your calculations are the same as mine although I would express them without the conversion to and from Men:

Recovery:

8 vehicles = 8 x .04 x 2 = .64

This gives a maximum of 1 vehicle.

Replacements:

96 vehicles = 96 x .02 = 1.92 

This gives a maximum of 2 vehicles.

So, 3 in total should be the highest number possible. I have tried repeatedly to replicate your result but only ever get 0, 1 or 2 vehicles on every attempt. I am using a quality A vehicle unit with 8 losses and a full strength of 96. I have used the J46 OOB and PDT. While I never got a value of 3 vehicles on a single turn, this is not surprising as the probability for this is low. 

To get a value of 5 vehicles, as you did, suggests that unless you happened to hit the maximum possible value, then even higher values are potentially possible. This is not consistent with my tests so there must be some important difference that is being missed. Any ideas?

John

Apologies, the mystery has been solved: the OOB replacement value is 2%. I opened a different OOB file when checking the Replacement rate than the one that was actually used. 5 vehicles would still be a pretty high end result, but it's possible.

I guess the maximum would be either 5 or 6 Vehicles: 96 x 0.4 =3.84 or 2 x 1.92 from Replacements and 1 or 2 from Recovery, depending on if there are separate checks of 2 x 0.32 or a single check for 0.64.
Quote this message in a reply


Forum Jump:


Users browsing this thread: 2 Guest(s)