War Supplies during first year

[Gray_Lensman]

deleted

[Major Tom]

Pocus finally did give us an answer, so now I can work with the replacement requirements in a more informative way. Seems the chance of using a replacement chit is 50% of the proportional loss of HPs, meaning a unit with 50% health (or HPs) would have a 25% chance of using a replacement chit. Since I can't really work "chance" into the number tally, the next best thing is to do a 50% proportional calculation for every unit that starts the scenario with reduced HPs.

That approach certainly makes sense.

That was an interesting tidbit from Pocus. I don't think it's been revealed before what the chance is of a chit being used when repairing a unit.

The fact there's essentially a 50% "discount" for replacements versus new units makes it even more painful than befor to lose an entire element in combat.

I think the discount makes a lot of sense -- because the replacements are going into pre-existing regiments, the regimental infrastrucure and support are already there and paid for.

[77NY]

Since I can't really work "chance" into the number tally, the next best thing is to do a 50% proportional calculation for every unit that starts the scenario with reduced HPs.

Here's a quick-and-dirty but reasonably accurate suggestion to get at replacement chit costs using Pocus's probability. Divide the force pool units into four quartiles based on % damage/incompleteness and do a unit count for each group.

# damaged 1% - 25% (multiply total # units by 6.25%)
# damaged 26% -50% (multiply total # units by 18.5%)
# damaged 51% - 75% (multiply total # units by 31%)
# damaged 76% - 100% (multiply total # units by 43.75%)

The % multipliers are nothing more than the average probability for each quartile range. This gets muddied by depot deployment and any limits (if any) put on replacements per turn. But barring any serious limits, this should be helpful to you.

[Major Tom]

Here's a quick-and-dirty but reasonably accurate suggestion to get at replacement chit costs using Pocus's probability. Divide the force pool units into four quartiles based on % damage/incompleteness and do a unit count for each group.

# damaged 1% - 25% (multiply total # units by 6.25%)
# damaged 26% -50% (multiply total # units by 18.5%)
# damaged 51% - 75% (multiply total # units by 31%)
# damaged 76% - 100% (multiply total # units by 43.75%)

The % multipliers are nothing more than the average probability for each quartile range. This gets muddied by depot deployment and any limits (if any) put on replacements per turn. But barring any serious limits, this should be helpful to you.

I'm not a mathemetician, but I don't see how this could give a more accurate cost estimate than Gray's simpler approach. Why divide the units into quartiles when you know the exact number of hits each unit needs replaced?

[77NY]

I'm not a mathemetician, but I don't see how this could give a more accurate cost estimate than Gray's simpler approach. Why divide the units into quartiles when you know the exact number of hits each unit needs replaced?

The larger the sample size, the more accurate the calculation will be.

Edit: I am assuming there is a distribution of hit %. If 100 line inf units are all at 50% hits to start, then it is fine to assume 25 line inf chits will be needed.

[Major Tom]

The larger the sample size, the more accurate the calculation will be.

Edit: I am assuming there is a distribution of hit %. If 100 line inf units are all at 50% hits to start, then it is fine to assume 25 line inf chits will be needed.

Yes, some units start with more "damage" than others, but there's no need to estimate it. The scenario file has setup info for every unit, which includes how much health they have at start. For instance, most of the CSA fort garrisons and batteries start at 50% health.

[Gray_Lensman]

If you think about it, resource value wise, the units provided as partial HP reinforcements represent less value than a full strength unit of the same type, and to (theoretically) build a unit up to full strength from 0 HP would cost 1/2 of a comparable replacement point. Not sure yet if I want to quibble over such details.

Redone CSA Tally (adding in some 100% events, but NOT accounting for partial strength units)

1862 Campaign Scenario (1862 Early March):

$6842, 3690 conscripts, 1871 war supplies

1861 April Campaign Scenario (plus all the reinforcements):

$4938, 2690 conscripts, 1439 war supplies

Resource amount differences: (does not take into account the 1861 per turn additions yet):

$1904, 1000 conscripts, 432 war supplies

(Non-Random) Resources generated in the 22 turns from 1861 Early Apr thru 1862 Late Feb:

$2023, 671 conscripts, 1256 war supplies

Net differences: (by 1862 Early March)

+$119 (surplus)
[color="red"]-329 conscripts[/color] (shortfall)
+824 war supplies (surplus)

[Major Tom]

If you think about it, resource value wise, the units provided as partial HP reinforcements represent less value than a full strength unit of the same type, and to (theoretically) build a unit up to full strength from 0 HP would cost 1/2 of a comparable replacement point. Not sure yet if I want to quibble over such details.

It's not that small of a detail -- my three test games showed an average cost of $564, 270 conscripts, and 147 WSu for bringing those partial HP units up to full health. This may be a little high if I miscounted the number of full elements that were used to fill incomplete units (I deducted those costs from these totals).

But -- in the overall scheme of things, with size of the numbers you're looking at, maybe these numbers don't make all that much difference. The WSu doesn't, and the concripts don't really, either, since with the volunteer and conscription options the CSA could bring in lots more conscripts in 1861 -- so the conscript shortfall is probably not a real issue. But, that's a pretty big pile of money. You could make it easier on yourself by ignoring this completely, and offsetting it by also not counting the free additions to the 1861 infantry/cav/militia replacement pools.

Two other things to consider:

In your $6842, 3690 conscripts, 1871 war supplies balaances for 1862, are you including the starting pools of these items -- $250, 100 conscripts, 85 WSu?

And finally - I think you're still missing the effect of several significant 100% occurance events that you can find on the Events_Scripts tab of the AACW_Setup1861 file:

May 1, 1861 -
+$75
+50 conscripts
+15 WSu (only gave me 10, but my setup file is v1.12a, so maybe this was changed in the last patch?)
+10 Rail Pool
+3 River Pool

June 1, 1861 -
+$75
+75 conscripts
+15 WSu
+10 Rail Pool
+3 River Pool

July 1, 1861 -
+$50
+25 conscripts
+5 WSu (I did not get this -- again, maybe the databsae file is out of date)
+5 Rail Pool
+2 River Pool

[Gray_Lensman]

deleted

[Jarkko]

So these calculations in no way take into consideration the cost of the replacements that have been used to fill up units that take damage along the way (combat losses, attrition), or the fact that militia converting to conscripts *also* use resources? Incidently, how many militia units are there on the field in the 62 scenario compared to what you get from events and if you'd be "historically" recruiting militia?

[Gray_Lensman]

deleted

[Jarkko]

I suppose the above numbers assume that the (un-verified?) comment from Pocus is not true:

For example, on average, a regular infantry replacement will give back 10 hits to your regiments (anywhere on the map, if supplied).

but that 1 replacement actually gives back 20 hits on the average as has been believed? I've tried to test this, but so far my results are mixed...

[Gray_Lensman]

deleted

[Jarkko]

Interesting fact about the replacements. Their value is thus much higher than I have believed; one replacement point is worth two elements in the long run. What makes this interesting is that I have been in the belief the replacing was not that efficient during ACW (I have to admit I have believed usually understremtgth veteran units were in fact merged with each other rather than recieving replacements), while the numbers in the game prove the exact opposite -> replacements not only are worth two rookie elements, they also reach any (supplied) forces anywhere on the map, and they also automatically recieve the experience of veteran troops (if joining such an element).

No wonder militia spamming has been the smart thing to do. With one regular replacement point we've been able to convert eight militia elements to conscripts.

[Gray_Lensman]

deleted

[Gray_Lensman]

deleted

[Jarkko]

Did you calculate in the fact that Union does conscript three times during this period (or the use of any other political means)?

[Gray_Lensman]

deleted

[PeterMac]

Sorry to throw another spanner in the works, but think it does need accounting for. I have seen no sign of any consideration for replacement of losses (Expected / Historical), even through 1861 these losses were, and are likely to be in any game, quite significant. This is especially true if played with historical attrition rates, due to disease, and weather effects. For CSA only over this period I have seen combat losses of around 20,000, in some games with attrition losses, difficult to asses, but probably half as much again, probably costs around 300 conscript points, 6-700 money, and 80-100 WS, to replace these.

[PeterMac]

Oops, Sorry Gray, just noticed your note 1: