Unreal Torpedo Spotting

[theCarthaginian]

I'm a bit confused as to how an OPFOR destroyer - which is making smoke - can notice that I am - while also making smoke - am launching torpedoes several kilometers away and immediately react by going hard over to let a spread of five torpedoes whizz by totally ineffectually. It seems that the AI is as acutely aware of our actions during combat as it is in how our ships are designed in order to counter our tactics.

Exactly what penalties in spotting is the AI suffering due to two smokes and at what range are the torpedoes of each type visible both in and out of smoke?

[Drenzul]

There are sonar/hydrophones as well. Torpedeo 'spotting' via mk1 eyeball was generally a last resort 'oh its too late' kinda thing

[theCarthaginian]

13 hours ago, Drenzul said:

There are sonar/hydrophones as well. Torpedeo 'spotting' via mk1 eyeball was generally a last resort 'oh its too late' kinda thing

Sonar/hydrophones in WWII didn't work at the kinds of speeds that destroyers in a fleet action would be moving... flow noise will blind modern sonars at moderate speeds on submarines - which exist in a much more stable environment that a surface ship's hull-mounted sonar. Remember, not only is it being plowed through the ocean at speeds in excess of 30 knots; it's also being slapped and slammed by wave impacts, picking up the flexing of the hull by conduction, ect. Sonars in WWII were not very advanced at all in the grand scheme of things. I mean, think about a voice recording of the era. Those were made under as near to perfect conditions as could be contrived under the circumstances of the time's technology. Now, translate that into State 4 or State 5 seas. It's amazing that sips could hear anything AT ALL. Indeed, sometimes, they couldn't - and that just from the environment, while they were doing everything they could to control ownship noise generation!

Edited September 26, 2021 by theCarthaginian

[Commander Reed]

I would like to see other uses for the Hydrophones/Sonars other than torpedoes.

[madham82]

Read the Scharnhorst's history. While going flank in bad seas is a fair call, passive sonar was used several times in WW2 to detect torpedoes at cruise speeds. 

Back to your issue, at what ranges was this taking place? Smoke in game is not handled very realistically, but that wouldn't hide a torpedo launch if you were close enough IRL anyway. 

[arkhangelsk]

Quote

The Americans were very interested in this system. In trials in the spring of 1946 run by the Philadelphia Navy Yard, underwater noise was detected at a distance of 27km despite the hammering of the Prinz Eugen’s steam turbines as she made 20 knots! It was found that all torpedoes fired from beyond 2km could be avoided in time (further trials at maximum speed were not possible because of the poor condition of the cruiser’s machinery). A link-up involving both amplifiers and their operators could provide a listening field 40km wide, although this was not carried out in practice. At long range, noise interference increased substantially and a very highly experienced operator was needed for reliable interpretation.

Koop, Gerhard. Heavy Cruisers of the Admiral Hipper Class: Warships of the Kriegsmarine (p. 47). Pen & Sword Books. Kindle Edition. 

During the battle with HMS Glorious, Scharnhorst started at 1700 at 19knots. 7 minutes later, she's at 24 knots. It then took her until 1726 to reach 26 knots (but only two more minutes to work to 29), finally reaching 30 knots at 1738.

At 1745, she heard and avoided a torpedo. She continued to hear more torpedoes throughout the battle.

[RedParadize]

10 hours ago, arkhangelsk said:

Koop, Gerhard. Heavy Cruisers of the Admiral Hipper Class: Warships of the Kriegsmarine (p. 47). Pen & Sword Books. Kindle Edition. 

During the battle with HMS Glorious, Scharnhorst started at 1700 at 19knots. 7 minutes later, she's at 24 knots. It then took her until 1726 to reach 26 knots (but only two more minutes to work to 29), finally reaching 30 knots at 1738.

At 1745, she heard and avoided a torpedo. She continued to hear more torpedoes throughout the battle.

By the times of ww2 torpedo detection was indeed good enough to detect incoming torpedo. One thing that isnt taken into consideration is that there was basically only one sonar, with one guy pointing it in a specific direction. With the sonar you get the incoming direction of the noise pretty accurately, but for the bearing and distance you can only rely on guesswork. The talent of the sonar operator could vary greatly and in a intense battle, where there is multiple boat and multiple incoming torpedo, things get much more complicated. Reality is that sonar equipped boat were indeed getting torpedoed in ww2.

As for the game its more of a balance issue I would say. Atm past the 20s DDs are more or less pointless, specially if the boat is AI controlled.

[RedParadize]

Oh, I should say one more thing to clarify my trough. While I do not think I will build much TB/DD and CL in campaign, if I build any really, I will certainly build Torpedo equipped CA. These are fantastic weapons.

[HistoricalAccuracyMan]

On 9/26/2021 at 4:04 PM, theCarthaginian said:

Sonar/hydrophones in WWII didn't work at the kinds of speeds that destroyers in a fleet action would be moving... flow noise will blind modern sonars at moderate speeds on submarines - which exist in a much more stable environment that a surface ship's hull-mounted sonar. Remember, not only is it being plowed through the ocean at speeds in excess of 30 knots; it's also being slapped and slammed by wave impacts, picking up the flexing of the hull by conduction, ect. Sonars in WWII were not very advanced at all in the grand scheme of things. I mean, think about a voice recording of the era. Those were made under as near to perfect conditions as could be contrived under the circumstances of the time's technology. Now, translate that into State 4 or State 5 seas. It's amazing that sips could hear anything AT ALL. Indeed, sometimes, they couldn't - and that just from the environment, while they were doing everything they could to control ownship noise generation!

While that may/may not be true (idk because I haven't read up on that subject at all nor do I have first hand knowledge of that matter), it's like I've said countless times before in my other posts...this game isn't supposed to be a "realism simulator" or "incredibly and historically accurate." It's more of a sandbox mixed with grand strategy (from what we can assume or guess about the campaign so far) that is historically based. Now while there are some things that are/were completely absurd--like some of the gun ranges and shell weights--the devs have fixed what they can so far.

I brought up a similar issue with rangefinders a while back and how coincidence/stereoscopic rangefinders were used and what their strengths and weaknesses were...and the best answer I got was (I'm summarizing here): the choice between coincidence and stereoscopic rangefinders in this game is just a design choice you make to decide whether your guns aim faster or if they are more "accurate" at long range. My guess is that it's a similar situation with the hydro/sonar modules...it's just an option that you can put on your ship if you want, and as the years go on you unlock the better versions or new technologies.

Ultimately...very few things, if anything at all, is historically accurate or works the way it would in the real world as far as game mechanics go. Just design what you want, the way you want it

[DougToss]

10 hours ago, HistoricalAccuracyMan said:

Ultimately...very few things, if anything at all, is historically accurate or works the way it would in the real world as far as game mechanics go. Just design what you want, the way you want it

Let’s hope not. There has been a lot of really fantastic feedback here from people like @akd, @Steeltrap, @Draco, @ColonelHenry and others pushing very hard to ensure accuracy, and @Nick Thomadis has been pretty receptive to that feedback, especially in the past year or so.

[HistoricalAccuracyMan]

@DougToss you are right and I completely agree with you: there has been a lot of good feedback and Nick Thomadis has seemed very receptive as of late. I too, would like to see a higher level of accuracy in the game...owing mostly to my history buff nature. What I was trying to get at is: in the game's current state, there isn't an overwhelming amount of "historical accuracy" outside of the hulls, superstructures and gun models. While I would love to see more and more historical accuracy, or as close as possible to reality without causing a balancing nightmare for the devs, I'm not going to be overly concerned with "getting it right according to what we know" or get extremely nit-picky about exact values or details until the game starts to get more to the point of being increasingly historically accurate.

[DougToss]

For sure. I think the difference between rivet counting accuracy and more simplified authenticity   is presenting the same dilemmas, choices and solutions present in history, through gameplay. It’s about feeling right, if you know what I mean.
 

The example of this I am most proud of our feedback meaningfully guiding game design is capital ship secondary armaments. A lot of people seemed to believe these guns could, did and therefore should, hit and sink torpedo boats, effectively making capital ships wholly able to defend themselves from torpedo attack. 
 

Now, right off the bat, I could tell you a million reasons why that’s not true - and trust me we did! Accuracy and penetration tables, gunnery tests, accounts of surface actions, all pointing out the minuscule chances of a 3,4 or even 6 inch gun wholly disabling a torpedo boat or small destroyer before entering torpedo range. It seemed just about impossible to convince people that everything they thought they knew (cough WOWS cough) was wrong. 

But I think it was by reframing it in terms of authenticity that was convincing:

- There needed to be a reason for players to choose destroyer and cruiser screens

and

- There needed to be a reason to include the cost, weight and crew of a secondary battery on capital ships

 

It was making the case that making attacks riskier (a 1-3% chance of being hit is still going to give you pause before putting your life on the line), and more difficult under fire, rather that accurately hitting, let alone sinking torpedos boats, was the main reason for secondary batteries. A deterrence value, in other words.

 

 Effective protection - if deterrence was not enough, and at night, poor weather, low visibility or if the ship was otherwise engaged, or against a determined or desperate attacker it wouldn’t be - required a screen.  

Ultimately, I’m with you in that in a perfect world @Nick Thomadis would just call up the guys who made Jutland Pro and give it a 2021 facelift with ship builder, but that’s not the only way. 
 

All of that to say, I would gently suggest giving @Nick Thomadis a suggestion for gameplay reasons why the difference between coincidence/stereoscopic rangefinders matters. Okay, the historically accurate choice was not between aiming faster and being more accurate, I understand that, I’m sure Nick does too, but how can the player authentically experience the distinction? 
 

What’s the simplest way you could explain it to someone who knew nothing about range control, to say nothing of optics? 
 

There was a passage in Rules of the Game, I believe Chapter 7: The Battlecruiser Duel, that goes into it. Maybe you could pull from that?

 

I recommend reading the book and taking notes, in examining decision making at Jutland it seems ready made to provide ideas for gameplay.

[RedParadize]

4 hours ago, DougToss said:

The example of this I am most proud of our feedback meaningfully guiding game design is capital ship secondary armaments. A lot of people seemed to believe these guns could, did and therefore should, hit and sink torpedo boats, effectively making capital ships wholly able to defend themselves from torpedo attack. 

 

Now, right off the bat, I could tell you a million reasons why that’s not true - and trust me we did! Accuracy and penetration tables, gunnery tests, accounts of surface actions, all pointing out the minuscule chances of a 3,4 or even 6 inch gun wholly disabling a torpedo boat or small destroyer before entering torpedo range. It seemed just about impossible to convince people that everything they thought they knew (cough WOWS cough) was wrong. 

But I think it was by reframing it in terms of authenticity that was convincing:

- There needed to be a reason for players to choose destroyer and cruiser screens

and

- There needed to be a reason to include the cost, weight and crew of a secondary battery on capital ships

I remember that discussion! I was on the side of the "justify the existence of secbat" side. They may have gone a tad too far trough, screening is effective but not that essential, specially late game. 2" and 3" are still useless, 4" can help... But if you really want a effective secbat 6" to 8" are the way to go.

I would say that compounded by the nerf on speed accuracy malus, secbat buff made the TD and DD a much lesser treat. Sonar may also be too good at range... it make evading torpedo really easy. It would be better if we would only get the warning sign until they get in visual range.

Edited October 1, 2021 by RedParadize

[madham82]

7 hours ago, RedParadize said:

I would say that compounded by the nerf on speed accuracy malus, secbat buff made the TD and DD a much lesser treat. 

If by nerf you mean the lessoning of impact of the pure fantasy concept that speed somehow makes a ship harder to hit, then yes that "fixed" some of the unrealistic tactics/designs being used when it came to TDs and DDs (as well as the AIs infamous BCs). 

It still has been removed completed, which it should, but we don't see the huge issues with speed like before. 

[theCarthaginian]

On 9/28/2021 at 4:32 AM, arkhangelsk said:

Koop, Gerhard. Heavy Cruisers of the Admiral Hipper Class: Warships of the Kriegsmarine (p. 47). Pen & Sword Books. Kindle Edition. 

During the battle with HMS Glorious, Scharnhorst started at 1700 at 19knots. 7 minutes later, she's at 24 knots. It then took her until 1726 to reach 26 knots (but only two more minutes to work to 29), finally reaching 30 knots at 1738.

At 1745, she heard and avoided a torpedo. She continued to hear more torpedoes throughout the battle.

I'm yet to hear any example of 'perfect sonar' like Prinz Eugen managed in this account.
In fact, every other accessible account of sonar (including the one sonarman I know) use seems to contradict it... indeed, if sonar was so incredibly and amazingly clear as this alleged account states, then why does modern sonar require computer assistance for noise-canceling or target tracking or a host of other  duties? Until shown other documented episodes of similar instances with similar results (which I haven't found), then such an aberration can only be put down to the fact that someone is stretching the truth.

If you can corroborate the capabilities of ANY OTHER SITUATION where this feat is repeated in the face of the thousands of times that contact loss due to flow noise...
well, I await your convenience.

EDIT: also, IIRC, the same basic hydrophones fitted to P.E. were fitted to U-boats... which suffered from flow noise interference at speed.

Edited October 1, 2021 by theCarthaginian

[madham82]

3 minutes ago, theCarthaginian said:

indeed, if sonar was so incredibly and amazingly clear as this alleged account states, then why does modern sonar require computer assistance for noise-canceling or target tracking or a host of other  duties?

1. Submarines are a hell of a lot quieter than a torpedo, which is the primary threat sonar is used against today.

2.The depths they can operate are much different than a torpedo on the surface.

3. Related to the depths they have to take into account thermal layers and there impact on sound.

4. Today's sonar can hear ships at great distances, even in the 1940's subs would submerge to listen for ships because they could hear them farther away than they could see them in many cases. Ships are actually quieter than torpedoes at the distances we are discussing. 

5. Most importantly, you don't need precise information to dodge an incoming torpedo. It doesn't take much analysis to determine the bearing and whether you are closing/opening the range. That's all that is needed to take evasive action. The key with torpedo avoidance is detecting them as early as possible. 

Now here's what is too "gamey" about UA:D. We should only get a bearing and direction from sonar detection. The game should not render an actual wake until visually detected. This would make evasion more realistic and make for more calculated decisions on when to take evasion action. 

[theCarthaginian]

10 minutes ago, madham82 said:

1. Submarines are a hell of a lot quieter than a torpedo, which is the primary threat sonar is used against today.

2.The depths they can operate are much different than a torpedo on the surface.

3. Related to the depths they have to take into account thermal layers and there impact on sound.

4. Today's sonar can hear ships at great distances, even in the 1940's subs would submerge to listen for ships because they could hear them farther away than they could see them in many cases. Ships are actually quieter than torpedoes at the distances we are discussing. 

5. Most importantly, you don't need precise information to dodge an incoming torpedo. It doesn't take much analysis to determine the bearing and whether you are closing/opening the range. That's all that is needed to take evasive action. The key with torpedo avoidance is detecting them as early as possible. 

Now here's what is too "gamey" about UA:D. We should only get a bearing and direction from sonar detection. The game should not render an actual wake until visually detected. This would make evasion more realistic and make for more calculated decisions on when to take evasion action. 

1.) Submarines lost contact with surface ships up to and including fast-moving escorts... negating the 'subs are quiet' argument.
2.) Indeed, it's quieter and more stable for a sub... and, barring something like a sound channel or a thermocline, more conducive to them getting BETTER overall performance in the general area under any given variable.
3.) But for a surface ship, this is moot... and both the extra performance and the interference that go with it.
4.) EXACTLY - part of the reason was the degradation of sonar performance on the surface versus that of sonar submerged. And the whole 'hear them before you can see them' was a very common thing at night and in bad weather - when sonar performance on the surface was trashed by the very conditions I mention. 😉
5.) Yeah... but then you run into the actual distance at which you'd be able to make that out over ownship noise - and whether or not that warning might be more than you'd get visually under any given circumstance, and that would vary by torpedo type. It would be a really big reason to love that crew experience variable, that's for sure.

[madham82]

4 minutes ago, theCarthaginian said:

1.) Submarines lost contact with surface ships up to and including fast-moving escorts... negating the 'subs are quiet' argument.
2.) Indeed, it's quieter and more stable for a sub... and, barring something like a sound channel or a thermocline, more conducive to them getting BETTER overall performance in the general area under any given variable.
3.) But for a surface ship, this is moot... and both the extra performance and the interference that go with it.
4.) EXACTLY - part of the reason was the degradation of sonar performance on the surface versus that of sonar submerged. And the whole 'hear them before you can see them' was a very common thing at night and in bad weather - when sonar performance on the surface was trashed by the very conditions I mention. 😉
5.) Yeah... but then you run into the actual distance at which you'd be able to make that out over ownship noise - and whether or not that warning might be more than you'd get visually under any given circumstance, and that would vary by torpedo type. It would be a really big reason to love that crew experience variable, that's for sure.

Rather that arguing over semantics, are you saying you do not believe the logs from Scharnhorst or the US Navy's test of Prinz Eugen’s (which is the same passive sonar system) that they were able to detect torpedoes? I mean the US Navy liked the unit so much they mounted it one of their submarines. 

Or are you just arguing the game handles detection unrealistically? 

The concept clearly has merit, it's implementation in UA:D may not so much IMO. 

For reference, this is from the wiki on the GHG sonar used by the German ships. 

"The group listening device (‘’Gruppenhorchgerät’’), abbreviated "GHG", consisted of two groups of 24 sensors (one group on each side of the ship). Each sensor had a tube preamplifier. These 48 low frequency signals were then routed to a switching matrix in the main unit. The sonar operator could determine the ship's side and the exact direction of the sound source. To improve the resolution, there were three switchable crossover with 1, 3 and 6 kHz center frequency. A disadvantage of the side mounting, was a dead zone of 40 ° to fore and aft. Range: 20 km to individual drivers, 100 km against Convoy" 

[RedParadize]

5 hours ago, madham82 said:

If by nerf you mean the lessoning of impact of the pure fantasy concept that speed somehow makes a ship harder to hit, then yes that "fixed" some of the unrealistic tactics/designs being used when it came to TDs and DDs (as well as the AIs infamous BCs). 

It still has been removed completed, which it should, but we don't see the huge issues with speed like before. 

It isn't pure fantasy. Fast were harder to hit, but certainly not as hard as it was before the nerf. Anyway, this is beside the point I was making. The speed nerf AND the secbat buff, once combined, make the DDs a much lesser treat. So much so that there is little point in building them. Before, I use to make super fast BC with no secbat. Now I make slow, high resistance, BB with a pair of high caliber secbat... I wont build DD and CL as I do not need them. On the other hand, I will make close range torpedo CA.

As @DougToss pointed out, its more about authenticity.

[theCarthaginian]

3 hours ago, madham82 said:

Rather that arguing over semantics, are you saying you do not believe the logs from Scharnhorst or the US Navy's test of Prinz Eugen’s (which is the same passive sonar system) that they were able to detect torpedoes? I mean the US Navy liked the unit so much they mounted it one of their submarines. 

Or are you just arguing the game handles detection unrealistically? 

The concept clearly has merit, it's implementation in UA:D may not so much IMO. 

For reference, this is from the wiki on the GHG sonar used by the German ships. 

"The group listening device (‘’Gruppenhorchgerät’’), abbreviated "GHG", consisted of two groups of 24 sensors (one group on each side of the ship). Each sensor had a tube preamplifier. These 48 low frequency signals were then routed to a switching matrix in the main unit. The sonar operator could determine the ship's side and the exact direction of the sound source. To improve the resolution, there were three switchable crossover with 1, 3 and 6 kHz center frequency. A disadvantage of the side mounting, was a dead zone of 40 ° to fore and aft. Range: 20 km to individual drivers, 100 km against Convoy" 

Yes, I doubt everything - skepticism is the foundation of Science.
Your pointing to the wiki article tells me nothing.
It regards a submarine, which you yourself admit has different characteristics than a surface ship. It doesn't talk about the number of hydrophones in the array, whether it was fixed or mechanically steered, what conditions were during the test, whether the sub was surfaced or submerged, the speed bands across which the testing was conducted, the depth of the submarine if it was submerged, the presence of a thermocline, salinity of the water, time of the year or any of 10,000 germane variables that could have yielded a skewed result. It also only cites a single data point, whereas there are multiple hundreds of submarines exhibiting lower performance from the same set... indicating a possible statistical anomaly in this case.

The US Navy did indeed test the set on Flying Fish... much to the detriment of her appearance, and used the system in the development on the sonar system that they installed on the Tang class - and it bears noting that the installation was not repeated in their nuclear-powered cousins the Skates, nor in any following vessel class (only Tullibee, which was more of an open-ocean operational test platform than a fleet submarine). One of the things this hints at is that the performance wasn't worth the trouble after all, and that more conventional arrays were effective enough when it mattered.

So, yeah... I have doubts.
Many, many doubts.
And that wiki just makes for a lot more questions.
If I get a chance I might look for the actual British tests - though the odds of finding them are slim. Until then, I have the data that is gathered from reading accounts of dozens of commanders of vessels both surfaced and submerged that talk about sonars losing contacts at remarkably short ranges (some using that EXACT hydrophone) and acquiring them at excessively unbelievable ones under differing conditions. This tells me factually that operator skill, external conditions and pure dumb luck was at least as important as equipment in acquisition and retention of a sonar contact during the 1940's and well into the Cold War... possibly even more important than the equipment.

Edited October 1, 2021 by theCarthaginian

[theCarthaginian]

On 9/29/2021 at 8:45 PM, HistoricalAccuracyMan said:

I brought up a similar issue with rangefinders a while back and how coincidence/stereoscopic rangefinders were used and what their strengths and weaknesses were...and the best answer I got was (I'm summarizing here): the choice between coincidence and stereoscopic rangefinders in this game is just a design choice you make to decide whether your guns aim faster or if they are more "accurate" at long range.

Yeah... they used two items grossly similar in actual performance to separate focuses on 'Field of View' (obtaining sight picture) versus 'Precision Optics' (point-target accuracy).
I'd like them to change those names, personally, as coincidence and stereoscopic rangefinders are pretty much 'six of one, half dozen of the other' in end result.

[madham82]

1 hour ago, RedParadize said:

It isn't pure fantasy. Fast were harder to hit, but certainly not as hard as it was before the nerf. Anyway, this is beside the point I was making. The speed nerf AND the secbat buff, once combined, make the DDs a much lesser treat. So much so that there is little point in building them. Before, I use to make super fast BC with no secbat. Now I make slow, high resistance, BB with a pair of high caliber secbat... I wont build DD and CL as I do not need them. On the other hand, I will make close range torpedo CA.

As @DougToss pointed out, its more about authenticity.

My bad, my sarcasm was lost. Wasn't arguing about it existing in game, was just being sarcastic that it was in the game to start with. In general hit rates are much higher than IRL already, so no surprise there are issues with balance. 

[RedParadize]

There is no meaningful difference between coincidence and stereoscopic range finding. Both system are only marginally different on a technical level only. They both rely on the difference between two point of view to extract trigonometrical estimation of range. Both system have a similar number of lens and mirror, thus are subject to the same amount of inaccuracy.

Not to claim any form of authority on the subject but my field of expertise is the stereoscopic treatment of image (in the movie industry), My job is basically to fix the deformation caused by lens, then extract accurate positioning data out of the image, then create a landscape out of it. I often use what could be described as a "stroboscopic range finding". Basically I alternate right and left image and offset one of the two until it converge to the desired point. That could have been a 3rd kind of range finding in ww1&2, with its own advantage and inconvenient... but not physically more accurate than the two other.

Having said all that, I am ok with designer choice.

@madham82 All good, no offense taken. Hopefully my answer did not feel to harsh to you.

[madham82]

1 hour ago, theCarthaginian said:

Yes, I doubt everything - skepticism is the foundation of Science.

Was pretty sure observation is the foundation of science.

The Germans on two separate ships (probably more if we dig through more records) and the US Navy both validated it's ability to detect torpedoes. Scharnhorst was not in calm perfect seas, it was in the North Sea during winter. Not at all ideal weather/seas. I believe I also heard Yamato had passive sonar for this purpose as well. 

Originally the game did not feature sonar on BBs/BCs until people posted evidence.  

But you are free to disagree. I'm not going to argue the effectiveness/testing done on the units. 

[theCarthaginian]

25 minutes ago, madham82 said:

Was pretty sure observation is the foundation of science.

The Germans on two separate ships (probably more if we dig through more records) and the US Navy both validated it's ability to detect torpedoes. Scharnhorst was not in calm perfect seas, it was in the North Sea during winter. Not at all ideal weather/seas. I believe I also heard Yamato had passive sonar for this purpose as well. 

Originally the game did not feature sonar on BBs/BCs until people posted evidence.  

But you are free to disagree. I'm not going to argue the effectiveness/testing done on the units. 

You are focusing on a single documented instance to fit your desired outcome.
While I acknowledge this exists I am able to point to a plethora of other documented instances by ships of multiple nations that had a much broader range of results. Convoy escorts saw torpedoes strike without ever hearing them (or the launching sub) on their sonars. Sub skippers saw ships before they heard them, and vice versa.
Observation that depends on a single instance and the mediocrity principle isn't the best policy.