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

[LGN1]

Hi h.sie,

maybe it's a good idea to make a small modification and include the constant probability in the equation:

p = 1 ............................................for TD < A*WS
P = 1 - (1-p0)/width*(TD - A*WS) .....for A*WS < TD < A*WS+width
p = p0 ...........................................for TD > A *WS+width

So, we need values for A, width, and p0(until June '40, June '42, and rest).

A = 0.3 seems fine from looking at the numbers I've posted before. A good value for the width might be 2m

Furthermore, I suggest:

p0(9.'39 - 6.'40): 15%
p0(6.'40 - 6.'42): 8%
p0(6.'42 - ): 0%

What I like a lot about this approach is

a) The player must take the weather into account when attacking ships and setting the torpedo depth.
b) In bad weather the player might not be able to attack a ship at all if it's too small. This adds to the frustration
c) It encourages players to use the magnetic pistol because with this you can set a deeper torpedo depth. Since using the magnetic pistol increases the failure rate due to the already existing premature explosion, I have reduced p0 a bit.

Regards, LGN1

@tekai: If you are using the contact pistol the failures must be because of the impact angle. If you set the depth too deep, the torpedo hits the curved part of the ship's hull and therefore, the impact angle is no longer 90°. In this case it's important to keep in mind that you are dealing with a three-dimensional problem. In general, the impact angle is the angle between the torpedo and the normal of the hull surface (i.e., the direction perpendicular to the hull) where the torpedo hits.

[h.sie]

@tekai: Your comment was understood as being 100% constructive. And my reply was 100% without anger - tried to be factual.

When I say "don't use this mod" - this is neutral and without negative emotions. Preferences are different.

But the guys who want the higher dud rates are also experienced players, and according to their experiences dud rates are too low. So we have 2 different opinions and 2 different options: Enable or disable this mod.

Your input is welcome!

[h.sie]

@LGN1: In your formula you assume a linear dependency between windspeed and waveheight. Is that applicable? From my "feeling" I'd say that the dependency is of higher order than 1. This can surely be found somewhere....

[reaper7]

Originally Posted by LGN1

Hi h.sie,

maybe it's a good idea to make a small modification and include the constant probability in the equation:

p = 1 ............................................for TD < A*WS
P = 1 - (1-p0)/width*(TD - A*WS) .....for A*WS < TD < A*WS+width
p = p0 ...........................................for TD > A *WS+width

So, we need values for A, width, and p0(until June '40, June '42, and rest).

A = 0.3 seems fine from looking at the numbers I've posted before. A good value for the width might be 2m

Furthermore, I suggest:

p0(9.'39 - 6.'40): 15%
p0(6.'40 - 6.'42): 8%
p0(6.'42 - ): 0%

What I like a lot about this approach is

a) The player must take the weather into account when attacking ships and setting the torpedo depth.
b) In bad weather the player might not be able to attack a ship at all if it's too small. This adds to the frustration
c) It encourages players to use the magnetic pistol because with this you can set a deeper torpedo depth. Since using the magnetic pistol increases the failure rate due to the already existing premature explosion, I have reduced p0 a bit.

Regards, LGN1

Like the look and sound of those calculations

Originally Posted by h.sie

Regarding torpedo pistol: Unfortunately, until now I was not able to analyse the individual pistol setting of each torpedo, thus, ATM I cannot fulfil your wish, even if it makes sense to me.

Hi Mate I think I came across that Value before, I have a load of these saved for a Project I'm working on - I know where the Main I/M switch Value is Stored.
Don't think I have the state per torpedo saved but can find it easy enough.
I know that Torpedo selected is stored as a int value so if all 6 tubes selected = 63 This in Binary = 11111. And if only tube 3 is selected then value = 000100 or Int = 4.
Most likely this is the same storage method for Mag/Impact contact settings for each torp so 1=Magnetic 0=Impact.
That would mean a value of 100101 = 37 would be Torp 6,2 and 1 are set for Magnetic.
I'll load up sh3 and run some checks to confirm this - If this is the case I'll supply you the code and memory address to access these variable

Originally Posted by h.sie

But the guys who want the higher dud rates are also experienced players, and according to their experiences dud rates are too low.
Your input is welcome!

Excuse me ? Did you just called me a noob in SH3 ?

I'm with this game for 6 years and counting and I went through fire and hell improving my skills, knowledge, tactics, etc. This would probably make me an experienced player as well, but I never complained about dud rates are being too low. I literally ENJOY the view of explosion of the torpedo and then watching the ship sinking if I made everything right in TDC. Thus I'm getting full emotional pack of dissapointment when the torp goes dud even with perfect angle and TDC. Even if it was my last torpedo that went dud and I have to watch the ship passing by and we can not stop him. This was life and this is already in the game. No need to fix anything that is not broken.

Experienced players are usually bored, because they know how the game works, so they want to mess with their boredom playing SH3 by looking for some fresh new play feelings, which could re-triger their bored interest towards SH3 with fresh updates.

So if the experienced player thinks he does too much of tonnage, it doesn't mean he has to make his life complicating in game with something that would directly influence the gameplay by making it too complicated like more duds, more bad weather, more malfunctions etc etc.

If this is the exitement for experienced players as you say to come back with as little as possible tonnage sunken due to "bad luck" factor then of course I can't make you think otherwise

And Yes the fix like this is becoming some sort of masoschizm, which I think is just a fruit of simple boredom that came from too experienced players looking for new sense.

[h.sie]

@reaper: If we are really talking about the individual settings for the torpedoes and not about the pistol switch, your finding could be very helpful, because we then could differ between A and M pistol and give them individual failure rates....

[LGN1]

Originally Posted by h.sie

@LGN1: In your formula you assume a linear dependency between windspeed and waveheight. Is that applicable? From my "feeling" I'd say that the dependency is of higher order than 1. This can surely be found somewhere....

At least this source http://en.wikipedia.org/wiki/Beaufort_scale suggests a roughly linear dependency in the area of interest to us.

[h.sie]

@LGN1: Great. Linear dependency means easy programming. But when calculating wave height from windspeed, we should use a different factor than your table uses, since in-game 15m/s is storm, while in reality 15m/s is moderate.

[reaper7]

Originally Posted by h.sie

@reaper: If we are really talking about the individual settings for the torpedoes and not about the pistol switch, your finding could be very helpful, because we then could differ between A and M pistol and give them individual failure rates....

Well did my test, and even better I found Individual Memory locations for each Torpedo Tube for Impact/Magnetic.
And yes these are the values for the Torpedoes not the current switch position.
Float value = 12702.00098 Or displayed as byte value = 1 (Depending on variable type used)
Float value = 12702 Or displayed as byte value = 0

Now just need to Trace there DMA address and Pointers then use this to find that code and Memory address for ollydebug

Will post the results once I get this done for you mate.


Edit: Darn there not Fixed but temp values - I'll keep searching

[LGN1]

Originally Posted by h.sie

@LGN1: Great. Linear dependency means easy programming. But when calculating wave height from windspeed, we should use a different factor than your table uses, since in-game 15m/s is storm, while in reality 15m/s is moderate.

Hi h.sie,

as I've already written I'm reluctant to interpret the in-game wind speed differently because I think that storms are not as common as 15m/s wind speed in SH3 is. I think stronger winds are just missing in SH3 and you cannot fix this by interpreting 15m/s as a real storm. Such a procedure screws up the whole weather model.

In addition, with a factor of 0.3 we already have a minimum torpedo depth of 4.5m at 15 m/s. Take into account the 2m width and you end up with 6.5m. There are already quite a few ships in SH3 that you cannot attack anymore with the contact pistol setting at this depth. Maybe it's possible to increase the factor a bit, but if you increase it too much, players will hardly be able to attack any ships at 15m/s. And I think this is not realistic.

In my opinion, 15m/s wind should make things more complicated, but shouldn't be a show-stopper.

Regards, LGN1

[reaper7]

Hi H.Sie here is the First One, I'll wait till you confirm this is Ok before I trace the remaining 4.
Not sure If 212F2CEC is fixed memory location but code push below may be sufficient for your use.

004B6808 8850 24 MOV BYTE PTR DS:[EAX+24],DL
DL=01
DS:[212F2CEC]=00 or 01 (Impact or Magnetic)

[h.sie]

@reaper: Thanks very much. Please don't put too much time & effort into it, since the information you provided is completely sufficient for me to continue. I anyhow would re-engineer it myself, since I tend to trust noone.

Great finding.

[reaper7]

Originally Posted by h.sie

@reaper: Thanks very much. Please don't put too much time & effort into it, since the information you provided is completely sufficient for me to continue. I anyhow would re-engineer it myself, since I tend to trust noone.

Great finding.

Ok that one is just for Torpedo 1, you sure you don't need the rest - if you can find them from that one, that's great .

Glad to be of assistance, any addition I can make to yours and stieblers work is time well spent

[h.sie]

@LGN1. I agree. I think I shall put some values into your formula and see what failure rates result from that. This will make judging easier. But your formula looks very promising, since it approximates real physical behaviour, while my formula was only the mathematical description of the consensus between Hitmans and your data.

Due to reaper's finding its very likely that we are now able to differ between different pistols settings, thus, we could choose different values for p0, depending on pistol setting.

[LGN1]

Hi h.sie,

using a factor of 0.33 the difference between a linear fit and the values given in the link are below 3/4m (up to a wind speed of 20m/s). If one assumes that 15m/s in SH3 correspond to 20m/s one could use a factor of 0.33*20/15=0.44 Maybe randomize it between 0.3 and 0.45 Just kidding. I don't think it's that crucial which one chooses.

Concerning the pistol switch: Supposing you will figure out how to determine it (I'm pretty sure you will ), how do you plan to take it into account? Via p0, i.e., make p0 pistol-dependent?

EDIT: Cross-post. Question already answered. Thanks!

Cheers, LGN1