Matrix Games Forums

Forums  Register  Login  Photo Gallery  Member List  Search  Calendars  FAQ 

My Profile  Inbox  Address Book  My Subscription  My Forums  Log Out

Under the Hood

 
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] >> The Operational Art of War IV >> Mods and Scenarios >> Under the Hood Page: [1] 2   next >   >>
Login
Message << Older Topic   Newer Topic >>
Under the Hood - 1/13/2019 11:49:11 AM   
cathar1244

 

Posts: 537
Joined: 9/5/2009
Status: offline
An interesting aspect of simulations is that they can be prodded to provide information about their function. TOAW is no different.

I am interested in casualty rates as determined by losses in TOAW. So, I've set a "lab" scenario with a small map, fair weather, identical opposing armies, proficiency and supplies topped off at 100%, etc., to see what the casualties look like at the end of the game. I let the PC play itself. Scenario is 10 turns, all terrain is open, each army has five regiments of 200 light infantry squads.

I ran the scenario 10 times. In five scenarios, one side was designated first; in the other five, the other side was first to move. In Elmer vs Elmer, at least, that seems to make a big difference: the first side moving won eight times out of ten. Equally interesting is that four times out of five, the first side moving had losses of about 50% by the end of the scenario, while the second side moving took losses more like 70 to 75%. There were two outliers in which the first side moving took heavier losses. I'm not sure what that means [if anything], but it seems counter-intuitive. By moving first, the first side arrived near the second force with some readiness and supply degradation, yet they almost always broke out a can of whoopin' on the second force.

I'll post some more comments as I alter the scenario. I may try next round with the regiments equipped with assault squads vice light infantry squads. Be interesting to see if casualties go up or the game ends earlier.

Cheers
Post #: 1
RE: Under the Hood - 1/13/2019 12:05:13 PM   
larryfulkerson

 

Posts: 38206
Joined: 4/17/2005
From: Tucson, AZ
Status: offline

quote:

ORIGINAL: cathar1244

An interesting aspect of simulations is that they can be prodded to provide information about their function. TOAW is no different.

I am interested in casualty rates as determined by losses in TOAW. So, I've set a "lab" scenario with a small map, fair weather, identical opposing armies, proficiency and supplies topped off at 100%, etc., to see what the casualties look like at the end of the game. I let the PC play itself. Scenario is 10 turns, all terrain is open, each army has five regiments of 200 light infantry squads.

I ran the scenario 10 times. In five scenarios, one side was designated first; in the other five, the other side was first to move. In Elmer vs Elmer, at least, that seems to make a big difference: the first side moving won eight times out of ten. Equally interesting is that four times out of five, the first side moving had losses of about 50% by the end of the scenario, while the second side moving took losses more like 70 to 75%. There were two outliers in which the first side moving took heavier losses. I'm not sure what that means [if anything], but it seems counter-intuitive. By moving first, the first side arrived near the second force with some readiness and supply degradation, yet they almost always broke out a can of whoopin' on the second force.

I'll post some more comments as I alter the scenario. I may try next round with the regiments equipped with assault squads vice light infantry squads. Be interesting to see if casualties go up or the game ends earlier.

Cheers

Thank you for doing these tests. It's illustrative, informative, and interesting. I would like to encourage you.

_____________________________

A drunkard stumbles into a confessional booth and closes the door. The priest waits for a minute and then knocks on the wall. The drunkard says: "No use knocking, there's no paper on this side either."

(in reply to cathar1244)
Post #: 2
RE: Under the Hood - 1/13/2019 12:39:37 PM   
cathar1244

 

Posts: 537
Joined: 9/5/2009
Status: offline
Larry, thanks for the encouragement. I'll post what I find. Honestly, I know very little about TOAW internal function. As you're well aware, there are a bunch of variables that could be altered even with this simple scenario regarding supply, proficiency, etc. I also want to vary the timescale. It is set to full week turns now, maybe I'll drop it to full days. I would not expect that to matter, but it will be mildly interesting to see if it changes anything. I think the key to useful results will be keeping the scenarios as simple as possible to reduce the number of interactions from variables. That is why I've got the units, formations, and forces at 100% setting.

Cheers

(in reply to larryfulkerson)
Post #: 3
RE: Under the Hood - 1/13/2019 4:44:33 PM   
cathar1244

 

Posts: 537
Joined: 9/5/2009
Status: offline
Okay, color me wrong. Setting the time scale to full days vice full weeks changed the scenario outcomes. A reminder that in TOAW, certain hex size vs time-represented-by-a-turn relationships are better than others.

The change to time scale resulted in less unit mobility in a turn. This meant that both forces met in the middle of the map. Combat was both less bloody and less decisive. "Less bloody" means this time, the losses of both sides were roughly equal with only a couple of outliers in which the loss of squads had a discrepancy of about 150. "Less decisive" means fewer entire units were destroyed at the end of the scenario. Full day turns resulted in four instances of this over ten repetitions of the scenario; full week turns, by comparison, produced 25 instances of such over ten repetitions, or more than six times as many.

In both full week and full day turns, unit movement is too fast. The units are steadily covering between 40 and 50 kilometers a day; bear in mind these units are foot infantry only. Of course, force-marching a unit in the real world can match this performance, but to see it every turn indicates that a scenario designer will want to compare the advance of units in the game to historical advance rates for a given campaign. Dampening of unit movement may be indicated, as in this case.

Cheers

(in reply to cathar1244)
Post #: 4
RE: Under the Hood - 1/13/2019 10:39:18 PM   
r6kunz


Posts: 1019
Joined: 7/4/2002
From: near Philadelphia
Status: offline
Greetings cathar1244,
Thank you very much for your input in the forum. You are looking at things with an unbiased eye and you seem to pick up on a lot of details.

A couple of my observations; I used TOAW III to design Korea 2013 and ported it over to TOAW IV (scenarios are not reverse compatible, and IV was only a beta.) The same scenario, but gave remarkably different results- draw vs overwhelming victory for Kim Jung-um. Of course one would say "different supply" but I am still surprised.

I too have looked at turn length. In Red Thunder II, I thought the half-day turns in Red Thunder I were a bit drawn out and found that full-day turns in II gave expected results.

Re: Unit movement. It has seemed like the rate of march was a bit optimistic. (you sound like you have had military experience). But I found TOAW quickly degrades this. For example, In Road to Moscow: I. Crossing the Border, In my hands, German Panzer Group II could not even reach the Dnepr River whereas Guderian historical established several bridgeheads. I tinkered with increasing Movement Bias to allow the Germans vs po Sovies, to have a chance to achieve the historic results.

Another case in point is the Allied counteroffensive in the Ardennes Offensive 1944. Historically Germans to almost reach the Mosel in eight days. Using the same parameters in the scenario, the Allies can push the Germans back to the other side of the West Wall. Historically, however, the Allies, even with overwhelming supply, reinforcements, and air power, cannot close the pocket much past Bastogne (at Hoffalize) by game's end. Fortuantely, TOAW IV comes to the rescue with variable Movement Bias as an Event...

Looking forward to your comments and research.

Cheers,
RAK




_____________________________

Avatar image was taken in hex 87,159 Vol 11 of
Vietnam Combat Operations by Stéphane MOUTIN LUYAT aka Boonierat.

(in reply to cathar1244)
Post #: 5
RE: Under the Hood - 1/14/2019 1:49:41 AM   
mussey


Posts: 650
Joined: 12/2/2006
From: Cleve-Land
Status: offline
Variable movement bias.

I need to investigate this. If I change for example, from 3.5 days to 1 day/turn, will the entry dates of reinforcements convert accordingly? Will the date conversions take place Events?

This is a great topic. Game scale/ unit density/ days per turn/ MP’s. Change just one and the ensuing results can break a game.

_____________________________

Col. Mussbu

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


(in reply to r6kunz)
Post #: 6
RE: Under the Hood - 1/14/2019 2:24:01 AM   
Curtis Lemay


Posts: 11426
Joined: 9/17/2004
From: Houston, TX
Status: online
Don't forget Attrition Divider. That's the main adjustment between turn intervals.

_____________________________

My TOAW web site:

Bob Cross's TOAW Site

(in reply to mussey)
Post #: 7
RE: Under the Hood - 1/14/2019 9:02:49 AM   
cathar1244

 

Posts: 537
Joined: 9/5/2009
Status: offline
Thanks much for the thoughtful comments. Bob and RAK, your comments re: movement bias and attrition divider are well taken, and give designers options to correct some things. Long ago, I made a scenario called "Rhine 44" and was frustrated the British forces in the game could not match the historical advance to Brussels and Antwerp in September 1944. Certainly do-able now.

Mussey, IIRC, if you change the days-per-turn setting, events will not automatically adjust -- beware, as I have seen such changes, in complicated event set-ups, trigger nasty chain reactions. Be sure you have a clean copy of the events structure before changing anything.

On unit movement, I would be interested in what you fellows think about transport vehicles. I've seen the idea in these forums they are best treated as a relative measure of transport assets available versus the actual number of trucks, halftracks, or whatever. My sense is that "measure of transport assets" approach is better in terms of game effects (if only because not all trucks are the same), but it would be interesting to hear the thoughts of experienced scenario designers.

Cheers

(in reply to Curtis Lemay)
Post #: 8
RE: Under the Hood - 1/15/2019 3:24:05 AM   
rhinobones

 

Posts: 873
Joined: 2/17/2002
Status: offline
Wonder if maybe the side which moves first (and attacks first?) has an initial advantage of selecting the units to be attacked, numbers of attackers vs defenders and the number of hex sides to attack from. Factors which are crucial in calculating losses.

After gaining an initial lead on total losses the side attacking first would then be expected to compound the advantage over the following nine turns.

The fact that several trials had the side moving second as winning would seem to show that the success probabilities built into the game engine due in fact work.

Regards, RhinoBones


_____________________________

Colin Wright:
Comprehensive Wishlist Forum #467 . . . The Norm (blessed be His name, genuflect three times and accept all values in the program as revealed truth)

Pre Combat Air Strikes # 64 . . . I need have no concern about keeping it civil

(in reply to cathar1244)
Post #: 9
RE: Under the Hood - 1/15/2019 3:30:34 AM   
larryfulkerson

 

Posts: 38206
Joined: 4/17/2005
From: Tucson, AZ
Status: offline
quote:

The fact that several trials had the side moving second as winning would seem to show that the success probabilities built into the game engine due in fact work.

I concur. Also, I would suspect that the size of the scenario might effect the results. FITE2 has so many units that loosing a few extra on turn one might not make all that much of an impact given the length of the game and the number of units available. I may have to do some tests to make sure. I'll run FITE2 100 times and tabulate the losses for each side and compare them to the toy scenario you guys are using for your test. I'll be back in about 10 years.

_____________________________

A drunkard stumbles into a confessional booth and closes the door. The priest waits for a minute and then knocks on the wall. The drunkard says: "No use knocking, there's no paper on this side either."

(in reply to rhinobones)
Post #: 10
RE: Under the Hood - 1/15/2019 11:27:46 AM   
cathar1244

 

Posts: 537
Joined: 9/5/2009
Status: offline
The latest variation of the scenario was full day turns with assault squads. Each force had 1000 assault squads divided among five regiments. Following are averaged force casualty rates for the various runs.

Situation: Full week turns using light rifle squads. Force 1 = 68%, Force 2 = 54%. 25 instances of unit destruction (as of end of scenario).

Situation: Full day turns using light rifle squads. Force 1 = 50%, Force 2 = 45%. Four instances of unit destruction.

Situation: Full day turns using assault squads. Force 1 = 60%, Force 2 = 56%. 24 instances of unit (regiment) destruction.

I'm not sure why the assault squad variant caused so many more unit destruction events, but only raised the overall casualty rate by 20% or so. This leads to a basic question ... defense strength. The defense strength of all of these squads is "10". I am not certain why but suspect that represents the number of men in a squad, and wonder if that is a valid measure of the unit's strength in the defense.

I am open to being shown why the number of men in a squad should be its defense factor. After all, infantry combat, whether attacking or defending, involves firepower. So why is the defense strength not equal to the antipersonnel firepower of the squad? I suppose it revolves about the TOAW combat model, and perhaps, the defense strength is not so important in the resolution of how many squads are lost in a given infantry battle.

The casualty rates seem stiff, but the units in the game are "perfect" with attributes at 100%.

I will bring up movement rate again, because I misspoke earlier. In the full day scenarios, these foot infantry regiments are moving 80 kilometers a day. Wow! In the full week scenarios, I think the movement rate was something 55 kilometers a day (averaged). Not only is the full day scenario movement very rapid, but it is considerably faster than the same units when the turns represent full weeks (which is okay as I would not expect a unit to keep up full bore forced marches for seven days in a row.)

As always, comments are welcome.

Cheers

(in reply to larryfulkerson)
Post #: 11
RE: Under the Hood - 1/15/2019 1:59:09 PM   
mussey


Posts: 650
Joined: 12/2/2006
From: Cleve-Land
Status: offline
quote:

IIRC, if you change the days-per-turn setting, events will not automatically adjust -- beware, as I have seen such changes, in complicated event set-ups, trigger nasty chain reactions. Be sure you have a clean copy of the events structure before changing anything.


Thanks for heads-up.

_____________________________

Col. Mussbu

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


(in reply to cathar1244)
Post #: 12
RE: Under the Hood - 1/17/2019 2:19:07 PM   
cathar1244

 

Posts: 537
Joined: 9/5/2009
Status: offline
Ran five iterations of a technology discrepancy infantry fight. One force had musket squads, the other had light rifle squads. The light rifle squad force inflicted the worst casualties every time, so the model performed as expected. Averaged losses were 64% for the musket force and 30% for the light rifle force. The musket force lost at least one regiment (of five) by the end of every iteration; the light rifle force lost no regiments.

Cheers

(in reply to mussey)
Post #: 13
RE: Under the Hood - 1/19/2019 5:56:39 PM   
cathar1244

 

Posts: 537
Joined: 9/5/2009
Status: offline
Set up a different situation this time. Both sides were given five infantry divisions, equipped almost identically to a U.S. 1943-type infantry division. Each division was also reinforced by a battalion of medium tanks. One turn was half a day. Ten iterations were run and rifle squad casualties noted.

The object of the testing was to look at how many more casualties would result if the number of machine guns in each division was doubled. So, this was done, and the scenario was again run over ten iterations.

With the number of machine guns standard, the loss rate of rifle squads on average 61.5%. When the machine gun quantity was doubled, the average loss rate of rifle squads went to 70%. Summarized in a different way, a 100% increase in machine guns caused a rise in rifle squad loss rate of 15%.

Appropriately, while the machine guns contribute to infantry losses, they are not dominant (IIRC, in World War II, artillery and mortars were the most important drivers of battlefield casualties).

Cheers

(in reply to cathar1244)
Post #: 14
RE: Under the Hood - 1/24/2019 6:24:42 AM   
cathar1244

 

Posts: 537
Joined: 9/5/2009
Status: offline
I've been looking at volume (equipment characteristic) as, as far as I can tell, the use of this setting is only partially defined. We are told it is the cubic meter volume of a piece of equipment, and that it plays a role in determining the chance of a hit upon an armored target. Yet, there is also the figure "999" used as volume for equipment as a different as rifle squads, helicopters, ships, artillery pieces, and fixed wing aircraft. My SWAG is that "999" is a flag of some kind for software, but for what, I have no idea.

Looking further at the relationship of volume and armor rating among equipment items led to the discovery of some odd bits -- such as the "Barak" SP-SAM. It has a volume rating but no armor rating, and it is not flagged as armored equipment. The use of an actual volume figure for unarmored equipment made me wonder if that could induce odd combat results, so I set up the lab scenario with one side equipped only with Barak vehicles and the other side with Sherman tanks. Even given the discrepancy in technological era, I expected the Shermans to clean up ...

... because among other settings, the Barak has zero for anti-personnel and zero anti-tank ratings. Per the database, it has only anti-aircraft ratings at both low and high altitudes. So what happened? The answer is, it varied.

The first lab run had the Baraks outnumber the Shermans 10 to 1. But with no AP or AT capability, I still expected the Barak force to be beaten on by the tanks. Didn't happen ... the tanks took up to 50% losses!

Later lab runs set the number of Baraks and tanks equal. This worked out better for the tanks, but the lab runs make me wonder what is going on in the software's model for combat resolution.

Way back in TOAW I, IIRC, there was an issue with cumulative weight of fire being used to resolve combat. This issue led to a more nuanced system in which equipment engages other equipment in combat that is not artillery bombardment.

Yet, the tanks vs Barak engagements produced very odd results considering that the Baraks have neither anti-personnel nor antitank ratings. The question is, precisely what characteristic of the Baraks is destroying the Shermans of the tank force? Is this some side effect of the anti-air ratings? Or is there some kind of "swarming" effect when one side strongly outnumbers the other in a unit vs unit engagement?

This makes me want to run further labs along these lines. For example, what happens when the "truck army" confronts a smaller (by 10X) of rifle squads or tanks?

Cheers

(in reply to cathar1244)
Post #: 15
RE: Under the Hood - 1/24/2019 7:32:53 AM   
larryfulkerson

 

Posts: 38206
Joined: 4/17/2005
From: Tucson, AZ
Status: offline
quote:

...the Barak has zero for anti-personnel and zero anti-tank ratings. Per the database, it has only anti-aircraft ratings at both low and high altitudes.

Are the AA guns 88mm GP's, mounted on a truck? I ask because sometimes the Germans would lower the barrel and engage tanks and other AFV's with it, to good effect. I doubt that's what's happening though. I'd like to know how they have any offensive power at all. Please keep up posted on this effect.

_____________________________

A drunkard stumbles into a confessional booth and closes the door. The priest waits for a minute and then knocks on the wall. The drunkard says: "No use knocking, there's no paper on this side either."

(in reply to cathar1244)
Post #: 16
RE: Under the Hood - 1/24/2019 5:03:37 PM   
cathar1244

 

Posts: 537
Joined: 9/5/2009
Status: offline
quote:

Are the AA guns 88mm GP's, mounted on a truck?


Larry, just to keep my own sanity I checked the equipment definition for the 88mm DP gun that represents the German wartime FlaK piece. It has both AP and AT values with the AT projectile defined as kinetic. Good so far. The Barak system is an AA missile and the definition shows it with an AA rating only. So, that is good, too.

Messing about with the lab scenario again, I think the tank losses are coming not from direct combat losses, but in instances where the unit has been forced to retreat so much that it can't retreat any farther and evaporates. I understand that happens in TOAW, but again the question is, how is a zero-attack-strength SPSAM-equipped unit causing tanks to retreat?

I am not familiar with the TOAWlog.txt file, but it may be the way combat power is calculated is causing some of this -- check out how defense strength figures into the overall strength figure:

quote:

Overrun :
Overrun :Overrun check begins.
Overrun :Defender anti armor strength: 311
Overrun :Attacker anti armor strength: 0
Overrun :Defender anti personnel strength: 176
Overrun :Attacker anti personnel strength: 0
Overrun :Defender defense strength: 73
Overrun :Attacker defense strength: 1766
Overrun :Defender armor defense strength: 73
Overrun :Attacker armor defense strength: 0
Overrun :Defender armor ratio: 10
Overrun :Attacker armor ratio: 0
Overrun :Defender size: 2
Overrun :Attacker size: 1
Overrun :Defender MP: 9
Overrun :Attacker MP: 9
Overrun :Defender strength(Total): 996 = d73 + at0 + ap176
Overrun :Attacker strength(Total): 1412 = d1766 + at0 + ap0
Combat :Removing available equipment from combat inventory.
Combat :Greys 1st Formation, 5 equipment added to combat inventory.


It seems that the larger number of Barak vehicles (each with defense strength = 12) generates enough overall strength for an overrun to occur in many combat instances, at least when the AI is playing itself.

So I'll do what I mentioned earlier with the "truck army" and see what results. I'm not sure what to think of these results. Yeah, I guess at 10 to 1 odds, the Baraks could "overrun" the tanks ... theoretically. I think in reality it would be simply be a target-rich environment for experienced tankers, since the Baraks don't have any weapons useful against targets on the ground. Hmm.

Cheers

(in reply to larryfulkerson)
Post #: 17
RE: Under the Hood - 1/24/2019 9:07:01 PM   
larryfulkerson

 

Posts: 38206
Joined: 4/17/2005
From: Tucson, AZ
Status: offline
quote:

Overrun :
Overrun :Overrun check begins.
Overrun :Defender anti armor strength: 311
Overrun :Attacker anti armor strength: 0
Overrun :Defender anti personnel strength: 176
Overrun :Attacker anti personnel strength: 0
Overrun :Defender defense strength: 73
Overrun :Attacker defense strength: 1766
Overrun :Defender armor defense strength: 73
Overrun :Attacker armor defense strength: 0
Overrun :Defender armor ratio: 10
Overrun :Attacker armor ratio: 0
Overrun :Defender size: 2
Overrun :Attacker size: 1
Overrun :Defender MP: 9
Overrun :Attacker MP: 9
Overrun :Defender strength(Total): 996 = d73 + at0 + ap176
Overrun :Attacker strength(Total): 1412 = d1766 + at0 + ap0
Combat :Removing available equipment from combat inventory.
Combat :Greys 1st Formation, 5 equipment added to combat inventory.

That looks like what you said to me. I concur. So the fix seems to be to establish a more realistic defense value for the unit. Right now it's Kryptonite.

_____________________________

A drunkard stumbles into a confessional booth and closes the door. The priest waits for a minute and then knocks on the wall. The drunkard says: "No use knocking, there's no paper on this side either."

(in reply to cathar1244)
Post #: 18
RE: Under the Hood - 1/25/2019 2:12:41 PM   
mussey


Posts: 650
Joined: 12/2/2006
From: Cleve-Land
Status: offline

quote:

ORIGINAL: cathar1244

Set up a different situation this time. Both sides were given five infantry divisions, equipped almost identically to a U.S. 1943-type infantry division. Each division was also reinforced by a battalion of medium tanks. One turn was half a day. Ten iterations were run and rifle squad casualties noted.

The object of the testing was to look at how many more casualties would result if the number of machine guns in each division was doubled. So, this was done, and the scenario was again run over ten iterations.

With the number of machine guns standard, the loss rate of rifle squads on average 61.5%. When the machine gun quantity was doubled, the average loss rate of rifle squads went to 70%. Summarized in a different way, a 100% increase in machine guns caused a rise in rifle squad loss rate of 15%.

Appropriately, while the machine guns contribute to infantry losses, they are not dominant (IIRC, in World War II, artillery and mortars were the most important drivers of battlefield casualties).

Cheers


Good work. After seeing your work I don't feel as bad removing them from my Orbats. Using your provided TO&E's from previous posts, I removed the MMG's & HMG's since they were mostly already embedded in their Infantry squads and vehicles. The freed-up slots were used for other equipment.

_____________________________

Col. Mussbu

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


(in reply to cathar1244)
Post #: 19
RE: Under the Hood - 1/25/2019 8:13:28 PM   
Cabido

 

Posts: 237
Joined: 12/11/2017
Status: offline

quote:

ORIGINAL: cathar1244

quote:

Are the AA guns 88mm GP's, mounted on a truck?


Larry, just to keep my own sanity I checked the equipment definition for the 88mm DP gun that represents the German wartime FlaK piece. It has both AP and AT values with the AT projectile defined as kinetic. Good so far. The Barak system is an AA missile and the definition shows it with an AA rating only. So, that is good, too.

Messing about with the lab scenario again, I think the tank losses are coming not from direct combat losses, but in instances where the unit has been forced to retreat so much that it can't retreat any farther and evaporates. I understand that happens in TOAW, but again the question is, how is a zero-attack-strength SPSAM-equipped unit causing tanks to retreat?

I am not familiar with the TOAWlog.txt file, but it may be the way combat power is calculated is causing some of this -- check out how defense strength figures into the overall strength figure:

quote:

Overrun :
Overrun :Overrun check begins.
Overrun :Defender anti armor strength: 311
Overrun :Attacker anti armor strength: 0
Overrun :Defender anti personnel strength: 176
Overrun :Attacker anti personnel strength: 0
Overrun :Defender defense strength: 73
Overrun :Attacker defense strength: 1766
Overrun :Defender armor defense strength: 73
Overrun :Attacker armor defense strength: 0
Overrun :Defender armor ratio: 10
Overrun :Attacker armor ratio: 0
Overrun :Defender size: 2
Overrun :Attacker size: 1
Overrun :Defender MP: 9
Overrun :Attacker MP: 9
Overrun :Defender strength(Total): 996 = d73 + at0 + ap176
Overrun :Attacker strength(Total): 1412 = d1766 + at0 + ap0
Combat :Removing available equipment from combat inventory.
Combat :Greys 1st Formation, 5 equipment added to combat inventory.


It seems that the larger number of Barak vehicles (each with defense strength = 12) generates enough overall strength for an overrun to occur in many combat instances, at least when the AI is playing itself.

So I'll do what I mentioned earlier with the "truck army" and see what results. I'm not sure what to think of these results. Yeah, I guess at 10 to 1 odds, the Baraks could "overrun" the tanks ... theoretically. I think in reality it would be simply be a target-rich environment for experienced tankers, since the Baraks don't have any weapons useful against targets on the ground. Hmm.

Cheers


As far as I know, the Barak system shouldn't inflicting any loss to the tank unit due to direct fire. At least, if all the info I have collected so far from the forum for the explanation of combat I have included in my charts file.

It seems it comes from the unit retreating due to being flanked, as for the rule that follows:

"The first time you order a unit to Attack any particular
enemy location in any given Turn, there is a chance
that the defenders will retreat before combat. If there
is a large difference between the combined Attack
Strengths of your current unit when the Attack is
ordered and the combined Defense Strengths of the
enemy units in the defending location, the computer
will check to see if the defending units immediately
retreat rather than face your assault.

When checking for retreats before combat, a
Flanking Check is made, which is based on the
quantity of “active defender” equipment in the
defending and attacking units relative the Scenario
Map scale. If the defender has less than the amount
of equipment necessary to fully cover his frontage
in the Scenario Scale, the Attacker has more
equipment, and the Attacker passes a Unit Quality
Check, the chance for a retreat before combat is
increased by a random fraction between one and:

Attacker Equipment Density / Defender Equipment Density

This makes is much more likely that “small” units
will retreat before combat with “large” units because
the large unit has succeeded in turning the flank of
the smaller unit."


But it seems that "combined attack strength" must be using the AA strength or adding, in contradiction to the manual, the defense strength to the total (your log info brings a total which sums ap, at and defense strengths; how it is used, I don't know) or the attacker strength would be 0. It would be nice to have it explained.

One may argue that this is reasonable, due to the fact that in combat the forces won't have perfect information on the other side, so that sheer volume difference can cause the retreat before combat, once the unit begins to be flanked. I don't know. More simulations are necessary to see if it works well with what would be less threatening pieces of equipment, like trucks.




Attachment (1)

(in reply to cathar1244)
Post #: 20
RE: Under the Hood - 1/26/2019 5:48:52 PM   
cathar1244

 

Posts: 537
Joined: 9/5/2009
Status: offline
Cabido and Larry, thank you for the comments.

From what the log file indicates, I think the defense factor [DF] plays a role in the resolution of overrun actions. As noted with the lab scenario and the SPSAMs versus tanks, this can generate unrealistic combat outcomes: not because of losses to enemy fire, but losses caused by a unit being repeatedly forced to retreat until it evaporates. As Larry suggested, the DF for the Barak SPSAM is set too high. This is one instance in which a scenario designer, by setting up a new piece of equipment with a lowered DF, can correct this issue.

Of course, this solution immediately poses the question as to what a realistic DF should be for a given item of equipment. And that question cannot be simply answered because of the logic employed by the game's software. First, let us consider a definition that was given for DF (this quote was attributed to James Matthews; my apologies to James if that attribution was not accurate.)

quote:

The defense factor is a number that, for the vast majority of equipment types, is scaled from 1-10 and represents a rough ability of that equipment to absorb firepower without being destroyed. Equipment is defined here, as items within the database, whether they are multi-man items like squads and teams, or singleton items like AFV’s, and Guns, which also may be multi-man operated.


But for a scenario designer trying to alter the DF of equipment, there is a serious caveat to bear in mind. And this caveat is that in many equipment item cases, the software ignores the DF entered into the equipment file, and calculates a DF value that the game uses for that equipment. Most of these autogenerated values concern guns and vehicles. Any towed, fixed, self-propelled, or railway gun will have an autogenerated DF. Any armored vehicle (as set by Flag 0) will have an autogenerated DF. This last observation also applies to SPSAMs if they are an armored target.

Vehicles like the Barak SPSAM fall into a category that one might call "soft vehicle without a ranged antipersonnel weapon". With these equipment items, the DF can be directly set in the equipment file, and for some reason, the Barak ended up with a DF of 12. Besides these kinds of vehicles, the DF can be directly set for trucks, helicopters, fixed-wing aircraft, ships, personnel squads, and the armored train.

I will put up another thread that explains which autogenerated values are used for SP guns, etc.

The use of DF in overrun combat can produce illogical outcomes. But a scenario designer could potentially use this quirk to advantage. Say we want to craft a force that has a talent for overruns and forcing retreats (Chinese infantry in Korea, November 1950?)

The autogeneration of DF values produced an extended discussion in a thread about the "Europa 1947" scenario; mentioning items such as SP-guns getting a higher DF than a King Tiger tank. In combat against armored targets, it does not matter. But the way combat power is calculated for overruns makes one wonder, once again, if these autogenerated values could provoke unrealistic combat outcomes.

Cheers

< Message edited by cathar1244 -- 1/26/2019 5:54:06 PM >

(in reply to Cabido)
Post #: 21
RE: Under the Hood - 1/26/2019 7:06:26 PM   
Cabido

 

Posts: 237
Joined: 12/11/2017
Status: offline
I think it is sensible to have the defense strength of the defender compared to the attack strength of the attacker as a criteria to define the chance of overrun (together with the density variable, also used), since the ratio (attacker attack strength)/(defender defense strength) defines the exposure to fire of the defender unit. Yet, using the sum (attack strength + defense strength) as the log suggests (here I'm not sure of understanding the role of these values in the log) isn't sensible, since then very high defense value equipment, with very low attack strength would have a high chance of causing enemy units RBC.
As I said earlier, perhaps because of imperfect information in the battlefield, such a unit should have some chance of causing RBC, based on density difference, due to flanking, but not a chance proportional to their defense strength; in that case, it should only be proportional to the density ratio. Anyway, I'm just speculating about the possibility that the defense strength of the attacker unit is being used as a factor, which would contradicts what is said in the manual.

(in reply to cathar1244)
Post #: 22
RE: Under the Hood - 1/26/2019 7:16:05 PM   
cathar1244

 

Posts: 537
Joined: 9/5/2009
Status: offline
quote:

ORIGINAL: Cabido

I think it is sensible to have the defense strength of the defender compared to the attack strength of the attacker as a criteria to define the chance of overrun (together with the density variable, also used), since the ratio (attacker attack strength)/(defender defense strength) defines the exposure to fire of the defender unit. Yet, using the sum (attack strength + defense strength) as the log suggests (here I'm not sure of understanding the role of these values in the log) isn't sensible, since then very high defense value equipment, with very low attack strength would have a high chance of causing enemy units RBC.
As I said earlier, perhaps because of imperfect information in the battlefield, such a unit should have some chance of causing RBC, based on density difference, due to flanking, but not a chance proportional to their defense strength; in that case, it should only be proportional to the density ratio. Anyway, I'm just speculating about the possibility that the defense strength of the attacker unit is being used as a factor, which would contradicts what is said in the manual.


Cabido, thanks again for your comments, this is an interesting discussion.

I think the use of DF in overrun attacks -by the attacking unit- is more than speculation, though. Not only does the log file indicate the DF is used by the attacking unit in its calculation ... but my lab scenario, in which units equipped with nothing but "Baraks" [with no AP and no AT rating] were able to regularly overrun defending tank units, also suggests that is the case.

Realistically, in most scenarios, that kind of unit encounter won't be seen, at least, not often. Another consideration is recalling that the DF, for an infantry squad, reflects its manpower count -- and that is something that definitely is realistic to account for in assessing a unit's combat power.

The DF for the Barak SPSAM can be changed in the equipment file (and should be IMO). More problematic are those DF values that are autogenerated by TOAW itself.

Cheers

< Message edited by cathar1244 -- 1/26/2019 7:17:05 PM >

(in reply to Cabido)
Post #: 23
RE: Under the Hood - 1/27/2019 4:46:12 AM   
cathar1244

 

Posts: 537
Joined: 9/5/2009
Status: offline
One more comment about the overrun strength calculations. It counts DF as part of the attacker's calculation. As mentioned for rifle squads and manpower counts, this is not necessarily bad. But the calculation also includes the DF of passive defenders (such as the Barak SP-SAM) in reckoning the attacker's strength. That seems off to me.

Cheers

(in reply to cathar1244)
Post #: 24
RE: Under the Hood - 1/27/2019 1:08:11 PM   
altipueri

 

Posts: 590
Joined: 11/14/2009
Status: offline
Does changing the random number seed affect any of these tests?

(in reply to cathar1244)
Post #: 25
RE: Under the Hood - 1/27/2019 1:26:47 PM   
larryfulkerson

 

Posts: 38206
Joined: 4/17/2005
From: Tucson, AZ
Status: offline

quote:

ORIGINAL: altipueri

Does changing the random number seed affect any of these tests?

I'm pretty sure it won't. The default, I believe, is the number used
for tests. You can enter a different number to try to get a different
game but I doubt if it makes all that much difference.

_____________________________

A drunkard stumbles into a confessional booth and closes the door. The priest waits for a minute and then knocks on the wall. The drunkard says: "No use knocking, there's no paper on this side either."

(in reply to altipueri)
Post #: 26
RE: Under the Hood - 2/9/2019 7:54:09 AM   
cathar1244

 

Posts: 537
Joined: 9/5/2009
Status: offline
As a partial wrap-up of the defense strength influence on overruns (RBC), I ran the same lab but replaced the Barak SP-SAMs with regular trucks, which only have a defense strength of one. Although the trucks vastly outnumbered the tanks, they did not provoke retreat before combat and took heavy losses.

Conclusion: Keep the defense strength low of vehicles that do not hold or take ground (like SP-SAMs). The question of passive defenders having their defense strength reckoned as part of overrun attacks is IMO questionable; maybe something for Ralph to consider modifying at some point.

Cheers

(in reply to larryfulkerson)
Post #: 27
RE: Under the Hood - 2/9/2019 8:40:57 AM   
cathar1244

 

Posts: 537
Joined: 9/5/2009
Status: offline
Another thing I've been wondering about is the antitank ratings of infantry squads. The AT rating of generic squads in TOAW tops out (I think) at 15, implying 15cm of potential armor penetration (the "AT+" rating). TOAW's equipment database correctly identifies this capability as being "1940's", IIRC, 15cm was something like the armor penetration of an early model Panzerfaust.

Of course, the performance of handheld antitank weapons has greatly increased since 1945. For scenario designers, there is much potential here to craft equipment definitions representing the various RPG's and other infantry AT weapons of postwar years. But my question was, "is it worth the effort?" That is, would the TOAW system produce results in such a way that there was clearly a link between increased AT capability of squads on the one hand and tank losses on the other?

Implied by the question was that such detail modeling in the equipment DB might be more an exercise in polishing the chrome of a scenario than anything else.

So, for the lab, I set up two forces. Both sides had five units. One force was made up of 500 rifle squads, the other side had 250 Sherman M4/76 tanks. I chose the Shermans deliberately because I didn't want the game accounting for ultramodern effects like Chobham armor plate or reactive armor. In other words, this was to be the classic confrontation: shaped charged warheads versus rolled homogeneous armor.

For the first five iterations, the rifle squads were considered to be equipped with antitank rifle grenades having a penetration rating of 7. In the last five iterations, I used a modified squad equipped with an RPG-2; the penetration rating deemed to be 18.

The results were gratifying. First, a snippet of the game resolving anti-armor combat (rifle grenades vs tanks), from the log file:

quote:

Combat : Blues weapons firing on Greys M4/76 Sherman.
ATCombat : Assault Squad R Gren current terrain/lighting/weather dependent accuracy: 10%.
ATCombat : M4/76 Sherman target cross section bias: 113%.
ATCombat : Assault Squad R Gren current effective accuracy vs. M4/76 Sherman: 11%.
ATCombat : No hit on Greys M4/76 Sherman by Blues Assault Squad R Gren. (shot missed)
Combat : Blues weapons firing on Greys M4/76 Sherman.
ATCombat : Assault Squad R Gren current terrain/lighting/weather dependent accuracy: 10%.
ATCombat : M4/76 Sherman target cross section bias: 113%.
ATCombat : Assault Squad R Gren current effective accuracy vs. M4/76 Sherman: 11%.
ATCombat : Chance of penetration on M4/76 Sherman [9 @ 73 degrees -> 9] by Assault Squad R Gren [7] is 59%.
ATCombat : Effective hit on Greys M4/76 Sherman by Blues Assault Squad R Gren.
Combat : M4/76 Sherman destroyed. (specificAttrit=7)


We see that first the game determines a base hit chance that reckons terrain, weather, and daylight. This is bumped up by a factor of 1.13 because in the equipment database, the Sherman is flagged as "boxy" (easier to hit) in terms of being a target. The "9 @ 73 degrees" blurb is telling us the game decided the grenade struck Sherman armor (rated at 9 in the equipment DB), with the armor presenting an angle of strike to the projectile of 73 degrees (close to a right-angle strike), leading to the game deciding the chance of penetration is 59%.

But what about the RPG-2? Here is the same look as above:

quote:

Combat : Blues weapons firing on Greys M4/76 Sherman.
ATCombat : Assault Squad RPG2 current terrain/lighting/weather dependent accuracy: 10%.
ATCombat : M4/76 Sherman target cross section bias: 113%.
ATCombat : Assault Squad RPG2 current effective accuracy vs. M4/76 Sherman: 11%.
ATCombat : No hit on Greys M4/76 Sherman by Blues Assault Squad RPG2. (shot missed)
Combat : Blues weapons firing on Greys M4/76 Sherman.
ATCombat : Assault Squad RPG2 current terrain/lighting/weather dependent accuracy: 10%.
ATCombat : M4/76 Sherman target cross section bias: 113%.
ATCombat : Assault Squad RPG2 current effective accuracy vs. M4/76 Sherman: 11%.
ATCombat : Chance of penetration on M4/76 Sherman [9 @ 84 degrees -> 9] by Assault Squad RPG2 [18] is 100%.
ATCombat : Effective hit on Greys M4/76 Sherman by Blues Assault Squad RPG2.
Combat : M4/76 Sherman destroyed. (specificAttrit=16)


Oh yeah, the increased penetration of the RPG-2 drives the penetration chance up to 100%, showing the game is accounting for the better penetration rating. But what about the results in aggregrate? In fact, the squads equipped with RPG-2's did considerably better. Of five runs of each situation:

One of the runs with rifle grenades ended with a (well-behaved) crash to desktop because the tanks succeeded in destroying the entire opposing force. The game did a formation check, complained that no units existed, and exited. Of the other four runs, 500 infantry squads lost on average 347 squads, or 69% of their force. The tanks opposing them (250 tanks) lost on average 31 tanks, or 12% of their force.

The situation improved for the rifle squads when they were equipped with RPG-2's. Squad losses were on average 235 squads, or 47% of their force. Tank losses increased to an average of 68, or 27% of their force. So up-gunning the squads with the RPG-2 dropped infantry losses by a third while doubling the losses of the tank force.

Conclusion: Including this kind of detail in scenarios is not "just chrome". The effects are easily visible and accord, at the least generally, with what one might expect. This is something for designers of post-1945 scenarios to consider, as the AT capability of the generic squads in the database seems to represent roughly 1945 at best ... while the equipment definitions for modern AFVs represents more accurately their tremendous jump in combat effectiveness. Put differently, how many soldiers today would want to take on a T-80 with a 2.36-inch Bazooka?

Cheers


(in reply to cathar1244)
Post #: 28
RE: Under the Hood - 2/22/2019 7:41:27 AM   
cathar1244

 

Posts: 537
Joined: 9/5/2009
Status: offline
LAB: AIR INTERDICTION AND AIR BOMBARDMENT IN TOAW

Introduction

For these tests, a lab scenario was devised in which an infantry battalion of 100 rifle squads was subjected to air attack in several situations. The attacking air unit had 20 SU-25's for equipment. Both the target unit and the attacking air unit were set to 100% proficiency. All attacks took place in clear terrain and fair weather.

The standard equipment definition of the SU-25 allows it to employ precision guided munitions (PGM). The lab scenario had a default force PGM setting of 20, out of a possible range of 1 to 999. The SU-25 also has the first "optics" improvement; that is, it has targeting capability slightly enhanced above that of equipment using iron sights.

The version of TOAW IV used for testing does not have the new equipment density rules option.

Three kinds of attacks were looked at. The first was a player directed bombardment strike. The second kind was bombardment made against a static unit by an air unit that had been assigned an interdiction mission. The final type of attack were interdiction missions carried out against the target unit while moving.

All air strike kinds include two attack phases against the target. The first is an attritional attack against soft targets and the second phase resolves anti-armor combat. In the lab scenario, no armored equipment was present in the target unit.

Of note is that even in a simple scenario like this, for all mission types there was a significant amount of variation as expressed by the losses inflicted upon the target unit. Loss figures mentioned in these comments should be considered with this variation, and the scenario's relative lack of complex interaction, in mind.

Results of the testing are presented with a brief description followed by percentages indicating the results noted in a series of five lab runs. In cases in which an interdiction strike did not occur, the notation "NA" is given. At the end of the results is an average figure for the five lab runs. The average figure was calculated by discarding the lowest and highest of the five percentages and taking the average of the remaining three figures. In cases where data was deemed insufficient, the average figure is presented as "++".

< Message edited by cathar1244 -- 2/22/2019 8:01:37 AM >

(in reply to cathar1244)
Post #: 29
RE: Under the Hood - 2/22/2019 7:45:31 AM   
cathar1244

 

Posts: 537
Joined: 9/5/2009
Status: offline
LAB: AIR INTERDICTION AND AIR BOMBARDMENT IN TOAW (Cont.)

Player directed bombardment strikes

These strike non-moving targets and appear to represent less intensive activity by the attacking air unit than strikes conducted while an air unit is assigned an interdiction mission. "Less intensive" refers to two characteristics. The first is that directed bombardment typically produces fewer losses than air interdiction strikes. The second is that multiple bombardments may be directed in a single turn. In the lab scenario, it was possible to direct bombardment of the target by the air unit up to nine times in a single turn, while the highest number of interdiction strikes by a single air unit was six.

Most individual bombardment strikes inflicted a losses of zero to two percent upon the target. On occasion the losses were greater. The maximum observed cumulative loss for the target in a single turn was 27 percent. Thus, it was possible to reduce the target's equipment inventory by roughly one-fourth in a single turn of bombardments.

Results:

5 km / hex; Bombardment, minimize losses, nine attacks in day turn:				07%; 07%; 09%; 06%; 02%		07%
5 km / hex; Bombardment, limit losses, nine-ten attacks in day turn:				14%; 07%; 05%; 10%; 10%		09%
5 km / hex; Bombardment, minimize losses, nine attacks in half week turn:			04%; 21%; 08%; 09%; 09%		09%

1 km / hex; Bombardment, minimize losses, nine attacks in day turn:				23%; 03%; 19%; 16%; 04%		12%
1 km / hex; Bombardment, limit losses, nine-ten attacks in day turn:				27%; 16%; 21%; 06%; 11%		16%
1 km / hex; Bombardment, minimize losses, nine attacks in half week turn:			05%; 17%; 14%; 04%; 18%		12%



< Message edited by cathar1244 -- 2/22/2019 7:56:31 AM >

(in reply to cathar1244)
Post #: 30
Page:   [1] 2   next >   >>
All Forums >> [New Releases from Matrix Games] >> The Operational Art of War IV >> Mods and Scenarios >> Under the Hood Page: [1] 2   next >   >>
Jump to:





New Messages No New Messages
Hot Topic w/ New Messages Hot Topic w/o New Messages
Locked w/ New Messages Locked w/o New Messages
 Post New Thread
 Reply to Message
 Post New Poll
 Submit Vote
 Delete My Own Post
 Delete My Own Thread
 Rate Posts


Forum Software © ASPPlayground.NET Advanced Edition 2.4.5 ANSI

0.156