Artillery Testing

[BigJ62]

1. How exactly are you calculating supply consumption, I only know one way and that is through using a custom logger and btw that is exactly how I deugged this code.
2. How much fuel and resources are there in the hex? You do realize there is industry in the hex.
3. You need to run several turns in an isolated area of the map such as Ceylon and remove all industry and to let dl simmer down because it starts the game at some random value.
4. Record starting supply requirements before any cbt as this represents the normal supply required with no shots fired. Shot count is reduced each turn if several things are in compliance plus randoms. Shot count increases the units supply requirement to as much as 2x so if a unit has 50 total devices and fires all 10 shots its supply req will be doubled for those devices which would make it 100. If you are the attacker this value can be increased further based on the number arty devices that actually fired so if you had 10 arty devices and fired all 10 shots and using the previous example your req would be 120. Keep in mind that the amount we muiltply by depends on the number of shots fired and is not displayed and effects the amount of supplies consumed.
5. It would be best to make a custom scenario with custom units that have the same weapons and the same numbers, same morale, same fatigue and disruption.

[pompack]

ORIGINAL: BigJ62

1. How exactly are you calculating supply consumption, I only know one way and that is through using a custom logger and btw that is exactly how I deugged this code.
2. How much fuel and resources are there in the hex? You do realize there is industry in the hex.3. You need to run several turns in an isolated area of the map such as Ceylon and remove all industry and to let dl simmer down because it starts the game at some random value.
4. Record starting supply requirements before any cbt as this represents the normal supply required with no shots fired. Shot count is reduced each turn if several things are in compliance plus randoms. Shot count increases the units supply requirement to as much as 2x so if a unit has 50 total devices and fires all 10 shots its supply req will be doubled for those devices which would make it 100. If you are the attacker this value can be increased further based on the number arty devices that actually fired so if you had 10 arty devices and fired all 10 shots and using the previous example your req would be 120. Keep in mind that the amount we muiltply by depends on the number of shots fired and is not displayed and effects the amount of supplies consumed.
5. It would be best to make a custom scenario with custom units that have the same weapons and the same numbers, same morale, same fatigue and disruption.

Doesn't industry fail to produce if there is an opponents army in the hex?

[BigJ62]

Not where HI and LI are concerned.

[pompack]

ORIGINAL: BigJ62

Not where HI and LI are concerned.

Thanks, didn't know that

[SqzMyLemon]

Shark7

Well done, I appreciate you taking the time to give us something concrete to sink our teeth into. I tend to agree with your suggestions on what could be changed to improve artillery effects. I don't think simply causing more casualites is the answer. I do think defending units should be suffering from negative disruption/fatigue/morale modifiers the longer a siege continues though, but the rate it is gained/recovered right now seems completely out of whack. Otherwise it appears that artillery bombardment actually helps your opponent in a siege situation.

Thanks again, and I hope the issue will be addressed quickly now, after your taking the initiative to see if there was indeed a problem by testing something.

[Shark7]

ORIGINAL: BigJ62

1. How exactly are you calculating supply consumption, I only know one way and that is through using a custom logger and btw that is exactly how I deugged this code.
2. How much fuel and resources are there in the hex? You do realize there is industry in the hex.
3. You need to run several turns in an isolated area of the map such as Ceylon and remove all industry and to let dl simmer down because it starts the game at some random value.
4. Record starting supply requirements before any cbt as this represents the normal supply required with no shots fired. Shot count is reduced each turn if several things are in compliance plus randoms. Shot count increases the units supply requirement to as much as 2x so if a unit has 50 total devices and fires all 10 shots its supply req will be doubled for those devices which would make it 100. If you are the attacker this value can be increased further based on the number arty devices that actually fired so if you had 10 arty devices and fired all 10 shots and using the previous example your req would be 120. Keep in mind that the amount we muiltply by depends on the number of shots fired and is not displayed and effects the amount of supplies consumed.
5. It would be best to make a custom scenario with custom units that have the same weapons and the same numbers, same morale, same fatigue and disruption.

1. Simple totals.
2. Industry is not supposed to work if the hex is under siege. If it is, then this is a bug as well. It also wouldn't matter since the supplies produced would have been the same in both cases, since no changes were made to anything other than the combat orders of the units in question. If supplies were being produced without combat, then the exact same amount would have been produced with combat. It would have had no effect on the end resulsts.
3. One can build a test scenario and run tests to either prove or disprove these results.
4. Supply required did not change.
5. Yes it would, however sometimes it is better to test in a real game setting. Also, sometimes it is not necessary to build such an isolated case. As I had stood down all aircraft on all sides, made no changes to the base at all, and was not allowing naval units to interfere, the units in question were just about as isolated as they could be without building a purpose built scenario.

As we see comparing the 2 results, without bombard the base used an average of 60 supplies per turn. With bombard the average was around 45. With everything else being the same, that is a 25% reduction with the bombardment taking place. I'd have only a 300 day siege if I sat there and never fired a shot, versus 400 by bombarding.

Just remember if you are Allied and not taking damage from Japanese artillery, when you go on the attack you'll face the same results for your artillery units. With japanese held islands like Guam, Saipan and Iwo Jima able to hold out against your attack for months or years when historically they generally fell in a matter of weeks. Inland bases out of the reach of your BBs will be even worse.

[treespider]

ORIGINAL: BigJ62

1. How exactly are you calculating supply consumption, I only know one way and that is through using a custom logger and btw that is exactly how I deugged this code.

He looked at the total in the hex the previous turn and compared to what was in the hex the following turn.

2. How much fuel and resources are there in the hex? You do realize there is industry in the hex.
3. You need to run several turns in an isolated area of the map such as Ceylon and remove all industry and to let dl simmer down because it starts the game at some random value.

Shouldn't matter as there were two tests run -in the same hex. The non combat test actually consumed more supply than the combat test.

4. Record starting supply requirements before any cbt as this represents the normal supply required with no shots fired. Shot count is reduced each turn if several things are in compliance plus randoms. Shot count increases the units supply requirement to as much as 2x so if a unit has 50 total devices and fires all 10 shots its supply req will be doubled for those devices which would make it 100. If you are the attacker this value can be increased further based on the number arty devices that actually fired so if you had 10 arty devices and fired all 10 shots and using the previous example your req would be 120. Keep in mind that the amount we muiltply by depends on the number of shots fired and is not displayed and effects the amount of supplies consumed.
5. It would be best to make a custom scenario with custom units that have the same weapons and the same numbers, same morale, same fatigue and disruption.

How do you explain the following:

1st test - Seven days - Opposing sides in the same hex - No combat

Supply on hand Day 1 = 18955.
Supply on hand Day 2 = 18892

Therefore the diff = Supply consumed = 63

Supply on hand Day 3 = 18829

Therefore the diff btw Day 2 and Day 3 = Supply consumed = 63

etc etc

During the Non-combat test during those seven days the defender consumed 60 supply points on average...360 points and total...single day high = 70 ....single day low = 36.

Daily differences: 63,63,61,36,70,67





2nd test - Seven days - Opposing sides in same hex. Attacker bombards with Artillery

Supply on hand Day 1 = 18936
Supply on hand Day 2 = 18898

therefore the difference = Supply consumed = 38

Supply on hand Day 3 = 18854

Therefore the diff btw Day 2 and Day 3 = Supply consumed = 44

etc etc

During the combat test the defenders supply cunsumption actually decreased by about 15-20 points per day... 259 total ...and 44 per day on average...

Through 5 turns - Daily Differences : 38, 44, 43, 50, 43, 41

EDIT: Interestingly the Single Day Low for Test 1 corresponds to the Single Day high in Test 2....makes me wonder if a divisor is multiplying or a multiplier divising...or it could be random.[:)]


Its the same hex in both cases - only difference - in one test the defender consumed less supply - and that is when the defender was attacked.



The only alternative explanation would be that the supply required increased during the attacks and the industry actually produced MORE supply when the attacker attacked the defender. And when there was no combat the industry produced LESS supply...so with that explanation Industry varies its production?

[BigJ62]

Not all industry functions and btw if is bug it has been one for a very long time.
When you first start a game not all units have exactly the right amount of supply some have less and some have too much.
If industry is mitidating supply totals then how do you know exactly how much supply is being consumed, simple totals will not work.
Shot counts can vary from turn to turn due to randoms, so once again how do you know how much supply each unit is consummoing - answer is you don't.
Who said the defender is supposed to use more supplies, where did that come from?

[Shark7]

ORIGINAL: treespider

ORIGINAL: BigJ62

1. How exactly are you calculating supply consumption, I only know one way and that is through using a custom logger and btw that is exactly how I deugged this code.

He looked at the total in the hex the previous turn and compared to what was in the hex the following turn.

2. How much fuel and resources are there in the hex? You do realize there is industry in the hex.
3. You need to run several turns in an isolated area of the map such as Ceylon and remove all industry and to let dl simmer down because it starts the game at some random value.

Shouldn't matter as there were two tests run -in the same hex. The non combat test actually consumed more supply than the combat test.

4. Record starting supply requirements before any cbt as this represents the normal supply required with no shots fired. Shot count is reduced each turn if several things are in compliance plus randoms. Shot count increases the units supply requirement to as much as 2x so if a unit has 50 total devices and fires all 10 shots its supply req will be doubled for those devices which would make it 100. If you are the attacker this value can be increased further based on the number arty devices that actually fired so if you had 10 arty devices and fired all 10 shots and using the previous example your req would be 120. Keep in mind that the amount we muiltply by depends on the number of shots fired and is not displayed and effects the amount of supplies consumed.
5. It would be best to make a custom scenario with custom units that have the same weapons and the same numbers, same morale, same fatigue and disruption.

How do you explain the following:

1st test - Seven days - Opposing sides in the same hex - No combat

Supply on hand Day 1 = 18955.
Supply on hand Day 2 = 18892

Therefore the diff = Supply consumed = 63

Supply on hand Day 3 = 18829

Therefore the diff btw Day 2 and Day 3 = Supply consumed = 63

etc etc

During the Non-combat test during those seven days the defender consumed 60 supply points on average...360 points and total...single day high = 70 ....single day low = 36.

Daily differences: 63,63,61,36,70,67





2nd test - Seven days - Opposing sides in same hex. Attacker bombards with Artillery

Supply on hand Day 1 = 18936
Supply on hand Day 2 = 18898

therefore the difference = Supply consumed = 38

Supply on hand Day 3 = 18854

Therefore the diff btw Day 2 and Day 3 = Supply consumed = 44

etc etc

During the combat test the defenders supply cunsumption actually decreased by about 15-20 points per day... 259 total ...and 44 per day on average...

Through 5 turns - Daily Differences : 38, 44, 43, 50, 43, 41

EDIT: Interestingly the Single Day Low for Test 1 corresponds to the Single Day high in Test 2....makes me wonder if a divisor is multiplying or a multiplier divising...or it could be random.[:)]


Its the same hex in both cases - only difference - in one test the defender consumed less supply - and that is when the defender was attacked.



The only alternative explanation would be that the supply required increased during the attacks and the industry actually produced MORE supply when the attacker attacked the defender. And when there was no combat the industry produced LESS supply...so with that explanation Industry varies its production?

Also, the game was saved the day before the test run starts. I then reloaded that turn to make sure that both tests were run over the same 7 days. this is not 14 consecutive days, this is 7 days run once with no combat, then the same 7 days run with bombardment. [;)]

[BigJ62]

How did the japanese come to be in HK?

[treespider]

ORIGINAL: BigJ62

Not all industry functions and btw if is bug it has been one for a very long time.

Maybe it has ...but a bug is a bug whether its Baby New Year or Methusalah.

When you first start a game not all units have exactly the right amount of supply some have less and some have too much.

Which is why I suggested he run the test over 7 days.

If industry is mitidating supply totals then how do you know exactly how much supply is being consumed, simple totals will not work.

So in the non combat tests he ended up with LESS total supply in the hex than in the combat test. In your theory then Industry actually produced more in one test and less in the other.... I though it was supposed to be constant.

Shot counts can vary from turn to turn due to randoms, so once again how do you know how much supply each unit is consummoing - answer is you don't.

But you can look at how much you had the previous day and how much you have now...if industry is constant then the difference is how much was consumed.

Who said the defender is supposed to use more supplies, where did that come from?

In Witp Mk I and in early AE testing I did extensive testing of supply consumed by an LCUs when under attack by air units. In WitP units that were attacked consumed more supplies than those that were not.

[Shark7]

ORIGINAL: BigJ62

Not all industry functions and btw if is bug it has been one for a very long time.
When you first start a game not all units have exactly the right amount of supply some have less and some have too much.
If industry is mitidating supply totals then how do you know exactly how much supply is being consumed, simple totals will not work.
Shot counts can vary from turn to turn due to randoms, so once again how do you know how much supply each unit is consummoing - answer is you don't.
Who said the defender is supposed to use more supplies, where did that come from?

If you do not believe my results, there is a very simple solution. Run your own tests. [:)]

It's not worth it for me to argue with people over it. I hid nothing, gave you all the information regarding how I did it, and shared the numbers so you can crunch them yourselves. If you don't believe the results, nothing I do can make you believe them.

I was not testing what happened to the attacker, nor was I overly concerned with what happened to the defending units, as they are able to repair at the end of each turn anyway. There is no way to prove or disprove if the combat report is 100% accurate due to that. What I was looking for was the amount of supply used by units not in combat over 7 days versus the amount of supply used by those very same units being bombarded over the same 7 days. The results are what they are; with bombardment they used less supply than just sitting there with no combat.

[Shark7]

ORIGINAL: BigJ62

How did the japanese come to be in HK?

Because the test starts on day 3 of the game and they marched in. I then ran 7 turns (to turn 10) with the units on defend. Then I reloaded turn 3 (which was saved at the beginning of the Japanese turn with the Japanese units in HK) and this time I turned the 6 artillery units on to bombard and ran 7 more turns up to turn 10.

[BigJ62]

nm

[treespider]

ORIGINAL: BigJ62

nm

38th Division
66th Infantry Regiment
19th Ind. Engineer Regiment
20th Ind. Engineer Regiment
2nd RF Gun Battalion
10th Ind. Mountain Gun Regiment
3rd Ind.Hvy.Art. Battalion
5th RF Gun Battalion
2nd Mortar Battalion
2nd Ind.Hvy.Art. Battalion
1st Hvy.Artillery Regiment
20th Ind. Mtn Gun Battalion

And if I understand the only units that bombarded were the bolded ones...

they march to HK on turns 1 and 2 ...

Tests run Turn 3 through 10

[Shark7]

ORIGINAL: treespider

ORIGINAL: BigJ62

nm

38th Division
66th Infantry Regiment
19th Ind. Engineer Regiment
20th Ind. Engineer Regiment
2nd RF Gun Battalion
10th Ind. Mountain Gun Regiment
3rd Ind.Hvy.Art. Battalion
5th RF Gun Battalion
2nd Mortar Battalion
2nd Ind.Hvy.Art. Battalion
1st Hvy.Artillery Regiment
20th Ind. Mtn Gun Battalion

And if I understand the only units that bombarded were the bolded ones...

they march to HK on turns 1 and 2 ...

Tests run Turn 3 through 10

You are exactly right treespider.

Also the units in red I set to reserve. The units in blue (Rapid Fire Gun Bns...IE Anti-Tank guns) are not capable of bombarding.

[USSAmerica]

I can only contribute one tidbit to this discussion, and maybe put one small question involved to bed.  The question of industry producing with enemy troops in the hex seems to be the same as it was described to me in WitP several years ago.  At that time I was told that Oil and Resource industries would stop producing raw materials with enemy troops in the hex, but HI and Manpower would continue to produce their output points, if there is still a supply of input requirements for them to run.  Since HI and LI in AE are "factories" and not producing raw materials, like Oil or Resource, then they are probably running off of stockpiled fuel and resources in the hex during the test.  I don't think this point is a bug.

[treespider]

ORIGINAL: USS America

I can only contribute one tidbit to this discussion, and maybe put one small question involved to bed.  The question of industry producing with enemy troops in the hex seems to be the same as it was described to me in WitP several years ago.  At that time I was told that Oil and Resource industries would stop producing raw materials with enemy troops in the hex, but HI and Manpower would continue to produce their output points, if there is still a supply of input requirements for them to run.  Since HI and LI in AE are "factories" and not producing raw materials, like Oil or Resource, then they are probably running off of stockpiled fuel and resources in the hex during the test.  I don't think this point is a bug.

But as to this test - the industry production question is not really at issue....unless its not working as well.

In test 1 with no combat:
The HK hex had a total of 18,955 supply points present in units and the base at the start. By the end of the seven day cycle there was only 18595 present - a drop of 360 points. So if the industry did produce here, it still only produced enough that would result in a 360 point shortfall.

In test 2 with combat:
The HK hex starts with 18955 and ends up with 18677 points a drop of 278 points.

For arguments sake - in test 2 if the units consumed more supply while under attack than at rest the industry must have produced more supply as well...for arguments sake lets say the attacked units consumed 800 points during the test...(18955-800 = 18155)... However there are 18677 present at the end of the test so the industry must have produced 18677-18155 = 522 points....assuming the attacked units consumed 800 points

So continuing - in test 1 the units were not attacked so theoretically should have consumed less supply...for this discussion we'll say 400 points... (18955-400 = 18555)...however there are 18595 present at the end of the seven day cycle...so in this example the same industry only produce 18595-18555 = 40 points...assuming the un-attacked units only consumed 400 points.


I would think that the industry should produce equal amounts in both tests.

[Mynok]

Interesting.

One point: I believe you have to add supply in the units to supply at the base to get the total supply for the defender. Certainly I'm prepared to be corrected on that, but I think that's kinda what BigB is driving at.

[Shark7]

ORIGINAL: Mynok


Interesting.

One point: I believe you have to add supply in the units to supply at the base to get the total supply for the defender. Certainly I'm prepared to be corrected on that, but I think that's kinda what BigB is driving at.

That is exactly how I did it. Took the supply at the base added to the supply level in each unit to get the Total Supplies # provided.

So yes, the supply in the units is accounted for. You will note that each report has a Hong Kong Supply XXXXXXXX level and a Total Supplies YYYYYYYY level, and a comparison of this turn versus the turn before.

So the total supply number for each turn is every last bullet accounted for. [:)]