About Torpedoes

[Kane]

Torpedoes have been an issue in this game, and that issue has morphed over time rather than going away.  They're also an issue I would really like to see addressed on several points that I'm going to outline here.

1)  Torpedoes are too weak.  This applies both from a realism, and a game play perspective.  I will tackle both individually.

A.  Realism. 

Getting hit by a torpedo should be a big deal.  As it stands, its really not such a big deal.  Yes torpedo belts protect, but the threat they are protecting from is too damn weak.

Destroyers and Torpedo boats should not be surviving hits from 24" torpedoes.  At that point you're talking a warhead powerful enough to blow that ship in half.  There really isn't too much more for me to say on this point.  Torpedoes are simply too weak, the damage they inflict, particularly for the torpedoes in the 21" and up range needs to be increased dramatically.  Of course this also ties into the problem of small ships being way too tough, but that's another issue.

Where you hit also seems to make very little difference for a torpedo.  Sure a sufficient torpedo belt can neuter a hit from a torpedo.  But the torpedo actually has to hit the torpedo belt for this to happen.  As far as I can tell, in game a torpedo belt seems to protect the entire ship.  Thus a hit to the forecastle deals the same reduced damage as would a hit to the midships where the torpedo belt lives.  This should not be the case.

Also I wish to specify that when I talk about lack of damage, I am referring to the structural damage they inflict on the ship.  Flooding is highly dependent on hit location, bulkheads, crew training, etc.  But when you get into your larger torpedoes detonating hundreds of pounds of explosives, that should cause a lot of damage to the structure of the ship if there's not enough torpedo belt to absorb it.  Or when the torpedo does not hit the belt in the first place.

B.  Game Play

   As it stands, small ships such as torpedo boats and destroyers are too easy to hit, but can absorb way more punishment than they should when hit.  That said, getting a torpedo on the target tends to be a high-risk, high-reward tactic.  Given the ability of ships to dodge, and the ridiculous failure rate on torpedoes, one often has to get very, very close to the target to get a hit.  (Which often results in a dead TB or DD)  That hit will probably be a dud, but when its not, the damage is just too small.  High-risk should be high-reward, and as it stands torpedoes are high risk-low reward weapons.  There is no situation where this is good for game play.  To justify the difficulties of using them, their damage needs to be increased.


2)  The failure rate is too high, and it stays too high as time goes on.  Bad for both realism and game play.  Again I will tackle them individually.

A.  Realism

   Not every torpedo is a Mark-14.  It makes sense for torpedoes to have a high failure rate in the early years.  But this high-failure rate is arguably too high and unarguably continues for far too long.  As I noted in the general feedback thread.  In my recent campaign, it was the mid 1940's.  I had researched torpedoes to the point that I was working on my second capstone for torpedo propulsion.  I slipped one of my destroyers right up to an enemy battleship.  That destroyer was armed with 2x, quad 21-inch torpedo launchers.  Seven of those torpedoes hit.  Every single one was a dud.  There is absolutely no excuse for this.  This happens frequently, even in late game where my torpedoes have a 100% failure rate for a battle. 

   This wouldn't piss me off nearly as much while playing, if not for the fact that I am convinced that the AI does not have the same failure chance as the player.  In a post talking about doom stacks, I shared the experience of one battle.  I was Germany fighting Japan.  I was Very Advanced technologically, while Japan was Very Behind.  Early in the battle one of my battleships took a single 18" torpedo which caused an ammo explosion.  Most of the fight was spent with me circling that ship with my others, trying to protect it from the doom stack.

   In one instance I watched a torpedo boat that slipped past me cruise up to my crippled battleship, and launch 5, 18" torpedoes into it.  Every single one of them detonated.  Every damn one.  All while the Japanese had more primitive torpedoes that should have been more prone to duds than mine.  How is it that torpedoes that belong in the 190X's or so are more reliable than my late 1940's torpedoes? 

   I can count on one hand the number of times an AI torpedo has hit one of my battleships, and that torpedo failed to detonate.  I keep a close watch on the reports as the battle comes in.  It seems to me the AI has about 1/10th or less the failure rate of player torpedoes.  This needs to be looked at.  If it weren't for bad luck I'd have no luck at all, but I'm that unlucky.  If I have to eat this shit-sandwich, the AI should as well.



B.  Gameplay

   From a gameplay perspective, something this infuriating should not be allowed to stand.  The failure rate simply does not diminish over time and research the way it should.  As it stands, torpedoes are a wasted weapon and a wasted line of research.  Fire 10 torpedoes at the enemy, at least half of them will veer wildly off course.  Of your remaining five, three will detonate before ever getting to the target.  (Why is this even happening without magnetic exploders?)  This will leave you two that might actually hit the target, if the target doesn't dodge, and more often than not both of them will be duds.  This is infuriating for a player and makes them want to quit playing the game.  Or at least quit using torpedoes, and effectively abandon one of the features of the game.  This needs to be addressed in a big way.

3)  Torpedoes are under-developed.

    I hate to suggest any additional features while the game is so bug-ridden.  But compared to guns in the game, torpedoes are severely under-developed, and addressing that issue could potentially address some of the problems.  This is for example no option to switch between magnetic and impact detonation.  Many of the failures to detonate would be more acceptable if it were a result of magnetic detonators that the player could choose to not use (thus decreasing dud change with it.)

  There is no option to incrementally increase torpedo size.  (A la HMS Nelson with her 24.5" torpedoes.)  Whereas one can do it with guns to create 16.1" and what have you.

   There is no option to research things like RDX or Torpex for torpedoes.  (And for whoever is going to come and try to tell me that RDX and Torpex are the same thing, [because there's always one] no they're not.  Torpex is a mix of TNT, RDX, and Aluminum powder).  There is no option to research and refine magnetic exploders, etc.  Torpedoes really need to be expanded upon.

I for example would suggest the ability to research magnetic exploders, and be able to switch between impact and magnetic on your command bar during combat.  Magnetic would give failure rates similar to what we have now (at least when first researched), while dealing huge damage to anything they hit.  Since, reasonably, nothing but the largest battleships could potentially survive it.  The alternative of contact exploders would deal reduced damage, but have a far lower failure rate.  In any case, torpedoes need some love that they haven't gotten.

Edited November 11, 2022 by Kane

[Skeksis]

@Kane There's another aspect, the AI side.

The whole time I've been playing with dud torpedo's that have misfire against me, I've thought, "wow, that was lucky, I got away with that one". But the truth is, it was my mistake, I sailed in a straight line, or I wasn't watching enemies torpedo launch sequence, or just was lazy with my tactics.

I should have been punished for those mistakes. That was the case before the introduction of duds, if you made mistakes, you lost ships, plain and simple, git gud or lose. The AI was punishing and effective.

So yes, the other aspect is the contest/challenge, or the lack of because of duds. I don't know why or if Dev's realize, but there's a systematic development towards a weaker AI. Features have to be thought through encompassing the bigger picture. 

Without cheating, the AI needs any advantage it can muster, including an effective barrage of torpedo's.

Edited November 12, 2022 by Skeksis

[Pappystein]

While I agree that tin cans should be at least MOSTLY destroyed by a big torpedo hit.   I can't agree with much of the premises of this post.

However it is a well written post, based on more historical fact that many posts that raise concerns about this or that in the game.   Your post however forgets that this game makes everything quite abstract... to ALLOW the customization we DO have.   If you want a HYPER Accurate game, you are better off playing Computer Harpoon3 or Command Modern Operations.   Sure both are strategy games vs this Tac-Strategy game... and are much less LEGO-able but both offer some ways to edit platforms.

Some of your concerns are valid... But others are specific to the history of one or two nations (the whole contact/magnetic fuse failure thing.)  

Most of the changes you seek would require an entirely new damage model to replace the existing one in the game.   The fact of the matter is, the game designers for good or bad decided to make Torpedoes a lower fidelity than guns...  If they were the same detail as guns, in my opinion, the the minutia for Torpedoes would outweigh that for gunfire combat.   That means the developers would have to spend a HUGE ammount of time "bringing to Torpedoes up to scratch"  

Between the two, the gun and the torpedo, we have much more quantitative and qualitative data on gun vs armor combat.   Torpedo combat is much more SUBJECTIVE data instead of quantitative.   That means a lot of the data is more anecdotal "Eye of the beholder" and "Blind stupid luck" data rather than empirical repeatable data.   There are just too many variables that have to align correctly in the syrup known as the OCEAN for a Torpedo's damage to be able to be quantitatively measured.  Air quality matters for guns but no where near as much as water quality does for torpedoes. 

Also the whole point of magnetic explosive fuse was for influence explosions below big ships... something you obliquely reference... but you forgot to mention needing another button added to "Run deep"    But this also introduces yet another reason the damage model would need to be replaced.

 

 

[Kane]

6 hours ago, Pappystein said:

While I agree that tin cans should be at least MOSTLY destroyed by a big torpedo hit.   I can't agree with much of the premises of this post.

However it is a well written post, based on more historical fact that many posts that raise concerns about this or that in the game.   Your post however forgets that this game makes everything quite abstract... to ALLOW the customization we DO have.   If you want a HYPER Accurate game, you are better off playing Computer Harpoon3 or Command Modern Operations.   Sure both are strategy games vs this Tac-Strategy game... and are much less LEGO-able but both offer some ways to edit platforms.

Some of your concerns are valid... But others are specific to the history of one or two nations (the whole contact/magnetic fuse failure thing.)  

Most of the changes you seek would require an entirely new damage model to replace the existing one in the game.   The fact of the matter is, the game designers for good or bad decided to make Torpedoes a lower fidelity than guns...  If they were the same detail as guns, in my opinion, the the minutia for Torpedoes would outweigh that for gunfire combat.   That means the developers would have to spend a HUGE ammount of time "bringing to Torpedoes up to scratch"  

Between the two, the gun and the torpedo, we have much more quantitative and qualitative data on gun vs armor combat.   Torpedo combat is much more SUBJECTIVE data instead of quantitative.   That means a lot of the data is more anecdotal "Eye of the beholder" and "Blind stupid luck" data rather than empirical repeatable data.   There are just too many variables that have to align correctly in the syrup known as the OCEAN for a Torpedo's damage to be able to be quantitatively measured.  Air quality matters for guns but no where near as much as water quality does for torpedoes. 

Also the whole point of magnetic explosive fuse was for influence explosions below big ships... something you obliquely reference... but you forgot to mention needing another button added to "Run deep"    But this also introduces yet another reason the damage model would need to be replaced.

 

 

I'm not sure what point you're making with your screen-shot, as I noted multiple times in my post that I had researched torpedo tech all the way to capstones, and still had ridiculous failure rates.

That aside.  I don't think addressing many of my concerns would require that big a change.  The fact that torpedoes don't do nearly enough structural damage seems like the kind of thing that could be easily handled with some numbers tweaking, and simply upping that value.  Unless the developers have used some kind of incredibly convoluted formula to calculate damage then this shouldn't be that big a change.

The failure rate as well is something that should be a relatively simple number-tweaking, that really shouldn't require all that much work.  It would also help if the AI had to play with the same failure rate we do, which I refuse to believe is the case.  As the old saying goes; "Once is coincidence, twice is happenstance, three times is enemy action."  Too many times now I've seen a wildly imbalanced success-failure ratio that heavily benefits the AI.  Again, this should not be that difficult to address.

Lastly on the idea of magnetic exploders, yes, I do imagine this one would require a fair amount of work on the developer's part.  (r not, depending on how its implemented) Incremental torpedo sizes would probably take some as well.  Work that I think would be worthwhile, but I'm an opinion of 1.  The idea of magnetic fuses was simply a way to justify the ridiculous failure-rate currently in game, in the event that some feel a high torpedo failure rate is needed for game balance purposes.  (Also goes back to the high-risk / high reward idea)

As it stands, every torpedo in game is a Mark-XIV (and from what I can tell actually has a higher failure rate than the actual Mark-XIV did).  This really needs to be dealt with, or there's not much point in ever bothering to put torpedoes on a ship.  There would be even less point if a player wanted to try to go the Japanese plan and build a torpedo-centric fleet.  Such a strategy would be effective suicide in game. 

Edit:  Now torpedo damage vs. hit location / torpedo belt probably would require some work to the damage model.  I'd argue that's work that should have been done in the first place, and if it requires tweaking to the existing damage model, so be it.

Edited November 14, 2022 by Kane

[Mundus_Dog]

See below i forgot to do the quotes

Edited November 15, 2022 by Mundus_Dog
text editing

[Mundus_Dog]

6 hours ago, Kane said:

I'm not sure what point you're making with your screen-shot, as I noted multiple times in my post that I had researched torpedo tech all the way to capstones, and still had ridiculous failure rates.

That aside.  I don't think addressing many of my concerns would require that big a change.  The fact that torpedoes don't do nearly enough structural damage seems like the kind of thing that could be easily handled with some numbers tweaking, and simply upping that value.  Unless the developers have used some kind of incredibly convoluted formula to calculate damage then this shouldn't be that big a change.

The failure rate as well is something that should be a relatively simple number-tweaking, that really shouldn't require all that much work.  It would also help if the AI had to play with the same failure rate we do, which I refuse to believe is the case.  As the old saying goes; "Once is coincidence, twice is happenstance, three times is enemy action."  Too many times now I've seen a wildly imbalanced success-failure ratio that heavily benefits the AI.  Again, this should not be that difficult to address.

Lastly on the idea of magnetic exploders, yes, I do imagine this one would require a fair amount of work on the developer's part.  (r not, depending on how its implemented) Incremental torpedo sizes would probably take some as well.  Work that I think would be worthwhile, but I'm an opinion of 1.  The idea of magnetic fuses was simply a way to justify the ridiculous failure-rate currently in game, in the event that some feel a high torpedo failure rate is needed for game balance purposes.  (Also goes back to the high-risk / high reward idea)

As it stands, every torpedo in game is a Mark-XIV (and from what I can tell actually has a higher failure rate than the actual Mark-XIV did).  This really needs to be dealt with, or there's not much point in ever bothering to put torpedoes on a ship.  There would be even less point if a player wanted to try to go the Japanese plan and build a torpedo-centric fleet.  Such a strategy would be effective suicide in game. 

Edit:  Now torpedo damage vs. hit location / torpedo belt probably would require some work to the damage model.  I'd argue that's work that should have been done in the first place, and if it requires tweaking to the existing damage model, so be it.

 

13 hours ago, Pappystein said:

While I agree that tin cans should be at least MOSTLY destroyed by a big torpedo hit.   I can't agree with much of the premises of this post.

However it is a well written post, based on more historical fact that many posts that raise concerns about this or that in the game.   Your post however forgets that this game makes everything quite abstract... to ALLOW the customization we DO have.   If you want a HYPER Accurate game, you are better off playing Computer Harpoon3 or Command Modern Operations.   Sure both are strategy games vs this Tac-Strategy game... and are much less LEGO-able but both offer some ways to edit platforms.

Some of your concerns are valid... But others are specific to the history of one or two nations (the whole contact/magnetic fuse failure thing.)  

Most of the changes you seek would require an entirely new damage model to replace the existing one in the game.   The fact of the matter is, the game designers for good or bad decided to make Torpedoes a lower fidelity than guns...  If they were the same detail as guns, in my opinion, the the minutia for Torpedoes would outweigh that for gunfire combat.   That means the developers would have to spend a HUGE ammount of time "bringing to Torpedoes up to scratch"  

Between the two, the gun and the torpedo, we have much more quantitative and qualitative data on gun vs armor combat.   Torpedo combat is much more SUBJECTIVE data instead of quantitative.   That means a lot of the data is more anecdotal "Eye of the beholder" and "Blind stupid luck" data rather than empirical repeatable data.   There are just too many variables that have to align correctly in the syrup known as the OCEAN for a Torpedo's damage to be able to be quantitatively measured.  Air quality matters for guns but no where near as much as water quality does for torpedoes. 

Also the whole point of magnetic explosive fuse was for influence explosions below big ships... something you obliquely reference... but you forgot to mention needing another button added to "Run deep"    But this also introduces yet another reason the damage model would need to be replaced.

 

 

you can literally change the value yourself for torpedo damage (take sapphires mods for example) its not difficult and i agree that the failure rate is way to high for the player and way to low for the ai i mean ive had torpedos not even get to target 90% of the time i fired them in a single battle and then the one that does is magically a dud but if a friendly ship that im not controling at the moment decides to sail in front of the torpedo ship it gets blown to smithereens by the whole salvo. 

unfortunately the research really does nothing and i no longer build any ships with torpedos its not worth the pain 

we arent asking for a hyper realism simulator we are asking for a reason to use torpedos and to make ships with torpedos

the whole system of where the torpedo hits could possibly be easy too but i dont know as i dont develop games and the custom sizes is another thing that sounds easy but probably isnt

i have also noticed that a lot of the percentages mean nothing in the game unles it is the cost of something then it makes all the difference

[Lastreaumont]

I remember previous versions where torpedoes were fully efficient. It was not my best remembrances. 😉

[Abuse_Claws]

A question to the OP: while I understand that your torpedo tech was maxed out, what was the status of your explosives research? IIRC, torpedo dud chance is increased by torpedo size and decreased by researching Torpedo Contact Exploder tech within Explosives research. So if you are using large torps with outdated exploders, you are kind of creating a dud-machine

Also, I think you might be overestimating the amount of damage that torpedoes should do. For example, I've looked at a few US destroyers that were sunk by 24-inch 'Long Lance' (Type 93) torps. Here are the results:

• USS Benham - 2250 tons (full displacement, which as far as I understand UAD uses) - got hit by 1 torpedo. Was it torn apart? Yes, the bow was ripped off from bridge forward. Was it sunk? No, it stayed afloat and was later scuttled
• USS Barton - 2474t - 2 torps - torn apart: yes, in half - sunk: yes
• USS Blue - 2325t - 1 torp - torn apart: no - sunk: no (scuttled)
• USS Strong - 2500t - 1 torp, rammed and detonation of depth charges - torn apart: yes - sunk: yes
• USS Gwin - 2395t - 1 torp - torn apart: no - sunk: no (scuttled)
• USS Chevalier - 2500t - 1 torp and rammed - torn apart: yes, bow ripped off from bridge forward - sunk: no (scuttled, including the torn-off bow, which was found still afloat and had to be scuttled separately)
• USS Cooper - 3515t - 1 torp - torn apart: yes, in half - sunk: yes

So even 24'' torps do not reliably sink or tear apart even destroyers. It may depend on whether or not destroyer's own munitions, boilers and engines explode, weather conditions and extra damage (i haven't included some destroyers that were hit by large-caliber shells as well as torps)

In my experience so far in UAD 1.09+ (Spanish campaign 1890-1925) torpedoes are a pretty reliable way of sinking even capital ships. My tactic is rather simple:

1. Find a separated enemy ship. The more ships the enemy has, the more chances they have to sink my incoming destroyers/TBs
2. Approach it preferably from the bow at flank speed. That way the enemy has minimal time to respond to the threat (as the relative approach speed is maximized) and also for most designs that puts your destroyer in a blind spot for most of the targets secondaries (and main guns have really low odds of hitting such a fast target)
3. Deploy smoke as late as possible to save as much smoke as possible for the retreat. Wait for the point where you are close enough for enemy to have a real chance of hitting you and deploy smoke only for ships the enemy is actually firing at, as first shots almost always go wide.
4. Ideally I prefer to have 2 pairs of destroyers on one target, both pairs moving in a tight ahead formation. The idea is to flank the victim from both sides, making any attempts to turn futile
5. Torps are disabled until the destroyer is almost (to give the tubes time to rotate on target) side-to-side with the victim, like 200m from it, to reliably achieve multiple hits on different sections of the target (preferably the destroyers should have at least two launchers not next to each other to get a proper spread)
6. If any destroyers are damaged in the attack, immediately detach them from the formation and turn them away so that they do not hinder other destroyers' movement

Usually such an attack ends with 1-2 of my destroyers lightly damaged and the enemy ship *very* dead (usually the target begins to sink after eating just one spread of torps, the second spread is to sink it quickly to prevent it from firing at my retreating destroyers and the last 2 destroyers are mostly just backup in case lead destroyers are hit or they somehow miss or there are a lot of duds)

Edit: the reason for such short distance of attack is that AI is exceptionally good at spotting and dodging torps. Maybe I'm just bad at it, but I seem to have more issues with doing the same

Edited November 15, 2022 by Abuse_Claws
typo

[o Barão]

Some quick tests I made now. BB vs DD at 1km range.

Half the AI torpedoes deviate from course away from my ship. Not against my ship.

6 torpedo hits - 4 duds.

 

This is a simple run test to show RNG works both ways. The problem is the human player natural reaction to only give importance to when things work badly against him and then reach the wrong conclusion that the AI is somehow having an unfair advantage.

 

There is one BIG thing I dislike about the torpedoes. These tests were made at 1 km distance, and for this close distance we can always see many torpedoes losing track and turn one side or another or maybe exploding. Now let's run another test where I set the AI DD at 20 km range with 15 oxygen torps.

Only one torpedo come close to me. 3 are spotted to the left.

I run again the same test and this time the results were worse.

I set the camera to follow the AI DD to see how many torpedoes were going to explode. I count four explosions from the total 10 torpedoes launched. 5 torpedoes from the mid ship were not used in this test. My sonar this time didn't pick any torpedoe. So the conclusion is from the total 10 torpedoes lauched all 10 were failures.

This is a 100% rate failure for torpedoes launched at long distances. RNG can be a bitch, but this clear show the failures rate is too high to the point that make it pointless to use them at long distances. This is not only unrealistic, but  it also makes the AI performance in battle weaker.

 

[Kane]

12 hours ago, Mundus_Dog said:

 

you can literally change the value yourself for torpedo damage (take sapphires mods for example) its not difficult and i agree that the failure rate is way to high for the player and way to low for the ai i mean ive had torpedos not even get to target 90% of the time i fired them in a single battle and then the one that does is magically a dud but if a friendly ship that im not controling at the moment decides to sail in front of the torpedo ship it gets blown to smithereens by the whole salvo. 

unfortunately the research really does nothing and i no longer build any ships with torpedos its not worth the pain 

we arent asking for a hyper realism simulator we are asking for a reason to use torpedos and to make ships with torpedos

the whole system of where the torpedo hits could possibly be easy too but i dont know as i dont develop games and the custom sizes is another thing that sounds easy but probably isnt

i have also noticed that a lot of the percentages mean nothing in the game unles it is the cost of something then it makes all the difference

"unfortunately the research really does nothing and i no longer build any ships with torpedos its not worth the pain " I'm glad I'm not the only one who sees this.

Edited November 15, 2022 by Kane

[Kane]

3 hours ago, o Barão said:

Some quick tests I made now. BB vs DD at 1km range.

Half the AI torpedoes deviate from course away from my ship. Not against my ship.

6 torpedo hits - 4 duds.

 

This is a simple run test to show RNG works both ways. The problem is the human player natural reaction to only give importance to when things work badly against him and then reach the wrong conclusion that the AI is somehow having an unfair advantage.

 

There is one BIG thing I dislike about the torpedoes. These tests were made at 1 km distance, and for this close distance we can always see many torpedoes losing track and turn one side or another or maybe exploding. Now let's run another test where I set the AI DD at 20 km range with 15 oxygen torps.

Only one torpedo come close to me. 3 are spotted to the left.

I run again the same test and this time the results were worse.

I set the camera to follow the AI DD to see how many torpedoes were going to explode. I count four explosions from the total 10 torpedoes launched. 5 torpedoes from the mid ship were not used in this test. My sonar this time didn't pick any torpedoe. So the conclusion is from the total 10 torpedoes lauched all 10 were failures.

This is a 100% rate failure for torpedoes launched at long distances. RNG can be a bitch, but this clear show the failures rate is too high to the point that make it pointless to use them at long distances. This is not only unrealistic, but  it also makes the AI performance in battle weaker.

 

 Give me a little credit and assume that I wouldn't be bitching about this unless I had a pretty big sample size to work with.  I may not be the most prolific poster on this board, but I think I've done a pretty good job of bringing logic, reason, testing, and history to the posts I've made when I've complained about something.  (Such as when I was making posts pointing out that until recently the weight of battleship radios could be measured in Tiger-tanks.)


I have run the very test you mention in this article, many, many, many times.  Do what I did, run this test about a hundred times and tell me if you still don't see a bias in favor of the AI.  If after all that you don't, well then I guess I'll have to admit to having incredibly bad luck. 

And when I am referring to failure bias for the AI, I am referring specifically to failure to detonate.  I am not counting deviations, etc.  (I should have been more clear on that.)  I've run this test a lot, and I'm sorry I just can't believe the AI is playing with the same failure rate on that that we are.

At least we can agree that either way, the failure rate is too damn high.

Edited November 15, 2022 by Kane

[Kane]

5 hours ago, Abuse_Claws said:

A question to the OP: while I understand that your torpedo tech was maxed out, what was the status of your explosives research? IIRC, torpedo dud chance is increased by torpedo size and decreased by researching Torpedo Contact Exploder tech within Explosives research. So if you are using large torps with outdated exploders, you are kind of creating a dud-machine

Also, I think you might be overestimating the amount of damage that torpedoes should do. For example, I've looked at a few US destroyers that were sunk by 24-inch 'Long Lance' (Type 93) torps. Here are the results:

• USS Benham - 2250 tons (full displacement, which as far as I understand UAD uses) - got hit by 1 torpedo. Was it torn apart? Yes, the bow was ripped off from bridge forward. Was it sunk? No, it stayed afloat and was later scuttled
• USS Barton - 2474t - 2 torps - torn apart: yes, in half - sunk: yes
• USS Blue - 2325t - 1 torp - torn apart: no - sunk: no (scuttled)
• USS Strong - 2500t - 1 torp, rammed and detonation of depth charges - torn apart: yes - sunk: yes
• USS Gwin - 2395t - 1 torp - torn apart: no - sunk: no (scuttled)
• USS Chevalier - 2500t - 1 torp and rammed - torn apart: yes, bow ripped off from bridge forward - sunk: no (scuttled, including the torn-off bow, which was found still afloat and had to be scuttled separately)
• USS Cooper - 3515t - 1 torp - torn apart: yes, in half - sunk: yes

So even 24'' torps do not reliably sink or tear apart even destroyers. It may depend on whether or not destroyer's own munitions, boilers and engines explode, weather conditions and extra damage (i haven't included some destroyers that were hit by large-caliber shells as well as torps)

In my experience so far in UAD 1.09+ (Spanish campaign 1890-1925) torpedoes are a pretty reliable way of sinking even capital ships. My tactic is rather simple:

1. Find a separated enemy ship. The more ships the enemy has, the more chances they have to sink my incoming destroyers/TBs
2. Approach it preferably from the bow at flank speed. That way the enemy has minimal time to respond to the threat (as the relative approach speed is maximized) and also for most designs that puts your destroyer in a blind spot for most of the targets secondaries (and main guns have really low odds of hitting such a fast target)
3. Deploy smoke as late as possible to save as much smoke as possible for the retreat. Wait for the point where you are close enough for enemy to have a real chance of hitting you and deploy smoke only for ships the enemy is actually firing at, as first shots almost always go wide.
4. Ideally I prefer to have 2 pairs of destroyers on one target, both pairs moving in a tight ahead formation. The idea is to flank the victim from both sides, making any attempts to turn futile
5. Torps are disabled until the destroyer is almost (to give the tubes time to rotate on target) side-to-side with the victim, like 200m from it, to reliably achieve multiple hits on different sections of the target (preferably the destroyers should have at least two launchers not next to each other to get a proper spread)
6. If any destroyers are damaged in the attack, immediately detach them from the formation and turn them away so that they do not hinder other destroyers' movement

Usually such an attack ends with 1-2 of my destroyers lightly damaged and the enemy ship *very* dead (usually the target begins to sink after eating just one spread of torps, the second spread is to sink it quickly to prevent it from firing at my retreating destroyers and the last 2 destroyers are mostly just backup in case lead destroyers are hit or they somehow miss or there are a lot of duds)

Edit: the reason for such short distance of attack is that AI is exceptionally good at spotting and dodging torps. Maybe I'm just bad at it, but I seem to have more issues with doing the same

My research was up to Triple-Base for explosives.  I'll also note that Torpedo Propulsion, also (at least claims) to reduce dud chance, and I was researching capstones on that one.

As to your examples.
USS Benham:  Would you imply that the British did not sink the Bismarck because the Germans scuttled her?  No.  The Brits killed her, the Germans scuttled her to make sure she went down.  Benham was mission-killed by torpedo, unrecoverable, and finished off by scuttling.  For all intents and purposes the torpedo sank her.  It definitely put her out of action.

Same to many of the others you listed, even if the ship did not literally sink, its fighting days were done once the torpedo hit.  There was no taking the torpedo, then merrily continuing the fight as though it were just a minor inconvenience the way destroyers do in this game.  (To say nothing of the buffet of 16" shells cruisers can eat and keep fighting.  But, small ships being way tougher than they should be is its own issue.)

Edit:  Also, all those examples are from one torpedo.  Unlike this game where a DD will take multiple hits from  24" torpedoes and still will continue fighting.

Note:  The steps you mentioned are similar to what I use, though I'll pretty much always have two separate groups of DD's on the same target in order to cross-torps, since as has been mentioned multiple times in this thread, the AI may be a little too good at dodging torpedoes.

Edited November 15, 2022 by Kane

[Abuse_Claws]

9 minutes ago, Kane said:

My research was up to Triple-Base for explosives.

Then the dud rate is genuinely puzzling. IDK, I don't have nearly as many duds, maybe just lucky
 

 

10 minutes ago, Kane said:

Same to many of the others you listed, even if the ship did not literally sink, its fighting days were done once the torpedo hit. 

Well, that's an interesting topic. The game does not simulate (in this regard) weather conditions or enemy pursuit ("Straggle" missions do happen sometimes, but definitely not all the time) which would force the ship to be scuttled, as it often happened in real life. But there are also examples of more lucky ships: SMS Seydlitz managed to get home safely even though anyone in their right mind would probably have scuttled this wreck right away.

In game we get "heavy damage", which takes months of expensive repairs to remove making it almost easier to build a new one. And if you get a Straggle mission while limping home, you can just surrender right away. I say it's a more fair representation of ships that would be scuttled IRL than straight up destruction, which would claim not just the ship, but also the crew, which might be really valuable, if it's a couple thousands of veteran sailors.

But nevertheless the DDs I mentioned certainly weren't ripped to shreds even by the 24" torps. Crippled - yes, sometimes broken in half (and I don't know whether in some of those cases there were also ammo explosions contributing to the scale of damage), but nothing drastic enough to say a 24" torp will absolutely surely kill a DD every single time. If weather conditions, enemy aircraft and pursuing forces were of no concern (like it basically is in UAD), some of those ships may have been towed to their bases and possibly even repaired later (certainly not reasonable, but technically possible), and most of the crews certainly were rescued even with all those aforementioned concerns in play

[o Barão]

54 minutes ago, Kane said:

 Give me a little credit and assume that I wouldn't be bitching about this unless I had a pretty big sample size to work with.

I had to read again the OP just to check if I miss something the first time.

"The failure rate is too high, and it stays too high as time goes on.  Bad for both realism and game play.  Again I will tackle them individually."

If you are talking about this, I agree 100%. Now...

 

"....if not for the fact that I am convinced that the AI does not have the same failure chance as the player."

This is the part where I disagree, and then you continue with this...

 

"I shared the experience of one battle.  I was Germany fighting Japan.  I was Very Advanced technologically, while Japan was Very Behind.  Early in the battle one of my battleships took a single 18" torpedo which caused an ammo explosion.  Most of the fight was spent with me circling that ship with my others, trying to protect it from the doom stack."

This is nothing. If you don't understand what is RNG that is another issue. One example doesn't mean nothing.

 

"In one instance I watched a torpedo boat that slipped past me cruise up to my crippled battleship, and launch 5, 18" torpedoes into it.  Every single one of them detonated.  Every damn one.  All while the Japanese had more primitive torpedoes that should have been more prone to duds than mine.  How is it that torpedoes that belong in the 190X's or so are more reliable than my late 1940's torpedoes? "

Same story, this is nothing. Bad luck with RNG, nothing more.

 

Now this...

"Do what I did, run this test about a hundred times and tell me if you still don't see a bias in favor of the AI."

I will not do that, sorry but 100 times is too much, however if you did notice my long range torpedo test you will see it was the AI firing against me. I only sailed in a straight line. If there was a BIAS to favor the AI we would will see at least some average perfomance rate right? Well 100% failure rate just shows how bad it is both to you and the AI.

Edited November 15, 2022 by o Barão

[Kane]

23 minutes ago, o Barão said:

I had to read again the OP just to check if I miss something the first time.

"The failure rate is too high, and it stays too high as time goes on.  Bad for both realism and game play.  Again I will tackle them individually."

If you are talking about this, I agree 100%. Now...

 

"....if not for the fact that I am convinced that the AI does not have the same failure chance as the player."

This is the part where I disagree, and then you continue with this...

 

"I shared the experience of one battle.  I was Germany fighting Japan.  I was Very Advanced technologically, while Japan was Very Behind.  Early in the battle one of my battleships took a single 18" torpedo which caused an ammo explosion.  Most of the fight was spent with me circling that ship with my others, trying to protect it from the doom stack."

This is nothing. If you don't understand what is RNG that is another issue. One example doesn't mean nothing.

 

"In one instance I watched a torpedo boat that slipped past me cruise up to my crippled battleship, and launch 5, 18" torpedoes into it.  Every single one of them detonated.  Every damn one.  All while the Japanese had more primitive torpedoes that should have been more prone to duds than mine.  How is it that torpedoes that belong in the 190X's or so are more reliable than my late 1940's torpedoes? "

Same story, this is nothing. Bad luck with RNG, nothing more.

 

Now this...

"Do what I did, run this test about a hundred times and tell me if you still don't see a bias in favor of the AI."

I will not do that, sorry but 100 times is too much, however if you did notice my long range torpedo test you will see it was the AI firing against me. I only sailed in a straight line. If there was a BIAS to favor the AI we would will see at least some average perfomance rate right? Well 100% failure rate just shows how bad it is both to you and the AI.

I'm well aware of what RNG is.  Do you know what statistics are?

I gave the examples I did because they're a good indicator of the statistics at work.  1 for 1 it could be brushed off as RNG.  AI gets good, I get bad.  When you start increasing the sample size and getting the same result, that's more than RNG.

So in short, you're going to continue arguing with me even though you refuse to do the same tests I did?  Ok....
Try this then, another thing I did.

Get a pen and paper.  Keep a log.  Every time you see an AI torpedo hit in the combat log, write it down, with a + for detonation or - for dud. Do the same for your own torpedoes.
After you've played a campaign or two, go back and average out the ones you wrote down to see the dud chance, player vs AI.

If you do decide to do that and get a different result, let me know, and I'll have to admit its just suspiciously bad luck.
Do that, and odds are you'll see what I'm seeing.  Namely that AI has about 1/10th the dud chance on impact that the player does.  Almost as if someone missed a decimal somewhere...

If you're not going to do either of those, then you really have no business trying to argue the point with someone who is doing the testing.

Edited November 15, 2022 by Kane

[Kane]

18 minutes ago, Abuse_Claws said:

Then the dud rate is genuinely puzzling. IDK, I don't have nearly as many duds, maybe just lucky
 

 

Well, that's an interesting topic. The game does not simulate (in this regard) weather conditions or enemy pursuit ("Straggle" missions do happen sometimes, but definitely not all the time) which would force the ship to be scuttled, as it often happened in real life. But there are also examples of more lucky ships: SMS Seydlitz managed to get home safely even though anyone in their right mind would probably have scuttled this wreck right away.

In game we get "heavy damage", which takes months of expensive repairs to remove making it almost easier to build a new one. And if you get a Straggle mission while limping home, you can just surrender right away. I say it's a more fair representation of ships that would be scuttled IRL than straight up destruction, which would claim not just the ship, but also the crew, which might be really valuable, if it's a couple thousands of veteran sailors.

But nevertheless the DDs I mentioned certainly weren't ripped to shreds even by the 24" torps. Crippled - yes, sometimes broken in half (and I don't know whether in some of those cases there were also ammo explosions contributing to the scale of damage), but nothing drastic enough to say a 24" torp will absolutely surely kill a DD every single time. If weather conditions, enemy aircraft and pursuing forces were of no concern (like it basically is in UAD), some of those ships may have been towed to their bases and possibly even repaired later (certainly not reasonable, but technically possible), and most of the crews certainly were rescued even with all those aforementioned concerns in play

"but nothing drastic enough to say a 24" torp will absolutely surely kill a DD every single time"

I ask you to quote where in any of my posts I said this.  But I will concede your points.  I'm going to have to be more conscientious about how I phrase things.  I pointed out that they should not be surviving hits from 24-inchers, and expanded on that in later answers to point out that the problem is that not only are they taking 24-inchers and surviving, they're remaining combat effective.  They're even taking multiple 24-inchers, and remaining combat effective.

Can we at least agree that this should not be the case?  That a destroyer eating a 24" torpedo should be a really big deal for that destroyer, and that it probably shouldn't be merrily going about its way in combat as if it only took a minor hit?

And you are correct, the game does not model things like scuttling ships.  (Wish it did).  Unless the game decides to start doing that, perhaps we should look at the effects of such a devastating hit being more....immediate.

[o Barão]

10 minutes ago, Kane said:

If you do decide to do that and get a different result, let me know, and I'll have to admit its just suspiciously bad luck.

Sorry 100 tests is just too much..

13 hits only 2 duds

14 hits and again only 2 duds

17 hits, one dud and one ammo detonation. Do you think 3 tests is enough for you? It is for me.

[Kane]

Edit:  Post deleted for being unnecessarily hostile.

 

Edited November 15, 2022 by Kane

[Kane]

To be frank, if I were to take your tests at face value and assume 3 is a respectable sample size, the most I could draw from that is that I'm wrong about what exactly the problem is.  I would actually hate for that to be the case since that would mean something else, somewhere else in the game code is bugged.  Which would mean have to run many more tests to find out whether its a bug in research, in AI, or something else.  Given those options I really hope its an issue in the dud rate the AI has since that would be a metric shit ton easier to fix.

Edit:  Though the more I think about that, the more I may have to look into that possibility given how bug-ridden the rest of the game is.

If anyone else is not seeing this problem the way @o Barãois not, please let me know.

Edited November 16, 2022 by Kane

[Lima]

I've never really liked torpedoes, in just a few campaigns I really focused on them, but it was in the old versions. After the changes with the torpedoes, I looked at how many of them explode before reaching the target by 10 km, and decided that this was not for me. At the same time, I can't say that I'm very sad about the previous torpedo soup.

But let's get back to the statistics. It's quite difficult for me to remember the last time torpedoes hit my ships, I try very hard to avoid it (War on the Sea habit). Now I install torpedoes only on destroyers, 3-4 of them and always with reduced ammunition. I use them to finish off damaged ships, in a skirmish at the beginning of a major battle and in battles like "Destroyer Ambush".

About 3/4 of torpedoes explode on damaged ships. I am not too interested in where the torpedoes went at the beginning of a big battle, their task is to confuse the enemy. And only in the last type of battles can I tell how many torpedoes hit in combat conditions.

In this battle, each of my ships carried 3 torpedoes, that's 12 torpedoes. 5 hit, 3 exploded.

In general, remembering about a dozen such battles and a lot of finishing ships, I can say that the chances of a torpedo explosion at the angle of impact about 60-90 degrees definitely more than fifty percent.

As for the problems with torpedoes, something really needs to be done here. Now it's easier for a destroyer to eat several torpedoes than to get one shell hit. In addition, the explosions of torpedoes on the way to the target make them of little use at long distances.

[Abuse_Claws]

13 hours ago, Kane said:

Can we at least agree that this should not be the case?  That a destroyer eating a 24" torpedo should be a really big deal for that destroyer, and that it probably shouldn't be merrily going about its way in combat as if it only took a minor hit?

100% agree on that. I just think it's the problem with destroyers being way too tanky rather than with torp damage, because on bigger targets (BBs and BCs) at least for me torps generally work as expected: 4-6 torp hits reliably sink a contemporary capital ship. Sometimes ships survive way more punishment than they should (like a DD eating 3 torps at once and somehow surviving that), sometimes they go down from just a single lucky hit to the magazine(my poor BB Espana, you are gone, but not forgotten) - just like in real life.

For example, let's take USS Sealion's track record: with 21" torps it took 2 or 3 hits to sink Kongo - an outdated basically BC, just 1 to obliterate (via ammo detonation) DD Urakaze and later 6 (!!!) to sink a lowly transport Mamiya, with the first 4 hits not sinking the ship and therefore Sealion being forced to shoot 3 more torps for 2 hits to finally put the thing down.

14 hours ago, Kane said:

And you are correct, the game does not model things like scuttling ships.  (Wish it did).  Unless the game decides to start doing that, perhaps we should look at the effects of such a devastating hit being more....immediate.

I too wish it did, as I think it's an important aspect of naval warfare. However, as trained crewmembers are a valuable resource in UAD, I don't think making it so that ships that would've been scuttled IRL are outright destroyed would be fair. So far I'm just waiting for scuttling ships to be implemented and putting extra torp salvoes on my targets

13 hours ago, Kane said:

"Sorry, 100 tests is just too much"

Well, if we assume (and I see no reason not to) that every torpedo has an equal and independent dud chance, from @o Barão's screenshots we have a Bernoulli trial sample with n = 44, k = 5. With 95% confidence we can place the dud chance between 3.8% and 24.6% with the actual prediction being around 11% (which as I understand is roughly historically accurate for 1940s)

13 hours ago, Kane said:

You also neglected to mention, what size torpedo?  Custom battle or campaign, what levels of research, etc.  (All things I noted in the original example I gave.)

If I understand correctly, you had explosives and torpedo propulsion research both capped, so unless they used a smaller torp size than you (which would be negligent of them, as you specifically mentioned you used 21" torps), I see no way their dud chance could've been lower than yours, unless there's an actual bug in UAD 

I say, it's a fair enough test for a forum discussion, though obviously not extensive enough by any scientific standards

[SpardaSon21]

9 hours ago, Abuse_Claws said:

unless there's an actual bug in UAD 

I wouldn't discount that.

[Kane]

@Abuse_Claws
I actually am wondering if that is the case, which is why I've asked people to let me know if they're not getting the same obscene failure rate that I am.  I'd like to see this fixed, so will probably spend a few hours this weekend trying to find if this is the case and where the bug is.  (If I can't do it in that time frame, then f- it.  Not like I'm getting paid to find these bugs)

Though based on some of what I've seen, I'm not quite convinced that every torpedo does have an independent dud chance versus some kind of batch operation being run on a salvo when its fired.  Part of why I wish the devs would answer questions and be a little more open on how they code some of this stuff.  I've been a modder for long enough to learn that is best not to assume developers did something in the most straight-forward or obvious way possible.  (Identifying possibilities like this is also one of the reasons I rely on a large sample size.)

Indeed small ships are way, way too tanky.  I've bitched about this in some other threads, and have been for awhile.  Its not even just destroyers.  The number of 16" rounds I routinely see cruisers survive reaches levels of stupid.  If your cruiser is tanking 16" fire better than Bismarck would have in its place, your cruisers are too tough.

That said, I do still think the amount of structural damage caused by torpedoes (particularly large ones such as 21 and up) is too small given the explosive yield such a weapon will carry.  Though addressing the over-tankiness of small ships could address that problem, depending on how well that problem is fixed.

 

[Mike L]

Its so bad for me, I have just put 1 twin launcher on my late game destroyers and load up on 4 or 5 triple turrets of 4.5 to 5.5 inch guns.  They can take down light cruisers pretty easily.  In an earlier version of the game, my cruisers were mostly torpedoes and I was wiping out fleets beyond visual range with 24 inch oxygen torpedoes.  Some balance between the 2 would be good.

 

[Abuse_Claws]

12 hours ago, Kane said:

I'm not quite convinced that every torpedo does have an independent dud chance versus some kind of batch operation being run on a salvo when its fired.

As a developer myself (not a game developer though), I would assume the existence of

 

individual dud chance for a couple of reasons:

• What exactly is a salvo from the code standpoint? Launchers can (and often are, if you take into consideration underwater ones) be fired individually, which would make a 'salvo' kind of a redundant entity in the code. What purpose would it serve? Trajectory, collisions and detections have to be (and from what I see, actually are, like if a torp veers off from a salvo towards you, you can detect it and not the rest of the salvo that goes away from you) calculated individually.
• Given that there seems to be a "torpedo dud chance" value in the game (as we can see from +-5% Dud Chance tech modifiers), rolling the RNG dice for every torp (rng.next() / rng.MAX < dudChance) seems easier than trying to implement a salvo roll, that would have to produce not just a yes/no result, but a "how many and which torps from this salvo are duds?".

Although getting a straight answer from the devs would certainly be nice

12 hours ago, Kane said:

The number of 16" rounds I routinely see cruisers survive reaches levels of stupid.

Yeeeeessss. Felt that yesterday, when Austrian (terribly-)armored cruisers ate 1-2 dozen 14" hits each before going down. For one thing, where in reality ships often gradually succumbed to damage and sank in the end, in UAD the anti-flood systems are some kind of magic which turns 3% floatability into 32% in a matter of minutes, just yeeting the water to /dev/null apparently and patching the hull with band-aids.

Same with structural integrity: IRL ocean waves would often over time finish off a crippled ship, but in UAD a ship with 0.5% structural integrity left can stay afloat indefinitely and even keep firing at you (ah yes, main caliber recoil certainly doesn't damage the already-wrecked hull) until you manually kill it.

And just plain and simple small ships have too much resistance IMO. One option is making resistance and flooding chance tied to the actual armor thickness. I think a heavy cruiser with 5" main belt armor should take more damage and flooding from a main belt hit than if it had 10" main belt armor. So far that seems not to be the case.