Critique my sub warfare (Allied vs AI)

[Korvar]

This thread is the result of a PM conversation I had with a veteran player; it was suggested that I post my description of my sub warfare for additions and comments.

As stated in the title, this is for an Allied campaign versus the AI; of course, these strategies are not designed for use against a human opponent (although some aspects may apply).

Alright, here is the outline:

(1) The US sub commanders need a major culling at the start of the war, as was historically the case. The commanders who may have done well in peacetime were not the ones needed to effectively execute a war. I use naval skill as the primary determination while keeping an eye for aggression outliers. The commanders of new boats have been good enough that I haven't micromanaged them.

(2) I use single-ship patrols, set reaction range to 1, and for the most part just use the 'patrol around target' feature. For specific chokepoints I might manually select patrol waypoints. I always set the loiter settings to zero.

(3) I use the 'list all active ships' button (shortcut 's') and filter for SS every turn, sorting by location to see which subs are in port. Those needing repairs are set to pierside/shipyard (where possible, to get them off the list) and I assign the rest to patrols. This keeps the number of idle subs to a minimum.

(4) For patrol route selection, aim for the DEI --> Japan supply routes as much as possible, but not exclusively. I have quite a few subs based out of Adak which patrol the southern coast of Japan (in particular), and it has been quite lucrative (as it was historically). I use the longer-legged boats for deeper patrols along supply routes and keep the shorter-legged S-boats and such to patrol closer to my bases. In my campaign, I've held on to Java and southern Sumatra, so the S-boats and Dutch boats have had quite a good time patrolling the Makassar Straight as well as the northern coast of Borneo (i.e. the primary points of origin for Japanese oil shipments since I've denied them Palembang). I also use a fair number of subs in the Solomons (based at Milne Bay) and the Central Pacific (based out of Johnston Island) - they don't bag much tonnage, but I like having the extra set of eyes around those areas. The few SSTs I have supply small garrisons so I don't have to risk transports in areas where torpedo-armed Bettys are operating.

(5) Sub patrol routes are kept in deep water when possible, but I'm not shy about putting them in shallow waters if I think the reward is worth the risk. For example, the Strait of Malacca is too narrow and too close to Singapore, but I don't mind operating in the shallow waters east of Singapore (a bit away from the Malaysian shoreline to cut down on Japanese aircraft on ASW patrol, such as around Groot Natoena).

(6) I put at least one AS ship in every port where I'm basing subs, and I try to have an AR for the major sub bases. For the most part, subs only need to be recalled to the larger ports when an upgrade requiring a shipyard is needed.


That's pretty much it - I haven't completely min/maxed my sub force, but I've been very happy with the results. I'm in mid Jan '43 and have just shy of 800 Japanese transports sunk (an average of about two per day) - the lion's share due to the subs. In the last three or four months I've seen a decent uptick in the number to TK/AO ships I'm sinking with 45 TKs and 12 AOs to date, almost all of those due to sub attack. I've only lost four subs in the war so far.

Seven Japanese subs have been sunk (+/= due to FOW) but part of that is because since the middle of '42 the Japanese subs have been eerily silent. Early in '42, they were quite active off the eastern coast of Australia, Fiji, etc. I set up my aircraft ASW patrols and formed ASW killer groups with the 8 ASW rated British DDs. Since roughly the middle of '42, the extent of Japanese sub activity is that I'll have an SSX try to sneak in one of my larger ports. Most sink before taking any shots, but one actually managed to hit the Illustrious in Colombo and a freighter was hit in Madras.

FYI, I also list some stats regarding my sub attacks in this post.

If I have any possible self-criticism it would be with the number of attacks (and hits) against warships, but on the other hand it doesn't seem out of place that those attacks would be relatively rare. Capital ships tend to be escorted, and escorts as a rule tend to be hard to hit. Also, a good chunk of the Japanese battle line has already been sunk via a combination of air and surface attack (11 carriers of all types, 5 BBs, 14 CAs, 20 CLs).

[Yaab]

ad 1 ) Do you mean you spend PPs at war's start to change sub commanders? I used to do that until I read some Clay Blair. The culling was done after the first combat patrol had been completed. Now, I employ the starting, default commanders (which are mostly poor or average), complete their first patrol, review them and eventually replace them. This is done purely for role-play reasons and also to give some leeway to Japs at start. I try to replace powergaming with role-play whenever I can.

[Korvar]

I like that idea.

I do similar 'role-play' with aircraft squadrons. Squadrons recalled to the ETO are sent with the aircraft type and all experienced pilots intact. I also like to equip training squadrons with a variety of aircraft including some with the latest models. Brewster Buffaloes are perfectly fine for learning basic flying maneuvers, but they would be useless for learning the air-to-air techniques that would make a successful Hellcat or Corsair pilot.

[witpqs]

(2) I use single-ship patrols, set reaction range to 1, and for the most part just use the 'patrol around target' feature. For specific chokepoints I might manually select patrol waypoints. I always set the loiter settings to zero.

The matter of patrols has even more aspects to it which are worth exploring.

I have not used the 'patrol around' feature (or much) so can't comment on how well it works out, but I imagine for many cases it should be fine.

Loiter is one I also have almost never used but I have heard people find it helpful, so I might try it some in appropriate circumstances. I'm thinking especially where there is a single or two-hex choke point which could (or each of which could) get a loiter command. See the comments on subs' DL (next).

A major factor in subs' effectiveness is their own DL. With good DL on a sub it can be avoided altogether or thwarted from attacking effectively even when targets are in-hex with it. Patrol zones are very helpful in limiting a sub's DL. Regardless of whether you have a sub patrolling an area or just a single hex, use one hex of the patrol zone to take the sub out away from enemy air cover (and potentially ASW surface groups) to where the enemy loses DL on the sub. It seems keeping DL on a sub is much easier than acquiring it in the first place, so use patrol zones to help your subs regain their stealth.

Another major factor in subs' effectiveness is DL of their prey. Obviously many areas where subs work are out of range or otherwise off limits to friendly air search but anywhere it can be provided to help subs it will make a big difference.

[Korvar]

Yes, from everything I've read DL for both the sub and potential targets has a lot to do with who gets to shoot first (if at all) in a sub-to-ship encounter; of course, lower DL on the sub and higher DL on potential target TFs transiting the area will increase the probability of the encounters.

I've also heard a lot about 'patrol lines', that is, having a sub patrol between two waypoints with the 'line' being aligned with a perceived shipping lane. With your concept of a "DL recovery waypoint", if you will, the result would be something like the following (with 'waypoints' meaning the patrol end-points, not normal navigation waypoints):




The loiter may or may not be needed depending on how far waypoint 3 is; the transit time alone may be sufficient to do the job. AFAIK, even a sub/ship which remains in a single hex still burns fuel as it is assumed to still be moving within the 40nm hex.

I've also seen some talk about aligning the length of the patrol 'legs' to the single-turn hex movement of the sub, but I'm not sure why that would be advantageous other than knowing your sub will be guaranteed to start on a specific hex (which may be the specific chokepoint you're aiming after).

[MakeeLearn]

Keeping subs around the targets of air attacks will aid in pilot rescue.

[Macclan5]

ORIGINAL: Korvar

(2) I use single-ship patrols, set reaction range to 1, and for the most part just use the 'patrol around target' feature. For specific chokepoints I might manually select patrol waypoints. I always set the loiter settings to zero.


(4) For patrol route selection, aim for the DEI --> Japan supply routes as much as possible, but not exclusively.

(5) Sub patrol routes are kept in deep water when possible, but I'm not shy about putting them in shallow waters if I think the reward is worth the risk.

(6) I put at least one AS ship in every port where I'm basing subs, and I try to have an AR for the major sub bases. For the most part, subs only need to be recalled to the larger ports when an upgrade requiring a shipyard is needed.

Very nice summary and thank you.

Valuable for all players no doubt and rookies such as myself in particular [8D]

A couple of points / clarifications and questions... [:D]

2) I am very certain that Ser Alfred and Bullwinkle have demonstrated a few times - that the 'manner in which the engine works' - Single sub patrols are superior to 'wolfpack task forces'.

However I do often have single sub patrols overlap patrol zones in many of the places I deem important. Say Port Moresby early in the game... do you also overlap certain patrol zones with multiple single sub patrols ?

4) I think this is interesting and perhaps could be expanded upon. Obviously patrol zones vary depending upon the stage of the war. However in early game I tend to concentrate a number of sub patrols in and about: (1) Gilberts (2) Marshalls (3) Truk especially (4) Port Moresby as indicated.

Is your experience - early in the war - such that DEI corridors are effective ? When you mention other locations does it vary widely?

One of the 'lessons' I have draw upon is the 'time in patrol' is equally or more important than location per see. Of course early in the war sub bases tend to be back not forward and therefore sub patrols spend inordinate amount getting to patrol zones as opposed to actually patrolling - such as getting to the DEI lanes be it from Pearl or Brisbane for example.

6) Yes agreed.

As per above where do you place your early major sub bases ? Of course time and war conduct will change. Obviously once one captures the Mariana's - you move your ports forward.

However in light of the need to maximize patrol - where do you find it is effective for a base or do you think it varies based upon opponent moves ?

I tend to err to the side of history. I use Pearl, Brisbane, and perhaps a small base at Darwin for low endurance Dutch subs that survive (assuming its not conquered by a human player) ; however Dutch subs while effective do end up needing time in dry dock be it Sydney or Brisbane - so Darwin is at best 'partially effective'.

I have never found Midway, Johnson, nor Canton for example to be of much use. Largely as it takes time to build up the ports, and to stockpile supplies. Further HQs carrying torps are at a premium early in the game.

--

Final question.

Disclaimer : in my few games I have always played with realistic US Torps on so to speak.

Do the experienced players really find that changing out Sub skippers is an effective tactic given the performance of US Torps early in the war / game ?

Certainly the higher aggression level helps - but does it really make a difference until say mid 43 when the better US Torps enter the game - and 'after the USN Sub Radar upgrades' ??

My anecdotal experience would suggest not. Save the better skippers till mid 43. But I have no especial empirical evidence to support that conclusion ...

Cheers

[btd64]

ORIGINAL: Macclan5

ORIGINAL: Korvar

(2) I use single-ship patrols, set reaction range to 1, and for the most part just use the 'patrol around target' feature. For specific chokepoints I might manually select patrol waypoints. I always set the loiter settings to zero.


(4) For patrol route selection, aim for the DEI --> Japan supply routes as much as possible, but not exclusively.

(5) Sub patrol routes are kept in deep water when possible, but I'm not shy about putting them in shallow waters if I think the reward is worth the risk.

(6) I put at least one AS ship in every port where I'm basing subs, and I try to have an AR for the major sub bases. For the most part, subs only need to be recalled to the larger ports when an upgrade requiring a shipyard is needed.

I have never found Midway, Johnson, nor Canton for example to be of much use. Largely as it takes time to build up the ports, and to stockpile supplies. Further HQs carrying torps are at a premium early game.

Cheers

AS ships provide torpedoes for subs. As well as a level 4 port or higher. I think it's a level 4 or higher....GP

[Macclan5]

Yes of course... doh... I do know that... [:)]

Although again with flotation damage / engineering damage etc - I still find it much quicker to get subs into a dry dock such as Pearl or Brisbane despite the presence of an AR for example at Midway / Johnson etc.

Perhaps that is my poor admiralship and or to risk patrol zones.[:D]

[HansBolter]

Getting a 3k capacity ARD to your major sub hubs helps to supplement the AS and AR.

I experimented a while back with 2 sub TFs and found that although both would never attack in any given encounter, I did anecdotally, seem to get more follow up interceptions of the same target TF with the second sub in the TF.

Haven't used this extensively, but did note its apparent effectiveness when I tried it.

I use the Patrol Around setting quite a bit, only giving specific targets for patrol hexes where necessary.

It seems to work fine.

[Korvar]

ORIGINAL: MakeeLearn

Keeping subs around the targets of air attacks will aid in pilot rescue.

That's a neat feature I recently discovered. I knew surface ships would attempt to pick up pilots such as when a carrier goes down (or ground troops if a transport goes down, assuming the rescuing ships have some carrying capacity).

I searched around and couldn't find anything other than apparently there is a hex check for presence of friendly forces. I'm curious if transiting a particular hex is sufficient or if the sub has to start or end a turn there; also, does reaction range count for anything?

[Korvar]

ORIGINAL: Macclan5

ORIGINAL: Korvar

(2) I use single-ship patrols, set reaction range to 1, and for the most part just use the 'patrol around target' feature. For specific chokepoints I might manually select patrol waypoints. I always set the loiter settings to zero.


(4) For patrol route selection, aim for the DEI --> Japan supply routes as much as possible, but not exclusively.

(5) Sub patrol routes are kept in deep water when possible, but I'm not shy about putting them in shallow waters if I think the reward is worth the risk.

(6) I put at least one AS ship in every port where I'm basing subs, and I try to have an AR for the major sub bases. For the most part, subs only need to be recalled to the larger ports when an upgrade requiring a shipyard is needed.

Very nice summary and thank you.

Valuable for all players no doubt and rookies such as myself in particular [8D]

A couple of points / clarifications and questions... [:D]

2) I am very certain that Ser Alfred and Bullwinkle have demonstrated a few times - that the 'manner in which the engine works' - Single sub patrols are superior to 'wolfpack task forces'.

However I do often have single sub patrols overlap patrol zones in many of the places I deem important. Say Port Moresby early in the game... do you also overlap certain patrol zones with multiple single sub patrols ?

I have also read that about wolfpacks and have done some experimentation. The interesting thing is that a Sub TF can have up to 25 subs, but I now see that more as a feature to easily redeploy entire submarine squadrons. The specific reason why wolfpacks don't work the way you might think is the game engine selects only a single 'best' sub per sub TF for an attack (based on several factors like ammo, damage, etc.). The confusion about multiple sub attacks seemed to come from the fact that a sub TF can contact the same target multiple times in the same turn. So players would see their wolfpack sub TF attack multiple times with different subs firing. It was in fact multiple encounters where the 'best' sub changed each time due to the depletion of ammo by the earlier shooters.

That said, the multiple contact possibility works quite well with overlapping (or close proximity) patrols. Depending on the number of subs you have to work with as well as the properties of the patrol area (total area to cover, presence or lack of chokepoints), it is possible to predict and cover the most likely shipping lane(s) quite well. I like to string line patrols as shown in the diagram earlier in the thread along these perceived likely routes and the effect is like a relay race. An enemy TF comes along and sub #1 fires. If it hits an escort, that TF just became a bit (or a lot) easier to attack for the next sub(s). Or if say a 'high value' target like a transport, oiler, or capital ship got hit - it will now either slow down the entire TF or will be spun out into its own Escort TF, perhaps taking an escort or two with it. In effect your subs are creating 'wounded gazelles' that the rest of the 'pack' can pick off further down the line. Enemy TFs become batons that your subs can pass to one another. The net effect is that a concentration of subs in a given area has a better chance of achieving many of the advantages of historical wolfpack tactics, even though the subs in game aren't technically operating as a single group.

ORIGINAL: Macclan5

4) I think this is interesting and perhaps could be expanded upon. Obviously patrol zones vary depending upon the stage of the war. However in early game I tend to concentrate a number of sub patrols in and about: (1) Gilberts (2) Marshalls (3) Truk especially (4) Port Moresby as indicated.

Is your experience - early in the war - such that DEI corridors are effective ? When you mention other locations does it vary widely?

One of the 'lessons' I have draw upon is the 'time in patrol' is equally or more important than location per see. Of course early in the war sub bases tend to be back not forward and therefore sub patrols spend inordinate amount getting to patrol zones as opposed to actually patrolling - such as getting to the DEI lanes be it from Pearl or Brisbane for example.

As per above where do you place your early major sub bases ? Of course time and war conduct will change. Obviously once one captures the Mariana's - you move your ports forward.

However in light of the need to maximize patrol - where do you find it is effective for a base or do you think it varies based upon opponent moves ?

The selection of patrol zones has to do with (1) logistics and (2) the behavior of your opponent. It's important to view the war through a strategic lens - what is the war all about? Fundamentally, it's about oil - for the Japanese to secure it and bring it to the Home Islands and for the Allies to deny it to them. The US oil embargo of Japan is what gave the Japanese their sense of urgency to attack the US in the first place; it was either that or heavily curtail their military adventuring. Oil was also heavily stockpiled by the Japanese in anticipation of war. The perceived threat of the US also made the Philippines a great threat to Japanese supply lines from Indochina (and DEI once the war began). So I see the DEI ---> Japan supply lines as the "holy grail" of things to target with sub warfare, as denying Japan oil is the best way to collapse their war economy. Consider how much of the IJN, in particular their remaining capital ships, were forced to remain in port mid/late war for lack of oil.

That said, you make a great point that "ToT" or "time on target" is an important consideration when planning patrol routes. You want to minimize transit time to/from and maximize time on station. Shortening the transit time to/from the patrol routes also increases the cycle time of subs heading back for fuel and ammo, effectively increasing the force multiplier of the subs you do have. Finally, damaged subs obviously have a better chance of making it home with a shorter distance to travel.

So effectively attacking the DEI shipping routes may or may not be feasible depending on the strategic situation of a particular campaign. Patrolling the oil 'points of origin' around Sumatra and Borneo will in particular depend on how far the Japanese have advanced their defensive perimeter. The Aleutians serve as a good base for patrolling Japan with the longer-legged subs, but it's certainly possible for the Japanese to deny this area to the Allies.

The other thing to remember is that the best feature of subs is their covert nature - not only with the manner in which they conduct their operations but even in how they are based. It's quite possible to set up a covert sub base by parking the right auxiliaries (an AS as well as a fuel-carrying ship) in a dot base much closer to your intended patrol routes. The riskiest part for the auxiliaries is getting there and eventually pulling out, but otherwise it is quite easy for them to fly 'under the radar' being disbanded into the dot base. Your enemy will pretty much have to recon the base to discover it as your subs will leave very few, if any, 'breadcrumbs' leading back to it. Sure, it won't be an 'evergreen' base unless you can bring in new fuel/supplies - but the good thing about keeping fuel and supplies on the auxiliaries is that you won't suffer the spoilage of a tiny dot base. It really opens some devious possibilities for placing subs where your opponent would not expect them (better chance of finding unescorted targets, etc.).

Which brings us to the other not so minor item - your opponent. Against the AI, it's not so much a consideration because it tends to repeat the mistakes of history. More or less predictable shipping lanes, disregard of chokepoints, and insufficient escorts or use of a convoy system. A human Japanese player will quickly learn to adapt, and it becomes apparent that the degree of historical success of US subs was in large part due to Japanese mistakes. From what I've heard, a skilled human opponent can largely frustrate Allied attempts to sink the Japanese merchant marine. In such a case, you would adapt too - realizing that your sub warfare is by necessity less about sinking merchant tonnage and more about modifying the behavior of your opponent. You've now forced him to carefully plan and escort every merchant/tanker convoy. He's dedicating resources (longer supply routes, more time to cycle fuel/supplies, more fuel spent on running escorts with transports over longer distances) that could otherwise be spent fighting you. Perhaps that's where you'll find value in patrolling some of the Central Pacific or other lesser trafficked areas - engendering the feeling in your opponent that nowhere is safe. Nevertheless, I've heard from veteran PBEM players that a side effect of careful protection of transports by a Japanese player can lead to an ironic situation where its more lucrative for the Allies to pursue Japanese sub doctrine - that is, to specifically target warships.

ORIGINAL: Macclan5

I tend to err to the side of history. I use Pearl, Brisbane, and perhaps a small base at Darwin for low endurance Dutch subs that survive (assuming its not conquered by a human player) ; however Dutch subs while effective do end up needing time in dry dock be it Sydney or Brisbane - so Darwin is at best 'partially effective'.

I have never found Midway, Johnson, nor Canton for example to be of much use. Largely as it takes time to build up the ports, and to stockpile supplies. Further HQs carrying torps are at a premium early in the game.

--

Final question.

Disclaimer : in my few games I have always played with realistic US Torps on so to speak.

Do the experienced players really find that changing out Sub skippers is an effective tactic given the performance of US Torps early in the war / game ?

Certainly the higher aggression level helps - but does it really make a difference until say mid 43 when the better US Torps enter the game - and 'after the USN Sub Radar upgrades' ??

My anecdotal experience would suggest not. Save the better skippers till mid 43. But I have no especial empirical evidence to support that conclusion ...

Cheers

As to base selection, the calculus changes depending upon if you have a human opponent as well as what the balance of power is. If you have a human opponent in the early stage of the war where the KB reigns supreme in the Pacific, maybe investing in sub bases at Midway or the Aleutians is too risky for your taste. One possible adaptation is to switch to an auxiliary-based logistics system where you set up some early warning triggers and pull them out at the first sign of trouble.

As to replacing the skippers early in the war, I would say there are two advantages: (1) better skippers are better able to preserve the assets (subs & the experience of their crews) and (2) they will be better at prosecuting deck gun attacks as well as taking shots and registering hits against targets. The more your subs successfully tangle with the enemy (even if less than ideal tonnage is being sunk), the more they will gain experience. Better to have a fleet of experienced crews ready to go in mid '43 than to wait till then to maximize their education potential.

[Korvar]

An illustration of using line patrols of multiple subs against a known shipping lane or chokepoint - the individual end points of the patrols may overlap or share the same hex, but I kept them separate in the illustration for the sake of clarity:




Sometimes a lot of possible shipping vectors converge to a single point, such as a major port. The line patrols can then be staggered as follows, adjusting for the area to be covered and the number of subs on hand - each possible lane has less coverage, but it prevents the enemy from completely circumnavigating your patrols:




In either case it is still possible to achieve the "baton passing" or "thinning of the herd" efficiency I explained earlier - with each hit either sinking a ship or making it easier for subsequent attacks by other subs. Covering probable lines of retreat (planning in advance, of course) can allow you to also catch battle survivors attempting to make it back to port. It can be a bitter pill for your opponent to have a damaged carrier survive a battle only to be torpedoed within sight of a friendly shipyard, sometimes days or weeks later.