RE: P-38E/F - Super plane?!?
Posted: Fri Jun 19, 2009 4:26 pm
ORIGINAL: herwin
ORIGINAL: rtrapasso
ORIGINAL: rominet
To my knowlegde (but it is only my knowlegde[8|]), the ratio in Salomons-New Guinea in 42-43 is about 1 allied fighter against 3 jap fighters.
Perhaps, but this counts the P-39, P-40, and even the Wirraway... [8|]
The P-38 was also extensively used as a fighter bomber or even pure bomber (it was sometimes fitted with a Norden bombsight)... the times the "kill ratio" of the P38s in the Pacific/CBI fell to around 3 to 1 was when it got jumped during these types of missions... normally it was much greater. As i have mentioned, even against the Luftwaffe when it was fighting on the wrong side of the odds, its A-to-A ratio was around 3 to 1.
EDIT: Personally in my PBEM, the P-38 might get a 1-1 ratio against Zeros, but i doubt it even does that well... but the war is yet young. [:D]
You might enjoy the attached.
Preliminary results from a regression analysis of World War II air combat records
H R Erwin
[Draft written mid-1978]
I have been attempting to determine the technology-dependent components of World War II fighter combat effectiveness. The measure used in this study is the ratio of exchange rates in combat against identical opponents. It is not clear that this is a transitive relation. Insufficient information is presently available to verify this. [This assumption remains unverified, although the underlying data for this study was suggestive in its general consistency.]
Training, experience and effectiveness of ground control have been factored out of the analysis in the following way. World War II British experience indicates that an effective air defence system tripled the effect of fighter aircraft, and combat experience multiplied a pilot’s effectiveness some 4½ times. Finally, American experience in the Pacific indicates an approximately 4-fold increase in pilot effectiveness in the interval from 100 to 400 hours of flight training. [In other words, combat data in the open literature were adjusted for these factors when known. The data I used actually classified the type of engagement.]
To simplify the analysis, the actual number used [below] is 2 log2(ratio) rather than the ratio itself. This has the effect of linearising the relationships between multiplicative factors. [All factors were multiplicative.]
The nine factors analysed in this study were wing loading, rate of turn, power loading, maximum airspeed, altitude of maximum airspeed, service ceiling, firepower, and maximum rate of climb. Four studies were made. The first study was of overall effectiveness. The significant factors were as follows: [These are tables with three columns, separated by |.]
Factor|Increase of 1|Mean Value
Speed|+22.5 mph|371 mph
Rate of turn|x2|0.19 radians/sec @ 240 ft/sec
Ceiling|+25000 feet|36140 feet
Climb|+7000 feet/minute|2968 feet/minute
The second study was of air superiority aircraft only.
Factor|Increase of 1|Mean Value
Speed|+20 mph|381 mph
Rate of turn|x2.5|0.20 radians/sec @ 240 ft/sec
Altitude of maximum speed|-20000 feet|19500 feet
Ceiling|+13000 feet|37000 feet
Rate of climb|+25000 feet/minute|3200 feet/minute
The third study was of interceptors.
Factor|Increase of 1|Mean Value
Speed|+30 mph|369 mph
Rate of turn|x2.5|0.20 radians/sec @ 240 ft/sec
Altitude of maximum speed|+8000 feet|19770 feet
Rate of climb|+2000 feet/minute|2906 feet/minute
The fourth study was of escort fighters (8 types with limited data)
Factor|Increase of 1|Mean Value
Speed|+28 mph|375 mph
Rate of turn|x4|0.19 radians/sec @ 240 ft/sec
Altitude of maximum speed|-5500 feet|21500 feet
Ceiling|+3000 feet|38000 feet
Climb|+1500 feet/minute|2941 feet/minute
A number of conclusions resulted. First, the easiest way of increasing effectiveness was to increase speed. This was normally handled by decreasing drag losses rather than increasing power-to-weight ratio (i.e., acceleration and climb). Only marginal increases in effectiveness were possible by decreasing wing loading or increasing power. Generally, speed, power and wing loading (in that order) contributed most to the effectiveness of fighters with speed by far the most important.
A note should be included here as to the method used to determine fighter ratings. The set of fighters on which I had [open source] effectiveness data [what shot down what by individual engagement] for World War II was ordered by this data. This resulted in a rather complete partial ordering. [I was able to derive ratios for almost all Allied aircraft against their Axis opponents of the same period and vice versa.] This was converted into a well-ordering by a process of cutting and fitting. [In other words, transitivity was assumed to hold.] Next numerical data [exchange ratios] were used to derive aircraft ratings. The statement in Morse and Kimball that the Spitfire 9 had twice the exchange rate of the Spitfire V provided useful data. Likewise the relative losses of the P-51 and P-47 against the German Lufwaffe provided more data. Finally statements that two aircraft were equal in combat or one was marginally superior (rated as +1) were used. This resulted in a numerical scale. The first regression analysis on the whole population demonstrated the existence of two distinct subpopulations—interceptors and (air superiority) fighters. This led to further analyses. The difference between fighters and interceptors appears to be that fighters fight fighters by making diving attacks. Interceptors don’t fight interceptors, and only fight fighters as an adjunct to making an interception. If a fighter makes a diving attack on an interceptor, it succeeds in stopping the interception only about 20% of the time (when it succeeds in shooting down the interceptor). Thus it must engage in a dogfight. If the interceptor has superior manoeuvrability, it may win enough of these dogfights to hold its own in terms of exchange rate despite a marked inferiority in speed.
Hopefully more data will be available in the near future to allow a significant extension of this study. [It wasn’t.]
[Additional Notes, I don't have the underlying statistical data any more.]
1. The standard aircraft used was the FW190A4, since it fought almost everything Allied during the late period of the war, and there were plenty of open-source data available. It was rated a 11.
2. The Spitfire V was inferior to the Bf109G, which was inferior to the FW190A4, so those were rated 9, 10, and 11.
3. The P-47 had an even exchange ratio against the Bf109G, and the P-51 had a 2-1 exchange rate, so those were rated 10 and 12.
4. The Spitfire V was marginally superior to the A6M3, so that model of the Zero was rated as an 8.
5. The F4U1 was marginally inferior to the P-51, so it was rated an 11.
6. The F6F was somewhat inferior to the F4U1 and definitely superior to the A6M3, so it was rated a 10.
7. The Bf109F, Bf109E, and Spitfire I were brought into the scale by noting that the latter two were an even match, the Spitfire V was a marginal improvement of the Spitfire I, but still superior to the Bf109E. The Bf109F was about equal to the Spitfire V, with superiority only at higher altitudes. Hence the ratings were Spitfire I—8, Bf109E—8, and Bf109F—9.5.
8. The Hurricane IIB was the equal of the A6M3, and the F4F was slightly inferior. So the former was rated an 8, and the latter a 7.
9. From this point on, adding planes involved carefully considered arguments. For example, the P-40B was marginally inferior to the A6M3, so it was rated a 7. The P-39 was definitely inferior to everything on the list, so it was rated a 6. The Hurricane I came in at 7.
10. The F2A was rated a 6 since it was marginally inferior to the F4F. The P-35 was also rated a 6 for similar reasons.
11. European data indicated the P-38 was inferior to the GAF fighters and Pacific data better than the Zero, so it was rated an 8.
12. The Typhoon and Tempest were rated as a match for the FW190A4 and superior to the Bf109G. The C202 was a match for the Spitfire V. The C205V was almost a match for the P-51, so it was rated 11.5. For similar reasons, the Ki100 was rated 11.5.
13. Since good Shinden (N1K1-J) pilots found the F6F a fairly easy kill, despite the speed difference, it was rated a 10.
14. The Ki61 was superior to the faster Bf109E, so it was rated a 9.
15. The Ki43 was marginally inferior to the P-38, so it was rated a 7.
16. The MiG 3 was rated an 8 based on its performance against the early Bf109 models. The Yak 3 was rated a 10 because it was an overall match to the Bf109G.
17. The Bf110 was rated a 6 based on the problems it had with the Hurricane I.
18. The Ta152H was markedly superior to the P-51, so it was rated a 14.
19. There were other data used very carefully to supplement these. For example, interceptors were known to shoot down bombers and escorts in about equal numbers, so the performance of P-40s in northern Australia during 1942 intercepting Japanese attacks (4-1) meant their performance against the escorts was about 2-1. That then had to be adjusted for the superiority of an air defence system (3-1), resulting in the conclusion that the P-40s were slightly inferior to the Zeros, so confirming other analyses.
Very interesting. Thanks.
