Planning LCU transport: LCU-Ship Allocation Calculus

Matrix Games Forums

Forums Register Login Photo Gallery Member List Search Calendars FAQ

My Profile Inbox Address Book My Subscription My Forums Log Out

Planning LCU transport: LCU-Ship Allocation Calculus

View related threads: (in this forum | in all forums)

Logged in as: Guest

Users viewing this topic: none

Printable Version

All Forums >> [New Releases from Matrix Games] >> War in the Pacific: Admiral's Edition >> Planning LCU transport: LCU-Ship Allocation Calculus

Page: [1] 2 next > >>

Message

<< Older Topic Newer Topic >>

Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 1:23:47 AM

glyphoglossus

Posts: 71
Joined: 9/26/2016
Status: offline

So, I've been through the manual closely, as well as scoured these forums, but it appears I still do not get it.

My understanding is that, each LCU has a troop and cargo cost, that both must be fulfilled by the transports.
There is a complex algorithm as to how these units --- troops, devices, etc. --- are distributed amongst the transports, but as long as my transports have sufficient independent troop and cargo capacity, I do not have to worry about cross-loading etc., I can leave all other worries to the (AI) ship masters?
Furthermore, if it is a Transport TF mission, then the LCU's must be in strategic move mode, while if it is an Amphibious TF mission, the LCU's must be in combat mode.
Crucially, as far as calculating loads go, if it is an Amphibious TF mission, then it is a combat load, where: transport troop/cargo capacity is reduced by 80%, and an extra cargo space if required to carry 3 days supply per unit.

In addition, and this very, very, very, very important thing is something that I missed ENTIRELY the first time: each ship can only accommodate a SINGLE unit, either in whole or in part.
Alternatively: even if a single ship is not filled to capacity by a unit, it cannot take another unit, either in whole or in part, even if there is room to spare!

OK, all well and good.

So, after very carefully planning the absolute most efficient collection of ships to provide transport for a COMMERCIAL load of LCU's ... I find that I have underestimated.

Here are the troops I want to transport, and their associated load costs:

-----------------------+-------+-------
 LCU                    | Troop | Cargo
 -----------------------+-------+------
 121st USN Base Force   | 1280  |   728
 131st USA Aviation     |  900  |     0
 3rd USN SeaBees        |  388  |  1200
 11th USN SeaBees       |  388  |  1200
 -----------------------+-------+-------
 Total                  | 2956  |  3128 
 -----------------------+-------+-------

And here is the TF that I have assembled, and their supposed/expected loadings:

+----------------------------------------+---------------+-------+-------+
| Load                                   | Ship          | Troop | Cargo |
+----------------------------------------+---------------+-------+-------+
| 121st USN Base Force Tr: 1280 Ca:  728 | - xAP Adelong |   500 |   750 |
|                                        | - xAP Manunda |   900 |   800 |
| 3rd USN SeaBees      Tr:  388 Ca: 1200 | - xAP Rhesus  |   500 |   750 |
|                                        | - xAP Mungana |   500 |   750 |
| 11th USN SeaBees     Tr:  388 Ca: 1200 | - xAP Murada  |   500 |   750 |
|                                        | - xAP Barwon  |   500 |   750 |
| 131st USA Aviation   Tr:  900 Ca:    0 | - xAP Maori   |   900 |   800 |
+----------------------------------------+---------------+-------+-------+

All looking good? Room to spare?

Apparently not.

After laboriously assembling the TF in one port and moving it to the base with the LCU's, the load cannot be accommodated:

http://imgur.com/a/C66N4

The manual describes the load allocation procedure in Pg. 122. The largest unit will try to be fitted into the smallest ship that can take it. If this does not work, it will be fitted into the largest ship, and then the cycle repeats with the excess.

The discrepency here seems to be how the game handles the remainder after the 121st is loaded on the largest available. After the Manunda takes on as much of the 121st, I have the remaining 380 troops (and 0 cargo) going on to the 500-troop capacity Adelong.
But, instead, the game selects the 900-troop capacity Maori for this, which is excessive.
As a result, there is insufficient capacity remaining to load up the SeaBees.

Why did the game select the 900-troop capacity transport for the remainder of the 121st, when any of the 500-troop capacity transports could have served?

How do you folks plan shipping required for LCU's?

Attachment (1)

< Message edited by glyphoglossus -- 11/5/2016 2:12:57 AM >

Post #: 1

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 2:23:14 AM

btd64

Posts: 9973
Joined: 1/23/2010
From: Mass. USA. now in Lancaster, OHIO
Status: offline

I know this is going to sound weird but, I calculate like this;

For every 1000 of troop capacity needed I add 1000 troop cargo space, rounding up. So for the 121st USN BF you would need 2000 troop cargo space.
So, for your 4 LCU's you need 5000 troop cap. Plus I add 1000 for every 10K troop cap. Same for Equipment. For amphibian transport I add 2000 for every 10K. 95% of the time I don't have a problem.

The reason I do it this way is that when I first started with WITP and then AE, following your method I never got all units loaded. So as I learned more I found that it's not so much Have the troop space greater than the number you have, It's about having enough ships to load everything. For example, You could have the Queen Mary with around 8K troop space and 2 units totaling 2K troop space needed and only 1 unit will load. No more than 1 unit per ship. Hope this helps....GP

_____________________________

Intel i7 4.3GHz 10th Gen,16GB Ram,Nvidia GeForce MX330

AKA General Patton

WPO,WITP,WITPAE-Mod Designer/Tester
DWU-Beta Tester
TOAW4-Alpha/Beta Tester

"Do everything you ask of those you command"....Gen. George S. Patton

(in reply to glyphoglossus)

Post #: 2

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 4:40:42 AM

glyphoglossus

Posts: 71
Joined: 9/26/2016
Status: offline

quote:

ORIGINAL: General Patton
For every 1000 of troop capacity needed I add 1000 troop cargo space, rounding up. So for the 121st USN BF you would need 2000 troop cargo space.
So, for your 4 LCU's you need 5000 troop cap. Plus I add 1000 for every 10K troop cap. Same for Equipment. For amphibian transport I add 2000 for every 10K. 95% of the time I don't have a problem.

Thanks. Trying to follow here. For the 121st USN BF, the troop space is 1280. So, if I were to "For every 1000 of troop capacity needed I add 1000 troop cargo space, rounding up", that would would be 2000 + 1000 = 3000? That is, I add 1000 to the 1280 of the 121st, for a total of 2280, and rounding up is 3000. How did you get 2000?

Continuing on, based on (my obviously mistaken interpretation) we would get: 2000 for the 131st USA Aviation and each of the two Seabees, which would mean a total of 9000 troops for the 4 LCU's, not 5000 as you calculated. What am I missing here?

quote:

ORIGINAL: General Patton
For example, You could have the Queen Mary with around 8K troop space and 2 units totaling 2K troop space needed and only 1 unit will load. No more than 1 unit per ship. Hope this helps....GP

Yes! I discovered this early: basically, a single ship cannot load more than one LCU, even if there is lots of capacity left over.

(in reply to btd64)

Post #: 3

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 4:40:47 AM

glyphoglossus

Posts: 71
Joined: 9/26/2016
Status: offline

What is weird is that I take the original TF, disband them, and then create a bunch of "mini-TF's", each consisting of just the ships I had originally planned out, then all the LCU's load fine!

So, for e.g., a 2-ship TF of just the Adelong and Manunda, with a total troop capacity of 1400 loads the 121st fine by themselves. And a 2-ship TF of just Rhesus and Mungana loads the SeaBees fine. Continuing this way, we get all the LCU's loaded into the original ship set, exactly as calculated! They can then, presumably, be merged into a single TF once en route. A bit of a pain, but maybe worth it for a more efficient packing scheme?

(in reply to glyphoglossus)

Post #: 4

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 5:52:54 AM

Bullwinkle58

Posts: 11302
Joined: 2/24/2009
Status: offline

In a GC you're going to load thousands of TFs. Thousands.

I just allocate a plug of ships, load them, and add some if they need more space.

Then make a new one to load the left-behind Motorized Support. That always gets left behind.

Your way will lead to insanity.

_____________________________

The Moose

(in reply to glyphoglossus)

Post #: 5

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 6:49:04 AM

BBfanboy

Posts: 18046
Joined: 8/4/2010
From: Winnipeg, MB
Status: offline

Your problem in the original situation was that you were looking only at the weight of the load and the load capacity (by weight) of the ships. The ship capacity is for a fully used hold capacity which it could achieve with stack of palletized/crated goods. But for amphibious load of military equipment it must be loaded "ready to use", so you cannot, for example, stack trucks or guns. That means unused hold space - inefficiency in loading.

There are no figures in the database describing the space that such equipment takes (per device), so something "under the hood" in the AI calculations must determine how much equipment will actually fit on each ship. Although it is possible to cut it very close or to fiddle with adding ships to the TF after the load does not quite fit, I find it easier to just multiply the total load weight X3 and that almost always takes care of the equipment and provides enough supply for the troops to operate for a little while.

For the troops the penalty for Amphib loading is about 15-20% so I use the 20% figure to ensure I have enough troop capacity in the ships. If I have plenty of transports I use more than needed to get faster unloading of the troops.

_____________________________

No matter how bad a situation is, you can always make it worse. - Chris Hadfield : An Astronaut's Guide To Life On Earth

(in reply to glyphoglossus)

Post #: 6

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 7:26:24 AM

glyphoglossus

Posts: 71
Joined: 9/26/2016
Status: offline

quote:

ORIGINAL: BBfanboy

Your problem in the original situation was that you were looking only at the weight of the load and the load capacity (by weight) of the ships. The ship capacity is for a fully used hold capacity which it could achieve with stack of palletized/crated goods. But for amphibious load of military equipment it must be loaded "ready to use", so you cannot, for example, stack trucks or guns. That means unused hold space - inefficiency in loading.

There are no figures in the database describing the space that such equipment takes (per device), so something "under the hood" in the AI calculations must determine how much equipment will actually fit on each ship. Although it is possible to cut it very close or to fiddle with adding ships to the TF after the load does not quite fit, I find it easier to just multiply the total load weight X3 and that almost always takes care of the equipment and provides enough supply for the troops to operate for a little while.

For the troops the penalty for Amphib loading is about 15-20% so I use the 20% figure to ensure I have enough troop capacity in the ships. If I have plenty of transports I use more than needed to get faster unloading of the troops.

In the original situation, I am using a commercial load (Transport TF), which does not have the inefficiency.

As I noted in the original post, according to the manual an amphibious mission uses a combat load, which reduces the troop and cargo capacity to 80% of original (for non-landing ship classes such as xAP/AP/AK ship classes), and, in addition, an amphibious mission requires an extra cargo load of 3 days supply per LCU. When I use this (i.e., calculating each ship's capacity as 80% of stated), the numbers all do actually work out, and I am able to load the LCU's as predicted.

However, the issue I ran into is that the commercial load is allocated in a different way. In fact, the numbers are correct, as evidenced by the separate TF's accommodating the load, but the ship allocation seems to (mis?)use the ship capacity. I'm trying to understand why.

(in reply to BBfanboy)

Post #: 7

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 7:31:54 AM

Lokasenna

Posts: 9297
Joined: 3/3/2012
From: Iowan in MD/DC
Status: offline

Ignore everybody else and just do the following.

Amphibious load means you only load up to 80%. Because of this, you need 25% additional troop space. However, devices can't be split between ships (there are no "remainders" and "carry the treads" in this long division), so you must either:
1) be loading a unit that is sufficiently small and/or has sufficiently small devices only, or
2) allow for even more additional total space

In addition to this, you must allow for a minimum allocation of supplies for the unit. If you do not, you'll end up leaving a fragment behind. My general rule is to look at the estimated free space and shoot for around 30% "troop safety" and closer to 40% "cargo safety", but variable depending on the number and size of the units I'm trying to load. Fewer units = need less free space. More units = need more free space (due to the no-remainders thing and only 1 unit per ship).

Also, it will help you a whole ton if you only load like-size units in the same TF. It's even easier if you just load one unit at a time, especially for big ones (think divisions, but also regiments). It's possible to quickly and easily load whole divisions in multiples in the same TF - so long as they are roughly the same load cost for both troops and cargo. You can stray from this a little bit without too much of a problem, but if you are trying to load a USMC Division and say 3 Seabee battalions in the same TF with only a minimum of extra space, you're going to run into issues because the game will use "too many" ships for the Seabees and too few for the Marines - leaving some behind. It's better to just make one TF for the USMC Division, and then another TF for the 3 Seabee units. You can combine them later if you wish.

And yes, Bullwinkle, I've done this without leaving the Motorized Support behind.

(in reply to BBfanboy)

Post #: 8

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 7:32:48 AM

Lokasenna

Posts: 9297
Joined: 3/3/2012
From: Iowan in MD/DC
Status: offline

quote:

ORIGINAL: BBfanboy

Your problem in the original situation was that you were looking only at the weight of the load and the load capacity (by weight) of the ships. The ship capacity is for a fully used hold capacity which it could achieve with stack of palletized/crated goods. But for amphibious load of military equipment it must be loaded "ready to use", so you cannot, for example, stack trucks or guns. That means unused hold space - inefficiency in loading.

There are no figures in the database describing the space that such equipment takes (per device), so something "under the hood" in the AI calculations must determine how much equipment will actually fit on each ship. Although it is possible to cut it very close or to fiddle with adding ships to the TF after the load does not quite fit, I find it easier to just multiply the total load weight X3 and that almost always takes care of the equipment and provides enough supply for the troops to operate for a little while.

For the troops the penalty for Amphib loading is about 15-20% so I use the 20% figure to ensure I have enough troop capacity in the ships. If I have plenty of transports I use more than needed to get faster unloading of the troops.

Except in cases of cross-loading Troops/Cargo space with Cargo/Troops, load cost is load cost is load cost is load cost...

(in reply to BBfanboy)

Post #: 9

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 7:36:21 AM

glyphoglossus

Posts: 71
Joined: 9/26/2016
Status: offline

quote:

ORIGINAL: Bullwinkle58

In a GC you're going to load thousands of TFs. Thousands.

I just allocate a plug of ships, load them, and add some if they need more space.

Then make a new one to load the left-behind Motorized Support. That always gets left behind.

Your way will lead to insanity.

I can imagine!

Yes, I can see this not working in the GC ...

Though it should be noted that I am not doing these calculations manually. What I am working on is actually a script that, given the available shipping and LCU's that need to be transported, comes up with a solution. This script hits the CSV files exported by WitPTracker for the ship availability, locations, capacities, etc. Then, based on the LCU's that need to be transported (specified by the user), it searches for an optimal configuration of ships for each TF, taking into account whether it is a commercial load or a combat load (the latter resulting in 80% reduction in capacities). The optimization algorithm is currently a brute force approach, and the optimization criteria is to minimize the total troop/load capacities while meeting the troop/load costs as minumum.

For small scenarios, this works out fine because the brute force approach is fast enough (trying out all the configurations) and it's not too big a deal for me to enter the LCU load costs myself (because WitPTracker does not report the LCU load costs).

Both of these issues will become objectionable on the scale of the GC.

However, a more efficient optimization algorithm is pretty easy to write and if I am lucky, by the time I am ready to take on the GC, maybe WitPTracker will report LCU load costs? Or I can figure out a way to extract them myself? Or, worst comes to worst, I'd have picked up a sufficient intuition to throw together TF's and hope for the best!

(in reply to Bullwinkle58)

Post #: 10

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 7:44:38 AM

glyphoglossus

Posts: 71
Joined: 9/26/2016
Status: offline

quote:

ORIGINAL: Lokasenna
Amphibious load means you only load up to 80%. Because of this, you need 25% additional troop space. However, devices can't be split between ships (there are no "remainders" and "carry the treads" in this long division), ...

Interesting. Does this hold for commercial loads as well?

quote:

ORIGINAL: Lokasenna
In addition to this, you must allow for a minimum allocation of supplies for the unit.

I read that combat loads for amphibious TF requires 3 days of supply, and factored this into the script when calculating load costs for amphibious TF's. But this is not the case with commercial loads or am I mistaken?

quote:

ORIGINAL: Lokasenna
Also, it will help you a whole ton if you only load like-size units in the same TF. It's even easier if you just load one unit at a time, especially for big ones (think divisions, but also regiments). It's possible to quickly and easily load whole divisions in multiples in the same TF - so long as they are roughly the same load cost for both troops and cargo. You can stray from this a little bit without too much of a problem, but if you are trying to load a USMC Division and say 3 Seabee battalions in the same TF with only a minimum of extra space, you're going to run into issues because the game will use "too many" ships for the Seabees and too few for the Marines - leaving some behind. It's better to just make one TF for the USMC Division, and then another TF for the 3 Seabee units. You can combine them later if you wish.

Ah, this is good to know. Indeed, I found no problems in my loading up each LCU separately into its own TF, and, as I noted above, the original ship set were, in fact sufficient for the LCU's when separated out into single-LCU TF's. This may be the way to go, tedious as it is.

(in reply to Lokasenna)

Post #: 11

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 8:21:53 AM

GetAssista

Posts: 2732
Joined: 9/19/2009
Status: offline

quote:

ORIGINAL: glyphoglossus
Ah, this is good to know. Indeed, I found no problems in my loading up each LCU separately into its own TF, and, as I noted above, the original ship set were, in fact sufficient for the LCU's when separated out into single-LCU TF's. This may be the way to go, tedious as it is.

Game has problems with the optimal distribution of ships if loading several LCUs into one TF, yes. It chooses ships somewhat mechanically, hence allocating too large a load for to small a unit and running out of space in the end. TEdious way is the way of you want to transport stuff very efficiently.
Yet I advise to strive for more redundancy, no putting all eggs in one basket. Allies can afford ships

(in reply to glyphoglossus)

Post #: 12

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 1:12:24 PM

MakeeLearn

Posts: 4278
Joined: 9/11/2016
Status: offline

whether it's �Use Minimum Ships" or "Use All Ships" if looks like the AI may use the same algorithm when picking which ships to load a unit on.

121st USN Base Force 1280 / 2 = 640 troops, thus 2 900-troop capacity ships instead of a 900 capacity and a 500 capacity.

After picking it then loads according to the option selected.

< Message edited by MakeeLearn -- 11/5/2016 1:17:44 PM >

(in reply to GetAssista)

Post #: 13

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 1:27:13 PM

Leandros

Posts: 1740
Joined: 3/5/2015
Status: offline

quote:

ORIGINAL: Bullwinkle58

In a GC you're going to load thousands of TFs. Thousands.

I just allocate a plug of ships, load them, and add some if they need more space.

Then make a new one to load the left-behind Motorized Support. That always gets left behind.

Your way will lead to insanity.

This is much how I am tending to do it now, too. Same experience as yours.

Fred

_____________________________

River Wide, Ocean Deep - a book on Operation Sea Lion - www.fredleander.com
Saving MacArthur - a book series on how The Philippines were saved - in 1942! https://www.amazon.com/gp/product/B07D34QCWQ/?ie=UTF8&redirect=true&ref=series_rw_dp_labf

(in reply to Bullwinkle58)

Post #: 14

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 1:57:36 PM

MakeeLearn

Posts: 4278
Joined: 9/11/2016
Status: offline

Like in real life what the book says dont always matter. Depends on who in supply/transport is doing the loading. Some are stingy with space and others can put 10 gallons into a 5 gallon hat no problem. Ive seen some doozies.

Or like Parachute Rigging School where the book says NO TRUCKER'S HITCH!!!!

So what is one of the first knots they show you to tie.

And in low-altitude parachute-extraction (LAPES) rigging which knot is used that "no one knows" who tied it.

< Message edited by MakeeLearn -- 11/5/2016 2:00:06 PM >

(in reply to Leandros)

Post #: 15

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 2:07:44 PM

btd64

Posts: 9973
Joined: 1/23/2010
From: Mass. USA. now in Lancaster, OHIO
Status: offline

quote:

ORIGINAL: glyphoglossus

quote:

ORIGINAL: General Patton
For every 1000 of troop capacity needed I add 1000 troop cargo space, rounding up. So for the 121st USN BF you would need 2000 troop cargo space.
So, for your 4 LCU's you need 5000 troop cap. Plus I add 1000 for every 10K troop cap. Same for Equipment. For amphibian transport I add 2000 for every 10K. 95% of the time I don't have a problem.

Thanks. Trying to follow here. For the 121st USN BF, the troop space is 1280. So, if I were to "For every 1000 of troop capacity needed I add 1000 troop cargo space, rounding up", that would would be 2000 + 1000 = 3000? That is, I add 1000 to the 1280 of the 121st, for a total of 2280, and rounding up is 3000. How did you get 2000?

Continuing on, based on (my obviously mistaken interpretation) we would get: 2000 for the 131st USA Aviation and each of the two Seabees, which would mean a total of 9000 troops for the 4 LCU's, not 5000 as you calculated. What am I missing here?

quote:

ORIGINAL: General Patton
For example, You could have the Queen Mary with around 8K troop space and 2 units totaling 2K troop space needed and only 1 unit will load. No more than 1 unit per ship. Hope this helps....GP

Yes! I discovered this early: basically, a single ship cannot load more than one LCU, even if there is lots of capacity left over.

Sorry, I forgot to finish

In this case you round up and add together like so;

1280=2000
388=1000
388=1000
900=1000
total=5000+1000=6000

Sometimes add in a ship or two just for good measure.

This should help....GP

< Message edited by General Patton -- 11/5/2016 2:10:03 PM >

_____________________________

(in reply to glyphoglossus)

Post #: 16

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 2:26:11 PM

bush

Posts: 444
Joined: 10/30/2007
From: san jose, ca
Status: offline

I use General Patton's "round to 1,000" rule also. I rarely have issues.

(in reply to btd64)

Post #: 17

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 3:40:06 PM

mussey

Posts: 683
Joined: 12/2/2006
From: Cleve-Land
Status: offline

When in doubt, add a ship. Unfortunately in this case, I'm always in doubt.

_____________________________

Col. Mussbu

The long arm of the law - "The King of Battle"

(in reply to bush)

Post #: 18

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 3:59:36 PM

crsutton

Posts: 9590
Joined: 12/6/2002
From: Maryland
Status: offline

As an Allied player (and less than organized) my solution is just to use more ships than I think I will need. If I think 10 APs will do the job then I use 15. I use almost double the capacity amount with an amphibious TF as the troops will unload all the more faster anyways. I know it is wasteful but hey, I am the Allies and have got plenty of ships. Just do this whenever possible and save yourself a lot of time and worry. Japanese players probably have to be a bit more careful. I just hate loading up a TF and finding out that I left 20 support units behind...

_____________________________

I am the Holy Roman Emperor and am above grammar.

Sigismund of Luxemburg

(in reply to mussey)

Post #: 19

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 4:26:32 PM

glyphoglossus

Posts: 71
Joined: 9/26/2016
Status: offline

quote:

ORIGINAL: General Patton

....

In this case you round up and add together like so;

1280=2000
388=1000
388=1000
900=1000
total=5000+1000=6000

Sometimes add in a ship or two just for good measure.

This should help....GP

Got it!

Ok, this makes sense. Thanks for this. A useful heuristic to know what to allocate until intuition/experience takes over, and maybe even then!

(in reply to btd64)

Post #: 20

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 4:29:14 PM

glyphoglossus

Posts: 71
Joined: 9/26/2016
Status: offline

quote:

ORIGINAL: MakeeLearn

whether it's �Use Minimum Ships" or "Use All Ships" if looks like the AI may use the same algorithm when picking which ships to load a unit on.

121st USN Base Force 1280 / 2 = 640 troops, thus 2 900-troop capacity ships instead of a 900 capacity and a 500 capacity.

After picking it then loads according to the option selected.

This would explain the results!

Hmm, the AI needs some schooling in packing efficiency!

(in reply to MakeeLearn)

Post #: 21

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 5:22:37 PM

Bullwinkle58

Posts: 11302
Joined: 2/24/2009
Status: offline

quote:

ORIGINAL: Lokasenna

And yes, Bullwinkle, I've done this without leaving the Motorized Support behind.

I have too. My comment was in jest. But also, it doesn't always matter that much you leave the MS behind. It will catch up, and it often is the reason the initial wave has to hang around the beach an extra day.

More, I was commenting on the micro-management of the OP. In mid-1944 I'm forming 15-20 TF PER TURN. Trying to squeeze the last ton out of everything results in five hour turns.

< Message edited by Bullwinkle58 -- 11/5/2016 5:33:40 PM >

_____________________________

The Moose

(in reply to Lokasenna)

Post #: 22

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 5:30:41 PM

Bullwinkle58

Posts: 11302
Joined: 2/24/2009
Status: offline

quote:

ORIGINAL: glyphoglossus

quote:

ORIGINAL: Bullwinkle58

In a GC you're going to load thousands of TFs. Thousands.

I just allocate a plug of ships, load them, and add some if they need more space.

Then make a new one to load the left-behind Motorized Support. That always gets left behind.

Your way will lead to insanity.

I can imagine!

Yes, I can see this not working in the GC ...

Though it should be noted that I am not doing these calculations manually. What I am working on is actually a script that, given the available shipping and LCU's that need to be transported, comes up with a solution. This script hits the CSV files exported by WitPTracker for the ship availability, locations, capacities, etc. Then, based on the LCU's that need to be transported (specified by the user), it searches for an optimal configuration of ships for each TF, taking into account whether it is a commercial load or a combat load (the latter resulting in 80% reduction in capacities). The optimization algorithm is currently a brute force approach, and the optimization criteria is to minimize the total troop/load capacities while meeting the troop/load costs as minumum.

For small scenarios, this works out fine because the brute force approach is fast enough (trying out all the configurations) and it's not too big a deal for me to enter the LCU load costs myself (because WitPTracker does not report the LCU load costs).

Both of these issues will become objectionable on the scale of the GC.

However, a more efficient optimization algorithm is pretty easy to write and if I am lucky, by the time I am ready to take on the GC, maybe WitPTracker will report LCU load costs? Or I can figure out a way to extract them myself? Or, worst comes to worst, I'd have picked up a sufficient intuition to throw together TF's and hope for the best!

Well, I wish you luck. Not my cup of tea, but some people love to tumble the nums.

In six GCs to date, my experience is that hyper-optimization is fine for mega-port initial loads, such as WC to Pearl, but that as the game goes on and the island count goes into the hundreds, the constraint isn't efficient or inefficient loading, it's having the right mix of ships where they need to be, when they need to be, with fuel. Many times an LCU is ready to hop forward, but the sealift is twenty hexes away, maybe with the escorts repairing voyage damage. Optimizing the lift allocations is far more challenging, especially when you have to think about LBA coverage, LRCAP, likelihood of enemy surface raiding into the transit route, mines, subs, and reefs. The Allies normally have so much lift overhang that they don't have to be hyper-efficient. And few players have the lifespan to spend every turn to perform those thousands of clicks.

Every GC I've started I vow to have no LCU fragments. I'm still looking for that game.

_____________________________

The Moose

(in reply to glyphoglossus)

Post #: 23

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 5:39:29 PM

witpqs

Posts: 26087
Joined: 10/4/2004
From: Argleton
Status: offline

quote:

ORIGINAL: glyphoglossus

quote:

ORIGINAL: Bullwinkle58

In a GC you're going to load thousands of TFs. Thousands.

I just allocate a plug of ships, load them, and add some if they need more space.

Then make a new one to load the left-behind Motorized Support. That always gets left behind.

Your way will lead to insanity.

I can imagine!

Yes, I can see this not working in the GC ...

Though it should be noted that I am not doing these calculations manually. What I am working on is actually a script that, given the available shipping and LCU's that need to be transported, comes up with a solution. This script hits the CSV files exported by WitPTracker for the ship availability, locations, capacities, etc. Then, based on the LCU's that need to be transported (specified by the user), it searches for an optimal configuration of ships for each TF, taking into account whether it is a commercial load or a combat load (the latter resulting in 80% reduction in capacities). The optimization algorithm is currently a brute force approach, and the optimization criteria is to minimize the total troop/load capacities while meeting the troop/load costs as minumum.

For small scenarios, this works out fine because the brute force approach is fast enough (trying out all the configurations) and it's not too big a deal for me to enter the LCU load costs myself (because WitPTracker does not report the LCU load costs).

Both of these issues will become objectionable on the scale of the GC.

However, a more efficient optimization algorithm is pretty easy to write and if I am lucky, by the time I am ready to take on the GC, maybe WitPTracker will report LCU load costs? Or I can figure out a way to extract them myself? Or, worst comes to worst, I'd have picked up a sufficient intuition to throw together TF's and hope for the best!

I did something for amphibious convoys a few months ago. If you are working in Python 3 this might serve as a starting point or least a view of how I approached it.
https://drive.google.com/open?id=0ByVleYydeDVcQW5NVFc4V0g1Qkk

BTW, I have found the most reliable method is to create a separate TF for each LCU, then (if I wish) combine them after they are fully loaded. Yes it's a pain. It's the most reliable way.

And I use 30% overage for amphibious convoys rather than 25%. The tool I wrote allows you to specify what percentage you want. I do plan to make it better but I haven't touched it in a while.

EDIT: The Tracker guys have said they will report the load costs in the next round, but I don't know how long that will be. Tracker is huge and they are volunteers of course.

< Message edited by witpqs -- 11/5/2016 5:43:45 PM >

_____________________________

Intel Monkey: https://sites.google.com/view/staffmonkeys/home

(in reply to glyphoglossus)

Post #: 24

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/5/2016 11:37:08 PM

rustysi

Posts: 7472
Joined: 2/21/2012
From: LI, NY
Status: offline

quote:

Your way will lead to insanity.

Pfft, that ship has sailed long ago.

_____________________________

It is seldom that liberty of any kind is lost all at once. Hume

In every party there is one member who by his all-too-devout pronouncement of the party principles provokes the others to apostasy. Nietzsche

Cave ab homine unius libri. Ltn Prvb

(in reply to witpqs)

Post #: 25

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/6/2016 1:46:17 AM

glyphoglossus

Posts: 71
Joined: 9/26/2016
Status: offline

quote:

ORIGINAL: witpqs
I did something for amphibious convoys a few months ago. If you are working in Python 3 this might serve as a starting point or least a view of how I approached it.
https://drive.google.com/open?id=0ByVleYydeDVcQW5NVFc4V0g1Qkk

Ha! Great minds think alike! Mine is in Python 2/3 as well, though nowhere as complete as yours. Thanks for sharing yours. Will put up mine on GitHub (or, more likely, a gist) once it reaches a more reliable level (like, for e.g., an option to use the "round to the nearest 1000" algorithm of GP above!).

quote:

ORIGINAL: witpqs
BTW, I have found the most reliable method is to create a separate TF for each LCU, then (if I wish) combine them after they are fully loaded. Yes it's a pain. It's the most reliable way.

Yes, this seems to make most sense if we need to maximize efficiency of transports. Dominant opinion here seems to be not to sweat these details and just over-estimate. And that makes sense, coming from various voices of experience and all. And that might be the way I end up going in the GC. But in the mean time, the OCD geek in me really wants to get it right! Might come in handy if/when I play IJN?

quote:

ORIGINAL: witpqs
EDIT: The Tracker guys have said they will report the load costs in the next round, but I don't know how long that will be. Tracker is huge and they are volunteers of course.

THIS will be REALLY useful!!

(in reply to witpqs)

Post #: 26

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/6/2016 3:39:50 AM

geofflambert

Posts: 14863
Joined: 12/23/2010
From: St. Louis
Status: offline

I have sufficient ADD to have not been able to read all the posts on this thread (nay, almost none

) but if you are merely transporting an LCU from point A to point B you can order load troops only. I've never done that so I don't know how well that works. As far as being efficient, if you are preparing an amphibious assault, fuhgeddaboudit. Make sure to have more shipping than necessary. It really blows if you successfully land on an enemy controlled hex and run out of supplies. Additionally, when you spread out your troops across more platforms it will be less of a disaster if you lose a ship to torpedoes, mines, air attacks or coastal defense artillery. Splurge your butt off!

(in reply to glyphoglossus)

Post #: 27

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/6/2016 8:38:56 AM

BBfanboy

Posts: 18046
Joined: 8/4/2010
From: Winnipeg, MB
Status: offline

quote:

ORIGINAL: geofflambert

I have sufficient ADD to have not been able to read all the posts on this thread (nay, almost none ) but if you are merely transporting an LCU from point A to point B you can order load troops only. I've never done that so I don't know how well that works. As far as being efficient, if you are preparing an amphibious assault, fuhgeddaboudit. Make sure to have more shipping than necessary. It really blows if you successfully land on an enemy controlled hex and run out of supplies. Additionally, when you spread out your troops across more platforms it will be less of a disaster if you lose a ship to torpedoes, mines, air attacks or coastal defense artillery. Splurge your butt off!

"Load troops only" is good for those occasions when you want to remove some of the combat troops that just took an Atoll of 6000 stacking limit so that you can land engineers to build the base up. You generally do not want to take off the supply you just landed and the combat troops you remove will go to a well-stocked base on a larger land mass.

_____________________________

No matter how bad a situation is, you can always make it worse. - Chris Hadfield : An Astronaut's Guide To Life On Earth

(in reply to geofflambert)

Post #: 28

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/6/2016 4:20:08 PM

Alpha77

Posts: 2116
Joined: 9/24/2010
Status: offline

What I do is more trial and error method (I am already insane)

Look up troops ready to load, look which ships, check speed and range, select ships with same speed and enough range, try to load the troops I "estimated" would fit. Often they donot fit. I do not use the "add more ships" button (as it would add ships I do not want eg. slower or faster ones), I need to go back and de-select all loading. Quit the screen go to TF screen and manually add more ship(s). Then hopefully it fits. If not same again. Or if no ships available (of proper speed, range etc), leave the unit(s) behind and load later

Yes, this method is not recommended for important assault TF. You do not want to leave 90% of your main infantry div left in the port and the rest get trashed at the destination :)

Also I now found some organisation trick, normally I do not use much "planning for" small guards, snlf..for defense or better "speedbumbs". But I now see it helps, eg. I have 3 x nav guard at Truk, each shall go to a different island, I now set the "plan for" before the load, so can see which one should go to where in the load screen.

< Message edited by Alpha77 -- 11/6/2016 4:30:16 PM >

(in reply to BBfanboy)

Post #: 29

RE: Planning LCU transport: LCU-Ship Allocation Calculus - 11/23/2016 1:28:28 PM

Chris21wen

Posts: 6249
Joined: 1/17/2002
From: Cottesmore, Rutland
Status: offline

quote:

ORIGINAL: General Patton

I know this is going to sound weird but, I calculate like this;

For every 1000 of troop capacity needed I add 1000 troop cargo space, rounding up. So for the 121st USN BF you would need 2000 troop cargo space.

So, for your 4 LCU's you need 5000 troop cap. Plus I add 1000 for every 10K troop cap. Same for Equipment. For amphibian transport I add 2000 for every 10K. 95% of the time I don't have a problem.

I've used something similar but I do per unit.

In this example I want to load a US Inf Div (9045T 12761C)into an Amphib TF. This equates to 12K troop and 15K for cargo. So I create a TF of 11 APs (13250T, 23920C). All good.

At the dsame time I want to load a Corp HQ(2160T) which requires just 3K. This one has 3 APs (3000T, 2880C). Again all OK

Now the tricky part doing it together. The figures simply do not work? The pic shows what happened when I merged the two TFs with 14 APs (16250T, 26800C). Exactly the same combined figures as before.