Resupply Issue

[Dave 'Arjuna' O'Connor]

I've now added the moon phase to the weather popup, so you can now at least see what it is. I've also reviewed the code that determines the effects on visibility due to moon phase. In essence, the moon phases are wrapped up inside the Visibility Night Modifiers, which can be set in the MapMaker. Currently these default to"

Daylight - 100
Dawn/Dusk - 75
Full Moon - 30
3/4 Moon - 25
Half Moon - 20
1/4 Moon - 15
New Moon - 10

Given these apply to a default 10,000m visibility range during daylight, I think we should reduce these to:
Daylight - 100
Dawn/Dusk - 50
Full Moon - 20
3/4 Moon - 15
Half Moon - 10
1/4 Moon - 05
New Moon - 02

What do you think?

 

[simovitch]

Does "New Moon = 02" mean the visibility will be 10,000*.02=200m? if so I agree with your latter suggestions. It would be tough to see anything beyond 200m in complete darkness.

Also, what about scenarios that are 10 turns long? that would mean a change from new moon at start to just past 1/4 moon...

 

[Dave 'Arjuna' O'Connor]

Yes it does mean 200m vis range. I still think the full moon vis range is too far even at 2,000m. Maybe the values should be reduced to:
Daylight - 100 - ie 10,000m
Dawn/Dusk - 50 - ie 5,000m
Full Moon - 10 - ie 1,000m
3/4 Moon - 08 - ie 800m
Half Moon - 06 - ie 600m
1/4 Moon - 04 - ie 400m
New Moon - 02 - ie 200m

 

[Dave 'Arjuna' O'Connor]

Implemented the DetermineMoonPhase function today and tested it against the historical records. I'm happy with that.

 

[Dave 'Arjuna' O'Connor]

I tracked down a 1964 study conducted by the U.S. Army Training Center Human Research Unit Presidio ~f Monterey, California into MOONLIGHT AND NIGHT VISIBILITY.
http://www.dtic.mil/dtic/tr/fulltext/u2/438001.pdf

This was a pretty good study. It focussed on four experiments/studies conducted by different organisations over the preceding thirty years. One was by the Japanese Army prior to WW2. Another was Canadian and two American. The good thing about them was they were focussed on detection, not just perception. Here's what I gleaned:

• Twilight detection ranges average around 400m
• LNS or light of the night sky under no moon detection range is around the 40m mark
• Full Moon is around 200m for detecting an infantry platoon in level grassy ground. It's 300m if they are silhouetted.
• So our table should look like:
• Daylight - 100 - ie 10,000m
  Dawn/Dusk - 04 - ie 400m
  Full Moon - 03 - ie 300m
  3/4 Moon - 02 - ie 200m
  Half Moon - 01 - ie 100m
  1/4 Moon - 01 - ie 100m
  New Moon - 01 - ie 100m

We run into a fidelity issue caused by our terrain grid size of 100m. Hence why there is no effective difference between a New Moon and a half moon.

Unless someone can come up with some more data, I'm going to run with these values for the conversion of the maps.

 

[simovitch]

design by committee be damned - run it!

How about the supply interdiction bug?

 

[Dave 'Arjuna' O'Connor]

I have just run it after modifying the vis effects in the ResupplyTestMap. Here's the APerFP Influence overlay with the new values. Note the moon phase is a half moon, so a 0.01 modifier applies or equivalent to 100m. The only concern I have is that I'm not sure the influence calculator is factoring in the unit radius correctly. The German Inf Coy at the crossroads has a unit radius of 150m, which is about right. Alas, this rounds down to 100m when factoring in the grid size. The net effect is that it really only influences the area it occupies. I'm thinking we should perhaps round up rather than down for the unit radius.


Anyway, now that I've sorted the moon phase effects I'm going to test to ensure the resupply column between the corps base and the brigade base is now not interdicted. Then I'll check to see if all my other code changes ensure that the resupply column between the brigade base and the 40th Arm Bn down south gets interdicted. We're getting there.

Actually, I stepped through the code again and can confirm the unit radius for the German unit was in fact 77m. So I got that wrong and the influence map looks good.

 

[Dave 'Arjuna' O'Connor]

OK, I've made the code and data changes and run through the supply test scenario. This time the hapless A Coy 40 Arm Bn comes up short. This screenshot shows the unit at the start of the scenario.

Note the supply line status defaults to open at the start. A short while later it requests a resupply.



Note its supply route is drawn in green at this stage. Then when the resupply event occurs it suffers casualties. In this run through it only lost one vehicle before it decided to abort. In the previous run through it lost eight before it aborted. There is a 20% chance of aborting after each hit. So on average a large column like this was ( 20 trucks ) will abort after three hits.

Note the report in the unit's log and on the battlemap. A short while later A Coy requests another resup. But then the base suspends its resupply.


That seems to work now. I'm happy. But this is only a simple test. We need to keep a close eye on the resupply situation as we test.

 

[Dave 'Arjuna' O'Connor]

Jim,

I haven't tested your saved game but I suspect it will be addressed with these latest changes. Trouble is the maps will have to be converted. I will do this for the new build. You will then need to run a new scenario from scratch and test to see if the same condition arises.