Juardis Posted September 16, 2000 Share Posted September 16, 2000 I was curious as to what the different spotter qualities and mortar team qualities meant as far as results. Also, I wanted to investigate any leader bounuses. So I setup a flat map with light buildings 700m away and 1200m away. First the 81mm mortar team. I had 6 teams, ranging from conscript to elite, all in C&C of a leader. I had another 6 teams, conscript to elite, not in C&C. I learned that leaders impart no additional benefits to mortar teams (other than being able to act as a spotter when a mortar is not direct firing). This makes sense when you think about it. I also learned that the higher the quality of the troops, the more tight the shell pattern (again, makes sense). The conscript teams were horribly imprecise. The elite teams very precise. One last observation (which I read in a Fionn AAR) is that mortar teams spray their shells front and back of target, not really side to side. So if you have a 20m horizontal distance from a mortar target, that unit should be OK. Next, I did the same with 105mm FOs. 6 observers, ranging from conscript to elite, in C&C and 6 FOs not in C&C. I really wanted to know if leaders sped up the time on target. What I found was surprising and if anybody has an explanation I'd appreciate it. I was so surprised I ran the test again, switching buildings to make sure distance wasn't the culprit (it wasn't). I'll give the average of the 2 tests for each troop quality. Hope this comes out in a readable form. CON GRN REG VET CRK ELT 6:12 5:00 2:57 3:01 2:00 2:02 in C&C 6:53 4:45 3:32 2:40 2:12 1:46 not C&C OK, not in C&C was a penalty for conscript, regular and crack units relative to being in C&C. However, suprise number 1 is that green, vets, and elite FOs not in C&C had steel on target 15-20 sec sooner than the ones in C&C. Suprise number 2 is that, when in C&C, regular troops were faster than vets, crack troops faster than elite. What's not conveyed is that the standard deviation for the elite troops is +/- 1 sec while the +/- for crack troops is 7 sec. That is, the fastest of the two times for the crack troops was 2:50 while the slowest was 3:05. So in one case, the crack troops had steel on target 10 sec sooner than the elite troops, but only when in C&C. All the results for the FOs not in C&C were repeatable to within +/- 1 sec. Go figure. Lastly, with regards to preciseness, troop quality matters little and neither did being in C&C. The pattern is horizontal and not front to back like the mortar teams. So if you're being fired upon by arty (not mortars) moving forward or back 20m or so should be get you out of danger. I did one last test, this with 240mm shells. Accuracy did not improve with troop quality, but time on target mimicked the results above for the 105mm results. Hope this helps and looking for explanations if anybody has them. Thanks ------------------ Jeff Abbott Link to comment Share on other sites More sharing options...
L.Tankersley Posted September 16, 2000 Share Posted September 16, 2000 It makes sense that "direct-fire" (i.e.on-map) mortars would be more accurate in azimuth than in range; it's relatively quite easy to see whether your first shot is aligned with the target or not, and if not, how far off you are. Range error is much more difficult to estimate. With off-map artillery, the firers don't have eyes on the target so have to rely on the reports of the FO, who may not be (and likely is not) aligned directly between the battery and the target. So I'd expect a less dramatic deviation from a normal distribution of shell impacts. Also, I seem to recall that off-map artillery is modelled as coming from the friendly board-edge, so even if there IS a bias one way or the other, it may well not be related to the FO-target axis except by chance. No idea what's up with the anomalies in delay times. But I'd want to see more than 2 trials before any definitive conclusions are drawn. ------------------ Leland J. Tankersley [This message has been edited by L.Tankersley (edited 09-15-2000).] Link to comment Share on other sites More sharing options...
RudeLover Posted September 16, 2000 Share Posted September 16, 2000 You said the time to FFE was consistent +/- 1 second in all three trials (105x2, 240)? That surprises me; I thought I'd seen some randomness in arty time to target as well as varation with caliber. Hmm. Not sure I understand your standard deviation. It looks like you're saying mean - stdev = 2:50 and mean + stdev is 3:05 for the Crack FO, but mean in the table above is about 2 minutes. Do you mean regular? Or am I misreading the table? Is N=3? (how many trials?) More tests needed! Very interesting data! Perhaps the National Science Foundation will fund a study.... Link to comment Share on other sites More sharing options...
Wolfe Posted September 16, 2000 Share Posted September 16, 2000 There's a lot of variability in Arty fire when you don't use TRPs (target reference points), particularly for conscript FOs. And it gets even worse when you're firing at something out of LOS. Being in C&C doesn't affect FOs, however. I did some testing and compiled the info in some charts: http://users.erols.com/chare/cm/ They're in Works format (sorry, it's all I have). If you need xls format, I converted them, but they don't print as well. http://users.erols.com/chare/cm/cm_xls.zip - Chris [This message has been edited by Wolfe (edited 09-15-2000).] Link to comment Share on other sites More sharing options...
Juardis Posted September 16, 2000 Author Share Posted September 16, 2000 <BLOCKQUOTE>quote:</font><HR>Originally posted by RudeLover: You said the time to FFE was consistent +/- 1 second in all three trials (105x2, 240)? yes, that's what I'm saying for the FOs that were not in C&C. One clarification, the 240mm results were run twice, but only for those troops not in C&C. The 240mm times to shells landing were different than the 105s though. I somewhat expect that, somewhat don't. GRN REG VET CRK ELT 3:45 3:30 3:00 2:30 2:00 for 240mm 4:45 3:32 2:40 2:12 1:46 for 105mm Hmm. Not sure I understand your standard deviation. It looks like you're saying mean - stdev = 2:50 and mean + stdev is 3:05 for the Crack FO, but mean in the table above is about 2 minutes. Do you mean regular? Doh! You're right, meant regular, not crack. My 2 times for the regulars in C&C were 2:50 and 3:05. Avg is 2:57.5 with a deviation of +/- 7.5 sec. My 2 times for the regulars not in C&C were 3:35 and 3:30. Avg is 3:32.5 with a deviation of 2.5 sec. Yeah, I know, it's hard to draw up statistically significant conclusions based on 2 tests. But I was hoping someone could save me the time of setting it up again and running with different crews on a different map. More tests needed! Very interesting data! Perhaps the National Science Foundation will fund a study.... <HR></BLOCKQUOTE> Agreed. Make checks payable to me Link to comment Share on other sites More sharing options...
Bullethead Posted September 16, 2000 Share Posted September 16, 2000 Juardis- Interesting results you've got here. They are different from what I obtained earlier as to shape of pattern. Perhaps this was changed in 1.05? I'll have to run my own shooting range again . FWIW, last time I did the test, mortars and arty had the same shape of pattern in the same conditions. If there was an LOS to the impact area, the pattern was ovate with the long axis E-W. If no LOS, it came in 2 sizes: a wide cirlce with a TRP, and a very wide cirlce with no TRP. TRPs didn't affect impact pattern if there was an LOS. And rockets ALWAYS made a VERY wide circle, regardless of LOS or TRP. My test range is about 4000x2000m of grid-textured grass. Every 600m is a grain field tile surrounded by 100m of brush tiles. This enables me to accurately determine the size and shape of each pattern, plus obtain average figures for them over many runs. ------------------ -Bullethead Visit the brand new Raider Operations message board at www.delphi.com/raiderops Main site www.historicalgames.bizland.com/index.html Link to comment Share on other sites More sharing options...
Juardis Posted September 16, 2000 Author Share Posted September 16, 2000 <BLOCKQUOTE>quote:</font><HR>Originally posted by Wolfe: There's a lot of variability in Arty fire when you don't use TRPs (target reference points), particularly for conscript FOs. And it gets even worse when you're firing at something out of LOS. Being in C&C doesn't affect FOs, however. They're in Works format (sorry, it's all I have). If you need xls format, I converted them, but they don't print as well. http://users.erols.com/chare/cm/cm_xls.zip - Chris [This message has been edited by Wolfe (edited 09-15-2000).]<HR></BLOCKQUOTE> Interesting data Chris. I took the liberty of copying the German 105mm data below. The LOS row (no TRP) is comparable to my 105mm numbers. They are different again. So I guess what we can take from this is that time on target is different for each battle, even for the same munitions. I think that's good, but in no way do I think that regulars should be as fast as vets, or cracks faster than elites. But with this knowledge, it's good to know that you can spend less yet get the same results almost. 105mm Arty (60) - German CON GRN REG VET CRK ELT 1:29 1:02 0:45 0:38 0:31 0:27 for TRP 2:58 2:04 0:28 1:15 1:01 0:52 for TRP no LOS 6:09 4:59 2:50 2:51 2:21 2:01 for LOS 10:45 9:28 5:45 5:40 4:19 3:33 for no LOS 6:12 5:00 2:57 3:01 2:00 2:02 in C&C 6:53 4:45 3:32 2:40 2:12 1:46 not in C&C [This message has been edited by Juardis (edited 09-15-2000).] Link to comment Share on other sites More sharing options...
Wolfe Posted September 16, 2000 Share Posted September 16, 2000 <BLOCKQUOTE>quote:</font><HR>time on target is different for each battle<HR></BLOCKQUOTE> Nope. Each FO has a different ETA. Take 6 conscript 170mm German FOs. Have them fire at a spot they can't see. Don't use a TRP. Here are the times it took each FO to begin firing. Remember there's nothing at all different about these FOs, and they're all exactly the same and participating in the same battle. 1st FO begins firing at 9:48 2nd FO 11:24 3rd FO 11:46 4th FO 12:27 5th FO 12:44 6th FO 13:08 The first FO actually *finished* shooting off all his rounds before the 6th one even began firing. It varies from FO to FO not from battle to battle. And the less experienced the FO, the more variability in ETA you get. Of course, if you use a TRP, then the numbers you'll see for the ETA are *always* the same from FO to FO. If I had used a TRP in the test above, all of the FOs would have begun firing at exactly 2:58 (well, the clock would show it is in turn 3, but it takes 2 minutes and 58 seconds to get the 'firing' message). <BLOCKQUOTE>quote:</font><HR>in no way do I think that regulars should be as fast as vets, or cracks faster than elites.<HR></BLOCKQUOTE> They're not. The numbers I generated are *NOT* averages; I only tested each once. What you're seeing are outliers, not hard consistent data. I only did the tests to get a rough idea of when to expect arty to arrive. And the numbers in my charts are indeed very rough. In order to get a more accurate view of how long it takes (say) a conscript 170mm FO to fire at a target he can't see, you'd have to run dozens of tests to get the overall mean time it takes to begin firing and then do a standard deviation to see how far off the ETA for each FO can be. That's the only way to get reasonably accurate ETA numbers. I didn't do this because it would be one Hell of a lot of work and I'm just not that bored. - Chris [This message has been edited by Wolfe (edited 09-16-2000).] Link to comment Share on other sites More sharing options...
Bullethead Posted September 16, 2000 Share Posted September 16, 2000 Well, folks are working on the timing aspects so I decided I'd do the pattern aspects in 1.05. I have 8 bullseyes on my range. This enables me to test all combinations of the following variables: gun/mortar/rocket type, LOS, TRP, and wide vs. regular spreads. Each test consisted of setting up 8 FOs, 1 per target, and giving them each 100 rounds (200 for mortars because they seem a bit wilder). Then I'd change FO type and/or nationality and do it again. All FOs used were regulars. Here are the averaged results: GUNS Regular Spread LOS & TRP: 100x50m, oriented E-W Blind & TRP: ditto LOS only: 140x60m, oriented E-W Blind only: 200x100m, oriented E-W Wide Spread LOS & TRP: 250m diameter circle Blind & TRP: ditto LOS only: ditto Blind only: 400-450m diameter circle MORTARS Regular Spread LOS & TRP: 150x60m, oriented E-W Blind & TRP: ditto LOS only: ditto Blind only: 250x100m, oriented E-W Wide Spread LOS & TRP: 300-350m diameter circle Blind & TRP: ditto LOS only: ditto Blind only: 400-450m diameter circle ROCKETS Wide Spread Unavailable, Only Regular All LOS/TRP combinations: ~600m diameter circle A few things to note: 1. These results are very similar to, but not quite the same as, those I posted a while back with some earlier version. 2. All impact patterns were centered on the point of aim. That is, there is no "scatter" to the over-all pattern regardless of LOS or TRP. This makes sense because the FFE isn't going to start until the FO has adjusted it onto the target. 3. If you have a TRP, you always get the same pattern regardless of LOS. 4. For the most part, if you have a either a TRP or an LOS, you get the same pattern. Only when you have neither do you get a bigger pattern. The exception is with guns, which have 3 steps of pattern size on the regular spread: TRP (LOS or blind), LOS only, and blind only. 5. Guns shoot slightly tighter patterns than mortars under the same conditions of LOS and TRP. This should be no surprise, because guns are controlled and fired by artillerymen, whereas mortars are done by grunts 6. Rockets are a special case. They only have the regular spread option, and they always shoot a HUGE circular pattern regardless of LOS and TRP. The only thing these variables affect is the time to FFE. On the subject of time for FFE, I make the following general observations: 1. All FOs have a base time that shows up in their info window when you call for fire. This base time is a function of a) weapon type (mortars are faster than guns and rockets) and nationality. The base time always shows UNLESS you're shooting at a TRP, which reduces the base time. But if the actual time is greater than the base time, you still only see the base time. 2. If the target is within the magic distance from a TRP, this base time is halved. 3. If the FO has no LOS, it will take at least twice the base time to get FFE. This is cumulative with the TRP bonus, so a blind TRP shot takes the same amount of time as a non-TRP shot in LOS, ie, approximately the base time. 4. Regardless of what the base time says, when it finally starts counting down like a timer to FFE, its time can run slower than the game turn clock. How much slower seems to depend on the existence of LOS or TRP. Troop quality probably also affects timer speed, but I didn't test this. 5. Besides starting FFE faster, TRP shots usually do not use spotting rounds before the FFE starts. Just BOOM, there it all is, so there's no warning to the enemy. ------------------ -Bullethead Visit the brand new Raider Operations message board at www.delphi.com/raiderops Main site www.historicalgames.bizland.com/index.html Link to comment Share on other sites More sharing options...
KiloIndiaAlpha Posted September 16, 2000 Share Posted September 16, 2000 Did you know that 78.45% of all statistics are made up on the spot? Seriously chaps, to make any assertions from this kind of numerical data you really must do lots of trials, I mean 100's at the very least and 1000's to get in the realms of having confidence in the results. There are perhaps 2 ways of doing this, first is to ask BTS for the raw algorithms... Somehow I don't think so. The other alternative would take a little organising but may work. First create several test cases, ie make some battles with the arty and FOO's etc. Post these on a web site and allow people who want to participate in this analysis to DL them, may be good idea to write instructions etc in the scenario briefing. So the result is 10, 20 30, 100 guys or more running test cases. Nominate someone to collect the raw data and then after a week or so? some serious stats can be done. Just my 2 pence worth Link to comment Share on other sites More sharing options...
KiloIndiaAlpha Posted September 16, 2000 Share Posted September 16, 2000 . Link to comment Share on other sites More sharing options...
pritzl Posted September 16, 2000 Share Posted September 16, 2000 KIA, you need to check your information my friend. Perhaps your source is out of date? Currently 82.37% of all statistics are made up. Link to comment Share on other sites More sharing options...
Guest Michael emrys Posted September 16, 2000 Share Posted September 16, 2000 Bullethead, did you run a test on each of the seperate calibers? Or are you saying that it doesn't matter? My casual observation in the game seems to suggest that smaller calibers in each category fire tighter patterns. I'm also a bit surprised at your observation re guns vs mortars, but for now I'm willing to take your word for it. Michael Link to comment Share on other sites More sharing options...
Bullethead Posted September 16, 2000 Share Posted September 16, 2000 Michael said: <BLOCKQUOTE>quote:</font><HR>Bullethead, did you run a test on each of the seperate calibers?<HR></BLOCKQUOTE> Not this time. I picked 3 of each type (gun, mortar, and rocket), at each end and in the middle of the size scale. I saw no noticeable difference due to caliber or nationality within each type. This corresponds with my test of all arty types in an early version, so I didn't bother doing them all this time. However, I did not test the very biggest things like 14" NGFS, which I consider too rare to worry about anyway. Besides, with a kill radius vs. light AFVs of over 100m, who really cares what the 14" pattern is? <BLOCKQUOTE>quote:</font><HR>I'm also a bit surprised at your observation re guns vs mortars...<HR></BLOCKQUOTE> So was I. Last time I did this test, there was no difference between guns and mortars. Both gave exactly the same pattern under the same LOS/TRP conditions. <BLOCKQUOTE>quote:</font><HR>...but for now I'm willing to take your word for it.<HR></BLOCKQUOTE> Well, you don't have to be. I think KIA is right, we do need a bigger sample size. So how about I send you my arty range "scenario", so both of us together can thoroughly test all types of arty. State your choice of being Axis or Allies and I'll set it up that way. Then all you have to do is change the type of FOs and their ammo. My email is jtweller@delphi.com. Let me know if you want it. ------------------ -Bullethead Visit the brand new Raider Operations message board at www.delphi.com/raiderops Main site www.historicalgames.bizland.com/index.html Link to comment Share on other sites More sharing options...
Wolfe Posted September 17, 2000 Share Posted September 17, 2000 Bullethead: very interesting stuff. I did a bit of testing myself and it looked to me like the overall patterns are similar between different size rounds as well as different types. Though I didn't try to measure them. I took 10 German 81mm mortar FOs and 10 210mm gun FOs and aimed all of them at the same spot (no TRP, and both in and out of LOS). It looked to me like, in total, both the 210mm and 81mm mortars covered the same area. The outer boundaries (farthest distance from the target) where shells fell seemed to be the same for both weapon types. Both seemed to have the same "tightness" of shell dispersal when firing at a target within LOS and also the same "looseness" when fired out of LOS. I did the same with the 210 and 75mm and it looks like similar results. Although I would actually give the edge to the 210mm rounds for their accuracy on target (I only did this one within LOS). BTW, I do like the East-West oval pattern that that many rounds creates around a target area. The one thing that didn't seem to matter was the experience level of the FO. Conscript, Regular, and Elite all produce the same size pattern. This was tested out of LOS with each target area getting 10 FOs a piece. Again, no TRPs were used. BTW, does CM have it correct when you're unable to adjust fire for targets out of LOS? That came as a shock in a game when I discovered I wasn't able to simply adjust fire on a target out of LOS, and instead had to start a new area fire countdown. - Chris Link to comment Share on other sites More sharing options...
Bullethead Posted September 17, 2000 Share Posted September 17, 2000 <BLOCKQUOTE>quote:</font><HR>BTW, does CM have it correct when you're unable to adjust fire for targets out of LOS? That came as a shock in a game when I discovered I wasn't able to simply adjust fire on a target out of LOS, and instead had to start a new area fire countdown.<HR></BLOCKQUOTE> Hmm, I hadn't noticed this. I've been able to adjust (actually should be called shift) blind FFE with the short time as normal. FWIW, you SHOULD be able to adjust blind FFE. The fact that the FFE starts at all indicates the FO has some way to get adjust the spotter rounds on target (via sound and/or smoke). Once on for FFE, though, it's a simple matter to say "drop 100, repeat" even blind. BTW, I don't agree with the area in which "adjusted" FFE is possible. I think it should be in thin donuts at 100m and 200m from the initial point of aim, and not everywhere inside a small area around the initial target. But that's another story. ------------------ -Bullethead Visit the brand new Raider Operations message board at www.delphi.com/raiderops Main site www.historicalgames.bizland.com/index.html Link to comment Share on other sites More sharing options...
Slapdragon Posted September 17, 2000 Share Posted September 17, 2000 This type of testing that Abbott started and Bulletthead and Wolfe also do is excellent for those of us who want to "kick" the tires and really know what is happening, not just use antecedents (your Stuart killed my Tiger, fix the Stuart so that cannot happen). For those of you that want to try scientifc testing (experiment): here is how it is done: 1) Come up with a problem 2) Figure out what factors effect the problem then choose one and only one to vary. For example: how are rear shots on Tigers effected by distance. If you let other factors creep in (different types of guns, experience, smoke was over part of them but not over others etc) then you will not have a valid test. 3) Run the test and measure the results. 4) In any system that has any random elements, you need to run the same test 40 times for each variable difference. For example, when I tested how often M4A3's used tungsten, I did two sets of 40 firings of the first shot only at a Panthe dug in which accounted for all the variables. You need at least 40 so you can say with some confidence that this is happening (if you get into advanced stats you can know exactly how confident you are.) 5) The best stats to use are the simplest: mean, percentage, and standard deviation. If you know stats, you can use correlation and regression. If you really want to get weird, you can use nonparametic stats, but no one will understand you. 6) Once you discover something, tell everyone else but tell them exactly how you did the test. The ideal thing is for someone else is to see how you did the test and go test it themselves. When they come back with the same or similar numbers then you know you were right. If they come back with different numbers then you have to figure out what you did differently. If you don't tell people how you came up with the information, they wont be able to see if you were right. 7) If this sort of accuracy is not your cup of tea, don't worry. Steve Jackson [This message has been edited by Slapdragon (edited 09-17-2000).] Link to comment Share on other sites More sharing options...
Juardis Posted September 17, 2000 Author Share Posted September 17, 2000 <BLOCKQUOTE>quote:</font><HR>Originally posted by Bullethead: Well, folks are working on the timing aspects so I decided I'd do the pattern aspects in 1.05. MORTARS Regular Spread LOS & TRP: 150x60m, oriented E-W Blind & TRP: ditto LOS only: ditto Blind only: 250x100m, oriented E-W Wide Spread LOS & TRP: 300-350m diameter circle Blind & TRP: ditto LOS only: ditto Blind only: 400-450m diameter circle <HR></BLOCKQUOTE> What are you saying about mortars? That the shell spread is to the left and right of the target? I'm assuming this is for offboard mortars, correct? My tests were conducted with on board 81mm mortars and I definitely saw a N-S shell pattern (front and back of target, not left or right of target). So we could both be correct and should be one way to tell what type of mortar is firing at you (onboard or offboard). If onboard, move laterally to get out of danger. If offboard, move forwards or backwards (if I understand what you're saying that is)? ------------------ Jeff Abbott Link to comment Share on other sites More sharing options...
Juardis Posted September 18, 2000 Author Share Posted September 18, 2000 One other question regarding the oval pattern. Was that for a area target (or wide area target) or direct fire? For FOs, I gave them heavy buildings as targets and got the e-w oval pattern. But my on-board mortar teams, when direct firing, gave me the n-s line pattern. Perhaps if you give an on-board mortar team an area target, then you get an e-w oval pattern? Link to comment Share on other sites More sharing options...
Bullethead Posted September 18, 2000 Share Posted September 18, 2000 Juardis said: <BLOCKQUOTE>quote:</font><HR>What are you saying about mortars? That the shell spread is to the left and right of the target? I'm assuming this is for offboard mortars, correct? My tests were conducted with on board 81mm mortars and I definitely saw a N-S shell pattern (front and back of target, not left or right of target).<HR></BLOCKQUOTE> I was using off-map mortars. Under certain conditions as noted above, they produce E-W patterns (left-right as you see the map in the editor). IOW, ahead of and behind the target, assuming the Allies are coming from the west. This is the same type of pattern, but slightly bigger under the same conditions, as off-map guns make. So if you want to get out of the way of it, move toward the flanks, not forward or backwards, again assuming the Allies are coming from the west. <BLOCKQUOTE>quote:</font><HR>One other question regarding the oval pattern. Was that for a area target (or wide area target) or direct fire? For FOs, I gave them heavy buildings as targets and got the e-w oval pattern. But my on-board mortar teams, when direct firing, gave me the n-s line pattern. Perhaps if you give an on-board mortar team an area target, then you get an e-w oval pattern?<HR></BLOCKQUOTE> I don't know. I haven't tested on-map mortars. Lack of interest, really. Being on-map just increases vulnerability and decreases flexibility. If you want 81mm mortars, get an FO and leave the tubes off the map. ------------------ -Bullethead Visit the brand new Raider Operations message board at www.delphi.com/raiderops Main site www.historicalgames.bizland.com/index.html Link to comment Share on other sites More sharing options...
Recommended Posts