fnord_disc

Because the arc for the guns is so high that gravity keeps pulling the shell down until the total speed is quite high. Meanwhile horizontal speed loss to drag depends on speed, so after a while the shell doesn't lose that much horizontal speed anymore but still gains speed from gravity. I'll edit in a speed plot to show you. This would be much less strong in reality because the air drag would slow it down, but I only modelled air drag on the x-axis, not y. edit: This is velocity over time for 10,000m. As you can see, the shell gains a whole bunch of speed from gravity because it's in the air for like 45 seconds and the guns shoot it very high up. My calculation gets unreliable once you get to the edge of the gun's historical range. 10k is about the range the gun has if you elevate it to 45°. But anyway, in reality this effect wouldn't be as strong, but it does exist. For example, this is the historical range table for Cleveland: Above 20km the shell starts getting faster than it was below 20km. South Carolina (experimental data from RichardNixon) Both ships set to 22° to use the same plot. Red curve is higher for the first point if 15° angle is considered.

All curves generated with rho=0.9. Budyonny (experimental data is WG video, therefore no impact angle in test data) Cleveland (experimental data from arczer25) USN 12.7cm / 38 (experimental data from arczer25)

I don't have either of those ships. As a constant multiplier: Hm, yes, perhaps... I'm getting pretty good results by just ignoring it, and since most shells are around 2400... edit: curves edited out and redone further down.

All this was considered during alpha and thrown out because it players wouldn't understand it.

Can you explain to me which units the formula takes? Because if I assume metric, this is what happens for South Carolina at 0 meters. 823 m/s MV 394.6 kg mass 2408 krupp 0.305 m caliber 823*sqr(394.6)/(2408*sqr(0.305)) = 12.3... m? mm? cm? none of these fit I've looked it up and apparently caliber and penetration are in decimeters. This would produce 390mm penetration @ 0m @ 0° for South Carolina, and by that measure she shouldn't be able to penetrate Montana's citadel, but she can. So this formula can't be 100% correct. Thanks. Also, if the Krupp formula you gave me is compared to the USN formula, it's easy to see that the exponents are pretty similar: 0.55 and 0.65 vs 0.5 and 1.1 vs 1, so it would be understandable if the Krupp formula is used and not USN and I still get similar results. I just need to get the scaling factor to work.

I'm assuming a homogenous material since the armor layers mined by GM3D don't have any properties attached to them except a name and a thickness. And an angle of course, but that's a part of the 3D model. What I mean is, there is no values inside the game telling it how good the steel quality is or whether something is cemented or not. At least as far as I know. I doubt the game takes these into account, but it could be the reason, yes. As for why there are differences, well: WG might include a second axis of drag I still don't know what krupp does (if it even does anything related to penetration) WG might use some kind of approximation for the ballistic arcs since the explicit solution of alpha(x_target) needs a bunch of messy functions, especially if you have xy-drag and not just x-drag

@ rho 0.9 (like with Budyonny and Cleveland, where it fit the data very well) without normalization. Also, I think you have a typo in the muzzle velocity for SC... should be 823, right? I'll look for some more penetration data to compare.

You think the hard part is behind you, huh? hfgl

Nigel Farage and Boris Johnson rumored as ingame captains. Battle starts and guns start firing, they'll turn around and sail the other way.

Whether or not the game takes angle into account isn't even the question though. There is no doubt that it does. The question is which formula tells you how many mm of armor the shell penetrates in the first place. Before the angle is calculated. You know, that thing where it just says "202mm" in WoT and where it says nothing in WoWs. WoT uses a linear energy loss model and WoWs definitely does not use the same (hyper-simple) model. edit: Including a second dimension of drag doesn't materially improve the accuracy of the calculation. For now I just need more drag coefficients. edit2: Comparison generated vs measured. % SHELL CONSTANTS % W = 59; D = 0.152; c_D = 0.321; V_0 = 762; @ rho 0.9

Okay, sorry for the double post, but trust me, it's worth it. I've expanded the calculation further up to include a simple ballistic arc with gravitational force and elevation angle. Since that is definitely too annoying for OpenOffice, the calculation has been shifted to MATLAB. The calculation does not include drag in the y-axis, only in the x-axis. This splits the two counterforces and makes calculating a firing solution easier for me, and since I'm not 100% convinced this is the right direction, I wanted to conserve time. I can expand it later, but for now, it means the speed at impact is somewhat too high and angle too steep, but it shouldn't be off by a lot. Here is an example of Budyonny firing at a 15km target. Budyonny is the only ship for which I have drag coefficients, data mined from Kutuzov (same shells IIRC). W = 55; D = 0.152; c_D = 0.321; V_0 = 950; I played around with the air density and if the air density is set to 0.9, then the following penetration values are obtained using this calculation. It's getting pretty close to the target.

Can someone data mine me the drag coefficients? edit: Yes, I know the angle matters. I'm not that stupid. And getting the impact angle is not hard once you nail down the ballistics.

It's definitely not similar to WoT's penetration mechanic in any way. WoWs handles it much more realistically.

The 5km penetration fits okay-ish, the longer ranged ones are lower than my calculation. It's a difference similar to the test data. I don't know. Afaik the normalization angle is just a subtraction. There is an additional speed loss due to the length of the ballistic curve. I thought this was negligible, but I'll include this in the formula and make another post.

The guns have a similar penetration constant compared to cruisers, but they are still heavier, so compared to 8" guns, their penetration is a little more than twice as high. It means she struggles against battleships, but it's more than enough against cruisers.

Oh, yeah, I was looking at stock Kongo. But as for Nagato: erm, is that unique to WoWs then? I have books with armor schemes and Nagato's remodelled armor is given as 300 + 75@45°. I was assuming this would be the same in WoWs, and I don't have an account at GM3D. As for Tirpitz: what armored deck? The citadel armor has 100@60-70°, hard to estimate the angle for me. I'm not really sure we're talking about the same armor parts. Also, Nagato does not have the V scheme I meant. Kii's and Tosa's side armor looks very different from Nagato. Nagato Tosa

Making a 45k ship work at T10 is almost impossible, sorry. Sure, 12x38cm, but Montana already struggles with its guns. Montana is also often criticized for its armor and Type 3 is still worse than Montana. Type 3 is tier 9. Tier 10 is fantasy.

I agree with this line of reasoning concerning the WW1 battlecruisers, but I can't agree with it for Scharnhorst. By the time of the 2nd world war, speed was not something you had to make extensive sacrifices for. If the Scharnhorsts had been larger than 30k or whatever it was, they could have mounted a larger caliber and retained both their armor and speed. In a time where fast battleships like Littorio, Bismarck, Iowa, etc, are all perfectly feasible, it seems nonsensical to classify battlecruisers by speed. Now, you said that a battlecruiser should be a ship that sacrifices something for speed, and that the Scharnhorsts sacrificed gun caliber. But I don't think they did. In my opinion they sacrificed nothing. They were the most logical battleship-like design at their displacement. The design logic is just different in my opinion. The WW1 ships were essentially battleships down-gunned to have more speed. This is the sacrifice made. But Scharnhorst was constructed to a smaller displacement than most contemporary battleships, and it retained all their design features. If anything, they should be called "small battleships". This is a logical descriptor in my opinion. edit: Of course, that prompts the question which ships newly-built from the 30s are battlecruisers. I think none are. Dunkerque is a small battleship just like Scharnhorst, albeit with relatively thin armor (but not thin enough to not be a battleship). The Alaskas are just scaled up Baltimores - calling them anything other than overgrown cruisers seems very silly to me.

Alsace would make a beautiful T9 and the French tree in general would be very appealing visually. As for the premium: I don't think it's a destroyer. Whenever WG has had the choice of implementing a DD premium or a larger premium, they went for the larger premium. Most of the DD premiums are from small navies.

This is a common misconception. IJN ships until Nagato often have the angled citadel armor, but it's usually so thin that you can't tell the difference. It's only 19mm on Kongo and even Nagato's 75mm don't really cut the mustard. If we ever get Tosa and Kii then this might change because they have a funny-looking 102mm V-shaped armor space behind the belt that should be pretty effective. Nikolai also has 75mm behind the belt, but she doesn't see 16" guns like Nagato.

Imperator won't return with a probability of 0.9 Gremys.

Because if you start speaking English in a game targetted only at Japanese players, they will lodge complaints with the mods to get you banned. If the mods don't ban you, the players will leave. They're pretty tolerant of foreign players if you speak Japanese, but otherwise, no chance. Read your PMs.

The Japanese 25mm was also awful and it's perfectly all right in WoWs.

I also think that battleships should win in any situation whatsoever, always, even horrible players. Because battleship. battleship. with a small b

Because their business model does not target them. Kancolle has two sources of revenue and the two companies behind Kancolle have split these sources, not the revenue. DMM gets all the money from ingame items and port slots and such since you need to buy DMM points to exchange it in the game. This is used to finance DMM's infrastructure. It's a very large amount and completely dwarfs all of DMM's other games (most of which is porn crap). Kadokawa gets all the money from the IP and leverages the popularity of the game to churn out all kinds of other stuff related to the game. DMM doesn't particularly mind foreigners playing since occasionally they do go to the trouble of buying DMM cards or using their credit card and in any case their load on the server is negligible compared to Japanese players, so they don't actively ban them. Still, DMM has nothing to really gain from promoting Kancolle abroad since most of those players would be unlikely to spend significant amounts on DMM points. The reason for this is mostly related to how much effort it is to buy DMM points outside of Japan, but a part of it is also that a lot of SEA countries (which love Kancolle) have much less disposable income than Japan. Kadokawa doesn't promote the game abroad because it would mean they need to expend significant effort translating a lot of the game's material. Books and comics won't sell abroad if they don't have a translation, and this is what makes most of the money for them. They don't feel they have much to gain financially from trying to expand the game beyond Japan's borders. All of this would be different if the novels or the comics had a licensor in the west, but they don't. So it would be up to Kadokawa to do all that themselves, and they don't feel it's worth the risk or effort. edit: Haha, okay, I could write a big essay about this. The essence of it is that Japanese companies don't really understand Western fans very well unless we're talking about the AAA games. It's too difficult for the Japanese companies to judge whether niche products will be successful in the West, so unless a Western licensor approaches them and basically slaps a wad of cash into their face upfront, they won't be interested. To them, the language and cultural barrier is so high that they would rather continue making games purely for the Japanese in peace. This might change in time if the Western market becomes a steady source of revenue for them, because there is nothing that scares Japanese companies as much as uncertainty. If they feel like Western consumers cannot be counted on to put their money where their mouth is, they will not license out more aggressively. If a Western company went to DMM/Kadokawa to license Kancolle and basically took over all financial risk for the localization and the Western infrastructure, then they would probably say yes. The same goes for the merch: Kadokawa would probably let you license their Kancolle novels, but only if you carried all the financial risk for this. But Kadoka would never do this themselves: the risk/reward is not good enough in their opinion. This applies to even large titles, though to a lesser degree. The Dark Souls series for example is self-published by From Software in Japan, but for the international release, they have Bandai Namco, which has much more experience with the Western market.