Bomber vs fighter

[CobberKane]

The reports I have read don't say a 50/50 chance of knocking down a B-17 with 20 rounds - merely that the estimate is 20 20mm rounds to knock out a B17, and that 4-5 20 mm rounds from the front would down a B-17.

That's 800 rounds fired for a non-frontal assault, or 180 rounds for a frontal assault.

I'd also guess the low chance to hit comes from pilots firing out of range or at least effective range, you more rookie pilots or those that are afraid to close.

More skilled pilots would get more rounds on target, less skilled of course less.

I think I remember reading where Saburo Sakai stated they changed to head on attacks against the B-17's.

I believe the report found that it took on average 20 x 20mm hits to down a B-17. That means, of a population of B-17s that each took that degree of punishment, 50% would go down and 50% would not. Or, as applied to an individual aircraft hit by that number of shells, there was a 50/50 chance of it going down. Also, assigning the low hit rate to the effect of rookie pilots doesn't work, because there were also experten who undoubtedly scored much higher hit ratios, pulling the stats in the opposite direction. The Luftwaffe arrived at their figure by examining the data for their entire force - rookies, aces and everyone in between - because they wanted their improvements to apply to the entire force. Cherry picking results wouldn't make sense, nor would removing the less expert pilots from the fray to improve the hit ratio.
I can certainly see how frontal attacks would greatly increase the lethality of the fighters, but I can also see some drawbacks, primarily the high level of skill required and that the requirement to turn and overtake the bombers for each attack might have a fighter running of of fuel before it ran out of ammo.
Given that the LW were in a position to know, I'm inclined to accept their findings. That said, they were collating data from a strategic viewpoint - they didn't care how many individual fighters hit given bombers how many times; they wanted to know how many shells from all the fighters hit all the bombers. The issue is how well that data can be applied to one fighter meeting one bomber

 

[stona]

The reasons given by Luftwaffe pilots for the use of a head on attack are several but break down to the blindingly obvious.
The attacking fighters were exposed to the fire of at most two defensive gun positions per bomber and largely avoided the supporting fire from other bombers in the formation (they never attacked singly but in larger formations to dilute the defensive fire.) Avoiding supporting fire from the other bombers features particularly in the accounts of the most senior and experienced pilots.
They were exposed to this fire for a very short time, most Luftwaffe recollections reckon between 1 and 3 seconds, much less than when making an attack from the rear or in a 'curve of pursuit'.
It may have taken an estimated twenty 20mm cannon hits to bring down a bomber, first hand accounts would suggest many more, but it only took one in the vicinity of the pilot(s).

Evasion was by diving down through the bomber formation or zooming over it, attempting to get out of range as quickly as possible, before formating for another attack. The immense relative speed of the fighters to the bombers helps put distance between them rather quickly.

The disadvantages are also obvious. Closing speed is very high and there is little time to line up, aim, fire and then avoid a collision. Inexperienced pilots simply couldn't do it.

Some units continued to make attacks from the rear as did many 'sturmgruppen'. To avoid scared and inexperienced pilots opening fire at extreme range the entire attacking formation would open fire on command. They would then dive away and reform to the left or right below the bombers, again on an order given prior to the attack.

Any fighter that had expended his ammunition was expected to stay in formation and, if called upon, make successive passes on the bombers 'unarmed'. They were not to leave the formation on their own. This was standard procedure for all the Gruppen for which I know the orders.

Cheers

Steve

 

[Garyt]

I believe the report found that it took on average 20 x 20mm hits to down a B-17. That means, of a population of B-17s that each took that degree of punishment, 50% would go down and 50% would not. Or, as applied to an individual aircraft hit by that number of shells, there was a 50/50 chance of it going down.

Dy definition of the word "average", that is not correct. Average means one plane may go down with 10 hits, another with 20, and another with 30. The "average" here would be 20.

If you have a 50/50 chance of bringing it down with 20 rounds, the average then would not be 20, it would be higher.

assigning the low hit rate to the effect of rookie pilots doesn't work,

I was not stating that the Japanese should do better - I was merely stating the fact that inexperienced pilots skew the stats lower, and if you notice I also said experienced ones would skew it higher.

Really depends on when in the war you are as to the accuracy of the opposing fighters. I would think in 1945 the hit rate would be lower, though there would be the occasional quality pilot there. Pilots from Kido Butai while it was around would do better.

 

[stona]

One fighter might well run out of ammunition before downing one bomber. There are several accounts of five of six Luftwaffe fighters failing to finish off a B-17.

As for the problems faced by fighters attacking defensive formations of heavily armed bombers I would suggest hearing them from the horse's , so to speak. These are just a few accounts from letters, memoirs and combat reports which illustrate the problems quite nicely. There are dozens more available, all saying more or less the same thing.

'During an attack from behind, we were under defensive fire from the bombers too long and at least three machine gun positions fired at us from each aircraft. In addition the escorting fighters had the job of keeping us away from the bombers. So, we had no option but to attack from head on. Everything went very quick in this tactic. Every second bought us 220m closer together. Of course we didn't want to collide but pull away over the bomber. For this pulling up and over the bomber one needed almost the whole last two seconds (=400m). Our guns were adjusted to 400m. Therefore we had two options: to fire too early, already at a distance of 600 or 500 metres, or to pull up half a second later. A very dangerous business. We didn't have one second to fire our guns. It is incredible when one thinks of all the efforts we made for just one second. One thing was absolutely necessary, aim very precisely for this short moment.'

Fw Fritz Ungar. JG 54 and JG 26

' The second viermot I bought down was over Paris on 20th December 1942. I managed to shoot down a Boeing in a head on attack, the machine turning over on its back and diving down with a lurch, and me being able to pull up and over it. I felt certain that I must have hit the pilots. A burst of the four 2cm cannons and two machine guns from our Focke Wulf into the cockpit (which was only built from sheets of glass or plastic) was guaranteed to be deadly, if we scored a hit.'

' ….it was sheer murder to attack the American combat boxes from the rear which we sometimes did to give ourselves more time to fire. This left us exposed to the defensive fire of the bombers for longer too. At a distance of some 1500m the American bombers opened fire with all barrels….. We normally opened fire at a distance of 300m and tried to close in to 30 or 40m. During all this time our fighters were exposed to the defensive fire of at least eight machine guns per bomber which, with a group of twenty bombers, meant the concentrated fire power of 160 heavy machine guns. Each group of twenty machines was also flying echeloned to the side and higher than the leading group, so we had to bear the full brunt of their defensive fire as well. Opening fire from behind at 300m and taking 5 to 6 seconds to overtake the bombers, these attacks indescribable in their sheer physical and mental stress…….We therefore changed tactics and started attacking from head on, which called for incredible dexterity, a good aim, and nerves of steel until the last second.'

Lt. Otto Stammberger. JG 26

'I opened fire at 500m with the twenties. At 300m I opened fire with the thirties. It was a short burst, maybe ten shells from each cannon but I saw the bomber explode and begin to burn. I flashed over him at at [by] 15m and then did a chandelle. When I turned around I was about 300m above and behind them and was suddenly mixed in with American fighters.'

Oblt Georg-Peter Eder. JG 2

'In Russia we had been accustomed to throw ourselves with gusto at the rear of any formation of Ivans and help ourselves as independent hunters. At first we thought it would be similar here in the west, and lost quite a few feathers in the process. The 'Fat Cars' flew nicely in formation making their firepower almost impenetrable. Our stout Gruppe had soon shrunk to the size of a strong Staffel . Soon we attacked the 'Fat Cars'only from head on, which meant of course that the chances of scoring hits was reduced considerably for both sides.'

Uffz Uwe Micheels. JG 3


Cheers

Steve

 

[Garyt]

Interesting, Stona.

It makes it clear that with a frontal attack you get 1 second to fire*, which is a lot less time to fire than if approaching from the rear. Well, at least the B-17 only has about a second to fire at you also!

* With an MG151/20, that's a chance for about 11 rounds on target. For the slower firing cannon from a zero it's about 8 rounds.

 

[mig-31bm]

head on attack doesn't always work , here there a b-17 facing 17 Zero and still win

View: https://www.youtube.com/watch?v=6Im086TCu3I

 

[mig-31bm]

to be fair allied fighter with gun on wing instead of nose and the lack of cannon make them really bad again bomber

 

[Juha]

to be fair allied fighter with gun on wing instead of nose and the lack of cannon make them really bad again bomber

Not so simple, Typhoons with 4 Hispanos had no problems with LW bombers same true to Beaus and Mossies with four Hispanos in nose.

Juha

 

[mhuxt]

The German report being discussed (at least, the one I have) says 20 hits from 20mm shells are what is required to be certain of bringing down a four-engine bomber. It notes that many bombers have been brought down with less, but as the intention is to decide on which armament combination is most effective, it's willing to accept a certain amount of overkill.

View attachment ammunitions.pdf

(Yes, it's in German.)

 

[CobberKane]

Dy definition of the word "average", that is not correct. Average means one plane may go down with 10 hits, another with 20, and another with 30. The "average" here would be 20.

If you have a 50/50 chance of bringing it down with 20 rounds, the average then would not be 20, it would be higher.

I'm sorry Gary, but this is incorrect. The average is the value required to reach the 50th percentile, so if a hundred bombers are each hit twenty times, fifty of them would go down and fifty would not, and when single bombers were attacked they would require, on average, twenty hits to be destroyed. Having the average, you can then construct a bell curve to calculate the chances of any given number of hits resulting in a kill, or how many hits are required to achieve a given probabilty of a kill, but the average remains a constant.
A couple of things regarding the effectiveness of head on attacks - unless the Luftwaffe deliberately discarded the data from bombers that were obviously destroyed this way, the effects of head on attacks in reducing the average number of hits required for a kill would already be imbedded in the figure of 20/1. That is to say, the figure of an average of 20hits/kill is arrived at by collecting data from all types of attack, including head on. Of course, we don't know the exact methodology used, but as the LW were collecting data in order to make decisions regarding their whole force, it is reasonable to conclude they would sample accordingly.
Also, consider this; if it required twenty rounds on average to score a kill from every other angle, but only two from a head on attack, we might assume that switching solely to head-ons would increase the effectiveness of the fighters tenfold. But this is a case of lies, damn lies and statistics, because the figure of a 2% hit rate no longer applies, as it was not arrived at by examining only the results of head-on attacks. A hypothetical switch to head-on attacks would certainly see the average number of hits per kill fall significantly, but given the skill required the ratio of hits/rounds fired would also fall significantly, diluting the gain. In short, you need less hits on average to down the bomber, but you need to fire more rounds to get the hits. The fact that the LW favoured such attacks when possible suggests that the net benefit was positive, though reading the previous posts it seems the desire to avoid the bombers defensive fire was as much an imperative as increasing lethality.
Okay, sorry about Applied Maths 101 everyone, but it is a subject a lot of people seem to have trouble getting their heads around - hence the propensity for Politicians to misuse it, no doubt.

 

[CobberKane]

The German report being discussed (at least, the one I have) says 20 hits from 20mm shells are what is required to be certain of bringing down a four-engine bomber

.

Aha! Could be a classic case of the dangers of relying on second hand information, although if the report says 20 hits are 'certain' to bring down a bomber that is obviously not so - they may be applying the term to a given high probability, like 95%. German translator anywhere?

 

[GrauGeist]

There was mention a little ways back about rockets being used on bombers.

This is a B-24M (44-50838) that was hit by a salvo of R4M rockets launched by a Me262 on 4 April 1945. Only one crewmember, Cpl. Charles Cupp, Jr., survived. This was exactly 69 years ago today.

 

[mhuxt]

.

Aha! Could be a classic case of the dangers of relying on second hand information, although if the report says 20 hits are 'certain' to bring down a bomber that is obviously not so - they may be applying the term to a given high probability, like 95%. German translator anywhere?

"Analysis and combat experience have shown that to be certain of shooting down a 4-motor Boeing with mine shells, circa 20 hits with 20mm or 5 hits with 30mm are required. These results seem to us to be quite high, however they mark the doubtlessly frequent shoot-downs with fewer hits as random results. So as not to calculate in a manner which is overly favourable, these hit numbers have been used as the basis for the analysis below."

(My translation.)

 

[Garyt]

I'm sorry Gary, but this is incorrect. The average is the value required to reach the 50th percentile, so if a hundred bombers are each hit twenty times, fifty of them would go down and fifty would not, and when single bombers were attacked they would require, on average, twenty hits to be destroyed.

O.K, answer me this - 5 B-17's are shot down. The take 12,15,20,21, and 26 rounds to down. What is the average # of hits to kock down a B-17?

We are looking for the Mean here, not the Median or Mode.

If you want to do percentile standing:

Order all the values in the data set from smallest to largest.

Multiply k percent by the total number of values, n.

This number is called the index.

If the index obtained in Step 2 is not a whole number, round it up to the nearest whole number and go to Step 4a. If the index obtained in Step 2 is a whole number, go to Step 4b.

4a.Count the values in your data set from left to right (from the smallest to the largest value) until you reach the number indicated by Step 3.

The corresponding value in your data set is the kth percentile.

4b.Count the values in your data set from left to right until you reach the number indicated by Step 2.

The kth percentile is the average of that corresponding value in your data set and the value that directly follows it.

Here we are getting the median really, or 20 in my example above.

There is not really a stat that is equal to the average number of hits to put it down 50% of the time that I am aware

Actually, thinking about it, your stat reminds me of a Yogi Berra quote: "Baseball is 90% mental, the other half is physical."

 

[CobberKane]

I'm sorry Gary, but this is incorrect. The average is the value required to reach the 50th percentile, so if a hundred bombers are each hit twenty times, fifty of them would go down and fifty would not, and when single bombers were attacked they would require, on average, twenty hits to be destroyed.

O.K, answer me this - 5 B-17's are shot down. The take 12,15,20,21, and 26 rounds to down. What is the average # of hits to kock down a B-17?

We are looking for the Mean here, not the Median or Mode.

If you want to do percentile standing:



Here we are getting the median really, or 20 in my example above.

There is not really a stat that is equal to the average number of hits to put it down 50% of the time that I am aware

Actually, thinking about it, your stat reminds me of a Yogi Berra quote: "Baseball is 90% mental, the other half is physical."

Yes, you have started out looking for the average, but employed method for finding the median - a different thing. A median is simply the middle number of a sample - 20 this case. In statistics a median is useful if you want to avoid a one or two highly unusual results influencing the conclusions you draw from a small sample. For example, one of the bombers in your small set of five went down after only a single hit, that would skew the average number of hits required disproportionately, making the bombers look easier to knock down than they actually were. By using the median, you effectively ignore that one atypical instance.
The average, or mean, is calculated by adding all the results in a sample and dividing by the number of results in total. In your example this gives a result of 18.8. For larger groups, an average is a better tool than a median because it is more accurate, you can do more with it, and any highly unusual results have a minimal effect. As the LW were in a position to get data from a large number of wrecks, this is would have been the obvious course.
Yogi Berra was obviously not a mathematician!

 

[CobberKane]

Great stuff, mhuxt. If 20 x 20mm shells were ''certain'' (in fact, highly likely) to knock down a B-17, then the average would be much lower, perhaps around ten. It would be great to know how the figure was arrived at, but that might be too much to ask (sigh).

oops - double post

 

[mig-31bm]

Great stuff, mhuxt. If 20 x 20mm shells were ''certain'' (in fact, highly likely) to knock down a B-17, then the average would be much lower, perhaps around ten. It would be great to know how the figure was arrived at, but that might be too much to ask (sigh).

oops - double post

i think they mean it will take average of 20 x 20mm shells that hit to bring down a bomber
btw there 3 kind of average : median , mean , and mode , it more likely that in the example they using either mean or mode

 

[mig-31bm]

Not so simple, Typhoons with 4 Hispanos had no problems with LW bombers same true to Beaus and Mossies with four Hispanos in nose.

Juha

oh yeah i kind of forgot them , just thinking about P-47 , p-40 , F4F , F6F ,p-51 , spitfire ..etc

 

[CobberKane]

i think they mean it will take average of 20 x 20mm shells that hit to bring down a bomber
btw there 3 kind of average : median , mean , and mode , it more likely that in the example they using either mean or mode

According to the translation 20x20mm are certain to bring down a B-17 - not on average. And no, median and mode are not types of average, regardless of how the term might be used in common vernacular; the mean is the average, nothing else.
I'll leave the subject of applied mathematics alone now. There is nothing I've mentioned that couldn't be verified by a simple google search, in the unlikely event anyone is interested

 

[mig-31bm]

According to the translation 20x20mm are certain to bring down a B-17 - not on average. And no, median and mode are not types of average, regardless of how the term might be used in common vernacular; the mean is the average, nothing else.
I'll leave the subject of applied mathematics alone now. There is nothing I've mentioned that couldn't be verified by a simple google search, in the unlikely event anyone is interested

actually all these 3 are average , but of different kind to deal with different sort of statistic
Unit 5 Section 2 : Mean, Median, Mode and Range

The mean, median and mode are types of average.

they have different advantages and disadvantage:
mean :All the data is used to find the answer but very large or very small numbers can distort the answer
Median : Very big and very small values don't affect it but takes a long time to calculate for a very large set of data
Mode :The only average we can use when the data is not numerical but there may be more than one mode or may be no mode at all if none of the data is the same also it may not accurately represent the data