ww2steel Posted June 20, 2006 Share Posted June 20, 2006 Okay, I have run it 300 times... (This is at 100m.) T-34/85M43 with APCR (tung) vs Panther A early: T-34 casualties = 89%. T-34/85M43 WITHOUT APCR vs Panther A early: T-34 casualties = 84%. Why does the tank do better when you take away the specialized ammo? I have theories, but I want to here other peoples' before I throw mine out there. 5% isn't much, but after 300 total tests, I'd think it would be significant. Ideas? Thanks, Mike [ June 20, 2006, 04:50 PM: Message edited by: ww2steel ] 0 Quote Link to comment Share on other sites More sharing options...
Kingfish Posted June 20, 2006 Share Posted June 20, 2006 What range? The effectivness of T rounds drops off after a certain distance / angle What are the other parameters? Both crews equal experience? Both tanks facing each other so spotting and LOS is not a factor? What do you mean by 89% casualty? 267 out of 300 times the T34 dies? 0 Quote Link to comment Share on other sites More sharing options...
ww2steel Posted June 20, 2006 Author Share Posted June 20, 2006 100m Everything identical in all respects except ammo loadout of the T-34, regulars both sides. Run exactly 200 times- 89% casualties (89% of the T-34s destroyed/ significantly disabled). Run more times because this figure seemed more suspect due to the rest of the graph. I have actually run 900 total engagements between these two vehicles. Run exactly 100 times- 84% casualties when the APCR is removed. Mike 0 Quote Link to comment Share on other sites More sharing options...
David I Posted June 20, 2006 Share Posted June 20, 2006 Luck. Try another 300 tests. DavidI 0 Quote Link to comment Share on other sites More sharing options...
ww2steel Posted June 20, 2006 Author Share Posted June 20, 2006 Don't really understand the rolling eyes... what's that all about? 0 Quote Link to comment Share on other sites More sharing options...
YankeeDog Posted June 20, 2006 Share Posted June 20, 2006 So you are comparing two sets of data; in one N=200; the other N=100. I'm pulling this out of my @ss, but I'm not at all sure a deviation of 5% is statistically siginificant. You'd have to do a formal statistical analysis to be sure. 15 years ago, when I was just out of prob/stat, I could do these calcs off the top of my head. Now, I'm lucky to be able to balance my checkbook. Maybe someone else here has fresher statistical analysis skills. I'm too lazy to go dig out my old textbook. . . 0 Quote Link to comment Share on other sites More sharing options...
ww2steel Posted June 20, 2006 Author Share Posted June 20, 2006 I was just alt tabbing out of my game to ask this. How can I calculate how many tests I need to do to eliminate the statistical noise down to maybe 1%, 2%, whatever? Will someone please help me? Thanks! Mike 0 Quote Link to comment Share on other sites More sharing options...
Kingfish Posted June 20, 2006 Share Posted June 20, 2006 Do what the Russians historically did. Place 4 T34/85s 100 meters in front of the Panther and then run the test 300 times. Ammo should work fine now. 0 Quote Link to comment Share on other sites More sharing options...
ww2steel Posted June 20, 2006 Author Share Posted June 20, 2006 LOL, very true. If I have time I was thinking of running a few of the more common combos like this in a two on one scenario. Mike 0 Quote Link to comment Share on other sites More sharing options...
zmoney Posted June 21, 2006 Share Posted June 21, 2006 If I understand you correctly, Tung rounds don't do as well vs curved or rounded targets. The flatter nose regular AP rounds are better because they tend to grip onto a target that is curved. Where as the Tung which has a very sharp point is more likely to slide off an equally curved surface. On the other side of the coin Tung with its sharper point is better at penetrating flatter surfaces than is the flat nose AP round. So maybe you should run a test between T-34/85 with Tung vs Tiger I’s which have a flatter surface. 0 Quote Link to comment Share on other sites More sharing options...
ww2steel Posted June 21, 2006 Author Share Posted June 21, 2006 Already done, the Tiger comes out normally. The Panther comes out normally as well, when you take away the APCR from the T-34/85, from ranges 300m - 1000m the T-34 suffers by about 8%... except at 100m, where it gets an ADVANTAGE by me TAKING AWAY the few APCR roudns it carries. I just don't understand why if it does better firing APBC rounds why it doesn't just disregard the APCR (it stops firing APCR at the Tiger under 300m and switches to APBC only). And yes, APBC especially does better than APCR when striking at an angle. I am using the later APBC ammo instead of earlier AP. Mike 0 Quote Link to comment Share on other sites More sharing options...
Andrew H. Posted June 21, 2006 Share Posted June 21, 2006 Zmoney is correct; there was a lot of discussion about this back in the day - maybe as far back as the CMBO days. Basically, tungsten's penetrating ability drops off more significantly vs. highly sloped targets than conventional munitions do. Meaning that in certain circumstances, conventional munitions may have greater penetrating power. 0 Quote Link to comment Share on other sites More sharing options...
ww2steel Posted June 21, 2006 Author Share Posted June 21, 2006 Okay, thanks. I think if I missed that I really would have been confused a long time ago. My point, again. 1) Even against the Panther, the APCR gives an 8% advantage at medium ranges. (Turret face) 2) At 100m, with BOTH APBC and APCR on board the AI fires APCR TO A DISADVANTAGE. WHEN I TAKE AWAY THE APCR IT DOES BETTER. WHY DOES THE AI FIRE APCR WHEN APBC DOES BETTER????? The last sentence is my main point, and has now been tested by 1160 engagements. (yes, that's 2320 tanks) Mike 0 Quote Link to comment Share on other sites More sharing options...
ww2steel Posted June 21, 2006 Author Share Posted June 21, 2006 Okay, This is one of many graphs going into my book: The CM part of my website (that badly needs updating!): http://www.ww2steel.com/Combatmission.html See how the T-34 does WORSE at 100, and a little at 300m? That's what I'm asking about. :confused: Mike 0 Quote Link to comment Share on other sites More sharing options...
Elmar Bijlsma Posted June 21, 2006 Share Posted June 21, 2006 A complete guess here but mightn't lethality be an issue here? At 100m both guns will have no problems penetrating the target so the 'T' round will not give you an advantage. But the regular round is a big ass round (with a charge?) while the 'T' round has just the Tungsten core penetrating thus lower lethality. That would produce the curious gap between the differing range at 100m and also why it closes again at longer ranges. There the normal AP round will start having difficulties penetrating while the 'T' round will still penetrate, thus copensating for it's lower lethality. The AI, only checking for the highest penetration chance when looking up which round to use, thus opts for the 'T' round. Pretty much the same thing happens when firing on HT and the like with a 75mm+ gun. The AI will insist on using AP while a HE round is far more likely to take it out. [ June 20, 2006, 11:28 PM: Message edited by: Elmar Bijlsma ] 0 Quote Link to comment Share on other sites More sharing options...
Bigduke6 Posted June 21, 2006 Share Posted June 21, 2006 That's great logic Elmar, and I think you have the round mechanics right, but we all have had plenty of cases where we're screaming at the T-34 in the movie to use the tungsten against the kitty, and the A/I nevertheless selects standard AP and gets a ping. (And the kitty kills the T-34 in return, usually.) I have no scientific evidence to support this, but my theory is that when the A/I has a choice between tungsten and AP it looks at penetration probability like you say, but then adds a randomizer to the decision, and I strongly suspect the randomizer is tied to crew quality. In other words, the better the T-34 crew, the more likely they are to use tungsten intelligently. I bet if BlueSteel were to run several hundred tests with veteran and green T-34s, along with his control group regulars, he could verify my theory. 0 Quote Link to comment Share on other sites More sharing options...
ww2steel Posted June 21, 2006 Author Share Posted June 21, 2006 An interesting idea duke. Would crack crews use AP instead of APCR and eliminate this 'wrong ammo gap'? (I doubt it though.) And Elmar, good point. That's what I was thinking too- about loooking for the highest penetration when the AP is fine and will have a higher chance of a kill. The light armor "let's all use AP" has always annoyed me. You sit there and watch 'em put a few rounds of AP into a BA64 or whatever when a single big HE hit would either break it in half or flip it over. Mike 0 Quote Link to comment Share on other sites More sharing options...
Elmar Bijlsma Posted June 21, 2006 Share Posted June 21, 2006 Originally posted by Bigduke6: That's great logic Elmar, and I think you have the round mechanics right, but we all have had plenty of cases where we're screaming at the T-34 in the movie to use the tungsten against the kitty, and the A/I nevertheless selects standard AP and gets a ping. (And the kitty kills the T-34 in return, usually.) I have no scientific evidence to support this, but my theory is that when the A/I has a choice between tungsten and AP it looks at penetration probability like you say, but then adds a randomizer to the decision, and I strongly suspect the randomizer is tied to crew quality. In other words, the better the T-34 crew, the more likely they are to use tungsten intelligently. I bet if BlueSteel were to run several hundred tests with veteran and green T-34s, along with his control group regulars, he could verify my theory. I don't think it's anywhere near as complicated as that. AFAIK the TacAI looks at two things when deciding when to fire the HC/T round: 1)How much of the special round has it got. The fewer the less likely it is to fire it. 2)The hit percentage. If it is low it will rather fire an AP round to gage the range then waste it's precious special. When that percentage gets raised to a point where it's reasonable to fire it will. Hence your better experiences with better crews. They will get a good firing solution faster thus will find it less objectionable to fire it's T/HC round. A conscript will need to bang away for several minutes before it's hit percentage gets good enough to chance a T/HC round. 0 Quote Link to comment Share on other sites More sharing options...
YankeeDog Posted June 22, 2006 Share Posted June 22, 2006 Originally posted by ww2steel: I was just alt tabbing out of my game to ask this. How can I calculate how many tests I need to do to eliminate the statistical noise down to maybe 1%, 2%, whatever? Will someone please help me? Thanks! Mike OK, you asked for it. . . Wikipedia Entry for "Standard Deviation" Ahhh. . . just skimming it brings me back to Sr. Year in High School. Who was that cute blond who sat in front of me in Prob/Stat. . . Julie? Jane? Janice? 0 Quote Link to comment Share on other sites More sharing options...
John Kettler Posted June 22, 2006 Share Posted June 22, 2006 Ah, Probability & Statistics! Sweated through 30 semester units, including that one, over 35 weeks while working full time. In the entire year after graduating, I had to compute one standard deviation. Grrrr! Regards, John Kettler 0 Quote Link to comment Share on other sites More sharing options...
Tagwyn Posted June 22, 2006 Share Posted June 22, 2006 This is completely and utterly a waste of time. rolleyes!! Now do you understand? Dorosh: Read Kettler posts. That is exactly how you used to sound. Tag 0 Quote Link to comment Share on other sites More sharing options...
ww2steel Posted June 23, 2006 Author Share Posted June 23, 2006 Well, I don't know why people want to give me sh*t for doing research on the game for the community. I wonder how many people emailed Chris Hare for his Excel spreadsheets and "rolled eyes" at him for spending hours and hours in front of the editor gaining information that 90% of you use regularly when playing the game. Maybe only 10% of you go in detailed enough to care about my research, but if you don't care just STFU! :mad: There's no need to discount my efforts just because you don't get into the game enough to care that your 85mm will do 8% (or whatever) better at 100m if it doesn't have APCR even though this is counter-intuitive. For those of you that are interested, thanks for the helpful posts. Mike 0 Quote Link to comment Share on other sites More sharing options...
David I Posted June 23, 2006 Share Posted June 23, 2006 :eek: DavidI 0 Quote Link to comment Share on other sites More sharing options...
ww2steel Posted June 23, 2006 Author Share Posted June 23, 2006 Guess I did kind of explode there. It just seems like 80% of the time I post a real question or one of my few graphical mods somebody gives me crap. Sometimes I'm sure I just take it wrong, sometimes it's meant to be rude. That's why I have relatively few posts and almost never do post any of the hundreds of pages of research I have done. Just makes me upset. Mike 0 Quote Link to comment Share on other sites More sharing options...
76mm Posted June 23, 2006 Share Posted June 23, 2006 ww2steel, downboy! I don't really see anybody giving you crap, even if no one is falling over themselves to encourage you. by all means continue (although I'm not staying up nights waiting for the results...)! 0 Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.