That depends on whether you consider a pilot landing at an airspeed that guarantees a crash to be an accident. Crashes caused by lack of training cannot be attributed to the aircraft in my opinion.
Recurring Theme in WW2 Aviation
- Thread starter Robert Porter
- Start date
Ad: This forum contains affiliate links to products on Amazon and eBay. More information in Terms and rules
MiTasol
1st Lieutenant
The Martin B-26 had an excellent safety record starting before D-day. Some sources I saw in about the 70's said the turning point was a petite female ferry pilot showed up at several bases flying a B-26 with one engine feathered and did low level aerobatics before landing with the engine still feathered. The good old if a small woman is not scared of this airplane then nor am I factor at work. Urban myth or true I do not know but I do know there were magazine adverts that said the B-26s did 4,000 missions during June 44 with not a single aircraft being lost.
There's a difference between a safety record and a combat loss record. A high performance plane with an atrocious accident rate early on in training has more than once matured into a combat survivor once training techniques and design modifications have caught up with its performance. The Martin Marauder is a good example. (In the heyday of the Twin Otter the Beech 99 was considered a "hot ship", as its Vref was [GASP!] OVER 100 knots! Better not send that one out in the boonies!)
A jump in aircraft performance often preceded the necessary jump in pilot performance by many lives lost, especially in wartime.
Cheers,
Wes
A jump in aircraft performance often preceded the necessary jump in pilot performance by many lives lost, especially in wartime.
Cheers,
Wes
Last edited:
Good question! Philosophically, I agree with you, you can't blame a pilot for the failings of his instructors, but it IS an operational loss, not combat, so statistically you've got to consider it an accident. And let's face it, if the airplane in question has quirks or performance parameters outside the comfort zone of the pilot community, those statistics will be needed in order to advocate for the necessary changes.
Cheers,
Wes
Peter Gunn
Master Sergeant
I'm confused (shocking I know) but looking over the link and reviewing the numbers where does the following stat come from?
"In Laura Hillenbrand's book, Unbroken, she notes that during WWII, "In the air corps, 35,946 personnel died in non-battle situations, the vast majority of them in accidental crashes."."
Either I'm really bad at math... I'm not, or I'm missing something in this equation. I looked over the table in the link pbehn posted and there's nothing like 35,000+ deaths, or am I looking in the wrong place? Or is Laura Hillenbrand looking in the wrong place?
I presume there were a lot of other causes of death that weren't training or in battle. Glenn Miller and Leigh Mallory are two famous examples. Losses on take off and landing may have been on a mission but they are non battle situations. Fog, running out of fuel, getting lost and wing icing caused many losses.
Last edited:
GregP
Major
The table only list accidents in the Continental USA. 1942 - 1945. There were a LOT of operational fatalities outside the USA and a startling number on delivery flights that happened after they left the USA and before they got to wherever they were going. As I recall, the losses on delivery flights were high. You might check the Statistical Digest of WWII. It is available as a pdf download on the web. I attached it below.
Check tables 99 and 100 as a good starting spot.
From it, I show 43,581 overseas fatalities, of which 22,948 were combat fatalities. That leaves 20,633 lost in non-combat accidents/incidents. If you add 13,621 fatalities in the Continental USA, you get 36,569 fatalities in accidents, both in the USA and overseas.
That's a lot of fatalities to accidents and would have shut down ANY peacetime Air Force. War makes casualties a bit easier to swallow as the result, but does not make them any less tragic.
Some in here aren't very enamored of the Statistical Digest of WWII, but it has tables and data I have seen no place else. If you can't find data to contradict it, then maybe these data are the best we have on some subjects. I believe the Digest on things like hours flown, casulaties, etc.
I kind of disagree on aerial victories, but that subject has been a bone of contention since the middle of the war and isn't likely to be settled anytime soon. I 'm NOT a revisionist, myself. If the scores were good enough for the people of the day, they are good enough for me.
I have no issue with trying to find out EXACTLY how many were shot down, but resist "updating" the scores for the U.S.A. only and ignoring the rest. Unfortunately, we are seemingly the only country that funded a study of same after the war, so the revisionists go after the US scores only and simply disparage the rest as fantasy. To me, when they correct ALL the scores, that would be a table worth having. Otherwise, let's compare what they had right when the war ended and be done with it. At least that way, we're comparing WW2 scores, instead of correcting US scores and comparing the corrected list against uncorrected scores from other countries. That is just NOT the way to do it.
That's why I still list Boyington at 28. That was his score when the war ended. When every other US pilot gets the same attention he did, then we might have something to talk about. There is an entire argument here that just doesn't fit the topic, so I'll forego it.
Check tables 99 and 100 as a good starting spot.
From it, I show 43,581 overseas fatalities, of which 22,948 were combat fatalities. That leaves 20,633 lost in non-combat accidents/incidents. If you add 13,621 fatalities in the Continental USA, you get 36,569 fatalities in accidents, both in the USA and overseas.
That's a lot of fatalities to accidents and would have shut down ANY peacetime Air Force. War makes casualties a bit easier to swallow as the result, but does not make them any less tragic.
Some in here aren't very enamored of the Statistical Digest of WWII, but it has tables and data I have seen no place else. If you can't find data to contradict it, then maybe these data are the best we have on some subjects. I believe the Digest on things like hours flown, casulaties, etc.
I kind of disagree on aerial victories, but that subject has been a bone of contention since the middle of the war and isn't likely to be settled anytime soon. I 'm NOT a revisionist, myself. If the scores were good enough for the people of the day, they are good enough for me.
I have no issue with trying to find out EXACTLY how many were shot down, but resist "updating" the scores for the U.S.A. only and ignoring the rest. Unfortunately, we are seemingly the only country that funded a study of same after the war, so the revisionists go after the US scores only and simply disparage the rest as fantasy. To me, when they correct ALL the scores, that would be a table worth having. Otherwise, let's compare what they had right when the war ended and be done with it. At least that way, we're comparing WW2 scores, instead of correcting US scores and comparing the corrected list against uncorrected scores from other countries. That is just NOT the way to do it.
That's why I still list Boyington at 28. That was his score when the war ended. When every other US pilot gets the same attention he did, then we might have something to talk about. There is an entire argument here that just doesn't fit the topic, so I'll forego it.
Last edited:
Then what's the point of doing any historical research? What's the point on looking in previously unopened or unavailable archives when they become available? In fact what's the point in history as a subject at all if it is left set in stone at some random point in time with no chance of correction in the light of new facts and factors?
It isn't a matter of ignoring new information on US figures and refusing to acknowledge the new data generated, it should be a matter of attempting to do the same for all other nations too, find the information, continue research, continue the debate.
Figures that are provably incorrect should never be good enough for anyone.
Cheers
Steve
Peter Gunn
Master Sergeant
I believe I have stated before that my reading comprehension seems lacking of late, right at the top of the table it CLEARLY states, and I quote:
"Table 214 -- Airplane Accidents in Continental US, By Principal Model of Airplane - Number and Rate: 1942 to 1945"
Emphasis mine.
Sorry guys, my error, many thanks to both of you for setting me straight.
"Table 214 -- Airplane Accidents in Continental US, By Principal Model of Airplane - Number and Rate: 1942 to 1945"
Emphasis mine.
Sorry guys, my error, many thanks to both of you for setting me straight.
Peter Gunn
Master Sergeant
Thanks for the link, I downloaded it but that's going to take some time to go through, just a little light reading. I'll just say I agree with you on the issue of revised scores for all nations, what's good for the goose and all. But I think if you can revise through postwar records where available I see no reason not to. Eh, that's all I'll say as you're right, not the thread for this topic. Thanks again for the info.
- Thread starter
- #231
Robert Porter
Senior Master Sergeant
With absolutely no offense to anyone intended, I think it can be definitively stated that there will NEVER be a 100% accurate and provable data record produced. Records produced in the last Gulf War are already being disputed and gaps found. During wartime every effort is made to generate good record keeping but it is never 100% nor is it even consistent. Different units often collected data using different methodologies. Sometimes primary records were destroyed in fire/water/loss even before those records made their way up the chain. Records retention was and is an issue. Often the source records are degraded beyond recovery by age and poor storage. Transcription errors creep in and sometimes even tertiary aggregate records, especially from WW2 and before degrade. I work in IT and my sub speciality for the last 25 years has been document management and archival management. I can tell you for a fact there is no all inclusive source for a lot of the information we seek. Speak to research historians and they will, grudgingly, tell you that most of the data we argue about is an educated guess with a lot of gaps filled in with statistics.
Where records do exist, there are often conflicting records sometimes even from the same source. Ask Dana about this I am sure he can attest. The example that comes immediately to mind was from the 80's when the Iran Contra scandal was in full bloom. Attempts to determine just what munitions may be involved failed mostly because of either missing records or outright lack of proper record keeping. In 1980 at FE Warren AFB my unit, the 90th MSS was tasked with performing an audit of hand guns, rifles, LAWS rockets and other weapons we had in our possession. Just our unit mind you, and at the end the best we could tell was we were either missing over 100 M-16's or we had 23 more than we should. Reason was bad record keeping. For instance when an armourer uses parts from a weapon slated to be destroyed to repair a weapon also slated to be destroyed and does not properly amend the records to show that combining two otherwise unserviceable weapon into one that was serviceable.
Then we have the fact that some of the records from the RAF and the Luftwaffe among others, were outright destroyed in battle. Or in the case of Germany, lost to deliberate destruction, or lost due to neglect before they could be properly cataloged. Not even mentioning the fate of the records held in the Soviet sectors prior to reunification. It is why I take records that do exist with a huge grain of salt and won't debate records accuracy because honestly there is NO accurate source now, and even in the day there was none. Human error, failure to file a report correctly, loss in transit, etc.
We have a lot of information, but to argue about numbers other than on a very large scale seems meaningless to me. I value anyones input, am I crazy or does it just seem a little disingenuous to argue if 34,519 people died in accidents or was 33,917? Honestly the best answer in my book would be prefaced by saying the best estimate is somewhere in the low to mid 30,000 range.
Where records do exist, there are often conflicting records sometimes even from the same source. Ask Dana about this I am sure he can attest. The example that comes immediately to mind was from the 80's when the Iran Contra scandal was in full bloom. Attempts to determine just what munitions may be involved failed mostly because of either missing records or outright lack of proper record keeping. In 1980 at FE Warren AFB my unit, the 90th MSS was tasked with performing an audit of hand guns, rifles, LAWS rockets and other weapons we had in our possession. Just our unit mind you, and at the end the best we could tell was we were either missing over 100 M-16's or we had 23 more than we should. Reason was bad record keeping. For instance when an armourer uses parts from a weapon slated to be destroyed to repair a weapon also slated to be destroyed and does not properly amend the records to show that combining two otherwise unserviceable weapon into one that was serviceable.
Then we have the fact that some of the records from the RAF and the Luftwaffe among others, were outright destroyed in battle. Or in the case of Germany, lost to deliberate destruction, or lost due to neglect before they could be properly cataloged. Not even mentioning the fate of the records held in the Soviet sectors prior to reunification. It is why I take records that do exist with a huge grain of salt and won't debate records accuracy because honestly there is NO accurate source now, and even in the day there was none. Human error, failure to file a report correctly, loss in transit, etc.
We have a lot of information, but to argue about numbers other than on a very large scale seems meaningless to me. I value anyones input, am I crazy or does it just seem a little disingenuous to argue if 34,519 people died in accidents or was 33,917? Honestly the best answer in my book would be prefaced by saying the best estimate is somewhere in the low to mid 30,000 range.
buffnut453
Captain
Great post. Mr Porter...and all of that occurs without factoring in those occasions when reports are deliberately fudged to hide an unpleasant truth or score a political point. It all comes down to what are we measuring/counting, how we are measuring/counting and why we are measuring/counting....and those all change over time, often in a very short space of time as opposed to the long-view historical perspective.
Valid points which all need to be taken into account when dealing with any kind of data, not just the historical kind. But none of it means that we shouldn't keep looking and trying to establish better numbers to work from.
The post also reminded me how meticulous and accurate the records of the various Einsatzgruppen were, they were used as a basis for the trials of some of the members of those units at Nuremberg. Almost every murdered man, woman and child accounted for. A year or so after the 'Aktions' through the Baltic States, Ukraine, Belorus etc., when the Germans sought to cover their tracks and destroy the evidence of their crimes they knew exactly where the bodies were buried, literally, but they never destroyed the incriminating records.
An extreme example admittedly, but it's why we dig in the archives and why history is an ever evolving discipline. I have books from the 1970s, written with the knowledge of the time, and representing a snap shot of our understanding at the time. Much of their content is as valid today as it was then, but a significant amount is not, having been superseded by developments over the last 40 years or so.
Steve
The post also reminded me how meticulous and accurate the records of the various Einsatzgruppen were, they were used as a basis for the trials of some of the members of those units at Nuremberg. Almost every murdered man, woman and child accounted for. A year or so after the 'Aktions' through the Baltic States, Ukraine, Belorus etc., when the Germans sought to cover their tracks and destroy the evidence of their crimes they knew exactly where the bodies were buried, literally, but they never destroyed the incriminating records.
An extreme example admittedly, but it's why we dig in the archives and why history is an ever evolving discipline. I have books from the 1970s, written with the knowledge of the time, and representing a snap shot of our understanding at the time. Much of their content is as valid today as it was then, but a significant amount is not, having been superseded by developments over the last 40 years or so.
Steve
- Thread starter
- #234
Robert Porter
Senior Master Sergeant
Agreed! My thrust was not so much that we just wring our hands and say oh well, we of course should do all we can to preserve and uncover as much history as we can. My real observation is that sometimes we get bogged down in arguments about whose particular numbers are correct and argue heatedly for and against specific sources when the sources may only differ by a very small amount. To me that is wasted time, energy, and emotion. I think it is our version of arm chair quarterbacking but with a defective instant replay system. Things like assuring men and women that served receive all recognition due them is one of those areas I think we should pursue vigorously, but I am not terribly impressed or worried about exactly what defines a non combat loss vs a training loss vs an accident vs an equipment failure. Since those terms are often redefined over time and since not all units necessarily agreed on how to apply the definitions at any given point of time, including purposely fudged records to get or keep a unit fitness report in acceptable bounds etc. It seems somewhat pointless to dismiss or accept a given author based solely on their use of available data. And often I see an author slammed over a trivial difference in numbers when the rest of their work appears to be well researched and presented.
Just cautioning against painting with a broad brush when the underlying data is suspect and incomplete to begin with. Also new information is constantly coming to light, which while it may cause, as mentioned above, a book written in the 70's to appear to be wrong, it does not mean the rest of the material in the book is instantly questionable or worthless. When I was a young lad growing up I was fortunate to have as a next door neighbor a retired History Chair from SUNY. He even though retired was constantly being sent history books for review etc. He taught me a very valuable lesson about history. It is 90% opinion, 8% bias and standing on 2% of fact. As an example he had me pick a civil war battle, which was his personal area of expertise. Then we went to his extensive library and he would pull out a dozen books written about that battle. He made me draw up notes from each book on specific aspects of the battle, which units were involved, the artillery used, the numbers of troops, the physical layout of the battle field, the commanders, the weather, casualties, etc etc. In the end it was pretty obvious that no 2 books agreed on much of anything. So I asked him, if they are all different which is correct. He just looked at me with a smile and said "None of them... and all of them."
Just cautioning against painting with a broad brush when the underlying data is suspect and incomplete to begin with. Also new information is constantly coming to light, which while it may cause, as mentioned above, a book written in the 70's to appear to be wrong, it does not mean the rest of the material in the book is instantly questionable or worthless. When I was a young lad growing up I was fortunate to have as a next door neighbor a retired History Chair from SUNY. He even though retired was constantly being sent history books for review etc. He taught me a very valuable lesson about history. It is 90% opinion, 8% bias and standing on 2% of fact. As an example he had me pick a civil war battle, which was his personal area of expertise. Then we went to his extensive library and he would pull out a dozen books written about that battle. He made me draw up notes from each book on specific aspects of the battle, which units were involved, the artillery used, the numbers of troops, the physical layout of the battle field, the commanders, the weather, casualties, etc etc. In the end it was pretty obvious that no 2 books agreed on much of anything. So I asked him, if they are all different which is correct. He just looked at me with a smile and said "None of them... and all of them."
Data reliability is also dependent on who compiled it. For example, as published by the Air Ministry in 1947, we can be fairly sure that the casualty figures for RAF Bomber Command are accurate, or as accurate as is reasonably possible.
For those interested, 47,268 were killed in action or died as prisoners of war, 8,195 were killed in flying or ground accidents and 37 died in 'ground battle action'. We also know the number of various wounded and injured, which brings the total number of casualties to 73,741, about 60% of those who served the Command. We can therefore safely say that approximately 85% of Bomber Command casualties were suffered on operations, and 15% in training and other accidents.
I'm actually surprised that the figures for the USAAFs are not so clear cut, I find it hard to believe that the data wasn't kept, maybe it was never processed?
Cheers
Steve
For those interested, 47,268 were killed in action or died as prisoners of war, 8,195 were killed in flying or ground accidents and 37 died in 'ground battle action'. We also know the number of various wounded and injured, which brings the total number of casualties to 73,741, about 60% of those who served the Command. We can therefore safely say that approximately 85% of Bomber Command casualties were suffered on operations, and 15% in training and other accidents.
I'm actually surprised that the figures for the USAAFs are not so clear cut, I find it hard to believe that the data wasn't kept, maybe it was never processed?
Cheers
Steve
Records of wrong data can be as important as the correct figures. For example the LW made decisions based on their incorrect estimates of RAF strength and losses during the BoB. Leigh Mallory also believed in the effect of his big wing based on incorrect claims
I remember an exchange of views on here discussing the courage (or lack of) a German ace pilot who bailed out when surrounded by allied fighters over Germany. With the losses in training and accidents on operations I think all pilots had a profound sense of their own mortality. The books I have read by renowned pilots who were exceptional at flying all conveyed somewhere or other that to survive the war they were above all lucky. Some, indeed many of the 35,000 US airmen and women killed in accidents would have been avoidable given more time training resources etc but some were not, in any case they were all accustomed to their fellow airmen dying long before they met the enemy.
GregP
Major
Actually, Stona, my point above was not to stop doing historical research at all. The point is to either give attention to the entire list or let it go. To pick and choose who you decide to "audit" smacks of bias. By "bias," I mean experimental bias, not emotional bias, though that is sometimes involved, too. Let's look at so-and-so because I want to discredit him or because I want to add to his score.
If you're going to review the record, then review the entire record and hold off on results until it is done, and publish all at once as a study. Also, don't review just the U.S.A. Either review aerial victories for some particular conflict or don't. If so, do the best you can. The rules should be, "if there is no solid evidence otherwise, then WW2 scores stand as recorded. No guessing."
Picking some individual(s) out for review is classic bias and skews the picture, which was complete with WW2 numbers and is now skewed when less than 1% of all awarded U.S. aerial victories receive hot attention but the rest are deemed "OK" without any scrutiny whatsoever. That's NOT OK. It's classic discrimination. By discrimination, I simply mean treating someone different from someone else. If Boyington is OK to look at with a critical eye and a change pencil ready, then everyone else is, too, and should absolutely get the same attention to detail.
If you're going to review the record, then review the entire record and hold off on results until it is done, and publish all at once as a study. Also, don't review just the U.S.A. Either review aerial victories for some particular conflict or don't. If so, do the best you can. The rules should be, "if there is no solid evidence otherwise, then WW2 scores stand as recorded. No guessing."
Picking some individual(s) out for review is classic bias and skews the picture, which was complete with WW2 numbers and is now skewed when less than 1% of all awarded U.S. aerial victories receive hot attention but the rest are deemed "OK" without any scrutiny whatsoever. That's NOT OK. It's classic discrimination. By discrimination, I simply mean treating someone different from someone else. If Boyington is OK to look at with a critical eye and a change pencil ready, then everyone else is, too, and should absolutely get the same attention to detail.
It's an impossible task to review every claim for every pilot, even of one nationality, in a lifetime. For many it's impossible because the material to support such a review does not exist or is lost. Cross referencing claims against enemy losses is a thankless task. It is inevitable that the high scorers and high claimers will fall under the spotlight. Basically nobody cares to investigate an RAF pilot with a 1/2 victory from the BoB, no matter how spurious the award is likely to prove, whereas a Luftwaffe ace with hundreds of claims acknowledged is far more interesting to those that actually care about these things.
ALL claims have to be taken with a huge bowl of salt, I don't see them as a reliable statistical data base for any serious investigation, nor do I pay much attention to them. The only way to gauge how a battle or campaign proceeded is to look at the losses acknowledged by each side. For example, generally the Luftwaffe Quartermaster General's reports give an accurate picture of Luftwaffe losses (though when reported, not when they occurred) as such a report was the only way a unit could acquire a replacement aircraft. Allied victory claims certainly do not. The same applies for other air force's which had equivalent systems.
Cheers
Steve
ALL claims have to be taken with a huge bowl of salt, I don't see them as a reliable statistical data base for any serious investigation, nor do I pay much attention to them. The only way to gauge how a battle or campaign proceeded is to look at the losses acknowledged by each side. For example, generally the Luftwaffe Quartermaster General's reports give an accurate picture of Luftwaffe losses (though when reported, not when they occurred) as such a report was the only way a unit could acquire a replacement aircraft. Allied victory claims certainly do not. The same applies for other air force's which had equivalent systems.
Cheers
Steve
