Glug, glug, glug in the Aleutians

[Kull]

NOTE: Scenario continues using the pre-patch version of AE.

This sure sounds familiar: I-5 encountered a Transport TF, did not attack, was discovered anyway and pounded by the escorts (2-1 escort-to-transport ratio), and was rattled and shaken multiple times, to include a hull penetration. But where this tale differs is that escorts only had one try at her (not the usual 2 or 3), plus the combat report is showing results that don't jibe with all the near misses seen on the screen.

AFTER ACTION REPORTS FOR Jun 11, 43
--------------------------------------------------------------------------------
ASW attack near Kiska Island at 159,51

Japanese Ships
SS I-5, hits 1

Allied Ships
LSI(M) Prince Henry
xAP Perida
DD Coghlan
DD Bache
DM Pruitt
DMS Chandler

I-5 diving deep ....
DD Coghlan fails to find sub and abandons search
DD Bache attacking submerged sub ....
DM Pruitt attacking submerged sub ....
DMS Chandler attacking submerged sub ....
DD Bache fails to find sub and abandons search
DM Pruitt fails to find sub and abandons search
DMS Chandler fails to find sub, continues to search...
DMS Chandler fails to find sub, continues to search...
DMS Chandler fails to find sub, continues to search...
Escort abandons search for sub
--------------------------------------------------------------------------------

[Kull]

Well, the combat report looks odd, but as we all know (from the previous graphic), I-5 DID take a hull penetration. Wonder how bad the damage is.....



FOUR points of damage? That hull penetration report must have been "fog of war"!

[Kull]

When the same TF arrived at Kiska, it encountered - and attacked - Ha-33. Maybe this TF has a leader who's more focused on troop transport than ASW, but again they didn't spend much time going after the midget. She took a few points of sys from the "hit" (a rattle), but lives to tell the tale:

AFTER ACTION REPORTS FOR Jun 11, 43
--------------------------------------------------------------------------------
ASW attack near Kiska Island at 157,51

Japanese Ships
SSX Ha-33, hits 1

Allied Ships
LSI(M) Prince Henry
DD Coghlan
DD Bache
DM Pruitt
DMS Chandler

DD Coghlan fails to find sub, continues to search...
DD Bache attacking submerged sub ....
DM Pruitt fails to find sub and abandons search
DMS Chandler fails to find sub, continues to search...
Escort abandons search for sub

--------------------------------------------------------------------------------

[Kull]

Continuing the odd late-game trend, I-175 is spotted and attacked near Dutch Harbor and this time the escort doesn't even fire a single depth charge!

AFTER ACTION REPORTS FOR Jun 13, 43
--------------------------------------------------------------------------------
ASW attack near Dutch Harbor at 169,50

Japanese Ships
SS I-175

Allied Ships
DD Lansdowne
PG Charleston
AM Oriole

DD Lansdowne fails to find sub, continues to search...
PG Charleston fails to find sub, continues to search...
AM Oriole fails to find sub, continues to search...
DD Lansdowne fails to find sub, continues to search...
PG Charleston fails to find sub, continues to search...
AM Oriole fails to find sub, continues to search...
DD Lansdowne fails to find sub, continues to search...
PG Charleston fails to find sub, continues to search...
AM Oriole fails to find sub, continues to search...
DD Lansdowne fails to find sub, continues to search...
PG Charleston fails to find sub, continues to search...
AM Oriole fails to find sub, continues to search...
Escort abandons search for sub

--------------------------------------------------------------------------------

[Kull]

And in the last action before the scenario ends, Patrol Aircraft attack Ha34 in Adak Harbor. The uncanny luck of the "canoe" occupants continued, and they escaped without a scratch (Final Sys damage is down to 15)

And so, the scenario ends. Next, the "Final Report".

[Kull]

But before we do. When the scenario ends, it gives you a last look at the force dispositions of the enemy. Imagine my shock at seeing actual Allied Subs on patrol, including one just NW of Attu. Keep in mind that I've now completed two Aleutians scenarios as the IJN, and NOT ONCE:
- Did a Japanese aircraft spot an enemy submarine conclusively enough to display it on the screen as an icon. Reports of wakes and periscopes and whatnot, but not enough to paint an icon.
- Did a single IJN TF ever spot an enemy sub (much less be attacked by one). So great was my confidence in the non-activity of US subs that by the end of this scenario I had 3 CVs in a TF with a single CL as escort.

[Kull]

So the scenario ends with an IJN "win". After you've played it once, that's not hard to do.

[Kull]

I gathered all the pertinent data involving the IJN subs and their contacts with the various Allied TFs. There's a LOT of information here, so let's begin by examining the meaning of each column of data:

1) Days: Sometimes there were more than one event on a given day, so this is how I kept a count of thedays versus the incidents.
2) Date: Although the scenario starts on 4/15, the first Sub-vs-TF action didn't begin until 5/6, three weeks into the scenario.
3) Sub: The identifier for each sub involved in an incident, if applicable.
4) Escorts: The number of escorts who encountered - and fired at - the submarine.
5) Hits: This is the count from the combat report. Possibly impacted by FOW, but probably reliable "in toto" given the number of data points.
6) Penetrations: This has nothing to do with the liberty activity of our intrepid sailors, but refers to the number of times an ASW attack actually "Hit" the submarine (counted by watching every attack on screen).
7) Damage: If the sub wasn't sunk in the attack, I'd count up the total amount of damage "points" (sys-flt-eng) afterwards, and would subtract those from the damage the subwas already carrying at the start of the attack (that was tedious).
8) Attack?: Did the Submarine attack the Task Force? Yes or No.
9) Torps Fired: If there was an attack (Y), this is the count of how many ships were fired at (not "number of torps actually fired").
10) Results: Description of the incident, or the results therefrom.
11) Patrol: Count of subs actually on patrol each day. The number changes as subs are sunk or forced to withdraw to repair damage. This doesn't count refueling runs, except the one case where the sub ran out of fuel and spent five days limping along, one hex at a time (plus a few more to repair the damage).
12) Patrol%: The percent of subs on patrol on any given day. The IJN starts with 10 subs, and then adds 5 more along the way - the percentage reflects the count of active subs vs. this "optimal" number. A few posts further on you'll see a chart that does a nice job of displaying this.

[Kull]

ORIGINAL: Kull
12) Patrol%: The percent of subs on patrol on any given day. The IJN starts with 10 subs, and then adds 5 more along the way - the percentage reflects the count of active subs vs. this "optimal" number. A few posts further on you'll see a chart that does a nice job of displaying this.

The data is fairly sobering. There were an average of 7.4 subs on patrol over the 39 days from initial contact through to scenario end, about half the committed total.

[Kull]

OK, so what is all this telling us? I played around with the numbers and separated the results into three categories that summarize the results from the entire scenario

1) ASW Effects: By far the largest category, there are ten different calculations here.
2) Escort Qty Impact: The Aleutians TFs are well escorted, perhaps more so than the player will encounter in the Grand Campaign. Is that driving results?
3) Sub Offense: Yes, there was some! Do the numbers tell us anything?

[Kull]

ASW Effects:

1) ASW per Sub: 15 subs and 18 attacks in 40 days. Ouch! Once a sub entered the "Death Box", it was nearly certain to encounter an Allied TF.
2) Subs Sunk per ASW: A very sobering number. If the Allies attacked an IJN sub, 33% of the time it was sunk. I don't have numbers from the real conflict, but that might be about right.
3) Sub Major-D per ASW: And even when a sub survived, almost 1/4 were damaged so heavily that "return-to-port was the only option. Now we're up to 55% of all attacks either sink the sub or force a return to port. That seems a bit high.
4) Sub Minor-D per ASW: This number is actually too high, since two of the attacks caused 4 or less damage which is effectively "got away scot free". But that still leaves 6 of 18 which shows that yet another 33% of all subs take noticeable damage. And if this were the GC, I'd have sent all of them home to repair. So that means 88% of all ASW attacks will either sink the sub or send it home for repairs. Whew!
5) Hits per ASW: Looking at all 18 ASW attacks, the average number of hits was 11.2. However, this number is skewed by the large number of hits taken by the three subs that were forced to surface before being sunk. Still, pulling 38 hits out of the total still leaves 9.1 per ASW attack - not a noticeable decrease. I believe this is the first major clue as to what is happening during ASW attacks. Being glued to the monitor during every ASW incident, it was apparent that the escorts were making multiple depth charge runs - 3 not being uncommon. Well, that is a LOT of depth charges, especially considering the number of escorts. My observation to the developers is this. High accuracy plus lots of escorts plus multiple ASW runs per escort equals high lethality. If all of the above were true in the Pacific War, then sail on. If not, this should be looked at in more depth.
6) Penetrations per ASW: When Depth Charges actually HIT the submarine, bad things happen. And the odds say that every ASW attack will generate at least one such hit. Seems a little high, but it's hard to say if this is driven by accuracy or just the huge number of depth charges being dropped during every attack (another possible impact from the multiple runs)
7) Damage per ASW: Even excluding the catastrophic damage incurred by the six sunk subs, the average for those that survived was 44.8 points. This is another "secondary metric" that simply reinforces the fact that multiple hits are going to do a lot of damage.
8) Subs Attacked: Of the total of 15 subs in the scenario, 14 of them encountered ASW at least once. Not a real shocker. All of them were patrolling the Death Box, a region heavily traversed by Allied TFs.
9) Subs attacked X2: Pity the poor souls who experienced two ASW attacks in the scenario. Almost surprisingly, only two of this group were sunk. On the other hand, the other two suffered major damage and were out of the scenario. Again, the key fact is that putting subs into the Death Box is inherently dangerous. No surprise here.
10) ASW per Day: 18 attacks occurring in 39 days is a 46% activity level. That sure sounds like a lot, but again it's more a function of the high volume of ships and subs occupying the Death Box then anything else.

[Kull]

Escort Qty Impact:

1) Escorts per ASW: Counting up the number of escorts listed in every combat report and dividing by the total number of ASW attacks (18). My original theory was that high escort numbers were driving the ASW lethality, but I no longer see that as the key metric. Rather, they provide a multiplier effect to the "multiple runs". There is an element of "chicken or the egg" when looking at escort qty versus number of runs per escort - which is the real "bad guy"? On the other hand - as the following stats seem to show - there's not a high correlation between escort quantity and the degree to which bad results happen to a submarine.
2) Avg Escorts - Sunk: In the six ASW attacks that led to a sinking, there were an average of 3.5 escorts per TF. Although high, it's not significantly different from the overall average of 3.3
3) Avg Escorts - Major: And here the metric goes out the window. There were only 3.0 escorts per TF when ASW caused major damage to four subs. On the other hand, that's not a large sample size.
4) Avg Escorts - Minor: And in these eight cases, the average is 3.3 - same as the ovcerall number, and even slightly higher than "major damage" escort average.

So there's no obvious correlation between numbers of escorts and damage incurred by submarines. Except that having 3 or more escorts for an Allied TF appears to be a very good way to keep your ships safe AND to inflict punishment on the IJN silent service.

[Kull]

Sub Offense:

That's right. In all the analysis of ASW effectiveness, it's easy to forget that, "Hey! These things are armed too!" The issue of ASW lethality is certainly worth looking into, but I'm not prepared to say it's evidence that anything is "broken". At most, the code should be looked at and *possibly* a bit of tweaking made. But submarine offense just looks wrong:

1) Sub Attacks per TF: In only four of 18 encounters did the submarines open the engagement by firing the first shot. And they weren't facing ASW or even Surface Combat TFs. All or almost all of the TFs were Cargo or Transport. And still, against these slow, freighter-laden TFs, submarines get the first shot only 22% of the time. Or more to the point - Transport-type TFs hunted down and attacked 78% of the submarines that crossed their paths, even though these subs did not fire first!
2) Torp Hits per TF: Only three times in the whole scenario did torpedoes hit ships, and two of those came in the same incident (described below).
3) Ships Sunk: Really a startling statistic, when you consider that both happened during one of the first incidents. On May 8th, I-21 sunk an escort and then was forced to surface after heavy ASW. And promptly sunk a transport with another torpedo before finally being blown out of the water. And that was it for the rest of the scenario. Not to beat a dead horse, but that kind of offense is way below "anemic".

Nobody else seems to be reporting results this pathetic, so maybe something else is going on here:
- Does the code keep submarines at bay when faced with large numbers of escorts?
- Is Allied ASW that much better in mid-1943,such that IJN submarines can only get a shot off 22% of the time?
- Or is it something to do with submarine crew skills and/or poor leadership in the Aleutians scenario?
Not really sure. Again, this is not a call to "do something", but simply a suggestion that something might be amiss.

[Kull]

Well, I hope you enjoyed the AAR, and didn't find the final deluge of stats and opinions too dreary! The real bonus is that now I can FINALLY download and install the AE Patch! That prospect has me eagerly looking forward to, in the words (and delightful imagination) of Gil Elvgren, a Splendid View!

[krupp_88mm]

hey i read your AAR it was funny ive never even played this game i dont think i could stand it i just like reading the aars.

- Does the code keep submarines at bay when faced with large numbers of escorts?
- Is Allied ASW that much better in mid-1943,such that IJN submarines can only get a shot off 22% of the time?
- Or is it something to do with submarine crew skills and/or poor leadership in the Aleutians scenario?
Not really sure. Again, this is not a call to "do something", but simply a suggestion that something might be amiss.

on this note i don't claim to be a sub expert.. but i  have played alot of sub simulation games including silent hunter 3 and 4.. and i can tell you a single jap sub attempting to attack an allied convoy.. with those capable destroyers in 43.. i think the results are pretty accurate.. assuming these things

1. these were pretty small groups of ships with lots of esocrts.. when attacking a convoy all the math and triangulation looking through that little scope all the time fearing for your life from hoards of escorts... it can be pretty hairy... an accurate shot against one or two ships is sometimes impossible... i have no idea if this is modeled but its not easy to sniper a moving ship with lots of escorts.. spray and pray usually works better with lots of escorts and i dont even think spray and pray would work with only 2 targets

2. the jap sub won't be able to approach the task forces from the front without slipping through the escort screen.. which can be alot harder than it seems.. anyone but an expert captain wouldnt stand much of a chance of hitting a destroyer.. so waiting through the screen for the  transports to show up.. i would guess most of the time it wouldnt work unless the weather was bad.. or if you got lucky.. which would explain why most of the time the subs got bushwacked before they were in range of the transports..

3. have you tried grouping the subs up for a wolf pack style attack.. if i were facing such a heavily guarded force the best thing to do would be to pull off a few escorts with one sub and have the other lie in wait for the juicy targets

4. what were the weather conditions?.. weather is extremely important in a sub attack.. its also extremely important in living.. also whats the depth in these waters

5. how well were the captains and crew experience.. that should make all the difference in the world for a sub... the japs are great at math so in theory they should be really fast with those shooting calculations lol

[Xxzard]

I would certainly like to know at what point in the GC the Japanese subs become inneffective, because certainly at the beginning, with the more aggressive AI control of sub patrols, those Japanese subs are to be feared. Of course in contrast, early in the war escorts are pretty sparse, and compared to the US subs, the Japanese torpedos actually explode.

[Kull]

ORIGINAL: Xxzard

I would certainly like to know at what point in the GC the Japanese subs become inneffective, because certainly at the beginning, with the more aggressive AI control of sub patrols, those Japanese subs are to be feared. Of course in contrast, early in the war escorts are pretty sparse, and compared to the US subs, the Japanese torpedos actually explode.

Bingo (emphasis added). There have been a few AARs in which folks suggest exactly this. Once they are able to group their shipping into well defended convoys, the IJN sub menace begins to fade. And if this scenario suggests anything, fade will turn to plummet as the war progresses.

[Kull]

ORIGINAL: krupp_88mm

hey i read your AAR it was funny ive never even played this game i dont think i could stand it i just like reading the aars.

It is a demanding game, but this scenario is a relatively simple introduction to the whole AE experience. And from here, Guadalcanal and Coral Sea are the next step up the cliff. This game isn't for everyone, but if you are fascinated by the era and the topic, there's none better.

- Does the code keep submarines at bay when faced with large numbers of escorts?
- Is Allied ASW that much better in mid-1943,such that IJN submarines can only get a shot off 22% of the time?
- Or is it something to do with submarine crew skills and/or poor leadership in the Aleutians scenario?
Not really sure. Again, this is not a call to "do something", but simply a suggestion that something might be amiss.

on this note i don't claim to be a sub expert.. but i  have played alot of sub simulation games including silent hunter 3 and 4.. and i can tell you a single jap sub attempting to attack an allied convoy.. with those capable destroyers in 43.. i think the results are pretty accurate.. assuming these things

1. these were pretty small groups of ships with lots of esocrts.. when attacking a convoy all the math and triangulation looking through that little scope all the time fearing for your life from hoards of escorts... it can be pretty hairy... an accurate shot against one or two ships is sometimes impossible... i have no idea if this is modeled but its not easy to sniper a moving ship with lots of escorts.. spray and pray usually works better with lots of escorts and i dont even think spray and pray would work with only 2 targets

2. the jap sub won't be able to approach the task forces from the front without slipping through the escort screen.. which can be alot harder than it seems.. anyone but an expert captain wouldnt stand much of a chance of hitting a destroyer.. so waiting through the screen for the  transports to show up.. i would guess most of the time it wouldnt work unless the weather was bad.. or if you got lucky.. which would explain why most of the time the subs got bushwacked before they were in range of the transports..

The TFs in the Aleutians scenario are VERY well escorted, and that may explain the relative infrequency of Japanese attacks. I'm still a bit concerned that even Transport TFs are so successful at preventing the IJN from firing the first shot. Your examples are quite good, however, and a coding version of that is believable.

3. have you tried grouping the subs up for a wolf pack style attack.. if i were facing such a heavily guarded force the best thing to do would be to pull off a few escorts with one sub and have the other lie in wait for the juicy targets

I didn't try that, but only because "conventional WitP wisdom" (the predecessor game to AE) says it didn't work. Should have tested it anyway. Doh!

4. what were the weather conditions?.. weather is extremely important in a sub attack.. its also extremely important in living.. also whats the depth in these waters

Believe me, the idea that subs might suck given the terrible weather conditions in the North Pacific is certainly true in RL. I didn't think weather was modeled to that degree in AE, but I could be wrong. All the ASW occurred in ocean hexes, so we can't blame shallows or coastal hexes, anyway.

5. how well were the captains and crew experience.. that should make all the difference in the world for a sub... the japs are great at math so in theory they should be really fast with those shooting calculations lol

I tried skipers of varying experience, and it didn't seem to matter much. I didn't collect that as a stat, unfortunately. And I never even looked at crew experience. I could probably go back and capture all that, but it would take time. Let me think about it (can I keep the "splendid view" of the patch at arms length any longer?)

[skrewball]

I am trying to find more numbers on actual IJN SS attacks...but here is a sobering note taken from www.combinedfleet.com

"Japan started the war with 63 ocean-going submarines (i.e., not including midgets), and completed 111 during the war, for a total of 174. However, three-quarters of these (128 boats) were lost during the conflict, a proportion of loss similar that experienced by Germany's U-Boats. Most of the surviving boats were either dedicated to training roles or were recently completed and never saw combat. Of those which saw significant combat, the toll was very grim indeed. For example, of the 30 submarines that supported the Pearl Harbor attack, none survived the war.

Compared to German submarines, Japan's huge boats were relatively easy to sight visually and with radar, slow to dive, hard to maneuver underwater, easy to track on sonar, and easy to hit. Japanese hulls were also not as strong as those of German boats, and therefore could not dive as deeply nor survive such rough treatment. Also, they lacked radar until the first sets were installed in June 1944, and never had sets as good as the Allies possessed."

So it looks like a JFB should us his subs for the occasional deep strike as opposed to dedicated commerce raiding.

[krupp_88mm]

I am trying to find more numbers on actual IJN SS attacks

i think the most inaccurate part of this simulation may be the fact that the subs even attempt to engage these forces.. no German sub captain in their right mind would attack two transports guarded by three destroyers.. it just isn't worth it(the more transports the escorts have to guard the less effective each of them become(i wonder if this is modeled in the game)).. so i guess the question is if you tell a Japanese sub captain to 'attack convoy forces' should they wait for an easy target or throw themselves to their death at the first heavily guarded force that shows up?