Realism mod (discussion)

[jenrick]

I think the appropriate question on soviet doctrine, is what other accounts of documents are available to so them using active sonar at a much higher incident then the West might have or not? If all the sources indicate they did, well then the normal pattern of historical research is to say that they probably did. Now two instances isn't a lot to base anything off of, but if it's all you have it's all you have.

I also don't view it as firing off an emergency flare personally. If I'm in the dark woods being hunted potentially by a wolf (we both have hearing, his is just significantly better then mine), I'm gonna use my flash light. It's the only chance I have. Please not the word potentially, that's the situation when we ambush them. It's wartime there's always a potential threat. A US sub is quieter, it has better passive gear. WTF do you do other then go active? Stay at home and don't fight a war was the historical answer, but beyond not even being there?

I agree that it runs absolutely counter to everything the West has developed about submarine warfare. This is also the country that launched counter attacks in WW2 with every man having a 5 round stripper clip of ammunition, and one in 5 or so having a rifle. The idea being that as men with rifles were shot, those without could pick them up. Losses against results, not a question of the human factor. There is a very different thought process in work due to cultural difference.

I also agree that their current torpedo evasion leads this to be a far more detrimental tactic then it might otherwise be.

-Jenrick

[Lanzfeld]

"Please work on the AI.

All it has to go on is one active ping and sends a torpedo in that direction, fair enough. What is not OK is the AI from that point on knowing exactly which direction you are heading, in real time, and correct the torpedo on every turn you make. Check this by modifying "SelfNoise" and "ActiveSonarReflection" to a very low number so there is no way you are detected through that, do active pinging until they launch, then go silent and deep. See the same with surface ships responding to a torpedo arriving, and from that knowing exactly where you are even tho you made a full 180 after launching them.
Seems to me like the AI knows exactly where you are but the devs have thrown in some randomizing on depth and distance until they have a more accurate sonar reading, in order for it to seem more legitimate. Problem is in the cases I've seen, they should have no reading to go on at all."

All of the above I copied this from another thread. If this is true I believe this should be addressed as soon as possible. I cannot stand it when the AI cheats in a game. Especially a simulator.

[Julhelm]

It doesn't cheat. The AI uses the same TMA and sensor algorithms as the player and generally shoots only on an >85% solution. What happens when they ping you is the same that happens when you ping them - if above the detection threshold, the solution % climbs rapidly, then each successive ping maintains it.

Enemy subs can launch wireguided torpedoes as well, and they will resteer them providing they have a >85% solution on you and can track you, just like you can.

[Lanzfeld]

Very good to know Julhelm and thank you for speaking up on the falsehood of this rumor.
I am glad I used the word "if" in my original concern

[Haukka81]

Originally Posted by Julhelm

It doesn't cheat. The AI uses the same TMA and sensor algorithms as the player and generally shoots only on an >85% solution. What happens when they ping you is the same that happens when you ping them - if above the detection threshold, the solution % climbs rapidly, then each successive ping maintains it.

Enemy subs can launch wireguided torpedoes as well, and they will resteer them providing they have a >85% solution on you and can track you, just like you can.

Maybe max "spawn" distance should be more when mission starts (least option to have say about 50, 25 is bit low) , now its easy to player to just shoot snap shot active torpedos because Ai can't run away. And then player would lost first contact more easily.

[Julhelm]

It likely is the Mk48 being OP. We used the 4000 yard sensor range quoted by Friedman, but some other sources say 1600 yards. Also it may not be noisy enough on its run-out. No figures are available so we had to guesstimate something that seemed reasonable.

[VizlaN]

Originally Posted by Julhelm

It doesn't cheat. The AI uses the same TMA and sensor algorithms as the player and generally shoots only on an >85% solution. What happens when they ping you is the same that happens when you ping them - if above the detection threshold, the solution % climbs rapidly, then each successive ping maintains it.

Enemy subs can launch wireguided torpedoes as well, and they will resteer them providing they have a >85% solution on you and can track you, just like you can.

Yes, active pinging gives them your solution which is the point. You then go silent after launch and watch the SNS COMP on passive drop to -30 or below, yet the torpedo keeps tracking by wire for several minutes. Why is that?

[jenrick]

Maybe max "spawn" distance should be more when mission starts (least option to have say about 50, 25 is bit low) , now its easy to player to just shoot snap shot active torpedos because Ai can't run away. And then player would lost first contact more easily.

My only concern here, is that we don't have any tools to help with the search. I have no clue how well or poorly the sonar is working at a given depth, and there's really no option but head in the direction of the last contact and do circles. I'm not advocating for a full DW type sonar experience, but I think it would take something else (not sure what) to make it work (for me at least).

-Jenrick

[Julhelm]

Originally Posted by VizlaN

Yes, active pinging gives them your solution which is the point. You then go silent after launch and watch the SNS COMP on passive drop to -30 or below, yet the torpedo keeps tracking by wire for several minutes. Why is that?

Don't know. I have to ask the programmer about it. Doesn't sound right.

[VizlaN]

Originally Posted by Julhelm

Don't know. I have to ask the programmer about it. Doesn't sound right.

SelfNoise+ActiveSonarReflection set to 1 and modified an added sonar(Los Angeles Flt II/III VLS mod) to keep him detected as much as possible.

[Julhelm]

Well that does look suspect. I'll forward the video.

[Lanzfeld]

Thank you! Let's at least start at the baseline of the AI not cheating please

[PL_Harpoon]

Ok guys. Here's the first version of "realism" mod.

Changes:

Campaign:
- reduced recon ranges - planes will pretty much detect only what's below them, satelites have a slightly bigger range but still half of what they had
- time 2x slower - that's just my personal preference
- bonus: Nato icons for the map - with the exception of satelites and airplanes, because in the game they rotate to face direction they're going


Movement:
- largely increased angular momentum for LA class sub (for testing) - the subs in CW feel a bit too weightless and responsive when it comes to turning. With this on they resemble behaviour from DW. It takes some time before the boat will start rotating and it will also take some time for it to stop. It's also more difficult to dodge torpedoes, but on the flip side, requires even more attention from the player. Right now only works for LA.

Sensors:
- MAD detection range reduced to 400 - even 1000 seemed to much for me
- Sonobuoys passive sensitivity reduced to 25 - I believe they were not as strong as BQQ-5.
- Dipping sonar sensitivity increased to 32 - I think the default 26 was a bit too weak
- decreased TMA rate for player and AI to 0.7 - before you could obtain a perfect TMA solutions very quickly. Now you'll have to work for them (change course, use ECM, etc.)

Weapons:
- increased warheads for all torpedoes and missiles - if known I used real tnt equivalent, if not I multiplied real warhead sizes by 1.8 (tnt ratio of Mk48 warhead materials) for later weapons and 1.3 (tnt ratio of Mk37 warhead materials) for earlier
- decreased acquisition range for Mark 48 to 1600. - based on a few sources on the internet - now it requires a proper solution or a bit of luck
- decreased sensor angle for Mark 48 to 60. - I think 80 was a bit too much
- increased acquisition range for Test-71 to 1500. - again, based on some internet sources
- decreased lifetime for knuckles to 3
- decreased noise (effectivnes) for knuckles to 50 - I think knuckles are a bit OP currently. With this mod they stay for very short time and work only when a torpedo is very close

Other:
- decreased player hull strength to 0.65 - to reflect single hull constructions - now most torpedoes will kill you
- increased AI hull strength to 1.25 - to reflect double hull. Takes 3 torps to sing a Typhoon, but for any other sub it's still pretty much 1 hit-kill.
- increased combat repair time to 2.5 - they felt too fast for me

Using MekStark's low underwater visibility mod.

One important thing: play it on HARD. On other some changes won't work.

Link: https://drive.google.com/open?id=0B5...2o4WDRIY1B0NHM

[Haukka81]

Lots of good stuff in your latest version. Hope that devs will take some of this to vanilla

[Nimmo55]

Thanks PL-Harpoon. Very much appreciated. It's one of the reasons I bought with confidence: the talented modders increasing the realism of the simulation.

Apoll