PDA

View Full Version : Question on Parameter Data



jztemple
25 Jun 08, 17:06
Well, give a man a few editors and he comes dangerous!! :cool:

These parameter data files are from Pacific War:
Main.pdt
Main-alt.pdt
MainHigh.pdt
MainHighNT.pdt
MainLight.pdt
MainLightHigh.pdt
MainLightHighNT.pdt
MainLightNT.pdt
MainNT.pdt

So I'm trying to figure out which is which so I can make some fun scenarios. Now, what's the meaning of "High", "Light", "NT" and "-alt" in the titles?

OK, since I'm not helpless I did find a text compare program (TextDiff) and compared "Main.pdt" and "Main-alt.pdt". The only difference is in line 23, where a "0" becomes a "7", but looking in the Parameter Data menu generated by F2 in the game, there aren't any sevens! So I suspect that line is a summary of binaries, perhaps, as in 0111.

Just to be thorough, I loaded up the game with "Main.pdt" as the parameter file for the scenario, pressed F2 to get the Parameter Data Dialog (can you tell I'm an engineer?), did a "Select All" from a right click menu, then pasted it in a text file. I did the same for "Main-alt.pdt" as the file for another scenario. When I compared the two using TextDiff, I found the difference was that "Main-alt" allowed Jungle Indirect Fire. Cool.

jztemple
25 Jun 08, 17:19
Ah, heck, since I'm doing the research I figure I might as well double post. Here are my findings:

These parameter data files are from Pacific War:

Main.pdt = default parameter file

Main-alt.pdt = only change is that indirect jungle fire is allowed

MainHigh.pdt = only change is that in "Main.pdt" infantry and wheeled vehicles cannot cross reef hexsides and tracked vehicles pay 6MP. In MainHigh, all three cross a reef hexside at no cost.

MainLight.pdt = one change is that indirect jungle fire is allowed. Another is that while in Main tracked vehicles cannot enter jungles hexes, in MainLight they can (note that infantry can always enter, wheeled may never enter).

MainNT.pdt = In Main, tunnel movement is allowed. In NT, it's not.

MainHighNT.pdt = Same as "MainHigh" plus "MainNT". effects.

I'm guessing that the last three files, MainLightHigh.pdt, MainLightHighNT.pdt, and MainLightNT.pdt, are just the combination of effects as mentioned in their titles.

Ok, that mystery cleared up! Only took an hour, but it was fun.

rahamy
26 Jun 08, 07:11
Love to see guys working in the files...that's meant to be one of the benefits of the series. However, with the latest update there is a parameter data editor...so you shouldn't have to be doing any copy & pasting or comparing to figure these things out!

jztemple
26 Jun 08, 08:50
Love to see guys working in the files...that's meant to be one of the benefits of the series. However, with the latest update there is a parameter data editor...so you shouldn't have to be doing any copy & pasting or comparing to figure these things out!

Yup, I've played with the parameter data editor. For my exercise above, I was trying to compare one file to another and the editor doesn't really help much with that. However, the editor does make changes much, much easier, unlike the old method of editing the pdt file, then loading up the game and looking at the F2 Parameter Data Dialog each time.

I'm pretty happy with my methodology, if I do say so myself. That TextDiff utility is pretty neat, I've used it in the past to check out code changes. It displays two windows side by side with the lines that have changes highlighted. It will also generate a report containing the changed lines.

With Winter War released, will the datafiles of AotR be unlocked in the next update? I think it's the last one that's still locked.

Also, in the Database Editor, that new function that lists the unused loads and vehicles weapons is pretty cool. I know it can be used to trim down the list, but it also flags me to check out those items which I might not have seen before in the game.

rahamy
26 Jun 08, 12:02
With Winter War released, will the datafiles of AotR be unlocked in the next update? I think it's the last one that's still locked.


Nope, sorry...its locked because the second title of the Eastern Front is still to come...



Also, in the Database Editor, that new function that lists the unused loads and vehicles weapons is pretty cool. I know it can be used to trim down the list, but it also flags me to check out those items which I might not have seen before in the game.

Cool...John has really added some value with the various new editors. Hope to see some new custom stuff from people before too long...this is a very moddable series, but few take advantage of it.

Bill Thomas
01 Jul 08, 14:41
...I loaded up the game with "Main.pdt" as the parameter file for the scenario, pressed F2 to get the Parameter Data Dialog (can you tell I'm an engineer?),

I too am an engineer. Numbers, statistics and graphs drive my world.

Your testing has inspired me to ask you if would investigate Assault Combat (concerning losses). I’ve already analyze Fire Combat routines by collecting test data and graphing the results which agree with the details as stated in the manual.

However, Assault Combat does not agree with the manual. I can’t complete an equation that duplicates observed results. Maybe you can.

The following was an email I sent to Ozgur about six months ago:



Ozgur,

I have a question concerning Squad Battles that I have been unable to determine through trials, observations and calculations. Before I send an email to John concerning my dilemma, I thought I would ask a veteran of SB. I believe that you have worked with John in the past and you have a high profile on both the Blitz and Gamesquad forums so I believe that you are more than qualified.

I’ve been playing Squad Battles since the release of The Proud & the Few (I’m a former USMC-R myself) and have added other titles as well. I wrote the program “Tactical Aid” for TP&TF that Frank Harmon posted on the ECHO-4 website. I play fairly often, however I’ve never played a head-to-head game. So I would not consider myself a novice but I can’t seem to grasp the intricacies of Assaulting.

Assault combat in Squad Battles as described in the user manual gives a general idea as to determining how casualties are calculated, however some further investigation is need to understand the seemingly complicated routine.

It is general knowledge that the casualties from assaulting are determined by the standard Combat Results calculation. This calculation involves the following:

1) Attacker_Casualties_Low = Defending_Assault_Value / Baseline_Combat_Value x Attacker_Low_Loss

2) Attacker_Casualties_High = Defending_Assault_Value / Baseline_Combat_Value x Attacker_High_Loss

3) Defender_Casualties_Low = Attacking_Assault_Value / Baseline_Combat_Value x Defender_Low_Loss

4) Defender_Casualties_High = Attackinging_Assult_Value / Baseline_Combat_Value x Defender_High_Loss


Where:
Baseline Combat Value = 10,000 (constant)
Attack Low Loss = 24 (constant per parameter file)
Attack High Loss = 240 (constant per parameter file)
Defend Low Loss = 12 (constant per parameter file)
Defend High Loss = 120 (constant per parameter file)
Attacking Assault Value: (variable)
Defending Assault Value: (variable)


The calculations results of 1) & 2) give a high to low range casualties for the attacker. A random number is generated between this range and is applied to the attacker. The calculations for 3) & 4) do the same for the defender.
So far everything is crystal clear. Now the hard part.

To determine the Attacking / Defending Assault Value includes many separate numeric values, to include;
* Squad assault value = squad strength x the Infantry Assault Parameter (24, constant per parameter file).
* Weapon assault values (per weapons parameter file)
* Vehicle assault values

However an “aggregate morale” is not defined as to how this value is used in the Attacking / Defending Assault Value calculation.

In an attempt to understand the assault combat system, I’ve collected data from two test combats composed of the following:

1st Test:
Attacker: one USMC squad, 10 str, 100% eff., A morale, no weapons, non-ground
Defender: one IJA squad, 10 str, 100% eff., A morale, no weapons, non-ground
Terrain: Clear, normal conditions
No optional rules in-play
Latest patch installed
Test Sample: 50

Results:
Attacker:
Max Casualty: 9
Min Casualty: 0
Average Casualty: 2.36
Median Casualty: 2
Mode Casualty: 0
Max Eff. Remaining: 100%
Min Eff. Remaining: 88%

Defender:
Attacker:
Max Casualty: 4
Min Casualty: 0
Average Casualty: 1.4
Median Casualty: 1
Mode Casualty: 1
Max Eff. Remaining: 100%
Min Eff. Remaining: 89%

2nd Test:
Attacker: one USMC squad, 10 str, 100% eff., A morale, no weapons, non-ground
Defender: one IJA squad, 10 str, 100% eff., F morale, no weapons, non-ground
Terrain: Clear, normal conditions
No optional rules in-play
Latest patch installed
Test Sample: 50

Results:
Attacker:
Max Casualty: 8
Min Casualty: 0
Average Casualty: 2.44
Median Casualty: 2
Mode Casualty: 1
Max Eff. Remaining: 100%
Min Eff. Remaining: 92%

Defender:
Attacker:
Max Casualty: 4
Min Casualty: 0
Average Casualty: 1.16
Median Casualty: 1
Mode Casualty: 1
Max Eff. Remaining: 100%
Min Eff. Remaining: 88%

For the 1st and 2nd Test the expected results (excluding the “aggregate morale”):

Attacker Casualties High = 4.8 or 5 randomly rounded up
Attacker Casualties Low = 0.48 or 0 randomly rounded down
Defender Casualties High = 2.4 or 3 randomly rounded up
Defender Casualties Low = 0.24 or 0 randomly rounded down

After reviewing the expected results compared to the collected data it’s apparent that the “aggregate morale” plays a considerable part in the Assault process.

I then attempted to remove the randomness from calculation by changing the attacker / defender high~low loss parameter with all four values set at 200. Insanely, this did not remove the randomness at all! Why?

So in conclusion, can you help me understand the calculation that concerns the Assault so I can sleep at night. I’ve just notice the length of this email and I apologize. I’m an a Engineer and I can’t help myself (maybe I should seek counseling ).

Thank-you,

William Thomas (AKA Rook & Cheeks)

Mike Cox
01 Jul 08, 22:47
Thank god that got addressed to Oz. I have now have a headache....!

Dazooz
01 Jul 08, 23:57
Wow! You engineers scare me! :)

I hope you don't go through these calculations in your head while playing, I play this game so much more by feel and less mathematical.

Ozgur Budak
02 Jul 08, 01:52
I think such math questions can be answered by John Tiller. He is the one that writes the code and algorithms. Besides I never approach the games from math angle. I love military history and wargaming. It is fine with me as long as it provides an accurate and enjoying experience. Contrary seeing the game as bunch of algorithms (not saying it is not) ruins my enjoyment. Sorry guys.

Bill Thomas
02 Jul 08, 08:53
Wow! You engineers scare me! :)

I hope you don't go through these calculations in your head while playing, I play this game so much more by feel and less mathematical.

Your "feeling" can be translated to the nerd language meaning "probability" :) . Everyone realizes that firing an M1 Garand at a PzIII is a futile gesture. However, employing a Bazooka and firing at the rear of a PzII is almost guarantying a kill while firing at the front of a PzIV at a range will result in a low probability of killing it.

New players may ask why this is the case. All the equipment in SB is given in a numerical format of variables & constants which is difficult to visualize. My goal is to show “what is possible”.

I would like to give an example to the SB community. I have created a “Quick Reference Graph” concerning Fire Combat that relates Fire Value, Casualties & Efficiency Loss to High, Nominal & Low probabilities for both men & vehicles. Let me state that this is my interpretation of the mechanics of SB. No one has back-checked the graph for correctness.

TheBigRedOne
02 Jul 08, 08:54
Math is why I went into organic chemistry and not engineering....

:)

Ozgur Budak
02 Jul 08, 09:22
Math is why I went into organic chemistry and not engineering....

:)

And why I am a social scientist :D

Dazooz
02 Jul 08, 10:36
Your "feeling" can be translated to the nerd language meaning "probability" :) . Everyone realizes that firing an M1 Garand at a PzIII is a futile gesture. However, employing a Bazooka and firing at the rear of a PzII is almost guarantying a kill while firing at the front of a PzIV at a range will result in a low probability of killing it.

New players may ask why this is the case. All the equipment in SB is given in a numerical format of variables & constants which is difficult to visualize. My goal is to show “what is possible”.

I would like to give an example to the SB community. I have created a “Quick Reference Graph” concerning Fire Combat that relates Fire Value, Casualties & Efficiency Loss to High, Nominal & Low probabilities for both men & vehicles. Let me state that this is my interpretation of the mechanics of SB. No one has back-checked the graph for correctness.

I appreciate all of the work that goes into making this a great series, including all that the community does outside of actual developers. Hope I didn't step on any toes with the math comment! :)

In fact, I majored in mathematics myself, however I was always drawn more to the theory side of things rather than the probability/statistics side.

I know deep down inside that these games are run completely by numbers, but I don't like to think about it. I prefer to lose myself in the game itself and hope it plays out like what I perceive a real battle would be like.

I dislike games that require a mathematical response (where every possible outcome is taken into account and the only feasible response is the one certain thing). The randomness of warfare is what makes it so challenging and interesting.

Squad Battles have always felt to me like I could try different things and while certainly some things are riskier than others, occasionally it pays off!

All this being said, I again understand that it is all numbers driven, I just don't like to think about it too much! :)

jztemple
02 Jul 08, 12:39
Obviously you must post your graph! We are waiting anxiously with all our nerdy synapses a-tingle.

FastPhil
02 Jul 08, 12:46
Wow! You engineers scare me! :)

I hope you don't go through these calculations in your head while playing, I play this game so much more by feel and less mathematical.

Yup, some days ya feel like dieing some days ya don't. Bill you ought to drop the engineering and join the Marines again-drop the slide rule and get down and dirty. To be honest I don't want assault to be calculable-or predictable.
Murphy needs to be somewhere in the calculation. JMHO:D

jztemple
02 Jul 08, 12:48
I do have to agree that knowing the numbers can be fun, but that come from my old Avalon Hill and SPI days, when you would be calculating which number of pips showing would be best for you.

Bill Thomas
02 Jul 08, 19:09
My appologies..., I thought I had this document in a PDF format and I didn't. Apparently I no longer have my adobe distiller either.

Anyways I present my PRELIM graphs as promised in MS word format.

Let the stone throwing begin :)

FastPhil
02 Jul 08, 19:23
I do have to agree that knowing the numbers can be fun, but that come from my old Avalon Hill and SPI days, when you would be calculating which number of pips showing would be best for you.

Been there and done that.:smoke: The old viipuri Defense, Afrika Corps, Bulge and final ASL.:D

FastPhil
02 Jul 08, 19:29
My appologies..., I thought I had this document in a PDF format and I didn't. Apparently I no longer have my adobe distiller either.

Anyways I present my PRELIM graphs as promised in MS word format.

Let the stone throwing begin :)

Why would we throw stones? Did SB go prehistoric?:clown: Why not hand Grenades?:ar15: (couldn't find a grenade smiley) Honestly why would we throw stones at your efforts. Now if you wanted me to do all this work, I might have to resort to violence.:argh:

rahamy
02 Jul 08, 21:34
Here it is in PDF if someone can't open the DOC file.

Dazooz
02 Jul 08, 23:13
Let the stone throwing begin :)

Certainly no stones Bill. I'm glad there are guys like you to help perfect those numbers! :)

Bill Thomas
03 Jul 08, 10:03
A clairification is needed when viewing my Fire-vs-Armor graph that Rich posted as a PDF. For some reason the X coordinate (Final Protection Values) is shaded black. These values are present in the DOC file.

If someone needs assistance in understanding this work, please ask. I'd be happy to help. I am also interested in feed-back, be it bad or good.