Realism- and gameplay-related hardcode fixes for SH3.EXE

[Robin40]

@h.sie

I wish to thank you also in the forum for having spent your time on the bug I posted

It was a corrupted save file andi it was my fault

Go on in your work

[Hitman]

Soo.. what makes you think it is the most reliable source ? This particular one ?

Are you joking?

[Hitman]

@Hitman: Thanks for your comments. Concerning your number

1) From my experience the in-game probability is close to zero if you stick to some basic rules. Therefore, it does not matter if the new probability is added on top of the present probabilities. In addition, if people don't like the addition they can switch of the realistic torpedo failure option in SH3.

2) As said in 1) I would simply add the new probability. In addition, as far as I know commanders had only sometimes the opportunity to choose the pistol type (because of technical constraints and BDU orders). So, the question which pistol the commander would prefer to use is somehow difficult to answer. Anyway, please make a suggestion for the failure rate depending on the sea state. What we need are values for the probability.

Although real Kaleuns were well aware of the impact angle issue (from what date on were they aware?), the failure rate was 26%.

The impact angle issue was known from the very beginning, because torpedo designers were well aware of it. In the US Navy more or less the opposite happened, doctrine called for 90º shots as best chance of pistol triggering, but commanders discovered by chance doing forced shots that it actually worked better at a steep angle (The reason being the needle bent in a harder impact, as it happens in a perpendicular collision, and didn't trigger the exploder). So, impact pistols were designed with 90º shots in mind and that was standard doctrine.

Regarding the pistol set, Stiebler will be able to tell that better than me, but IIRC you could not change it once in the tube. So yes, commanders would normally not have a lot of margin to switch such settings, as torpedoes would normally be preloaded and prepared with whatever pistol the standing orders said.

Regarding sea state, the influence it has in the succesful torpedo attack is felt in several areas:

1) Depth keeping issues. The torpedoes travel at high speed and have small finns/rudders, so they will have difficulties keeping depth.

2) Course keeping issues. For the same reasons as above, plus the shorter travel of the internal gyro compass stabilizator (Due to the confined area) they might easily veer off course.

3) Detonation issues. The torpedo might impact at weird angles against the ship's hull because the hull is curved and also moving up and down plus rolling due to heavy seas.

So, the first thing to consider is if some of those effects are already included in the already existing failiure rates. In that regard, note that the 26% failiure rate you have pointed out might very well include already torpedoes that malfunctioned due to being shot in heavier seas!

Then, you must ensure that the proportion of this effects increases with windspeed (I think H-Sie has repeatedly said that tying something to that parameter is not difficult for him, so that is a good thing).

The proportions are difficult to tell, but should go from absolute zero influence in calm seas, to a very high one in storms (75-85%). I have readed several accounts of uboats meeting a juicy target in a storm, and both looking at each other harmlessly due to the inability of the uboat to make an effective shot. (F.e. H.W. Schulz in his biography, "Über dem nassen Abgrund").

It is also important to note that the failiure curve would not rise linearly, but instead climb steeply up as weather worsens.

[h.sie]

@Robin40: Glad to know. Thus, V16A3 seems to be very stable, but will remain in beta state for some time.

@PapaKilo: If you are against a torpedo fix, just don't use it. There are others (including myself) who want that fix. If you are not interested, please let those who are interested, continue the discussion about that topic. Thanks.

Regarding your mod requests:

• Sonar guy on/off station fix, Sonar guy takes his postition on "surface cruise": Can be done with one mouseclick. Eye-candy.

• WO stands still (not moving) on bridge after nearest contact report request: Eye-candy. This is a job for an animation expert. Out of the area of my expertise.

• Ships do not react to premature explosions of the torps, Destroyers are deaf to music of gramophone and echolot/sonar ping: Interesting, already on my list for a long time, but I fear I cannot fix it, since sensors are complex, hard to understand and hard to fix (for me).

• Ships and player u-boat Buoyancy and Agility. Also already on my list.

By the way: My 1WO-Automove-Mod also is pure eye-candy. I made it only to see if it is possible for me. It's not an important fix, since the player can do it with one mouseclick.

[reaper7]

Originally Posted by PapaKilo

Read this part 5 times and couldn't get the logical mind flow.

90deg is 90deg if it IS a 90deg for real. It doesn't make sense how you get the angle - by fast 90 method or by calculating TDC with tools.

The fair angle of torpedo with impact pistol and detonation to the target is 80-110 deg.

At least with GWX, probability of dud torps are pretty good adjusted. In early war you can make a perfect TDC solution with briliant 90deg runto the target, and all you see is the torp bounces of the ships hull and sinks

Sorry noticed the error in my sentence - should have read as:
Would be easy to monitor if torp is set to 90deg and gyro is set to 0deg and hence add another 10% chance that it will be a dud to encourage solutions that have a gyroangle not equal to 0deg
Thats what happens when you don't proof read

So in other words add a 10% chance that Torpedo may be a dud if using the Fast90 technique thereby increasing the dud probability from max25% to max35%.
This would neutralize the current method of getting around the stock Torpedo Dud system.

[LGN1]

Hi,

I don't fully agree that the WO and a sonar guy fix are at the same level. Yes, in both cases you can solve the problem with a few mouse clicks, but you have to move the WO in order to have a correctly working watch on the conning tower. You do not have to do anything with the sonar guy to have a correctly working hydrophone. In this sense, the WO fix is more crucial.

Anyway, concerning the torpedo failures:

As I've written in post #2551 players' opinions and preferences concerning torpedo failures are quite different. But still we don't have any suggestions for the probability (except Papakilo's suggestion of zero). Just general comments along the lines 'take this into account...', '...should be higher for...' and so on. I know I repeat myself, but please offer hard numbers. We know what h.sie can do and offers to do, so please offer hard numbers that fit into his scheme. If you don't like the scheme or have a more detailed scheme then it's fine, but it does not really help at the moment.

Cheers, LGN1

[h.sie]

@reaper:

1) Could you please explain (in simple words) the stock Torpedo Dud system? I havn't played much, but I don't remember to ever have had a dud......except from those resulting in a too low impact angle (angle between torpedo course and ship course).

2) Also, please explain, why the method of calculating the solution should influence the failure rate. You should know that I never seriously learned manual targeting, and from my beginners view, the only important angle (regarding failures) seems to be the impact angle (I meant the angle between torpedo and ship course).

H.Sie

[h.sie]

The problem is now to represent a system of complex interdependencies using a simple algorithm.

Thus, I

- have to restrict on essential factors but

- have to neglect inessential factors.

It is impossible to represent the complexity of the real situation during the war. I can only offer a coarse approximation - as always.

[reaper7]

Originally Posted by h.sie

@reaper:

1) Could you please explain (in simple words) the stock Torpedo Dud system? I havn't played much, but I don't remember to ever have had a dud......except from those resulting in a too low impact angle (angle between torpedo course and ship course).

2) Also, please explain, why the method of calculating the solution should influence the failure rate. You should know that I never seriously learned manual targeting, and from my beginners view, the only important angle (regarding failures) seems to be the impact angle (I meant the angle between torpedo and ship course).
H.Sie

Hi mate as I understand it the stock dud system basically only has 2 types:
1. Premature detonation - ok
2. Impact Angle - Generally Torpedo hits with Impact angles greater than 90+/-15deg have a chance to be duds to model the pin not being pushed on impact. Again the Hull of the ship plays a vital part to this angle if hit to low in the hull (The rounded part even from 90deg the impact angle is greater then 90deg due to the curve).

The problem is that this is easily avoidable by the player.
One just needs to set up for a Fast90 Attack and set the torps to low depths.
So torpedo has a gyro of 0deg and AOB is 90deg and depth is set low to hit just under the waterline. Hence dud rate will be close to 0.

If you look at nearly all sh3 videos on youtube attacks are using this method. Yes quick and easy with minimal solution required but the same old attack over and over again with piratically zero dud rate .

By applying an extra chance of dud rate to this attack type it will increase the rate of duds from stock 0% to a max35% using your code
(That is if you settle for a max25% - so standard across the board 25% and an extra 10% chance if using fast90).
This extra 10% would model the Pin bending when hitting a target at 90deg due to increased force when hitting using this technique.
And finally we may see more players starting to use there TDC's to work out full firing solutions that require all parameters of the Torpedo setup.

So it would be a tradeoff:
Fast90 (AOB=90) attack (minimum solution) - more chance of a dud
AOB not equal to 90 attack (Full solution) - less chance of dud

[h.sie]

@reaper: Thanks for your input (and congrats for successfully solving the reload issue).

It seems I have to do a course in manual targeting before I fully can understand your words.

This extra 10% would model the Pin bending when hitting a target at 90deg due to increased force when hitting using this technique.

I never heard about the fact that 90 degree angle increases the risk of a dud. Until now I thought an angle of 90 degree is desireable. Where can I read details about that?

And even if there really is an issue with 90 degree: I cannot measure the impact angle, I can only analyse the AOB setting in the TDC, which isn't equal to the impact angle.

????

[LGN1]

Hi,

just a few comments:

- the crucial parameter is the impact angle. The gyro angle doesn't matter. Therefore, it doesn't matter whether you use the fast 90° method and shoot with a gyro angle of zero or use the TDC and have a non-zero gyro angle, but 90° impact angle.

- Hitman has written that the real commanders knew about the impact angle issue. Thus, I think it's not required to change anything about it in SH3 (they knew it, we know it, and we both try/tried to avoid it).

- I know that the US torpedoes had an increased failure rate at 90° impact. However, I've never read about a similar issue with German torpedoes.

- The probability for a premature detonation is much too low in SH3.

- As I've said previously, I think we should have a quite high failure rate no matter what the player does. That's how it was back then for the commanders and I think it will provide the main feeling they felt in the beginning of the war: frustration.

Cheers, LGN1

[reaper7]

Originally Posted by h.sie

@reaper: Thanks for your input (and congrats for successfully solving the reload issue).

It seems I have to do a course in manual targeting before I fully can understand your words.

This extra 10% would model the Pin bending when hitting a target at 90deg due to increased force when hitting using this technique.

I never heard about the fact that 90 degree angle increases the risk of a dud. Until now I thought an angle of 90 degree is desireable. Where can I read details about that?

And even if there really is an issue with 90 degree: I cannot measure the impact angle, I can only analyse the AOB setting in the TDC, which isn't equal to the impact angle.

????

Thanks Mate, still some more work to get it nailed down and working 100%.
Impact Angle is not required just the GyroAngle. I can mail you the Memory locations of these variable if you wish.
So this value can be read and if it = 0 then add another 10% chance of dud to equation.

Originally Posted by LGN1

Hi,

just a few comments:

- the crucial parameter is the impact angle. The gyro angle doesn't matter. Therefore, it doesn't matter whether you use the fast 90° method and shoot with a gyro angle of zero or use the TDC and have a non-zero gyro angle, but 90° impact angle.

- Hitman has written that the real commanders knew about the impact angle issue. Thus, I think it's not required to change anything about it in SH3 (they knew it, we know it, and we both try/tried to avoid it).

- I know that the US torpedoes had an increased failure rate at 90° impact. However, I've never read about a similar issue with German torpedoes.

- The probability for a premature detonation is much too low in SH3.

- As I've said previously, I think we should have a quite high failure rate no matter what the player does. That's how it was back then for the commanders and I think it will provide the main feeling they felt in the beginning of the war: frustration.

Cheers, LGN1

Hi LGN1, yes the Impact angle is the critical one - but I stated the gyroangle as this is one we are all use in Sh3 and when using the 90°.
Beside when using this technique Both gyroangle and Impactangle are going to be 0degrees.

Stock Sh3 does not have an impact angle dail, I have programed a fully working impact Angle dial for SH3 for the forthcoming U-boot_HAHD Mod though and hence why I'm suggesting this extra 10% for gyro=0deg solutions .
As it will give more reason to set up solutions where ImpactAngle = 90deg while GyroAngle is not = 90deg

Thanks for the info on the Us torps suffering from the 90deg Impact. I had read that some time ago and had it in my mind it was German Torps that suffered this fault

Agreed that we should have a higher torpedo dud rate, I suggested that 25% max should be used for the formula.
Just would like to see the extra 10% on top of the 25% for gyro=0deg shots to give more Incentive to the player to try more Solution orientated shots than what Fast90 attacks use.

[h.sie]

@reaper7: I already have re-engineered a pointer to the gyro-angle dial value, so no need. Thanks anyway.

In order to get things sorted:

1) Impact angle is a cruical value for dud torpedoes. So far I understood.

2) But AFAIK dud torpedoes caused by wrong impact angle are already modeled in sh3. So why consider it twice??

3) Your answer to 2) might be: In order to motivate the player to do a full solution instead of a fast one, by giving a penalty for fast solutions.

4) My question now is: Is a fast solution cheating?

My 2 counter-arguments to your possible 2 answers are as follows:

4a) If a fast solution is cheating -> "Let people cheat, if they want to cheat. It's not worth my effort to simulate/model a reaction/penalty for cheating".

4b) If a fast solution is not cheating -> "Players should benefit from their fast firing solution instead of being penalised".

In total, I have no good feeling to consider the gyro-angle for calculating the failure-rate.

If I make the failure rate dependent on time with a probalility of max. 25% and ADDITIONALLY consider the windspeed with a max. probalility of 50% for 15m/s, we have p=75% in total. That's enough I think and people will be glad to have a torpedo hit in 1941 at 15m/s.

[LGN1]

@reaper: I don't understand why you want to have an incentive for not using a gyro angle of 0° Historically, commanders wanted to have a gyro angle as close to 0° as possible because a) range hardly matters b) low chance of pistol failure c) lower chance of torpedo steering trouble.

@all:

I have carried out some tests concerning premature detonations in SH3. I shot 50 torpedoes (TI, Aug. '39) with a depth setting of 4m, speed 40 knots, magnetic pistol, and 15m/s wind. Here's what I got:

3 torpedoes exploded < 1000m
9 torpedoes exploded between 1000m and 1500m
7 torpedoes exploded between 1500m and 2000m
(for larger distances I did not care anymore)

So, it seems the premature detonation rate is actually not that low. However, for ranges below 1000m it is quite low (in real-life it seems most premature detonation happened at short ranges after fusing ).

Therefore, I suggest to reduce the probability from 25% to 20% early in the war (until June 1940) and then afterwards until June '42 to 5-10%.

Cheers, LGN1

PS: Maybe 15% for wind < 10 m/s and 25% for stronger wind

[LGN1]

Originally Posted by h.sie

...
2) But AFAIK dud torpedoes caused by wrong impact angle are already modeled in sh3. So why consider it twice??

...

4) My question now is: Is a fast solution cheating?

...

If I make the failure rate dependent on time with a probalility of max. 25% and ADDITIONALLY consider the windspeed with a max. probalility of 50% for 15m/s, we have p=75% in total. That's enough I think and people will be glad to have a torpedo hit in 1941 at 15m/s.

Hi h.sie,

dud torpedoes because of a bad impact angle are already modeled in SH3 (quite well if you ask me).

The fast 90° attack is no cheating! In most situations it's the smartest thing you can do.

I agree that stormy weather makes things worse, however, I'm not sure how much. If you take a look at the document about the torpedo crisis you'll see that it never mentions specifically the weather.

Cheers, LGN1

Here's the link to the document:

http://eaglescholar.georgiasouthern...._201005_MA.pdf

It gives a very good impression of how complex the real-life situation was.