I can't look at the files right now, but I'm almost certain there was a random seed variable defined somewhere in the My Documents/SH4 folder.

Maybe this gets reset every patrol, and thus accounts for wild fluctuation in matters of chance in SH4.

On reason I bring up the couple of known, nearly 100% (if not actually 100%) mk 14 failures is that in modding those in, you should really get the vast majority of the RL malfunctions.

The problem is that the error rate in stock SH4 assumes that the fish were basically OK, but then had a random chance of various and sundry failues that combined create the mk14 problem. In RL, there are 2 issues: Failures (real duds), and design errors. The bulk of the mk14 problems were fundamental to the design. They in fact functioned exactly as they were designed to. They were badly designed.

Design failures (the bulk of RL mk14 problems):
1. Deep runners should be 100% until the date they fixed them, this was not a random failure, they were built to run 10-12 feet below their set value. NO skippers had them run at the right depth, not ever. The magnitude of the depth error should be ~10-12 feet, though perhaps it might go 9 to 15 or something like that to take into account everyone setting them at min depth.

2. The contact exploder is another. The failure rate should be very very high at high speed within a range of impact angles. 90 to 70-80 degrees, perhaps, with a large % like 80-90%. At slow speed ideally it would be 0%, or very very low to represent actual duds instead of a design error.

3. The magnetic pistol. This one is the blend between design failures, and real "duds." The design was such that the results were actually unpredictable, and varied with the target hull, magnetic field in the target area, etc, ad naseum. The plus in game is that a random number actually works for this one accurately.

Real duds:
4. Failures of maintenance, material failures (bad batch of explosive in the pistol, flawed internal parts, etc), etc. This would be the circle runner %, and some of the "dud" problems.

I'm not sure what a realistic value for #3 should be, frankly. I think 1 and 2 should be 100% and 80-90% in the right angle range respectively. #4 will be a very small number.

That's for a historical baseline. For gameplay, perhaps lower on 3 by a bunch, but I think 1 and 2 should be realistic since the player can control them entirely (within reason). Perhaps for TM set the deep runners to 10-12, no slop. The player will set them shallow, so no effect. #2 they have to plan attacks not to T-bone the target with fast fish. 3 and 4 become the random element, but the %s should be very small for 4, and 3 is the big question mark.



tater

RE depth errors.
Remember, i have two possible settings for depth errors for the first era pistol.

Im expermenting with a 60% chance of a depth devation of 1.5 to 5 meters for the first one

and for the second one, a 40% chance of a deviation of 9 meters, to 10 meters. This particuarl one, im aiming at folks like myself, who always shoot shallow, and never bother with magnetic pistols. Idea being, *occasionally*, you get a real deep runner that thwarts even your (my) own smartassedness :rotfl:

I'd say that the minimum depth deviation should be 3m. X% at 3m-4m, and 100-X% at 3.5 to 6-7m.

With X maybe 60-70%.

I can see the point of really deep though, since they'd set for a CA plus 10 feet, for example, then the fish runs another 10 deep and it gos 20ft under.

tater

Im having trouble with the chance percentages.


Are they independant checks, or do they have to be a summation of 100%?

For instance

60% chance for 1 error, 40% chance for the other- thats 100%.

But.. 100% torpedo depth error is not what im getting.

It seems to be handleing each error seperatly.

Error number 1, 40% chance to pass?

error number to, 60% chance to pass?

What would happen if i set error number 1 to 99% and error number 2 to 40 or 50 percent? Would error number 2 even occur then?
.

Well I would guess that in the one out of 100 times that error 1 does not happen, error 2 would have a 40 or 50% chance of happening. So in effect error 2 has a 1 in 200 chance of happening.

That is unless they are compound, ie more than one error can happen at the same time and they add on to each other.

BTW may I ask does anybody know what range a magnetic exploder torpedo explodes at from a ship in-game? Is this moddable?

2 meters, and yes its moddable.

Think i almost have a decent test version going now.

Ok first, let me reiterate the stated goal in MS paint:

http://www.ducimus.net/sh4/release/torp_goal.jpg
My intent, is to get somewhere in the shade of grey. It's also not my intent to make another hardcore torpedo mod. There is already one out there, there is no need for me to reinvent the wheel. However, stock game, after reaping in 20,000 tons and not even trying, i decided that i have to do SOMETHING about these super reliable stock torpedo's.


So here's the file. Let me know your findings if you try it out.
http://www.ducimus.net/sh4/release/W...Torpedo_1.0.7z

Can someone else playtest this for me, before i do something rash and stick in a mod revision in the not to distant future?

I took a quick look into this and it seems you have the same problem with Mark 14 torpedoes as I do. They are too reliable when using magnetic pistol. Did 2 tests and every single one of them detonated ok

Also did 2 tests using a contact pistol and high speed. 8 torpedoes, one deep runner, one hit on a target. All the rest bounced off from the hull

Actually, in the evaluation of the Wahoo's 6th patrol, one of the criticism's of Morton's tactics was only firing a single fish at a target. The higher ups considered a 2 torpedo spread a minimum and a spread of 3 the recommended practice . Part of this was to compensate for an calculation errors, and the other part was because they new problems could arise with the torpedoes.

And on Peto's comment- some enterprising torpedo men would pull the pin's and machine their own instead of waiting for them to be supplied to the boat. Bubbleheads have always been an enterprising bunch.:)

One thing that I haven't seen in the game yet is a fish broaching. When that happend and they dove back into the water ANYTHING could happen. circles, veering off, diving down at a steep angle and so on. This was not as common a problem, but occasionally the cause of a fish going to deep, was because it actually started too shallow. Especially in rougher seas.

On at least one occasion of a circle-runner hitting a submarine (Wahoo, I believe), one of the eyewitnesses to the event said the torpedo jumped out of the water before circling back around. So yes, a torpedo breaching the water was no bueno all around.

Duci, I'm giving this a try. Will let you know what I find:up:

Edit:
I did the torpedo the bad guy tutorial (of sorts) and I found the mix to be pretty good. Basically 1 out of every four was a under runner or the magnetic pistol did not function. I had one that ran to fast. Don't as me how. This first of the four past the bow by a good margin. The remaining 3 were on target. With the solution the same for each torpedo the firsts fast should have been a hit. I did not experience any duds. On a second try I had a premie (calm seas). Two worked fine with magnetic exploder. The other was just contact and worked fine. Third attempt I had 1 under runner and the rest worked fine. My personal opinion, tweak the torpedoes just a hair worse than what you have here.