Russia and its inablility to cope with losses

[Beryl]

Siberian division saving moscow is a myth (from http://operationbarbarossa.net/Myth-Busters/Mythbusters3.html#an_3) :

The Soviet land model shows that 182 rifle divisions, 43 militia rifle divisions, eight tank divisions, three mechanised divisions, 62 tank brigades, 50 cavalry divisions, 55 rifle brigades, 21 naval rifle brigades, 11 naval infantry brigades, 41 armies, 11 fronts and a multitude of other units were newly Mobilised and Deployed (MD) in the second half of 1941. If Mobilized and Not Deployed (MND) units are included then this list is considerably higher.
Even if the few Siberian divisions exhibited a higher than average combat proficiency in the winter of 1941/42, their contribution was almost insignificant compared to the mass of newly mobilised units. There is no doubt that the 1941 Soviet mobilisation programme was simply the largest and fastest wartime mobilisation in history. The multitude of average Soviet soldiers from all over the USSR that made up these units saved the day, and definitely not the existing units transferred west after June 1941, or the mostly non-existent and mythical Siberian divisions.

 

[knightofni]

I really still think that AI USSR just needs some sort of event which replenishes their troops/adds a lot of new militia or outdated infantry divisions.

You can try this from FT

 

[PanzerMan7]

once again, its the build times (i really hate DH build times)

 

[Majorlee2000]

How do you think the AI would react if for INF the build time was halved and the IC cost was doubled? This would mean the same amount of IC used, thus keeping the number of divisions built prewar potentially at around the same number, but with the quicker build time it would allow the AI to field more divisions in a shorter time frame.

 

[Cybvep]

It's not that easy, changing build-times affects balance severely. Also, they are already not that long when you are mobilised.

 

[PanzerMan7]

as it is, fighters are cheaper than infantry

because supposedly it takes longer to train soldiers than pilots

 

[Shosho Tanaka]

as it is, fighters are cheaper than infantry

because supposedly it takes longer to train soldiers than pilots

Extremely valid point.

 

[Westphalian]

Another approach could be to adjust your tc modifiers. fe edit TC Occupied Province Load from 1.0 to 5.0 or TC Load from partisans from 12.0 to 50.0. Of course you also need an event to boost the red army.
I tried it myself, but i failed to make a proper event to enhance the red army after its defeat.

 

[Gort11]

Default value is 0.6. So changing it to 30 multiplies the cost by 50, without changing the time

Detailed explanation by Dichromate are here : Link.
Summary : Reinforcing one strength of an infantry div is 22.5 times cheaper (in IC days) than building one strength of new infantry. Kind if explains why it's so easy to always keep your units full strength.

Original data (in misc.txt)

# Reinforce cost
	0.6 # 0.6
# Reinforce time
	0.4 # 0.35

So if i want to increase reinforcement cost in IC days by 22.5 times linearly, i guess we should multiply each component by sqrt(22.5) = 4.75

Proposed data (in misc.txt)

# Reinforce cost
	2.85 #0.6 # 0.6
# Reinforce time
	1.9 # 0.4 # 0.35

Gonna try playing like this.

I will also try this and post a trip report.

Trip report: Jeez, get used to seeing "Reinforcement cost: 3000" on your industrial screen. It seems that this change makes planes in particular EXTREMELY expensive to keep around. I was using 8 tactical bombers as Japan versus China, and once they got down to 50% strength they were costing about 1500 to repair, which sucks away all the IC I had to spare. I wonder if they're really worth it at this cost. It seems to me as though they ought to be one of Japan's great advantages, but they seem just a liability as it is. I'll need to play more and figure it out.

I have a question though - if reinforcing is 22.5 times cheaper than building new things, shouldn't we be making reinforcements 22.5 times more expensive, not 50 times? I don't understand why we've gone so far the other way. It seems like you'd be a fool to reinforce at all if it's twice as expensive as building new stuff - you're better off just sending divisions to their doom.

 

[unmerged(455658)]

I just finished a game as England and the USSR performed slightly better than in history. Germany pushed within one province of Moscow on at least two sides, surrounded Leningrad in the north, but didn't quite make it to Stalingrad in the south. They barely held on until the allies took Italy in 1943 and pushed into France and the Balkans in 1944, then by the end of 1944 the UK and USSR were racing to Berlin. It was nice to see things play out so close to history, but maybe it was just me opening a second front earlier than in reality.

 

[knightofni]

I will also try this and post a trip report.

Trip report: Jeez, get used to seeing "Reinforcement cost: 3000" on your industrial screen. It seems that this change makes planes in particular EXTREMELY expensive to keep around. I was using 8 tactical bombers as Japan versus China, and once they got down to 50% strength they were costing about 1500 to repair, which sucks away all the IC I had to spare. I wonder if they're really worth it at this cost. It seems to me as though they ought to be one of Japan's great advantages, but they seem just a liability as it is. I'll need to play more and figure it out.

I have a question though - if reinforcing is 22.5 times cheaper than building new things, shouldn't we be making reinforcements 22.5 times more expensive, not 50 times? I don't understand why we've gone so far the other way. It seems like you'd be a fool to reinforce at all if it's twice as expensive as building new stuff - you're better off just sending divisions to their doom.

I've tried a Brazil game with the parameters i posted (# Reinforce cost 2.85, # Reinforce time 1.9), and though reinforcement was a bit high at times it was definetely under control (Maybe 10-15 at most, quite high considering Brazil's low IC, but acceptable nonetheless. The biggest change for me was time : with vanilla value, in most cases a retreating infantry would need 3-4 days to be fully reinforced. With those settings you have to wait longer, about 2 weeks i'd say. Of course my army was mostly infantry, with some TAC but not much air opposition.

Maybe you've used the settings proposed by rotten venetic ? (# Reinforce cost 30 # 0.6) That would explain your results. I'd suggest using mine (which you quoted). That would also solve your second question (about why 22.5 vs 50)

 

[Gort11]

Yeah, I used rotten venetic's numbers. I'll try yours and see how that is.

 

[unmerged(233955)]

I don't see the point of making reinforcement even more expensive then it would be to create a normal division. That's completely out of whack.

The base reinforcment cost/time values are for strength=0. Due to calculations in the game engine, real cost/time values are much smaller than that but it still means at strength=1 or something you'll be going through multiples of what it would take to raise a new division. Which makes no sense whatsoever.

 

[Schmitz]

yea, well reinforcement should definitely be cheaper than building a new division, in fact, that was one of Guderian's complaints, that Hiter wanted to create new divisions rather than reinforcing the older ones. having said that, at the moment, reinforcement is much cheaper and quicker than a new division, which doesn't make perfect sense since you still have to train new troops and build new weapons/tanks that were lost. so basically, IMO, the difference between reinforcement and creating a new division should be closer together than it is, but reinforcement should still be cheaper.

 

[Cybvep]

Start small - multiply reinforcement cost by 3 and reinforcement time by 1.25 and see what are the in-game results. I expect that we want reinforcements to be quite expensive, but not game-breaking, because with a 1:1 ratio the in-game economic system will collapse (~3000 IC for reinforcements is not sth we want to see).

 

[unmerged(193702)]

Russia had to keep a very serious army in the east because of the Japs
drop them back to Stalingrad through event or something

 

[unmerged(233955)]

I say put Reinforce cost at 1.0 and time at 0.8 myself. Of course, it's only 3x the current rate, but it should be much fairer while keeping reinforcement a viable option. Besides, I don't like going above 1.0 for now, since that effectively just doesn't make sense.

 

[knightofni]

I say put Reinforce cost at 1.0 and time at 0.8 myself. Of course, it's only 3x the current rate, but it should be much fairer while keeping reinforcement a viable option. Besides, I don't like going above 1.0 for now, since that effectively just doesn't make sense.

Actually going above 1 does make sense, because the final upgrade cost & time is the product of the values in misc.txt by the one present in the db/units/division/*.txt

Example for infantry, with vanilla numbers :

in Infantry.txt
reinforce_time = 1.1
reinforce_cost = 0.56

in misc.txt
# Reinforce cost
0.6 # 0.6
# Reinforce time
0.4 # 0.35

As far as i undertand, it means that the final multipliers are :
Cost : 0.56 x 0.6 = 0.336 original cost
Time : 1.1 x 0.4 = 0.44 original time

so in ICdays, gives : 0.336 x 0.44 = 0.147 of a new division

So today, reinforcing a infantry division in the middle of the russian winter cost 14.7% in ICdays compared to creating a new one, safely in the german heartland. Which seems to be slightly off the mark.

Personally i think that the overall cost (in ICdays) should be cheaper when reinforcing, but not that much (maybe 70 to 90% of the full cost, depending on softness), but it should be longer than creating new divisions, to simulate the difficulty/time of sending reinforcement to the front.

----
Applying the above calculation to the numbers i've suggested above

# Reinforce cost
2.85 #0.6 # 0.6
# Reinforce time
1.9 # 0.4 # 0.35

Cost : 0.56 x 2.85 = 1.595 original cost
Time : 1.1 x 1.9 = 2.09 orginal time
ICdays : 1.595 x 2.09 = 3.33 of a new division

So it's too much, as noted by posters above.

------
Anyway, i've just given a look to the 1.3 sneak peek [link], and look at what is planned

As part of this rework I went over most, if not all modifiercombination that affect different engine aspects. Whenever more then 1 modifier was used, I decided to set the general modifer in misc.txt to 1.

Just one example.
The upgrade modifier in misc.txt and an individual model upgrade time in the unit files.
But there are a lot more of such modifier pairs.

old:

misc.txt
# Upgrade cost
0.6
# Upgrade time
0.8
unitfile:
upgrade_time_factor = 0.75
upgrade_cost_factor = 1.25

new:

misc.txt
# Upgrade cost
1
# Upgrade time
1
unitfile:
upgrade_time_factor = 0.6
upgrade_cost_factor = 0.75

It's just a little thing, but makes it much easier to understand and adjust values of indivdual units or unittypes.

--------
Because 1.3 is still some time away, i'm going to try the following, as a quick patch :

# Reinforce cost
0.9 #0.6 # 0.6
# Reinforce time
1.5 # 0.4 # 0.35


Cost : 0.56 x 0.9 = 0.504 original cost
Time : 1.1 x 1.5 = 1.65 original time
ICdays : 0.504 x 1.65 = 0.83 of a new division
----

 

[Cybvep]

Because 1.3 is still some time away, i'm going to try the following, as a quick patch :

# Reinforce cost
0.9 #0.6 # 0.6
# Reinforce time
1.5 # 0.4 # 0.35


Cost : 0.56 x 0.9 = 0.504 original cost
Time : 1.1 x 1.5 = 1.65 original time
ICdays : 0.504 x 1.65 = 0.83 of a new division

With DH's buildtimes, it will be too long. For starters, I would use sth like this:

Cost : 0.56 x 0.9 = 0.504 original cost
Time : 1.1 x 0.85 = 0.935 original time

so in ICdays, that gives : 0.504 x 0.935 = 0.471 of a new division

This is still 3 times as much as the original value and can affect balance severely. It needs testing!

 

[Gort11]

To be honest, I found that raising the reinforcement time to anything like the amount of time it takes to build a division felt incredibly slow. I think perhaps making divisions simply take less time to build and making the reinforcement stats match that might be better.