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

[reaper7]

Originally Posted by h.sie

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

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.

Hi Mate, no problem (If you need info on Impact Angle Calculation let me know.
I've a working formula done and working.

Originally Posted by 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.

@Both, I guess the real reason why I was pushing for this is due to all the real footage I've seen on U-Boat operations, none I've seen show An Impact on a Target that look like they used a Fast90 attack.
All footage have the Target with an AOB greater or less than 90deg on Impact - whereas if it was a fast90 attack than AOB would have to be 90deg port or 90deg starboard when viewed from the scope at Impact.
Maybe this technique wasn't known to German U-boat operators at that time.

But yes I get your point that its not really cheating so why penalize the player.
And I for one will be happy with any way you Implement this, I'm just a gluten for punishment (Don't like things easy ).
Cheers Guys, Appreciate all your doing for Sh3.

[h.sie]

@LGN1: Thanks for that interesting document. Havn't read it yet completely, but it is about the Torpedocrisis until 1942, which mainly is caused by technical problems, not by the weather. So I'm not surprised that weather is not mentioned, since weather isn't the cause for the crisis. The influence of the weather on torpedo failures is a different problem. I'm also not sure about quantities, but, as Hitman says, it's surely not a linear dependency.

[Hitman]

To put specific percetages of failiure for each issue, we have currently:

- Player error when calculating solutions: That doesn't need to be implemented, and I personally would not even take it into account. A player that does manual targeting correctly excludes that error, and one that is learning, well, is learning and must accept that in real life they would not have given an UBoat to someone like that

- Prematures: Already in the game, with the results that LGN reported. Unless we can modify its behaviour, it's a given and will not enter in our calculations except for detracting the percentage it already means.

- Magnetic pistol failiure: This is set in the torpedo files, so H.Sie doesn't really need to do anything about it as it is already moddable.

- Impact pistol failiure: Same as previous

- Angle of impact too acute: Same as previous

Now this is H.Sie's territory:

- Torpedo mechanical depth keeping issues: It is well known that torpedo depth keeping mechanism failed until 1942, when the issue was discovered and fixed. It should add a 25% chance of the torpedo going down as implemented by H-Sie. Reason: The failiure appeared in part owing to torpedo maintenance and air pressure in its depth keeping mechanism after the Uboat submerging deep (As the pressure in the boat builds up, the torpedo mechanism would become discompensated as it was not built for such high external pressures). So it would not happen in boats that only had done routine dives for trimming and would instead happen more in boats that had dived deeper.

- Torpedo course alteration, porpoising, etc, due to heavy seas: This would all be packed together in the single effect of the torpedo going down as implemented by H-Sie, according to following table:

Windspeed 0 = 0%
Windspeed 5-7 = 5%
Windspeed 7-12= 15 %
Windspeed 12-15 = 75%

Ideally this should also be incremented as the set torpedo depth diminishes, meaning:

Depth 0-3 metres = 100% chance of the previous Windspeed random parameter effectively acting on the torpedo
Depth 4-8 metres = 75% chance of the previous Windspeed random parameter effectively acting on the torpedo
Depth 9-15 metres = 50% chance of the previous Windspeed random parameter effectively acting on the torpedo
Depth 16-Maximum metres = 25% chance of the previous Windspeed random parameter effectively acting on the torpedo

Reason: The deeper the torpedo, the less chances of the surface sea state acting upon it

-------------------------------------

To sum up we will add an ADITIONAL (To what the game or moders already put into the pistol and prematures) margin of failiure that goes from:

Minimum 25% aditional failiure rate due to faulty depth keeping mechanism (Until 1942)

Maximum 100% aditional failiure rate in heavy seas, low depth torpedo.

In between, many combinations are possible, but all of them are not in the hands of the player except those real Kaleuns already knew -like not shooting with too acute angles, not setting the torpedo too shallow in rough seas, etc.

How does this all sound?

[LGN1]

Hi Hitman,

just a quick comment: As far as I know the only implemented malfunctions in SH3 are premature detonations for the magnetic pistol and the discussed impact angle issue. Magnetic pistol failure is not implemented and cannot be modded without touching hard-code, I think. It's different for SH4 where you can set the parameters in the torpedo.sim files. Are you confusing SH3 and 4?

Cheers, LGN1

@h.sie: Weather is mentioned quite a few times (e.g., Doenitz trying to blame the weather for the failures when the pistol was changed,...). The problem is that we do not know exactly how often a certain malfunction happened (propulsion problem, pistol problem,...) and how it was influenced by weather. Certainly it's not linear, but what is it? And do we really need it?

[reaper7]

Originally Posted by Hitman

Windspeed 0 = 0%
Windspeed 5-7 = 5%
Windspeed 7-12= 15 %
Windspeed 12-15 = 75%

Ideally this should also be incremented as the set torpedo depth diminishes, meaning:

Depth 0-3 metres = 100% chance of the previous Windspeed random parameter effectively acting on the torpedo
Depth 4-8 metres = 75% chance of the previous Windspeed random parameter effectively acting on the torpedo
Depth 9-15 metres = 50% chance of the previous Windspeed random parameter effectively acting on the torpedo
Depth 16-Maximum metres = 25% chance of the previous Windspeed random parameter effectively acting on the torpedo

Reason: The deeper the torpedo, the less chances of the surface sea state acting upon it

-------------------------------------

To sum up we will add an ADITIONAL (To what the game or moders already put into the pistol and prematures) margin of failiure that goes from:

Minimum 25% aditional failiure rate due to faulty depth keeping mechanism (Until 1942)

Maximum 100% aditional failiure rate in heavy seas, low depth torpedo.

How does this all sound?

Sounds good to me

[h.sie]

@Hitman: Good point to consider the torpedo depth when calculating weather dependency.

I think about to rename the thread into

Masochism- and gameplay- related hardcode fixes.

@LGN1: I would say we need storm dependency, but this is assumption not knowledge.

-> Facts needed!

[LGN1]

@Hitman: I like your approach, but I think the numbers are too high. I would tune the parameters that (assuming an equal weather distribution) you get a failure rate of approx. 20-25% up to June '40 and then less until mid '42. I think the failure rate dropped considerably after the invasion of Norway. I would also avoid a failure rate of 100%.

Maybe let's try to first agree upon the weather-independent probability on which the weather part is added?

[SquareSteelBar]

Originally Posted by h.sie

...I think about to rename the thread into

Masochism- and gameplay- related hardcode fixes...

Some Sadism items would be much appreciated...

[h.sie]

@LGN1 & Hitman: Guys, "fight" about the quantities, while I'm now looking a movie.

[LGN1]

Just a quick estimate:

Most of the time in SH3 you have a wind speed > 12m/s and most of the time you shoot torpedoes with a depth of < 15m. In this situation the failure probability is

0.25 + 0.75*0.5 = 62.5% failure rate in quite common situations

I really think the 0.75 is over-kill.

Regards, LGN1

[reaper7]

Not sure everyone is on the same page when it comes to percentages.

What I mean is when a figure is mentioned eg 25% is it:

1. 25 % of torpedoes carried will be duds
2. a random numder between 0% and 25% carried will be duds

Just wondering as both the same figure but different dud amounts

Just we should agree on which we use in discussion so as not to confuse.

[Depth Charger]

Would be interested to know if it is definately Wb's orders causing this conflict and wether it can be fixed, so we may have both

Thanks again

slipper[/QUOTE]

Cant say for sure but it changes the way sh3 sends radio messages and so it has a high chance of conflicting with the wolfpack and radio message parts of h.sie's mod and so I took it out.

kindest

dc

[Depth Charger]

@ H.Sie

Just a short note to apologize in advance for cursing both you and LGN for the dud torpedoes...

As it is, I am already struggling enough out there with manual targeting and using this patch. Last night I lost a bitter and desperate battle against 2 destroyers because of your repair fix.

You are already killing me, and now you going to make my torpedoes bounce off the ship as well?

Well that's just cruel and so I thought it best to get my apology in up front because I am definitely not going to think or say any pleasant when it happens... Then again, I will never play the game again without it either and it sounds like an excellent addition to your folks patch...

Shout if you need any help from us testing...

DC

[h.sie]

@reaper: If I talk about failure probability p=25%, I mean that (in average) 25 of 100 torpedoes will be duds. If I remember correctly, this is the mathematical definition of probability.

@all: WB's mod and my hardcode fixes use different ways to manipulate messages. There is surely a way to make both compatible. WB's mod used SH3-Cmdr, right?

@LGN1: It's easier for me to program a continuous dependency function on windspeed instead of the discrete windspeed bands Hitman proposed, so that we get a lower failure rate for 12m/s than for 15m/s. The dependency on torpedo depth will also be a continuous function instead of discrete depth bands. Tomorrow I'll think about the character of these functions, and than we'll surely find a consensus regarding the quantities/parameters.

Gähn!

[h.sie]

Considering the posts above, I created an analytical formula for the torpedo failure rate p, which considers:

1) The failure rate p_crisis, resulting from the torpedo crisis until 1942
2) The failure rate p_wind, resulting from waveheight + torpedo depth

The resulting failure rate p is the sum of both components (to make calculations simple):

p = p_crisis + p_wind

This failure rate p is added on top of the already modelled torpedo failures (premature explosion, dud because of bad impact angle).

-----

For p_crisis I set

p_crisis = {

• 20%: until June 40

• 10%: until June 42

• 0%: else

}

These values are lower than the proposed 25%, since there are already some premature detonations modelled in sh3.

-----

For p_wind I set

p_wind = 50% * (Windspeed)^2 / 225 * (25m - TorpedoDepth) * 4/100

That means: Under worst conditions (windspeed=15m/s, torpedo depth = 0), wind can cause a maximum failure rate of p_wind_max = 50%. If windspeed sinks, p_wind drastically (sqarish) sinks below 50%. The deeper the torpedo, the lower the failure rate. This additionally motivates the player to use magnetic pistol.


Some Examples for failure rates in different situations:

1) Worst case (June 40, Wind 15m/s, TorpedoDepth=1m)

p = 20% + 50% * 1 * 0,96 = 68%


2) LGN1's situation in post #2606 (June40, Wind=12, TorpedoDepth=9)

p = 20% + 50% * 0,64 * 0,64 = 40%

3) After the crisis: (July 42, Wind = 7m/s, TorpedoDepth = 9)

P = 0% + 50% * 0,22 * 0,64 = 7%

4) Quasi best case (July 43, Wind = 1, TorpedoDepth = 12)

P = 0% + 50% * 0,005 * 0,52 = 0,1%


I noone disagrees, I'll implement the formula above into the forthcoming "More torpedo failures" Mod.