Matrix Games Forums

From my perspective, thats a yes and no. The mission generator itself can't do it, but scenario designers with some lua work can make that happen. If you start 100 nm behind a fleet, you will most likely not catch up. Running at 20 knots submerged is a good way to eventually get killed. Satelite, air, and human intel are all critical to modern fleet stalking. Without that starting point, its like the Japanese at Midway.

For a sub IRL or in CMO to ambush a surface fleet, it most likely has some strategic intel or warning. Even with that, a human would have a hard time getting in position if they are not in the right staring point.

With the tools available, a scenario designer can give the player enough latitude and warning to get a sub in position to ambush a fleet. This especially true any time before the 90s, ignoring aircraft. If a fleet has long range ASW air assets, or maybe an attack sub, available, then its a whole different story.

"A well trained crew in an Astute should sink a Burke nine times out of ten, but in CMO the sub loses that edge due to its mission behavior."

As I pointed out in the tests in the posted scenario, If I move the Burke back 100 nm, the subs detect it an average of 90 nm on ten play throughs. Considering the subs are crossing the Burke's bow, they can potentially blow it out of the water with an AShM, or wait for a torpedo shot. They have the time and the ability to make the first shot. Although with a modern Burke, getting a missile through is not likely. If its almost anything but a modern Burke, its no contest for a modern sub...nuke or diesel.

You're correct- but it's also the case that many scenario designers don't seem to do it.

And yes, if I'm already behind the fleet I don't expect the sub to catch up undetected. However, even if I'm 20 miles in front but 45 degrees off on either side, the sub will often get get detected while maneuvering into position.

Other than scenario-specific Lua, CMO simply doesn't have a C2 system to automagically coordinate an interception before the contact(s) arrives. And that's okay, we all knew that when we bought it, but I do think that this leads to a systematic disadvantage to subs that isn't reflective of real-world performance. Because the sub doesn't sit and wait, being where it needs to be ahead of time, it's typically going to be the underdog against a modern vessel. Not to say that it never wins its fights- but surely you have noticed how the limited AI logic is more often than not a death sentence for whoever has the shorter ranged weapon in these types of encounters.

Micro-management and Lua seem to be the only two solutions, which is only a problem because CMO is marketed more as macro-management than micro. When I want to micro manage units I play Wargame Red Dragon instead

Its only micromanagement if the scenario designer designs it that way, intentionally or unintentionally. The tools are there are free of charge.

But just like any PC game, its only as good as how the tools are applied.

btw, just altered the test scenario for the sub to be 45 deg off the bow of the Burke at just over 20 nm from the Burke. The sub transited to within 7 nm of the Burke's predicted path and got off four torpedoes within 5 minutes without the Burke getting a solid contact. It was a little unfair because the Burke goes to flank as soon as it detects the torpedo and it never has a chance to localize any contacts. So even with a modern Burke, a sub vs ship fight is a one-sided affair if the sub has general advanced knowledge of the ships path. That is with the sub traveling at 12 knots in the layer. I changed the ROE to fire at any surface contact.

So the short of it is, it can be done. btw, I never once touched the sub's controls once the initial orders were set. As hands-off as you can get.

I don't doubt it- nor do I have any saves to suggest otherwise- but over the time that I have played there have been a surprising number of occasions where the sub should have won but didn't. Most losses were LAs/Virginias to Udaloys or Sovremennys. Hell I even lost a creeping Seawolf to a lone Kirov once. Still bitter about that. Each time I went back afterwards in god mode and still couldn't make sense of it. Doesn't necessarily mean the balance is off but I do empathize with the "underwhelming" feeling that OP and others have put forward. I'm not suggesting that anything needs to be changed, only that I can relate to the feeling- I have experienced instances where I thought to myself "dang for all the creeping these subs sure do get sunk a lot"

If I find one such instance I'll be sure to post it here.

Here is the original test scenario modded to have all sides hostile, weapons free for the sub, WRA on the torps set to 5nm so the Burke can't escape.

I have also set the sub to travel at 12 knots in the layer. It ends up stopping near the Burke's path a few minutes before the Burke is in range. I would never do this in a real scenario because the Burke more than likely would have a helo out looking around.

Usually, the Burke only gets a hint of sub after the second salvo. And that's with active pinging, advanced towed array, and 2010 signal processing technology. Think about being a Perry class with relatively weak and short-ranged sonar capabilities. I would imagine an Akula would eat it for lunch.

ORIGINAL: thewood1

And this is why people ask for saves. The Burke has active sonar pinging away. No wonder its picking those subs up. The Burke's active sonar has a range of 40nm. The subs are at 7-8nm. It also has the VDS deployed. Both of those together and you couldn't hide a sleeping baby in a storm. This is a modern warship with every advanced piece of ASW technology you can think of. Equipped with powerful sonar with 2010 technology and a very modern towed variable depth sonar receiver. Not sure what the expectation was.

My amateur's view is this is working exactly as a I would expect from a detection standpoint.

No it isn´t working exactly as you would expect. Without background nor education its not a good idea to make statements and claims because one "believes", "assumes" "thinks" how the real deal might look like and sells them almost as facts. In academics and professional areas you might make yourself looking ridiculous by doing that but it is a wide-spread culture in wargaming communities, especially in CMO, not really helping its sim aspect at all.

Even when the advertisement reads active sonar range of 40nm it doesn´t mean you will detect submarines in that range, not even close.

Some years ago we had a single French Rubis taking out a carrier and almost its entire ASW screen by utilizing the acousting environment to the benefit of the hunter over the prey. These ships were using active sonar, these ships were also carrying "2010 tech" VDS, DTAS, CATAS, airborne with ALFS\ADS, HSS, (EER) buoys you call it and still didn´t manage to stop it. Same did other European nations achieve, same did our subs achieve. It can really hurt to see how this is handled in CMO if you know the reality. And before somebody mumbles something about classified, there is enough uptodate quality OSINT to be digged up with a good research. I am pretty sure the CMO devs have even access to more than that, it just requires some effort and motivation to do it right.

There is a reason why the Navy is rating destroying subs as one of the the hardest task in naval warfare, in CMO it is one of the easiest tasks.
It requires alot of ASW teamwork to produce results and still might yield nill. In CMO it is exactly the contrary.

Hyperbolic statement: Aerial and surface warfare are good but the subsurface battlespace including submarines and ASW are on World of Warships level of realism in CMO.

Deuce

ORIGINAL: MarechalJoffre

It is just quick test that you can do in the editor. Put a Burke on one side and two Virginias on the other. Sea state is as I described, Burke is at 20kts and cavitating with active sonar and Virginias are at creep speed. It literally takes 2 minutes to set up. But fine, I'll upload the file.

Here it is. You might point out that subs are showing their broadsides, but even when submarines are facing towards the Burke detection ranges are only down to 7NMI, barely inside the practical torpedo range (keep in mind that the Burke is moving at 20kts still). Kinematic doesn't matter as torpedoes peter out beyond 12NMIs in this build. One other funny thing is how Burke detects the bottom Virginia before the Virginia can get a passive detection. That thing is moving at 20kts!

Okay, let's run the numbers on this example a bit. For clarity's sake, I renamed the eastern sub as Virginia-1 and the western one as Virginia-2.

Local conditions: Heavy rain, 6C degress temperature, Sea State 5. Open sea (local depth 5275m). There is a 50% strong layer between 85 & 146 meters. CZs are forming at 29-58-87-116nm.

The Burke is cruising at 20 knots and cavitating. Virginia-1 is at 5 knots and 122m depth (within the layer). Virginia-2 is at 5 knots and 450m (deep under the layer). Both are beam-on to the Burke.

10:04:18 PM - AN/SQS-53C(V)1 (hull sonar - active mode) on Burke attempts to detect Virginia-1, at 7.14nm. The Virginia has a fairly large active sonar beam-on signature (signature ratio is 0.8). Own-noise modifier is 0.5 (suffers greatly from own-ship noise). No surface ducting. The thermal layer reduces effective range by ~25% (the sub is within the layer). No effect from deep sound channel. No reverberations (deep sea, no ice). No bottom clutter (local depth too great). The sonar has LOS to the sub. Direct-path range in these conditions is 5.68nm. No bottom-bounce possible (too great local depth). No CZs involved. Direct-path limit is less than the actual distance, so the sonar FAILS to detect the sub.

10:04:18 PM - AN/SQS-53C(V)1 (hull sonar - active mode) on Burke attempts to detect Virginia-2, at 7.61nm. The Virginia has a fairly large active sonar beam-on signature (signature ratio is 0.8). Own-noise modifier is 0.5 (suffers greatly from own-ship noise). No surface ducting. The thermal layer reduces effective range by ~49% (the sub is well under the layer). No effect from deep sound channel. No reverberations (deep sea, no ice). No bottom clutter (local depth too great). The sonar has LOS to the sub. Direct-path range in these conditions is 5.11nm. No bottom-bounce possible (too great local depth). No CZs involved. Direct-path limit is less than the actual distance, so the sonar FAILS to detect the sub.

10:04:23 PM - AN/TB-34 (towed array - passive) on Virginia-1 attempts to detect Burke, at 7.11nm. No surface ducting. The thermal layer reduces effective range by ~49% (the sub is inside the layer, but the TA is hanging _under_ it, so it is degraded far more than a hull sonar would). No effect from deep sound channel. No reverberations (deep sea, no ice). No masking factors. No CZs involved. Direct-path range in these conditions is just 2.64nm. The TA has a nominal ~18nm range against the Burke (fairly strong noise source, own-noise negligible), but the severely degraded direct-path limit prevents it from being utilized; the sonar FAILS to detect the ship.

10:04:24 PM - AN/TB-34 (towed array - passive) on Virginia-2 attempts to detect Burke, at 7.57nm. No surface ducting. The thermal layer reduces effective range by ~49% (the sub is well under the layer). No effect from deep sound channel. No reverberations (deep sea, no ice). Direct-path range in these conditions is 5.11nm. The TA has a nominal ~18nm range against the Burke (fairly strong noise source, own-noise negligible), but the degraded direct-path limit prevents it from being utilized; the sonar FAILS to detect the ship.

10:04:26 PM - AN/BQG-5A LWWAA [BSY-2] (flank array - passive) on Virginia-1 attempts to detect Burke, at 7.09nm. The sonar has a nice 24.53nm nominal detection range against the ship (fairly strong noise source, own-noise negligible; if conditions were a bit better it would be able to get a 1st-CZ detection). No surface ducting. The thermal layer reduces effective range by ~25% (the sub is inside the layer). No effect from deep sound channel. No reverberations (deep sea, no ice). No masking factors. No CZs involved. Direct-path range in these conditions is 7.49nm. The sonar has LOS to the ship. Since the target is within both direct-path range and sensor nominal range, the sonar DETECTS the ship successfully.

10:04:26 PM - AN/BQQ-10 (hull sonar - passive) on Virginia-1 attempts to detect Burke, at 7.09nm. The sonar has a 23.3nm nominal detection range against the ship (fairly strong noise source, own-noise negligible; if conditions were a bit better it would be able to get a 1st-CZ detection). No surface ducting. The thermal layer reduces effective range by ~25% (the sub is inside the layer). No effect from deep sound channel. No reverberations (deep sea, no ice). No masking factors. No CZs involved. Direct-path range in these conditions is 7.49nm. The sonar has LOS to the ship. Since the target is within both direct-path range and sensor nominal range, the sonar DETECTS the ship successfully.

10:04:27 PM - TB-37/U MFTA [AN/SQR-20] (towed array - passive) on Burke attempts to detect Virginia-1, at 7.09nm. Nominal detection range against the sub is just 4.2nm (the array avoids the own-noise degradation because it is offboard and under the layer, but the target is very quiet and inside the layer). No surface ducting. The thermal layer reduces effective range by ~25% (the sub is within the layer). No effect from deep sound channel. No reverberations (deep sea, no ice). The sonar has LOS to the sub. No masking factors. No CZs involved. Direct-path range in these conditions is 7.49nm. Since the nominal detection range falls short of the actual target range (and is not boosted by effects such as surface ducting or the DSC), the sonar FAILS to detect the sub.

10:04:27 PM - TB-37/U MFTA [AN/SQR-20] (towed array - passive) on Burke attempts to detect Virginia-1, at 7.55nm. Nominal detection range against the sub is 11.2nm (the array avoids the own-noise degradation because it is offboard and under the layer, but the target is very quiet). No surface ducting. No modification from thermal layer (they are both on the same side of it). The sub is under the DSC, but the array itself is squarely in it, so the practical range is boosted by 50%. No reverberations (deep sea, no ice). The sonar has LOS to the sub. No masking factors. No CZs involved. Direct-path range in these conditions is 14.81nm. Since the target is within both direct-path range and sensor nominal range, the sonar DETECTS the sub successfully.

10:04:28 PM - AN/TB-29 (towed array - passive) on Virginia-1 attempts to detect Burke, at 7.08nm. The sonar has a 23.3nm nominal detection range against the ship (fairly strong noise source, own-noise negligible). No surface ducting. The thermal layer reduces effective range by ~49% (the sub is inside the layer, but the TA is hanging _under_ it, so it is degraded far more than a hull sonar would). No effect from deep sound channel. No reverberations (deep sea, no ice). No masking factors. No CZs involved. Direct-path range in these conditions is 5.11nm. The sonar has LOS to the ship. Since the target is outside direct-path range, the sonar FAILS to detect the ship.

These are just a few of the detection checks happening continuously as the sim runs.

A general observation: You are really not doing your subs any favors by placing them inside the layer or under it. You are repeatedly overlooking a crucial ace that the Burke holds: Its very powerful, very modern towed array. The TA doesn't care about the weather and sea state, and it (mostly) ignores the ship's own noise (it's still affected by its own flow noise though). You need to put a strong layer between it and you if you want to have a (small) chance of getting close enough for a torpedo shot.

A second observation. You complain that subs are detected too easily. Okay. You then substantiate your claim by pitting them against arguably one of the most advanced surface ASW platforms in service, in near-ideal conditions for it (deep water - no reverb or bottom clutter, and your subs on the same side of the layer as its TA). This is like pitting Mike Tyson against Ali in their respective primes and expecting any of the two to walk out of the ring without a scratch. They're both great, and because of this they will both get bloodied.

ORIGINAL: MarechalJoffre
I have been checking out discussions online about this topic. Pretty much every bit of evidence (non-classified that is) hints at subs being a bit too easy to detect in CMO.

The stories you read in Reddit and elsewhere, are usually from fleet exercises where the deck is stacked in favor of the sub guys in order to give the tin cans a good workout. In addition, most of these cases revolve around shallow-water setups where the sub has several factors going for it (strong reverberations, bottom clutter, multiple false contacts - all of these are also modelled in Command). In the example save you have provided, it is deep water which makes things considerably simpler for the surface ASW guys.

Plus, most of the exercises in the last 2-3 decades have centered around shallow/littoral ops, with ultras-quiet diesel subs (on load, usually from the Swedish) standing-in as OPFOR subs. If you go back to exercises in the 70s/80s you will read a lot of stories about Spruance destroyers (and even Knox & Perry frigates) regularly sneaking up on OPFOR nuke subs and surprising them with their brand-new (at the time) towed arrays. Current-line Burkes are very much the "sons of Spruances" in both ASW hardware and lessons learned. You should not be surprised that they are supremely effective ASW hunters (in addition to highly capable AAW escorts).

ORIGINAL: MH-60Deuce
Some years ago we had a single French Rubis taking out a carrier and almost its entire ASW screen by utilizing the acousting environment to the benefit of the hunter over the prey. These ships were using active sonar, these ships were also carrying "2010 tech" VDS, DTAS, CATAS, airborne with ALFS\ADS, HSS, (EER) buoys you call it and still didn´t manage to stop it. Same did other European nations achieve, same did our subs achieve. It can really hurt to see how this is handled in CMO if you know the reality. And before somebody mumbles something about classified, there is enough uptodate quality OSINT to be digged up with a good research. I am pretty sure the CMO devs have even access to more than that, it just requires some effort and motivitation to do it right.

Since you (presumably) claim to "know the reality", I am making this offer in full earnest: Contact me in private with your (verifiable) details and credentials. If you are for-real, we'll put you under strict NDA & contract and you will be shown the sonar source code, line-by-line. You will show us which abstractions & assumptions are incorrect or off-the-mark, and how to adjust them. If the model is indeed improved, you will be compensated for your time.

Deal?

ORIGINAL: MH-60Deuce
Hyperbolic statement: Aerial and surface warfare are good but the subsurface battlespace including submarines and ASW are on World of Warships level of realism in CMO.

Hyperbole indeed. Command's sonar model currently takes these factors into account:

* Line-of-sight, with separate check for blockage from landmass (for example, SOSUS arrays can hear beyond the sonar "horizon" but are still blocked by e.g. an island or peninsula).
* Sensor parent platform own noise, broken down to powerplant/propeller and flow noise. (A boat may be on high speed but low throttle setting, e.g. drifting after a sprint, or conversely low on actual speed but high on throttle, e.g. accelerating from a slow speed.)
* Target noise, again broken down to powerplant/propeller and flow noise.
* Target noise signatures are compared against "reference maximums", e.g. for submarines the reference maximum is a November-class submarine at flank throttle. For each unit, we use four discrete acoustic bands: VLF (100Hz reference frequency), LF (3kHz), MF (7.5kHz) and HF (15kHz).
(Note: At frequencies greater than 200 Hz the Source Level falls by 6 dB when the frequency is doubled. This is reflected by the units' passive sonar signature. For example a sub with a sonar signature of 140dB at 1-200Hz, has a signature of 110dB at 3kHz, 102dB at 7.5kHz, and 94dB at 15kHz)
* Target aspect (critical for active sonar but also makes a difference for passive)
* Target and sensor-parent cavitation (boosts noise)
* Target submarine snorkelling (again increases noise)
* Sensor arc restrictions (e.g. hull sonar baffles)
* Onboard/offboard nature of sensor (offboard sensors get significant own-noise reductions)
* Sound propagation modifiers: Surface duct, thermal layer and deep sound channel (as described in manual)
* Reverberation modifiers: shallow water and under-ice conditions
* Direct-path range (nominally 9.5nm but actually varied by propagation modifiers) and convergence zones (as described in manual)
* Noise masking

I'd be curious to see how WoWs models these factors.

ORIGINAL: MH-60Deuce
Even when the advertisement reads active sonar range of 40nm it doesn´t mean you will detect submarines in that range, not even close.

Who ever claimed that the sim uses the raw DB values?

if I had a nickel for every time I've publicly reminded folks that "in-sim values != DB nominal values", I could already have retired.

"Progress in traditional sonar detection technologies continues to be slow and evolutionary (O’Hanlon, 2018). ... they face problems in the classification of submarine signals[/b] (Perkins, 2016). All this leads one expert to argue that traditional sensors “are rapidly approaching a point of obsolescence” (Perkins, 2016)."
NATO Parliamentary Assembly STC NATO ASW 2019.

Can´t be that that to do a proper (OSINT) research in order to get things right in CMO. Investing some digging into ASWEX like DYNAMIC MANTA or MONGOOSE and quickly find out whats wrong with the simulation. Otherwise ASW and sub ops in CMO will come nowhere close to reality.

ORIGINAL: Dimitris
Who ever claimed that the sim uses the raw DB values?

I didn´t and it is obvious but the other user made it appear as "everything would be correct by putting raw value against sub distance. So this needs to be clarified expescially for users that are not educated on this.

Despite great job avoiding the - way more important - rest of the post but I get it hard to argument against something that comes from 1st hand.

I think there is, in general, a confusion of two different aspects in those who claim that CMO is "too unforgiving" to subs. These two aspects to consider are: The underlying physics models (ie. mostly the sonar model), and the AI handling of submarines.

The physics appear to be solid. We have gone over the various models with numerous SMEs and the consensus has been that in the main "we get it right". There are always things to improve as SeaQueen and others have pointed out, and there is always the chance we are overlooking something crucial (which is why we don't outright dismiss any issue report like the OP's, but we do take the time to investigate), but there has been no "slam the brakes" emergency. As thewood1's save demonstrates, it is indeed perfectly possible to set up a good approach against even a very modern ASW hunter under the right conditions. To echo JMS, "context is everything".

Sub AI is another matter. While it has been continuously tweaked and adjusted over the years, it still falls short of a "human like" behavior. This should surprise only those who are unaware of the general limits of AI in complex games (we don't really need to bring up examples from the mainstream/AAA market, do we?).

I have a suspicion, that some of the people currently holding to the view that subs are "too easy to find and kill" in Command, are in for a rude shock when MP comes around and they find themselves eating torps for breakfast, lunch and supper, from folks here who know how to drive them properly. (I would suggest refraining from challenging e.g. SeaQueen to a duel unless actively craving humiliation)

But of course maybe I'm wrong.

ORIGINAL: MH-60Deuce

"Progress in traditional sonar detection technologies continues to be slow and evolutionary (O’Hanlon, 2018). ... they face problems in the classification of submarine signals[/b] (Perkins, 2016). All this leads one expert to argue that traditional sensors “are rapidly approaching a point of obsolescence” (Perkins, 2016)."
NATO Parliamentary Assembly STC NATO ASW 2019.

Can´t be that that to do a proper (OSINT) research in order to get things right in CMO. Investing some digging into ASWEX like DYNAMIC MANTA or MONGOOSE and quickly find out whats wrong with the simulaton. Otherwise ASW and sub ops in CMO will come nowhere close to reality.

ORIGINAL: Dimitris
Who ever claimed that the sim uses the raw DB values?

I didn´t and it is obvious but the other user made it appear as "everything would be correct by putting raw value against sub distance. So this needs to be clarified expescially for users that are not educated on this.

Despite great job avoiding the - way more important - rest of the post but I get it hard to argument against something that comes from 1st hand.

The offer I made stands, if you ever feel like making a real contribution.

ORIGINAL: Dimitris

ORIGINAL: MH-60Deuce

"Progress in traditional sonar detection technologies continues to be slow and evolutionary (O’Hanlon, 2018). ... they face problems in the classification of submarine signals[/b] (Perkins, 2016). All this leads one expert to argue that traditional sensors “are rapidly approaching a point of obsolescence” (Perkins, 2016)."
NATO Parliamentary Assembly STC NATO ASW 2019.

Can´t be that that to do a proper (OSINT) research in order to get things right in CMO. Investing some digging into ASWEX like DYNAMIC MANTA or MONGOOSE and quickly find out whats wrong with the simulaton. Otherwise ASW and sub ops in CMO will come nowhere close to reality.

ORIGINAL: Dimitris
Who ever claimed that the sim uses the raw DB values?

I didn´t and it is obvious but the other user made it appear as "everything would be correct by putting raw value against sub distance. So this needs to be clarified expescially for users that are not educated on this.

Despite great job avoiding the - way more important - rest of the post but I get it hard to argument against something that comes from 1st hand.

The offer I made stands, if you ever feel like making a real contribution.

I appreciate the offer, but you need to understand that I might not be allowed to or expect to get into trouble for that. To be honest I don´t know what is off-limits what not.

Isn´t giving directions and highlighting OSINT a "real contribution"? I don´t ask you to buy what I am telling because an internet forum discussion is no way a quality source, only thing I am asking is check the research because from what I´ve seen everything significant is attainable UC-OS.

Fair. Thanks for the pointers, we'll check to see if we are overlooking a significant factor.

ORIGINAL: Dimitris

Fair. Thanks for the pointers, we'll check to see if we are overlooking a significant factor.

And thanks for taking a look into it.

I read the whole thread and, while being far less knowledgeable than some (or most) of the commenters, I have some thoughts.

Firstly, there's a lot that can be construed as the obsolescence of torpedoes (versus modern surface warships) more than submarines, given their short range and the increasing range of ASW equipment. The thing is, a modern surface vessel may well also have good anti-missile defenses, which leaves submarines completely relegated to a surveillance role at best. There's also the fact that unless you're in an absolutely perfect position in everyone's baffles, launching a torpedo in the presence of helo-equipped warships is certain death.

Then there's the matter of ocean space, which is something someone mentioned earlier. Submarines generally have a whole swath of ocean to hide in, and the most effective ASW tools are focused and useful only in bursts. A warship won't have its active sonar screaming at all times, no matter how effective, because it gives away its position beyond its range and even if a perceptive submarine couldn't touch it, it could relay the information to forces which could. And ASW aircraft can't be operating around the clock at peak effectiveness: they'd run out of supplies before they've combed a sizeable portion of the real "here be dragons" zone.

So with that in consideration, enter CMO/CMANO with its focused scenarios. That's most of the difficulty stripped away from surface-borne ASW efforts: submarine threats will certainly be encountered within hours, maybe even less. What can a sub do in that context other than stay out of the way, act only against thorourghly defenseless targets and just relay intel about anything with teeth it can see?

As the test I posted shows, the sub detects the pinging ship many times further away than the ship detects the sub. If the scenario is set with reasonable positioning, proper ROEs, and a good understanding of how subs can hide, subs are a hard to manage for any fleet, but they are a nightmare for single ships. Even a modern Burke is easy bait for a short range torpedo if its actively pinging.