Underappreciated Aircraft of WWII

[buffnut453]

imho a damaged aircraft flying all the way back to base where they land but are subsequently written off as being beyond economic repair it's a loss

Yes, it's a loss but it's not a kill. There's a difference. We simply don't have the records to determine which aircraft were deemed beyond economic repair and hence it introduces huge scope for interpretation and subjectivity. I identified the criteria i used in evaluating Mohawk and Hurricane performance against the Ki-43 which I believe is fair, consistent and as objective as it possibly can be.

 

[Nikademus]

good data,
so C.R. 32 was more difficult for Hurricanes that most recent C.R. 42,

and for Mohawk someone has data?

I wouldn't say that. Rather the Cr-42 had a greater level of engagement vs. the older Cr-32. Hurricanes became prevalient closer to the later part of the campaign when Cr-42's were the primary Italian fighter. While intiially fairing well against the new British fighter, the Hurricanes began scoring more as time went on. Good example of how kill ratio estimates can be misleading.

I show two Mohawk IV op losses for East Africa.

 

[Nikademus]

imho a damaged aircraft flying all the way back to base where they land but are subsequently written off as being beyond economic repair it's a loss

I do count immediate write offs due to combat damage as a kill as long as it's clearly documented as such. Similar to a Deadstick landing situation.

 

[buffnut453]

I do count immediate write offs due to combat damage as a kill as long as it's clearly documented as such. Similar to a Deadstick landing situation.

Yes, but that's damned hard to define clearly. I remember a photo of a Spitfire with a broken back following a wheels-up landing which was ultimately returned to service. The repair of damaged aircraft depends hugely on available second and third-line maintenance resources which, again, may confuse the kill/loss picture rather than clarify.

 

[Nikademus]

Yes, but that's damned hard to define clearly. I remember a photo of a Spitfire with a broken back following a wheels-up landing which was ultimately returned to service. The repair of damaged aircraft depends hugely on available second and third-line maintenance resources which, again, may confuse the kill/loss picture rather than clarify.

I agree. Its tricky. It was fully possible to get a "kill" in more than one situation but have the aircraft in question be repaired and returned to service. One way to look at it is that a "kill" is a definition of success, not simply a matter of a destroyed aircraft that never again flies. To use another example,In the European definition, a Kill could and did include instances whereby a pilot surrendered by dropping his gear and landing. The pilot who forced the other pilot to conceed would be awarded the kill or score.

 

[buffnut453]

I agree. Its tricky. It was fully possible to get a "kill" in more than one situation but have the aircraft in question be repaired and returned to service. One way to look at it is that a "kill" is a definition of success, not simply a matter of a destroyed aircraft that never again flies. To use another example,In the European definition, a Kill could and did include instances whereby a pilot surrendered by dropping his gear and landing. The pilot who forced the other pilot to conceed would be awarded the kill or score.

Agreed, hence my definition of a kill being achieved if the victim aircraft doesn't reach a friendly (to him) airfield. In the WWII context, aircraft typically didn't surrender unless they've been fired upon so my first requirement still applies.

 

[Juha]

One more underappreciated a/c, Allison Mustangs. IIRC RAF liked them very much in army co-op work and would have kept all their Mustang I/IA/II sqns running at least to the end of 44 but when Mustang production switched to P-51B and USAAF wanted most of them source of replacement a/c dried up and RAF had to convert many Mustang I/IA/II sqns to Spits because of attrition. IIRC some sqns flew with Allison Mustangs to the end. Anyway, the plane suited well in low level tac recce role.

Juha

 

[kration]

One more underappreciated a/c, Allison Mustangs. IIRC RAF liked them very much in army co-op work and would have kept all their Mustang I/IA/II sqns running at least to the end of 44 but when Mustang production switched to P-51B and USAAF wanted most of them source of replacement a/c dried up and RAF had to convert many Mustang I/IA/II sqns to Spits because of attrition. IIRC some sqns flew with Allison Mustangs to the end. Anyway, the plane suited well in low level tac recce role.

Juha

Good suggestion. I'm impressed that this thread has as many posts as it has, and the forensic discussions above are excellent. But it's suggestions like the Allison Mustangs which go back to the reason i asked the question i.e. forgotten or disparaged aircraft which actually fulfilled a useful role.

In relation to the Allison Mustangs, I've always been a bit puzzled re. the army co-op work. They are often referred to as being good at low level, but were they any better than contemporary aircraft? And I would also have thought that visibility (a prime requirement for army co-op?) wasn't that brilliant in the non-teardrop canopy mustang. What were the characteristics that made them so suitable for army co-op/tec recce role?

 

[Markus]

In relation to the Allison Mustangs, I've always been a bit puzzled re. the army co-op work. They are often referred to as being good at low level, but were they any better than contemporary aircraft? And I would also have thought that visibility (a prime requirement for army co-op?) wasn't that brilliant in the non-teardrop canopy mustang. What were the characteristics that made them so suitable for army co-op/tec recce role?

The were very fast at low/medium altitudes. 390mph@18,000ft in case of the P-51A. They could also fly all the way to the Ruhrgebiet from base in the UK, an early P-47 could not.

 

[JoeB]

Which is still no reason to dismiss such a major battle in which the Hurricane was not only a major contributor, but the majority contributor. (it shot down the most planes on the British side and was present in larger #'s than the Spit though the latter gets all the glory from most accounts)
I'm showing an approximate 1.28:1 exchange ratio in favor of the Luftwaffe overall in just fighter vs. fighter exchanges. (1.33:1 in favor of the UK for total airframe losses)**

"Spitfire Special" quotes BoB Spitfire v Bf109 180:219 (1:1.2), Hurricane v 109 153:272 (1:1.8 ). And those figures are often seen, said to be known air combat losses in known encounters between those types. But, they are not AFAIK as robust as the kind of results we're debating otherwise, where books give all the results of each combat, even though the more detailed works rely on authors' interpretations of the records, and counters' interpretations of the books! The figure you quoted seems closer to the ratio of losses to all causes, which is potentially important of course, but obviously apples and oranges to compare to case by case results of head to head air combat.

I'm not 'dismissing' the BoB but just saying the data is less detailed in published works that I've familar with. But the commonly quoted 1:1.8' isn't all that different from BoF result, albeit very different from Malta result, of course.

And I'll repeat, I have not at any time said the Jagdwaffe's general advantage over British fighters (and everybody else's) in 1939-42 was due entirely to their a/c. That wouldn't be the starting assumption in any case of any air arm's fighter success and would seldom be provable even if true, if the a/c were at all comparable. So that's all a straw man argument, that nobody denies the LW tended to have the advantage in human factors in first half of the war... nobody including me so I don't see the point in mentioning it.The relevance of establishing that the Hurricane's record v Bf109E, while indeed somewhat variable, was in general on the short end of the stick by a considerable margin, is two fold: to illustrate that Hurricane travails in the Far East were worse on average but not outside the range of its record v Bf109E; and to illustrate that the Hawk 75's somewhat better record v Bf109E in the BoF is not being compared to a totally uncharacteristic record for the Hurricane in the BoF, though the Hurricane did do better (and a lot worse) in some other episodes v Bf109E than it did in the BoF.

Joe

 

[buffnut453]

Ok, here's the result of my analysis re Mohawks and Hurricanes over Burma. I hope it will prove interesting, even though slightly divergent from the thread theme. Firstly, the criteria I used to determine a kill was:

"A kill is caused by hostile gunfire from an enemy fighter aircraft which results in the victim aircraft failing to return to a friendly airfield. Details of losses on both sides must be available for the combat to be included in the evaluation."

This was applied to both sides, so an aircraft that crash-landed at its home base was not considered a kill. I used the same period for both the Mohawk and Hurricane (18 June 1942 thru 31 December 1943). I also only considered fighter-on-fighter combat so any kills against bombers or recce aircraft were not counted.

Now for the stats:

Mohawk:

Number of Days of Valid Engagements: 5
Kills achieved by Mohawk: 3
Kills achieved by Ki-43: 6

Hurricane:
Number of Days of Valid Engagements: 30
Kills achieved by Hurricane: 11
Kills achieved by Ki-43: 45

So the Mohawk achieved a kill-to-loss of 0.5:1 whereas the Hurricane's was 0.24:1. On this measure, the Mohawk did indeed perform better (roughly two times better than the Hurricane), and the better manoeuverability of the type (coupled with the performance-sucking Vokes filter on the Hurricane) may well have played a role.

BUT some other interesting statistics came out during the analysis.

Two of the 3 Mohawk kills were achieved when the RAF fighters outnumbered the opposition and had the height advantage (8v6 on 10 Nov 42 and 10v6 on 9 Nov 43), the third kill occurred when the Mohawks were outnumbered 6v15 but still had an altitude advantage (5 Dec 42): during these 3 combats 3 Mohawks were shot down.

The Hurricane combats typically involved an entirely different scale of fighting being routinely outnumbered by a much larger number of Ki-43s: on only 2 dates (26 Mar 43 and 15 May 43) did the Hurricane meet the Ki-43 with a numerical superiority. The Hurricanes also faced larger numbers of their adversary - on only 3 occasions were fewer than 10 Ki-43s engaged (with one instance of 101 Japanese fighters (Ki-43s and A6Ms) being engaged by both Spitfires and Hurricanes). It was not uncommon for a squadron of Hurricanes to meet 2 Sentais of Ki-43s - 30+ Japanese fighters against a dozen from the RAF, with the Japanese having the altitude advantage as the Hurricanes were scrambled (often late) to intercept.

Why were the combats so different? The Hurricanes were deployed right in the forward area to take the brunt of the IJAAF offensive, and so had less warning and their airfields were the targets for concerted IJAAF attacks by bombers and fighters. Although in the same operational theatre, the tactical employment and combat conditions faced by the Mohawks and Hurricanes were markedly different - the Hurricanes were engaged at an entirely different level of effort and, under combat conditions, routinely faced 4-5 times the number of Ki-43s than were engaged by Mohawks.

One final interesting point - on 9 Nov 43 2 Hurricanes shot down a Mohawk which the Hurricane pilot had misidentified as a Japanese fighter. I throw this in as a point of whimsy primarily because it seems that, thankfully, the Mohawk pilot survived.

Which aircraft was really better? I think it's still very difficult to tell. The Mohawk achieved better results but had a much more favourable combat environment. The Hurricane undoubtedly didn't perform as well by the kill/loss metric but it also faced, on a daily basis, considerably more enemy fighters and was frequently at a tactical disadvantage because of its more central forward location.

So, you pays your money and you takes your choice...

Sorry for monopolizing the thread but I hope the above was of interest. Now, back to the real topic which was under-appreciated aircraft. Have we had the Hudson? How about the Sunderland?

 

[rochie]

what about the Curtis C-46 Commando

 

[timshatz]

Good analysis Buff, had a feeling that was the case. That is the problem with the statistics of the situation. Putting too much stock in them without the supporting story makes the Mohawk a better airplane. While I have nothing against the Mohawk, I didn't think the stats rated the superior position given to it by the same stats.

You can rely too much on the numbers in some cases. I have found that out, to my own dismay in the Market. You have to know why the math is the way it is.

Good post.

 

[claidemore]

Buffnut:, I get 1 Ki43 on Nov 10/42 (2 if you count the pilot who died of wounds later).
1 Ki43 on Dec 5,
1 on Feb 16,
2 on April 20 and
1 on Nov 9.
Total of 6 kills by gunfire in 5 engagements. (There was one other engagement on Jan 19/43 where a probable is listed.)
I get two Mohawk losses on Nov 10, 1 Mohawk that crash landed on Dec 5 which according to your system used above would not be counted, and 2 Mohawk losses on April 20 for a total of 4 losses. What losses on what day am I missing?
If I'm not missing anything, that would be a 3/2 kill/loss ratio for the Mohawk?

 

[Nikademus]

=JoeB;658619

And I'll repeat, I have not at any time said the Jagdwaffe's general advantage over British fighters (and everybody else's) in 1939-42 was due entirely to their a/c. That wouldn't be the starting assumption in any case of any air arm's fighter success and would seldom be provable even if true, if the a/c were at all comparable. So that's all a straw man argument, that nobody denies the LW tended to have the advantage in human factors in first half of the war... nobody including me so I don't see the point in mentioning it.

Joe

I cited the Jagdwaffe's reputation sans 1940 because your past discourses re: the Hurricane suggest that there's something about the plane in your opinion that makes it an inferior fighting machine in comparison to other near-contemporaries (allied), such as the F4F based on kill ratio comparisons. My viewpoint, in past discussions and as now, is that the discussed kill ratio estimates are influenced by far more factors as well as plane attributes. You agree with this in principle as you wrote above but where we differ is on the emphasis and on outside conditions. While we don't have specific details on the Hurricane for this battle, the general kill ratio estimates still prove that the plane was competetive and performed admirably even if the 109E had the edge overall in performance, more so given it was the majority player on the RAF side.

From my viewpoint the differing kill ratio estimates in the career of the Hurricane bear out the influence of outside conditions and tacticss as well as pilot experience. The various P-40/Tomahawk/Kittyhawk experiences of the Commonwealth and USAAF also make for an interesting comparison as does relevent comments made by Shores in his Tunisia book (and interviews with current RAF personell listed in the back section of it)

The Hurricane experience in Burma was, and does remain controversial with differing professional opinions on why. I do personally believe that WgCmdr Richey's salient points had more to do with it than the local RAF commanders would admit and don't see why the Hurricane, better employed couldn't have done better.

 

[Nikademus]

Ok, here's the result of my analysis re Mohawks and Hurricanes over Burma. I hope it will prove interesting, even though slightly divergent from the thread theme. Firstly, the criteria I used to determine a kill was:

"So, you pays your money and you takes your choice...

Sorry for monopolizing the thread but I hope the above was of interest. Now, back to the real topic which was under-appreciated aircraft. Have we had the Hudson? How about the Sunderland?

Nice post. As i mussed prior, i feel that had the Mohawk been employed in similar fashion and scale to the Hurr squadrons their experiences would have been similar. From a technical standpoint there's not a huge amount to choose between them, though the Hurr's better speed should theoretically be better able to be capitalized on.

Has the P-39 been added yet? I'm starting to appreciate this plane more as i continue studying it under Russian tutiledge.

 

[JoeB]

1. I cited the Jagdwaffe's reputation sans 1940 because your past discourses re: the Hurricane suggest that there's something about the plane in your opinion that makes it an inferior fighting machine in comparison to other near-contemporaries (allied), such as the F4F based on kill ratio comparisons.

Based on kill ratio comparisons of the Hurricane to other Allied fighters against the same Axis opposition. For example F4F and Hurricane v same or similar Japanese oppostion, Hawk v Hurricane v the same German or Japanese opposition. So it has nothing to do specifically with how Axis air arms stacked up against Allied in respects other than a/c. That's why I see your point about LW and RAF as both obvious and not relevant to my point. The results of Hurricane v other Allies might also be affected, obviously, by differences among Allied air arms (though it would not be in case of Hurricane v Hawk or P-40 types in British/CW service). But there's less reason to believe that Western Allied air arms varied as greatly from one to the next as they did v Axis air arms.

And anyway we can examine how much other factors differed compared to how much the results differed, why not? IMO (I don't disagree with your characterizing it as just my opinion) there's a tendency to overstate the difference in circumstances between say F4F and Hurricane/Spitfire in PTO. And I've backed my opinion with facts about points like what USN/USMC formal tactics really were in 1942, whether there were really consistent big differences in warning times and tactical situations (not as much as some seem to want to believe) etc.

I didn't say I never draw any inference from combat results if *any* other factor besides airplanes differs even a little. I'm just saying your specific point or example, about LW v RAF, is obvious and at the same time not relevant because I'm not trying to draw an inference about Hurricane's v Bf109's, just using results against the Bf109 as a point of comparision, as in:
1. French Hawks did at least as well v Bf109 in 1940 than Hurricane (measureably better but not to a highly statistically significant degree).

2. Hurricane results in PTO are not as far out of line with Hurricane results v the LW as some seem to initially assume. So the Hurricane results in PTO are not as strange and in need of extraordinary explanations as they would be if the Hurricane units really had been typically fully competitive with the German fighter units, which is again seems to the common reputation among some.

To summarize my response re: Mohawk v Hurricane in PTO. I don't disagree that it's limited data nor do I find the specific points and/or nitpicks about the various incidents without any merit (as I posted above, I misread my notes about the Jan 22 case, my mistake on that one; otherwise I would stick by what should count as kill though I respect varying opinions). But, the arguments against considering that sample still contain a form of logical double counting. A relatively small sample can have relatively a lot of noise. That's what we can see, roughly, by calculating the binomial probability of achieving 7:7, in 14 'throws' if the real probability was .5 or .25. It's small in case of .25, but the chances of achieving 70:70 when the underlying probability was .25 would be nanscopic, that's the effect of sample size. But if we examine the sample and identify the incidents we think are the 'noise' then remove them and look at the smaller sample that's left and say 'well wow that's *tiny*, that could have all kinds of noise'...that's double counting the randomness. That sample is big enough statistically, *given how different it is from Hurricane's* to indicate an advantage for the Mohawk. I'll say for probably 5th time, it's not a big enough sample to say the Mohawk's advantage was as big as the difference in the ratio's; it is too small for that. But too small to be at all meaningful? I don't agree.

Joe

 

[buffnut453]

Buffnut:, I get 1 Ki43 on Nov 10/42 (2 if you count the pilot who died of wounds later).
1 Ki43 on Dec 5,
1 on Feb 16,
2 on April 20 and
1 on Nov 9.
Total of 6 kills by gunfire in 5 engagements. (There was one other engagement on Jan 19/43 where a probable is listed.)
I get two Mohawk losses on Nov 10, 1 Mohawk that crash landed on Dec 5 which according to your system used above would not be counted, and 2 Mohawk losses on April 20 for a total of 4 losses. What losses on what day am I missing?
If I'm not missing anything, that would be a 3/2 kill/loss ratio for the Mohawk?

I agree on the following:

1xKi-43 on 10 Nov 42
1xKi-43 on 5 Dec 42
1xKi-43 on 9 Nov 43

I don't count the death of the pilot of the damaged Ki-43 on 10 Nov 42 as a kill because it doesn't meet the criteria I applied (any aircraft that is witnessed leaving the flight with some damage is, at best, classified as a "probable" or, more likely, "damaged"). According to "Air War for Burma" there was no combat involving Ki-43s and Mohawks on 16 Feb 43 - indeed, there isn't even an entry for that date. Also, no Ki-43s were lost in the combat on 20 Apr 43 (although 2 Mohawks were shot down).

I'll stick with my measuring stick - by all means apply yours if it works for you.

Regards,
Mark

 

[buffnut453]

But, the arguments against considering that sample still contain a form of logical double counting. A relatively small sample can have relatively a lot of noise. That's what we can see, roughly, by calculating the binomial probability of achieving 7:7, in 14 'throws' if the real probability was .5 or .25. It's small in case of .25, but the chances of achieving 70:70 when the underlying probability was .25 would be nanscopic, that's the effect of sample size. But if we examine the sample and identify the incidents we think are the 'noise' then remove them and look at the smaller sample that's left and say 'well wow that's *tiny*, that could have all kinds of noise'...that's double counting the randomness. That sample is big enough statistically, *given how different it is from Hurricane's* to indicate an advantage for the Mohawk. I'll say for probably 5th time, it's not a big enough sample to say the Mohawk's advantage was as big as the difference in the ratio's; it is too small for that. But too small to be at all meaningful? I don't agree.

Joe,

I think we're in violent agreement. I was simply seeking to ensure the same measure was applied to both. Small samples can be significant - I absolutely agree. But, given the massive number of variables in air combat, they are also prone to being inadvertently skewed by one or two "unusual" events (like mid-air collisions) which have much less impact on a larger sample size.

My main gripe was the statement that the 2 aircraft operated by the same air arm against the same adversary provided a level playing field for comparison purposes. Tactical situation is pivotal and, again, can skew statistics one way or another. By ensuring a consistent (and relatively objective) definition of what constitutes a "kill" we are able to examine the tactical conditions for each combat against a common frame of reference and hence determine what factors were, or were not, relevant in influencing the outcome.

As always, it's a pleasure jousting with you Joe because you do think and research before you post. I have a tendency to "Ready--Fire--Aim" but I think I'm getting better!

Cheers,
Mark

 

[Nikademus]

Joe,

I think we're in violent agreement. Mark

I love that term.