What was the most effective naval anti-aircraft gun of WW2?

[Karl Heidenreich]

40mm Bofors with a proximity fused shells, thats my vote

[dunmunro]

..... For what it's worth, the rounds per bird figures for 5-inch AAA in the afore-mentioned report as follows -

Yes, and at least a two for one over claim would be typical, as these are based upon wartime kill claims. In 1942, the USN awarded about 202-246 "kills" for all calibres of AA, when the actual figure, based upon recorded Axis losses was 80-90. However, the USN recorded 60 5" kills (5/25+5/38) for 252 RPB, when the actual number was maybe, 5-10:

http://www.ibiblio.org/hyperwar/USN/rep ... index.html

but the increase in RPB as the war progresses, seems to indicate an increase in the accuracy of the 5" kill claims.

One thing to note, is that the 5"/38 could probably fire more rds in a given time than most other similar sized weapons.

[Byron Angel]

..... For what it's worth, the rounds per bird figures for 5-inch AAA in the afore-mentioned report as follows -

Yes, and at least a two for one over claim would be typical, as these are based upon wartime kill claims. In 1942, the USN awarded about 202-246 "kills" for all calibres of AA, when the actual figure, based upon recorded Axis losses was 80-90. However, the USN recorded 60 5" kills (5/25+5/38) for 252 RPB, when the actual number was maybe, 5-10:

http://www.ibiblio.org/hyperwar/USN/rep ... index.html

but the increase in RPB as the war progresses, seems to indicate an increase in the accuracy of the 5" kill claims.

One thing to note, is that the 5"/38 could probably fire more rds in a given time than most other similar sized weapons.

..... I have no doubt that the early war USN European theater AAA claims that you cite are completely bogus, but I'm not certain to what degree the late war claims are sinning; it will be necessary to review these reports and check the underlying analytical assumptions.

One factor which MAY militate for a more accurate assessment of late war 5/38 performance is the body of data available from the kamikaze engagements fought by isolated DD's on radar picket duty during the Okinawa campaign. Such data would eliminate the over-claim factor. On that score, I do recollect that the conventional wisdom was that the entire 5/38 battery under director control against a single a/c target was required to give a fair (50/50?) chance of shooting it down. Assuming a 250 kt target approach speed and fire opened at 8,000 yards with a battery of 4 or 5 guns firing a MT/VT ammunition mix at 20 or 25 rpmpg, the numbers do seem to fit approximately.


Byron

[Bgile]

I think that when it comes to claims of Kamikazes shot down the figures are likely to be very accurate. Individual ship claims might be in error, but the total number of planes shot down wouldn't be far off. Simple arithmetic. 10 planes attacked and 4 of them hit ships in the TF. 10 - 4 = 6 planes shot down. Most of the records are of the nature "Plane A came out of the clouds on a bearing of 300 deg at 8,000 yds. It was immediately taken under fire by ships A, B, and C. It was hit a number of times and was seen to begin a slow roll and passed over our ship before spashing 200 yds off the port side."

[dunmunro]

..... I have no doubt that the early war USN European theater AAA claims that you cite are completely bogus, but I'm not certain to what degree the late war claims are sinning; it will be necessary to review these reports and check the underlying analytical assumptions.

One factor which MAY militate for a more accurate assessment of late war 5/38 performance is the body of data available from the kamikaze engagements fought by isolated DD's on radar picket duty during the Okinawa campaign. Such data would eliminate the over-claim factor. On that score, I do recollect that the conventional wisdom was that the entire 5/38 battery under director control against a single a/c target was required to give a fair (50/50?) chance of shooting it down. Assuming a 250 kt target approach speed and fire opened at 8,000 yards with a battery of 4 or 5 guns firing a MT/VT ammunition mix at 20 or 25 rpmpg, the numbers do seem to fit approximately.


Byron

Those were both Pacific and European theatre claims. The problem with using the picket DD stats is that they are often unrepresentative of a typical engagement profile, where if possible, many ships would open fire at extreme range, and so both accuracy and ability to distinguish kills between ships, suffers.

AFAIK, the RN made overclaims in the same time frame, albeit, perhaps not to the same extent.
According to a USN AA study* made in early 1943, the RN (and AFAIK Commonwealth), claimed, by type,
Warships: 505
Warships (minor) 120
DEMS: 96
Allied Navies (non USN) 19

for a total of 740 kill claims from Sept 1939 to Dec 31 1942. My own guesstimate is that the RN shot down about 200 aircraft in 1942 and maybe 100-125 for prior years. That same study states that the USN shot down 447 aircraft from Dec 1941 to Dec 1942, but later studies seem to have revised this downward. Both navies made additional claims for "probable" kills.

AA Action Summary, July-Dec 1942, IB#22.

[dunmunro]

I think that when it comes to claims of Kamikazes shot down the figures are likely to be very accurate. Individual ship claims might be in error, but the total number of planes shot down wouldn't be far off. Simple arithmetic. 10 planes attacked and 4 of them hit ships in the TF. 10 - 4 = 6 planes shot down. Most of the records are of the nature "Plane A came out of the clouds on a bearing of 300 deg at 8,000 yds. It was immediately taken under fire by ships A, B, and C. It was hit a number of times and was seen to begin a slow roll and passed over our ship before spashing 200 yds off the port side."

The problem with this, is that the number of aircraft is also likely to be an overestimate, and this was certainly the case earlier in the war, and there are almost certainly cases were aircraft taken under fire escaped into cloud or at low altitude but were claimed shot down. I certainly don't think there was any intent to deceive, but it was probably impossible to accurately count the kills. A minimum 2 to1 overclaim seems to be pretty standard for every navy, throughout the war, when large numbers of claims are made.

[Bgile]

I think that when it comes to claims of Kamikazes shot down the figures are likely to be very accurate. Individual ship claims might be in error, but the total number of planes shot down wouldn't be far off. Simple arithmetic. 10 planes attacked and 4 of them hit ships in the TF. 10 - 4 = 6 planes shot down. Most of the records are of the nature "Plane A came out of the clouds on a bearing of 300 deg at 8,000 yds. It was immediately taken under fire by ships A, B, and C. It was hit a number of times and was seen to begin a slow roll and passed over our ship before spashing 200 yds off the port side."

The problem with this, is that the number of aircraft is also likely to be an overestimate, and this was certainly the case earlier in the war, and there are almost certainly cases were aircraft taken under fire escaped into cloud or at low altitude but were claimed shot down. I certainly don't think there was any intent to deceive, but it was probably impossible to accurately count the kills. A minimum 2 to1 overclaim seems to be pretty standard for every navy, throughout the war, when large numbers of claims are made.

But Kamikazes didn't escape. Once they arrived at a TF, they ALL went into the water or a ship.

[dunmunro]

But Kamikazes didn't escape. Once they arrived at a TF, they ALL went into the water or a ship.

The USN had no way to know if an aircraft was a kamikaze unless it attempted to crash a ship, and even then an aircraft may have started out on a regular bombing mission and then converted to a kamikaze after being fatally hit, but at longer ranges non fatal damage might cause the aircraft to return to base or the pilot may simply abort. The IJN also used decoy aircraft as well. It might be possible to take a subset of kamikaze attacks and be certain that they were kills, but this is almost certainly not the case when all claims are considered.

BTW, it was USN doctrine, and they stress this repeatedly, to open fire with the 5"/38 at the longest possible range and at these long ranges it must have been extremely difficult to verify kills and avoid multiple claiming.

[Bgile]

So you are saying that 10 aircraft attack, and 10 of them crash a ship or are shot down, and you think one of them might have been a regular bomber? I think that's pretty unlikely. As I understand it, the attack profile of a Kamikaze is quite different and recognizable.

[dunmunro]

So you are saying that 10 aircraft attack, and 10 of them crash a ship or are shot down, and you think one of them might have been a regular bomber? I think that's pretty unlikely. As I understand it, the attack profile of a Kamikaze is quite different and recognizable.

"Note.--In the foregoing figures the number of suicide planes includes not only those planes which definitely attempted to crash into ships but also a proportion of the planes shot down by AA. before they demonstrated their intentions. This proportion is determined by separating the uncertain kills according to the ratio of known suiciders to known nonsuiciders." http://www.ibiblio.org/hyperwar/USN/rep ... index.html , p2.

Also:
"Note.--All planes which miss ships in their suicide crashes are classified in this publication as AA. successes, whether or not they cause damage."
So an aircraft where the pilot misses through his own error is still considered a kill, but by the same token a non kamikaze that crashes close to a ship might be classified as a kamikaze.

But these are only two caveats when there's lots on information qualifying the definition of a kamikaze, and there was hundreds of claims, and there certainly must have been mistakes made.

This chapter of another report makes for interesting reading:
http://www.ibiblio.org/hyperwar/USN/rep ... ary-2.html

Finally, we know and I've demonstrated that overclaiming occurred on a large scale, and it certainly must have occurred with the Kamikaze attacks as well.

[Bgile]

Finally, we know and I've demonstrated that overclaiming occurred on a large scale, and it certainly must have occurred with the Kamikaze attacks as well.

So you don't see any difference between claims of Kamikaze kills and any other type. For example, they may have miscounted how many aircraft crashed into their ship.

[Byron Angel]

But Kamikazes didn't escape. Once they arrived at a TF, they ALL went into the water or a ship.

The USN had no way to know if an aircraft was a kamikaze unless it attempted to crash a ship, and even then an aircraft may have started out on a regular bombing mission and then converted to a kamikaze after being fatally hit, but at longer ranges non fatal damage might cause the aircraft to return to base or the pilot may simply abort. The IJN also used decoy aircraft as well. It might be possible to take a subset of kamikaze attacks and be certain that they were kills, but this is almost certainly not the case when all claims are considered.

BTW, it was USN doctrine, and they stress this repeatedly, to open fire with the 5"/38 at the longest possible range and at these long ranges it must have been extremely difficult to verify kills and avoid multiple claiming.

..... It's not clear to me that data overload was necessarily a large problem. The AAORG (AA Operational Research Group) report, covering a four month period Oct44-Jan45 (approx 123 days) lists the numbers of AAA claims by weapon type as follows:

5in-common.........22
5in-VT................17

3in-common...........5.5
3in-VT.................0.5

40mm...............115.5

1.1in...................0.5
20mm................78.5

50cal..................5.5
30cal..................2.0

Note: fractions imply that two or more calibers may have been responsible for the same kill.


Kamikaze attacks were actually not terribly chaotic. They typically involved a succession of discrete individual attacks spread over lengthy periods of time from various directions and altitudes by small handfuls of 3 to 5 aircraft. Even the "mass attacks" at Okinawa, officially involving several hundred sorties and aimed at swamping the defense, were actually spread out over periods of 24 to 36 hours. The "swamping" process did not involve a sudden onrush of a large number of aircraft in a single body, but a series of individual probes spread in time, approach direction, and altitude with the intention of evading radar detection and CAP interception Since about half of any attacking Japanese force would statistically fall to CAP before reaching the AAA zone, the AAA defenses were rarely confronted by large numbers of simultaneously attacking enemy aircraft. 5in claims in toto were only 39 aircraft over a period of 123 days - about one every 3 days. For these reasons I believe that over-claiming would be significantly less of a factor than in other cases.

I have been able to discover no evidence of any departure from standard tactics in attacks upon isolated picket ships. Prescribed Kamikaze attack tactics were fairly rudimentary: either an approach at altitude, followed by a 20-45 degree dive into the target ship from the broadside aspect, or a very low approach over the water at <50 ft altitude either into the side of the ship or followed by a last minute pop-up and dive into the target ship from above. Attacks were, to the extent that I have been able to discern from my reading, performed by individual aircraft in succession.

For what it's worth .....

[lwd]

Indeed my impression anyway is that after the Mariannas Japanese attacks tended to be in fairly low numbers. The one that got Franklin for instance was a single plane if I recall correctly. These small raids would be much easier to evaluate although assigning credit to individual ships or weapons could be problematic. Another thing that should be remembered is that the role of AA guns is to protect the ships. If a raid comes in and all the planes make it back to base it's still an AA success if none of the ships are hit. Long range AA fire particularly when radar controled and with VT fuzes certainly isn't going to help the aim or coordinaiton of the attackers.

[dunmunro]

We are really only concerned with the 5" kills and these are the ones where overclaiming are most likely:


"It is interesting to note that the percentage of AA. successes increased only against suicide planes. This may be attributed chiefly to the following:

1. Comparative lack of surprise attacks as a result of (a) greater ship alertness, and (b) early warning by DD pickets.

2. Enemy use of obsolete aircraft, which were vulnerable to AA. fire and were operated by unskilled pilots.

3. Ships' practice of opening fire at long range, using a high percentage of VT fuzes, thus increasing the percentage of 5-inch kills."
http://funsite.unc.edu/hyperwar/USN/rep ... index.html

Long range fire with multiple ships engaging the same targets = overclaiming. This happened throughout the war, and it certainly happened against kamikaze attacks.

[Bgile]

We are really only concerned with the 5" kills ...

Why?