Artillery Death Stars Post Patch Two Hot Fix

This new stand alone release based on the legendary War in the Pacific from 2 by 3 Games adds significant improvements and changes to enhance game play, improve realism, and increase historical accuracy. With dozens of new features, new art, and engine improvements, War in the Pacific: Admiral's Edition brings you the most realistic and immersive WWII Pacific Theater wargame ever!

Moderators: wdolson, MOD_War-in-the-Pacific-Admirals-Edition

jackyo123
Posts: 703
Joined: Mon Feb 04, 2008 8:51 pm

RE: More information needed...

Post by jackyo123 »

ORIGINAL: Canoerebel

I updated those figures in Post # 185 - Over four days of bombardment, the Allies at Akyab suffered 4,000 casualites, 40 infantry squads destroyed, 240 infantry squads disrupted, 80 non-combat squads destroyed, 280 non-combat squads disrupted, 24 guns destroyed, and 24 vehicles destroyed. So, the average per day was roughly 1,000; 10/60; 20/70; 6 and 6. This is roughly the same as the losses suffered at Changteh in a deliberate attack although the force engaged at Chengtah was considerably larger.

As noted previously, you would expect a pitched battle to be far more bloody than an artillery bombardment, but that wasn't the case with this particular comparison and I don't think players will find that to be the case in their games. Artillery is simply too powerful in the game.

ckammp, you've made it clear that you don't think there is anything wrong with artillery. Let's step aside and allow other players to weigh in with the results from their own games. A large sampling of data from a variety of ongoing games should shed more light on the issue.



Unfortunately Canoerebel, artillery seems only to be broken when used *en masse* such as the examples above. My island hopping campaigns, where 10k troops face off against 3 or 4k troops and less than 100 guns, look pretty good from an artillery standpoint. Maybe 30 or 40 casualties (though still WAY too many non-infantryman casualties - casualties need to be skewed more towards 'the sharp end' i.e the combat rifleman - and away from the rear area support troops.)

Fixing it for the mass death star attacks might break it for the island campaigns. I think supply requirements for bombardments need to be increased even more, and need to be skewed more towards DISRUPTION versus DESTRUCTION. And perhaps there needs to be strict stacking limits about guns in a hex.

Not sure what they can do without breaking the balance of the islands. My suggestion is to reduce, by a large amount, the victory points gained by taking cities in china, by reducing artillery effectiveness in jungle, mountain, and fortified hexes, and slowing down movement rates of artillery units. Movement of artillery units in jungle terrain should be EXTREMELY difficult; likewise for mountain terrain. Note how the Japanese in Maruyama (sp?)s attack in Guadalcanal - on the way to their jump-off points, they had to navigate the jungle to approach Henderson from the south. The infantrymen all carried a part of a disbanded artillery piece. On the way, almost all the heavy pieces were left behind on the trail. They were too difficult to move.
My favorite chinese restaurant in Manhattan -
http://www.mrchow.com

The best computer support firm in NYC:
http://www.thelcogroup.com

Coolest internet toolbar:
http://www.stumbleupon.com
User avatar
Sardaukar
Posts: 12513
Joined: Wed Nov 28, 2001 10:00 am
Location: Finland/Israel

RE: More information needed...

Post by Sardaukar »

I think also in this issue old saying from WitP forum holds true: "Ahistorical things cause ahistorical results". [8D]
"To meaningless French Idealism, Liberty, Fraternity and Equality...we answer with German Realism, Infantry, Cavalry and Artillery" -Prince von Bülov, 1870-

Image
User avatar
JWE
Posts: 5039
Joined: Tue Jul 19, 2005 5:02 pm

RE: More information needed...

Post by JWE »

jackyo123, you are exactly correct. The effects of artillery are applied across-the-board, and when used in appropriate ways and amounts, work just fine. But the algorithm is, of course, open ended, so things like 'Death Stars' are possible. If artillery effect is reduced to accommodate Death Stars, then effect will go to squat for simple, normal, things, because it an across-the-board proposition.

It is possible to program arbitrary limits onto artillery; total # that can shoot, etc.. But one would also have to determine the mix - how many 155s? how many 105s?, 75s, howitzers, or guns?, mortars? Implementing such a program limit is not an insignificant task. And then just whose limits will be used? If 7 is chosen, what about all those angry people, with historical results in hand, that demand 8, or 9, or even 6.

And what about the break point? Below the break, a player could pop 99 155 guns, but above the break, they might only get 50, along with 50 more 105s, 75s, mortars, etc, depending on how the limits are coded.

No, don't think that solution is the sharpest pencil in the box. Severely limits game flexibility in those nominal areas where it works just fine.

Best answer is a house rule for PBEM. Everybody knows about Death Stars, so a negotiated limit is a rational thing to do. It is better than an arbitrary hard code limit, because different players will negotiate to different levels - code limit is one-size-fits-all and take-it-or-leave-it proposition - will likely tick-off as many people as the present system.

I guess, from my perspective, I will have to plagiarize (perhaps paraphrase) ckammp and Sardaukar; arty works just fine if used as it was used, but gamey things cause gamey results.
User avatar
Canoerebel
Posts: 21099
Joined: Fri Dec 13, 2002 11:21 pm
Location: Northwestern Georgia, USA
Contact:

RE: More information needed...

Post by Canoerebel »

JWE, that's interesting insight and I wasn't aware of the dilemma.  If it can't be fixed it can't be fixed, but I hope the developers will try to come up with good fixes to problems whenever possible rather than relying on "just do a house rule."  House rules can be ticklish:
 
1)  Newer players may not be familiar with the problem.  It's tough on players to play deeply into a game, and to base months and months of strategy and movement, only then to find out that a key invasion/offensive/whatever will be challenged as "gamey."
 
2)  Many players - and I'm in this group - don't know enough about the war to know when many things are gamey or not.  With complete innocense and in the utmost good faith I've done things only to have an opponent say, "You shouldn't have done that; it couldn't be done; it was never done," etc.
 
Players are devoting hundreds of hours to the game and are justifiably anxious to see flaws addressed if humanly possible.  
 
"Rats set fire to Mr. Cooper’s store in Fort Valley. No damage done." Columbus (Ga) Enquirer-Sun, October 2, 1880.
herwin
Posts: 6047
Joined: Thu May 27, 2004 9:20 pm
Location: Sunderland, UK
Contact:

RE: More information needed...

Post by herwin »

ORIGINAL: Sardaukar

I think also in this issue old saying from WitP forum holds true: "Ahistorical things cause ahistorical results". [8D]

Use a cookie cutter model--that takes into account diminishing returns.
Harry Erwin
"For a number to make sense in the game, someone has to calibrate it and program code. There are too many significant numbers that behave non-linearly to expect that. It's just a game. Enjoy it." herwin@btinternet.com
jackyo123
Posts: 703
Joined: Mon Feb 04, 2008 8:51 pm

RE: More information needed...

Post by jackyo123 »

ORIGINAL: JWE

jackyo123, you are exactly correct. The effects of artillery are applied across-the-board, and when used in appropriate ways and amounts, work just fine. But the algorithm is, of course, open ended, so things like 'Death Stars' are possible. If artillery effect is reduced to accommodate Death Stars, then effect will go to squat for simple, normal, things, because it an across-the-board proposition.

It is possible to program arbitrary limits onto artillery; total # that can shoot, etc.. But one would also have to determine the mix - how many 155s? how many 105s?, 75s, howitzers, or guns?, mortars? Implementing such a program limit is not an insignificant task. And then just whose limits will be used? If 7 is chosen, what about all those angry people, with historical results in hand, that demand 8, or 9, or even 6.

And what about the break point? Below the break, a player could pop 99 155 guns, but above the break, they might only get 50, along with 50 more 105s, 75s, mortars, etc, depending on how the limits are coded.

No, don't think that solution is the sharpest pencil in the box. Severely limits game flexibility in those nominal areas where it works just fine.

Best answer is a house rule for PBEM. Everybody knows about Death Stars, so a negotiated limit is a rational thing to do. It is better than an arbitrary hard code limit, because different players will negotiate to different levels - code limit is one-size-fits-all and take-it-or-leave-it proposition - will likely tick-off as many people as the present system.

I guess, from my perspective, I will have to plagiarize (perhaps paraphrase) ckammp and Sardaukar; arty works just fine if used as it was used, but gamey things cause gamey results.


Agree completely. Some people are totally against house rules for maybe 'purist' reasons, but I think that's not wise; a game cannot be perfect, nor can it completely model history without being a totally rigid system without room for 'small' deviations. A house rule is simply nothing more than an agreement not to 'cheat' or 'game' the system - if you do NOT want a house rule, then you have to expect that there will be people who do indeed 'game' the system and figure out broken exploits. But to expect that all possible loopholes be closed is, IMHO, unrealistic. It's not possible to close all loopholes, i.e backdoors, in the most sophisticated programs costing tens or hundreds of millions of dollars (like banking systems) - so why is it expected that it could be done in a game costing a fraction of that?

My favorite chinese restaurant in Manhattan -
http://www.mrchow.com

The best computer support firm in NYC:
http://www.thelcogroup.com

Coolest internet toolbar:
http://www.stumbleupon.com
jimh009
Posts: 368
Joined: Sun May 15, 2005 10:54 am
Contact:

RE: More information needed...

Post by jimh009 »

ORIGINAL: herwin

ORIGINAL: Sardaukar

I think also in this issue old saying from WitP forum holds true: "Ahistorical things cause ahistorical results". [8D]

Use a cookie cutter model--that takes into account diminishing returns.

That seems the easiest way to prevent the Death Star problem. I agree that artillery, when not massed, seems quite historical. And that any "fix" to the Death Star problem might have unintended consequences everywhere else. By either having "diminishing returns" for "x" artillery units above "x" number in a hex - or by simply having a cut-off for the number of artillery guns in a hex that will actually fire - might be an easy way to reduce the death star problem.
bradfordkay
Posts: 8596
Joined: Sun Mar 24, 2002 8:39 am
Location: Olympia, WA

RE: More information needed...

Post by bradfordkay »

Canoerebel wrote: "Many players - and I'm in this group - don't know enough about the war to know when many things are gamey or not."

Dan... you should have paid more attention in Dr Ziemke's class!    [:-]



Just kidding... the level of knowledge exhibited on this forum goes way beyond that achieved in a upper level collegiate course - even one dedicated to WW2.
fair winds,
Brad
User avatar
Sardaukar
Posts: 12513
Joined: Wed Nov 28, 2001 10:00 am
Location: Finland/Israel

RE: More information needed...

Post by Sardaukar »

ORIGINAL: herwin

ORIGINAL: Sardaukar

I think also in this issue old saying from WitP forum holds true: "Ahistorical things cause ahistorical results". [8D]

Use a cookie cutter model--that takes into account diminishing returns.

Diminishing returns and increasing supply use would be quite good idea, problem is most likely implementation.

I think increasing the effect of fortifications and some terrain against artillery bombardments even more would, IMHO, do the trick. Through WWII, entrenched troops remained very resilient against artillery.

I'd drastically tone down casualties from even massed artillery bombardments and instead make effect more as unit DISRUPTION. This would accurately portrait the use of offensive artillery in WWII. While effect would be still considerable, you could not pulverize enemy with only artillery. You'd still have to deliberately attack the enemy with your troops to occupy the position. This would both allow historical use of massed artillery (albeit most of this happened in Europe/NA) and force the attacker to actually assault the position instead of destroying it with artillery (that didn't work even in WW I), because, while disrupted, enemy troops would still be there.

This would allow use of "artillery death star" for those who want to do it and allow more historical outcome from massive offensive artillery bombardments. This combined with drastically increased supply consumption by artillery set to Bombard would probably be more satisfactory solution than diminishing returns solution.

To add, I have nothing against troops in clear terrain without entrenchment to be mauled by artillery. That's what artillery is for and why entrenchments were invented in first place (not the only reason, but main one). Fire kills.

This was the main reason for example why Finns were able to stop Soviet onslaught during summer 1944 battles (Tali-Ihantala, Vuosalmi etc.). Most were won by advanced and massed use of accurate artillery fire against attacking enemy. Those battles were size of El Alamein, but rarely known outside dedicated military history circles.
"To meaningless French Idealism, Liberty, Fraternity and Equality...we answer with German Realism, Infantry, Cavalry and Artillery" -Prince von Bülov, 1870-

Image
herwin
Posts: 6047
Joined: Thu May 27, 2004 9:20 pm
Location: Sunderland, UK
Contact:

RE: More information needed...

Post by herwin »

Implementation of a cookie-cutter model is simple. Each shell has a lethal area, LAshell. Each shell takes out LAshell/BombardedArea, so the probability that a point target will survive the shell is 1.0-LAshell/BombardedArea. The probability that a target unit will survive the bombardment is the product of the probabilities that it will survive each shell, conservatively assuming independence. Group the shells based on lethal area. The probability of survival is the product over the groups of the probability that the unit will survive each group. The probability that the unit will survive a group is (1.0-LAshell/BombardedArea) to a power corresponding to the number of shells in the group. Take logs and compute the log of the overall survival probability, then exponentiate. Finally draw a U[0,1] random number and see whether the unit survives. Move to the next unit. A lot of the component probabilities need only be calculated once for a whole lot of units. You can even get by with expected values rather than testing for each unit. Supply is computed from the ammo expenditure. You quickly get diminishing returns.
Harry Erwin
"For a number to make sense in the game, someone has to calibrate it and program code. There are too many significant numbers that behave non-linearly to expect that. It's just a game. Enjoy it." herwin@btinternet.com
User avatar
Sardaukar
Posts: 12513
Joined: Wed Nov 28, 2001 10:00 am
Location: Finland/Israel

RE: More information needed...

Post by Sardaukar »

ORIGINAL: herwin

Implementation of a cookie-cutter model is simple. Each shell has a lethal area, LAshell. Each shell takes out LAshell/BombardedArea, so the probability that a point target will survive the shell is 1.0-LAshell/BombardedArea. The probability that a target unit will survive the bombardment is the product of the probabilities that it will survive each shell, conservatively assuming independence. Group the shells based on lethal area. The probability of survival is the product over the groups of the probability that the unit will survive each group. The probability that the unit will survive a group is (1.0-LAshell/BombardedArea) to a power corresponding to the number of shells in the group. Take logs and compute the log of the overall survival probability, then exponentiate. Finally draw a U[0,1] random number and see whether the unit survives. Move to the next unit. A lot of the component probabilities need only be calculated once for a whole lot of units. You can even get by with expected values rather than testing for each unit. Supply is computed from the ammo expenditure. You quickly get diminishing returns.

While I somewhat agree, I did mean implementation into existing program code would be a beast. [;)] And I think your example is very "inflexible". [8D]
"To meaningless French Idealism, Liberty, Fraternity and Equality...we answer with German Realism, Infantry, Cavalry and Artillery" -Prince von Bülov, 1870-

Image
User avatar
Grotius
Posts: 5842
Joined: Fri Oct 18, 2002 5:34 pm
Location: The Imperial Palace.

RE: More information needed...

Post by Grotius »

So I'm about to start a PBEM, and we're wondering what a fair house rule is to cover the arty issue. Do you all have any suggestions?

Also, more generally, is a "quiet China" house rule a good idea? Is it feasible? (Maybe this question deserves its own thread?)
Image
User avatar
Canoerebel
Posts: 21099
Joined: Fri Dec 13, 2002 11:21 pm
Location: Northwestern Georgia, USA
Contact:

RE: More information needed...

Post by Canoerebel »

Thanks to the good work of the developers, artillery is no longer a problem with respect to fortified troops (IE, bases that have fortifiications). I don't *think* you need a house rule in that regards, though I would caution players to adjust to the situation if they inundate a hex with artillery and it does wildly horrible things.

I think artillery is still too effective against unintrenched troops. I would suggest something like this: No more than one artillery unit per 25,000 troops in a hex (but with a minimum of two permitted - IE, a player can always employ up to two units even if he has less than 50,000 troops in a hex).
"Rats set fire to Mr. Cooper’s store in Fort Valley. No damage done." Columbus (Ga) Enquirer-Sun, October 2, 1880.
User avatar
ny59giants
Posts: 9889
Joined: Mon Jan 10, 2005 12:02 pm

RE: More information needed...

Post by ny59giants »

Also, more generally, is a "quiet China" house rule a good idea?

Take a look at your garrison requirements and it will be relatively quite on it's own.
[center]Image[/center]
ckammp
Posts: 756
Joined: Sat May 30, 2009 4:10 pm
Location: Rear Area training facility

RE: More information needed...

Post by ckammp »

I would suggest that no house rules are needed for artillery. The devs have done a great job addressing all known issues with the artillery combat model.
Keep in mind, the AE artillery model is based on historic artillery usage of Japan and the Allies. Use historic tactics, and you'll get historic results. Create a stack of 3+ divisions and 6+ artillery units, and you'll get a-historic results.

User avatar
Canoerebel
Posts: 21099
Joined: Fri Dec 13, 2002 11:21 pm
Location: Northwestern Georgia, USA
Contact:

RE: More information needed...

Post by Canoerebel »

That doesn't work since many players won't actually know whether the Japanese used one or five or fifteen artillery units in Burma (or wherever).  Also, different players will have different ideas as to what is historical and what is within the realm of possiblity and what is flat-out ludicrous.

The artillery model for troops in non-fortified hexes is borked against massed artillery (based upon my experience facing six Japanese units at Akyab - a non-fortified jungle [or is it forest?] hex).

It will be better for everybody if this problem is fixed. 
"Rats set fire to Mr. Cooper’s store in Fort Valley. No damage done." Columbus (Ga) Enquirer-Sun, October 2, 1880.
ckammp
Posts: 756
Joined: Sat May 30, 2009 4:10 pm
Location: Rear Area training facility

RE: More information needed...

Post by ckammp »

ORIGINAL: Canoerebel

That doesn't work since many players won't actually know whether the Japanese used one or five or fifteen artillery units in Burma (or wherever).  Also, different players will have different ideas as to what is historical and what is within the realm of possiblity and what is flat-out ludicrous.

The artillery model for troops in non-fortified hexes is borked against massed artillery (based upon my experience facing six Japanese units at Akyab - a non-fortified jungle [or is it forest?] hex).

It will be better for everybody if this problem is fixed. 

The artillery model is broken, just because YOU say it is?
No other opinion matters?
Everyone who hasn't experienced the same results you have is wrong?

How many devs (not me, but devs) have to tell you - stacking 6+ artillery units is a-historical, and if you do, you get a-historical results?
OTOH, if you want to put 15+ artillery units in the same hex- go ahead. But when you get massive casualties, don't complain 'These results would never happen in real life! The game is broken!'

Have you, even once, thought 'maybe it's my tactics, not the game engine, that's wrong?'
Don't you find it odd, that you alone are having so many serious problems with AE?
Shouldn't that tell you something?

It will be better for everybody when you stop whining about this so-called problem.
User avatar
Canoerebel
Posts: 21099
Joined: Fri Dec 13, 2002 11:21 pm
Location: Northwestern Georgia, USA
Contact:

RE: More information needed...

Post by Canoerebel »

Come on, ckammp, half (or more) of the people that have chimed into this thread agree with the point.  Even some of the developers acknowledge that utilizing too many arty units in open terrain is a problem, but that they don't know (yet) how to fix it without perhaps messing up the model when just a few arty units are employed. So it is a problem, but we don't know yet if there's a satisfactory solution.
 
I started this thread to register my concerns and it has generated plenty of discussions, for and against.  If you don't like people voicing opinions contrary to yours...well, just ignore them.  But don't tell them to shut up (which is what you're doing when you accuse them of whining).
 
 
"Rats set fire to Mr. Cooper’s store in Fort Valley. No damage done." Columbus (Ga) Enquirer-Sun, October 2, 1880.
ckammp
Posts: 756
Joined: Sat May 30, 2009 4:10 pm
Location: Rear Area training facility

RE: More information needed...

Post by ckammp »

ORIGINAL: Canoerebel

Come on, ckammp, half (or more) of the people that have chimed into this thread agree with the point.  Even some of the developers acknowledge that utilizing too many arty units in open terrain is a problem, but that they don't know (yet) how to fix it without perhaps messing up the model when just a few arty units are employed. So it is a problem, but we don't know yet if there's a satisfactory solution.

I started this thread to register my concerns and it has generated plenty of discussions, for and against.  If you don't like people voicing opinions contrary to yours...well, just ignore them.  But don't tell them to shut up (which is what you're doing when you accuse them of whining).


What about the half that agree with me?
Our opinion doesn't matter?

And you totaly miss the point - yes, the devs agree that too many artillery units in one hex cause unhistoric casualties. Because the model was based on historical employment of artillery units; it was not based on "What if I put 15+ artillery units in the same hex?" So the 'problem' is easily solved - don't put so many artillery units in the same hex. What is so hard to understand?

You started this thread because you had legitimate concerns about the artillery model regarding the effect on fortifications. In this I, as well as the majority of forum readers, and the devs, shared your concern. The devs responded, and have very effectively answered this concern.

Their answer, however, wasn't good enough for you, and you continue this quest to 'fix' what you perceive as a 'problem'. In this quest, you stubbornly refuse to acknowledge that there might be other explainations for the ONE bad combat result you have experienced since the patch. NO, it is clearly a fact - the artillery model is hopelessly broken because Canoerebel's LCUs were milling around outside Akyab and got themselves shot up.
The fact that your opponent had more artillery in that one hex than ever happened in real life; well, that just doesn't matter, does it?
And the units' morale, fatigue, disruption, leadership, prep - all factors taken into account in the combat model - they don't matter either, right?
Of course, bad die rolls - never had them in your game?

As for not liking contrary opinions - you are the one who, five days ago, suggested that both of us step away from this issue. I agreed, and complied. You, however, have continued, on a daily basis, to press your opinion on this issue. In other words, you just wanted me to shut up. You are the one who is attempting to shout down anyone who disagrees with you, not me. You are the one who insists the game is broken ,just because you say so. It seems that only you can answer questions or give advise in this forum, anyone else who does, gets attacked by you. A question was asked about house rules - you gave your opinion, why can't I give mine?

As for whining - well, that's what you are doing. If you don't believe me, read your own AAR, especially the post where you castigate yourself for whining about this vey problem!
User avatar
Chickenboy
Posts: 24646
Joined: Fri Jun 28, 2002 11:30 pm
Location: San Antonio, TX

RE: More information needed...

Post by Chickenboy »

ORIGINAL: Canoerebel
ckammp, you've made it clear that you don't think there is anything wrong with artillery.  Let's step aside and allow other players to weigh in with the results from their own games.  A large sampling of data from a variety of ongoing games should shed more light on the issue.
Any other players with data from their games that would provide more data to this discussion?
Image
Post Reply

Return to “War in the Pacific: Admiral's Edition”