jwduquette1

#1 is pretty similar to my own line of thinking. Just a straight across the board effective range as a function of a couple of small arms groupings. These could be a different range for each of: SMGs\MPs -- Rifles -- LMGs -- MMGs and HMGs. Or something along those lines. Thanks for the verification of the T-54 and T-55 examples. Dunno how deeply you have dug into NP&W...but I'll toss out another head scratcher'. I've been working through Dupuy's Air Interdiction example in Chapter 6 (or at least its chapter 6 in my edition). It outlines Allied air interdiction of German supplies prior to the Anzio Breakout\Link-up. Operation Strangle. Dupuy works through effectiveness of the interdiction campaign and its impact on German combat power. Dupuy provides the air interdiction details on a couple of Tables in chapter 6. He breaks-down German supply reduction to the Tenth and Fourteenth Armies as a function of Allied Air OLIs committed to the interdiction campaign. He states that the air OLIs are calculated per the rules he details in Chapter 2 of NP&W (ala some of the stuff we have been talking about above). Total air OLIs by aircraft type and their respective sortie numbers are shown on Figure 6-5, Pg 85. For example: March 21, 1944 he shows 11 sorties of P-47s. This -- according to the table on page 85 -- yields a total OLI contribution from P-47s of 671, and a per sortie OLI of 61. He uses an OLI of 61 for P-47s in all of his interdiction calcs for Operation Strangle. Here's my head scratching part: Dupuy has a table of example TLIs he's calc'd back in Chapter 2 (page 26-27 of my edition). One of the examples he provides is the TLI for a P-47. The TLI he gives is 1,245,789. Assuming a WWII dispersion factor of 3000 the resultant OLI is about 415. So what do you reckon is the reason for the huge contrast between the 415 OLI Dupuy calcs for the P-47 in Chapter 2 of NP&W and the 61 OLI Dupuy uses for P-47 Interdiction contribution to his Operation Strangle calcs?

capt luke All the above aside -- when you were playing with NP&W -- did you settle on a rule of thumb and rational for establishing an "effective range" for small arms within the context of calculating TLIs?

I don't think my edition is the revised edition. Just says copyright 1979 at the front end. But yeah I guess I see what you're saying -- or I guess what Dupuy is saying about the range factor for small arms. Presumably the potential lethality index is somewhat a function of the tendency of folks to go to the ground or take cover when being fired upon. Although suppression effects still seem inconsistent with his premise regarding theoretical weapon indices being based upon his theoretical massed formation with a spacing of 1sq-meter between soldiers (as well as all the other aspects of TLI calcs that I mentioned in my above post). As he says its a proving ground index. One would therefore think that the structure of the model would be to deal with the tendency of folks to grab cover within the dispersion factor Di. Or deal with cover grabbing within one of the numerous factors included in the QJMA. Otherwise the whole premise of the 1x1 meter spacing formation has to be tossed as you would now be dealing with a theoretical massed formation with dimensions that are constantly fluctuating as elements within the formation dive for cover. Use of a "effective range" -- as Dupuy does with his AK47, PKM etc examples automatically bumps the resultant TLI for small arms into an average of the two separately calculated range factors and their two TLIs. Conversely if the model considers ONLY muzzle velocity for direct fire weapons a different TLI results. So they are not as you say consistent formats for calc'ing TLI. In Dupuy's AK47 example the difference are TLI = 0.22 for the MV range factor and 0.41 for the effective range calc'd TLI. In one example he uses and effective range for an AK47 assault rifle of 2250-meters. In another example he uses an effective range for a M1 Garand rifle of only 600-yards. The way the Range factors are set up, the final TLI for just about all 20th century small arms will end up being an average of the TLIs determined from the two RFs -- MV and "effective range". Thus the questions regarding what or how is Dupuy defining "effective range" for small arms...I don't mean philosophically what is "effective range" -- I mean a consistent formula for small arms to establish Dupuy's "effective range" for small arms. And again this may sound like a storm in a team cup given the relative lack of importance Dupuy associates with small arms -- but it goes toward one of my initial questions regarding repeatability. LoL -- I've already experienced this in putting together a simple calculator. Three different people I'm working with have used the calculator and all come up with vastly different answers for the same small arms weapons...all of which boils down to what they decide to toss into the calculator for "effective range". Conversely when I limit their input choice to only muzzle velocity -- puff -- problem disappears. 3 different folks all get the same answer -- repeatable. Also -- Did the Punishment Factor for the T-54 and T-55 get changed in the revised issue -- my edition has a table on the first page of Appendix B showing both to have a PF of 54.

OkeeDoke -- thnx Rich.

If you want to go down to a finer level of modeling that Dupuy used, and want to stay consistent with Dupuy's thinking, I think you would model small arms without the MV computed factor at all. You would just compute a range based factor, but based off the furthest range that the small arm could be used to suppress an enemy unit, cause it to go to ground. You will get a much more accurate result, but now your work is much more complicated, and your range assumptions much more arguable, two things that Dupuy seemed to be trying to avoid. Which is somewhat contrary to what Rich just said above -- Dupuy settling on use of the Mv based range factor for direct fire weapons. And just in case -- that's Richard Anderson Jr. with TDI (or I guess retired now and drinking Daiquiris). . Which seems to make more sense than taking the average of the MV determined RF and the effective range determined RF. It also eliminates the guess work associated with determining a max theoretical effective range for direct fire weapons. The context of the model is determination of a "lethality index" for individual weapons. I think the QJM is ultimately supposed to be a casualty prediction model. But perhaps Rich -- if he's not making another blender full of Daiquiris and still reads this thread -- can comment on TLIs\OLIs being combined lethality + suppression effect indices). The nitty-gritty of the whole TLI equation entails -- among other things -- a relative incapacitating effect per blow\hit (RiE). In addition, PTS is potential targets per strike -- for small arms this is taken to be 1 per bullet. PTS for high explosive shells in Dupuy's words is "the theoretical lethality of high explosive shells". This appears to be related to effective fragmentation generated by high explosive shells (there's a curve for PTS as a function of projectile\gun caliber for HE on page 193 of NP&W). "Effective fragmentation" being numbers of fragments produced by a high explosive round that have sufficient energy to produce an incapacitating wound -- so I assume PTS is basically pretty much equal to effective fragmentation of a projectile(?). And finally, the units of TLI are in theory "casualties/hour" for a given battle area or I suppose a given dispersion factor\index (Di). Or at least that's what I've seen from a dimensional analysis study of TLIs. So my take is that lethality index is representation of theoretical casualty producing capability of a weapon rather than a combination of theoretical [suppression + lethality]. Anywhoo...be that as it may...Dupuy provides a number of OLI and TLI example calculations in the Appendix of NP&W (see page 226-227). For mobile weapons (bottom table) Dupuy includes an additional multiplier way over on the right of the table under the header "AM". Dig as I might, there's no definition for what "AM" is in NP&W. It's not included in Dupuy's write-up for determination of OLI\TLI for "Mobile fighting machines", nor does "AM" appear in the definition of terms at the start of the book (see pages xv to xvii). My guess is that maybe this is supposed to be "AME" (Amphibious Effect)? The other thing I noticed is that his calculation for the punishment factors for the T-54, T-55. From EQ4 on page 23 -- with removal of the Dispersion Factor (Di): PF = weight(tons)/4 x SQRT(2*Weight) He puts the weight at 36 tons so shouldn't that yield a PF = 36/4 x SQRT(2 x 36) = 76.4 Dupuy calc's PF as 36/4 x SQRT(36) = 54. Why did he drop the "2" in the sqrt portion of the EQ in his example calc? Is the "2" in the original EQ an error or is his PF example calc for the T54 and T55 in error?

LoL...Official Wargaming newsflash today.

Thanks Rich. TLI sure is artillery\tanko-centric. The model really seems to imply that there wasn't any real use for infantry other than cannon fodder. Particularly reinforced by your comment earlier that TDI often just ignored infantrymen TLIs. I wonder why Dupuy went through the trouble to try and model infantry weapons with the level of detail he does in NP&W. If TLI is taken to be proportional to a units combat power, something like a typical circa-WWII/Korea era infantry battalion would have about 95%(+) of it's combat power wrapped up in a handful of mortar tubes. Same again at divisional level where in an Infantry Division of the same era of warfare will have 95 to 99% of its casualty producing capability concentrated in 1 or 2% of the divisions weapon systems. And by weapon systems, I just mean anything that Dupuy defines as having a TLI -- rifles, MGs, mortars, artillery, etc.

Ideally. That would involve a function that reflects decreasing probability of attaining a hit on a man sized target as range increased. In terms of ranging error -- as I understand it, the calculation of TLI assumes idealized conditions devoid of human error. So shot dispersion includes only inherent weapon error. Which I suppose adds to the confusion associated with the term "effective range" as it's being applied to small arms in NP&W. I just read some operational report on small arms prepared by the US Army -- circa Korea -- the gist of which is the Army looking for a replacement for the M1 Garand. In it they provide all sorts of data on effective ranges of rifles during WWI, WWII and Korea. Basically 100meters is sort of the typical range at which rifle fire in combat was found to be most effective. It's effectiveness drops radically beyond 100meters such that around 300-meters leathality index is approaching zero...i.e. soldiers are rarely hitting things with a rifle at 300m and soldiers are rarely being hit by rifle fire at 300m+. There's some scatter of course -- like fighting on Bougainville max range has dropped from 300m to about 75m...jungle.. max visibility etc.

Kitsap – you're right up the road. I'm living in Portland. I'm just loving all the rain we've been having – not. Supposedly the wettest its been in 75-years. I saw that Chris Lawrence finally published his book on Kursk. http://www.amazon.com/Kursk-Prokhorovka-Christopher-A-Lawrence/dp/0971385254 Would love to buy a copy – but I'm also retired and the wife would probably poison me in my sleep if I purchased it. Out of curiosity is there any delving into QJM or TNDM in terms of testing historical data from the battle against QJM (or whatever QJM has evolved into)? As I recall from “Hitler's Last Gamble” there's only a very small appendix in the back where effectiveness statistics are touched upon. There's a number of older HERO reports available on line now from a variety of sources. Many of the reports I've looked at thus far seem to focus number crunching efforts on Division level engagements – ala the snippets I posted above for the 24th Infantry Division in Korea. Was QJM accuracy for say WWII, Korea engagements + checked mostly against Division level engagements. By that I mean to ask are there boundary conditions within which QJM is most effective at prediction? I don't recall reading anything in NP&W about constraints outside of which extrapolation becomes less and less accurate. In other words if someone (someone being me ) were looking at larger engagements -- say Corps, Army or Army Group -- or smaller engagements – say squad to company size – would they be pushing the models ability to predict results? I guess I am also asking about reproducibility given the equations and parameters Dupuy originally detailed in NP&W.

There are similar math errors (or more likely transposition errors) in NP&W. In the case you point out, I think somehow Col. Dupuy's wife thought dad's rate of fire of 7500 somehow looked like 125 (or whomever was reading his hand written notes and doing the typing -- remember this HERO report was prepared back in the day when stuff often got written by hand and than was subjected to someone slaving away on a typewriter) . That's my guess anyway. The PKM example Dupuy provided in NP&W also has a couple of math errors...or more likely transposition errors. Which is unfortunate in technical\mathematics or statistics based work. It makes it that much more difficult for those unfamiliar with QJM. But I am determined to remain positive toward the approach.

No -- years back I used to post frequently on TDI and that was the first place I thought to go. But the forum appears to be inactive. Tried to post on it with my above questions but it said something along the lines of not possible. I thought maybe someone on tanknet might have fiddled around with QJM and might be able to provide some additional insight. Right now I'm just trying to put together a calculator that produces repeatable numbers that are consistent with methodology detailed in NP&W...as well as various online HERO studies (and etc) and other studies I have found that utilize QJM for a variety of reasons. The equations themselves are not difficult. It's the assumptions behind the various input parameters that I am trying to pin down. Maybe I'll try emailing Chris Lawrence over at TDI.

Yeah -- I have that report -- but it doesn't break down how individual TLIs were being calc'd. The only thing it provides along those lines was the dispersion factor. And notice that the OLI for say the AK-47 differs between the report you cite and the example Dupuy provides in NP&W...even accounting for the contrast in dispersion factors utilized. And yes small arms TLIs are insignificant relative to how Dupuy models TLIs for AFVs, mortars and artillery. At least in battalion size (+) considerations. But they aren't simply ignored in various HERO reports I've seen. my intent is simple -- understanding the mechanics of how to solve for TLI values. The way my mind works, I need to see a few examples of the equation worked out so that I know I am on the right track with my own calcs. That becomes difficult when the requirements for input parameters are not clearly laid out. It's like I'm solving for F=ma -- except the author of the equation doesn't really define what "m" and "a" are supposed to be.

double post...

For anyone actually interested...was poking around google and came across a report Dupuy did using what looks like an early version of his QJM model (screen shots from this HERO report posted below). It seems relatively close to the QJM we see in "Number Prediction & War" (NP&W) -- or at least his descriptions of input parameters for calcing' TLIs sound pretty much identical. The value that's underlined at the bottom of each weapon type is Dupuy's TLI estimate. The effective ranges he is using in his calcs for circa-WWII\Korean era US-Army small arms are considerably reduced from the few small arms examples he lays out in NP&W. Not sure what his thinking was in the NP&W examples and their uber high effective range values. Even the values below seem on the high side. But they are at least somewhat more in the ballpark for what most of us might think of for "effective range" of small arms. Unfortunately I just realized that his sustainable rates of fire in the HERO report (images below) look out of keeping with his rules laid out in NP&W. The sustainable rates of fire for the QJM model are in rounds/hour. Dupuy's NP&W rules for establishing sustainable rate of fire for small arms is something along the lines of: hand and shoulder fired weapons -- Sustainable Rate of Fire (rounds/hour) = ~2 x the maximum rate of fire in rounds per minute. For crew served LMGs, MMGs and HMGs the sustainable rate of fire is something like 4 to 6 times the maximum rate of fire -- dependent upon whether its clip fed or belt fed. As a comparison the sustainable rates of fire Dupuy provides in NP&W are: AK-47: 1280 rounds/hour PKM: 2600 rph 12.7mm M38/46: 2280 rph

yeah -- that's kind of what I'm thinking. One of the go to rules Dupuy provides in NP&W is that effective range can be estimated by: effective range = max range x 0.9 But the context of the above Equation seemed to be fragment producing weapons -- mortars and artillery. Which is easier to wrap ones head around. Applying the same rule of thumb to small arms and calling it "effective range"...