There was a suggestion that aircraft ROC data might not be perfect. There was a suggestion that Bill Gunston's Combat Aircraft of World War II might help us in that regard. I sent to UK to get one - and it came complete with a wonderful photograph just stuck in it of a Walrus - presumably in a museum - this is a new photograph. It does indeed have more initial ROC data - so I have begun to check it against our data bases. Often we were forced to use "time to altitude" data vice initial ROC data.
The Zero data is unfortunately wrong - possibly a misprint. But I was able to dig out initial ROC data for A6M2 I didn't know I had - and it is indeed 102 feet per minute greater than we now use in RHS. That does NOT change its maneuverability - although it theoretically might have done. We went from about 27.8 to about 28.3 - both of which round to 28 for maneuverability. So we can change the ROC - but it changes nothing else. The A6M2 initial ROC data is 2750 feet per minute. That is 4% greater than we were using - and we may estimate from that other A6Ms probably should be 4% greater than their average ROC data as well. That would make A6M3 2746 fpm. That in turn gives us something I like: the maneuverability (might be better called air combat value) goes to 28.9 = rounds to 29 - I was not happy the 3 was rated the same as the 2 - you traded all that range for no increase in combat value. This will help show that you got a tiny advantage. On the other hand, the A6M5 then would be 2876, and that is NOT enough to increase its maneuverability rating - which is still 30 - but once again we show the most marginal of increases in maneuverability traded for a significant drop in range - and that is right. I like it. OOPS - Gunston HAS the A6M5 - at 3150 fpm. SO with that - we get maneuverability = 32 - which gives one a sense of why this model was so mass produced.
The Pete is different. Using initial ROC instead of average ROC should gain us one maneuverability point - to 19 from 18.
And as she was very maneuverable - and played fighter - that may be worth doing. Initial ROC should be 1969 fpm.
This process will take some time - and I am not inclined to retrofit it to Levels 5 and 6 (because they are suspended - but IF we change them for some other reason - the aircraft file is identical - and easy to include). I don't think we should implement the data until ALL planes are reviewed. But by posting my findings - we might attract still more initial ROC data in some cases.
Aircraft ROC Review
Moderators: wdolson, Don Bowen, mogami
-
el cid again
- Posts: 16983
- Joined: Mon Oct 10, 2005 4:40 pm
RE: Aircraft ROC Review
A6M7 is a different kettle of fish: it isn't a fighter any more - but a fighter bomber - trading the centerline drop tank for a 250 kg bomb. It has armor as well. BUT NO more power for all the added weight. It also carried two wing drop tanks - which were absent before. These will increase the logistical cost of using the aircraft - which I like. Using 3150 fpm does not increase the maneuverability value - which remains 31. [Gunston gives that for the 6c - and a 7 is a 6c with a different bomb rack] But endurance drops 100 minutes because we put on 2 50 gallon drop tanks (adding 660 pounds to load). Normally it carried 40 gallon tanks - which we don't have - but it could carry even 77 or 84 gallon tanks - so this is a compromise. This aircraft is much heavier and slower than the A6M5 due to having the same horsepower - so the air combat value (maneuverability) rating is less. But it was not intended to be a fighter in the same sense previous Zeros had been - rather an attack aircraft and secondary fighter - escorted by still better fighters. This rating seems reasonable.
A6M8 is a data nightmare. We didn't have a reasonable altitude climb data - and were forced to use average ROC to 19,685 feet (= 6000 meters). I estimated based on A6M7 data, and came up with 3341 fpm. Now MAYBE the rate was 4500 fpm - if the Gunston A6M2 data is not a misprint it must be for A6M8. Using 3341 we get a maneuverability rating of 33. This aircraft also adds 2 x 50 gallon drop tanks, loses 100 minutes of endurance (made up by them), and increases its load by 660 pounds (increasing its logistic cost). IF we assume the 4500 fpm value is right - that would add 6 more to the maneuverability value - which sounds excessive. But we will do that if it can be confirmed. Note that the A6M8 is one of those aircraft present in standard RHS scenarios but NOT in EOS (or AIO) - so EOS people - don't be looking for it.
A7M2 has initial ROC of 3621 fpm - previously we had to estimate from a 10,000 meter base - and average ROC to 10,000 m is never close to initial ROC. This changes the maneuverability rating to 36 - and shows why JNAF wanted and needed it. It could carry 2 x 77 gallong drop tanks - but these are too big for our code to handle - so I fitted 2 x 50 gallon tanks - dropping endurance by 100 minutes (which gets added back by the tanks - but it makes extended range missions cost more supply points - which is right).
A6M8 is a data nightmare. We didn't have a reasonable altitude climb data - and were forced to use average ROC to 19,685 feet (= 6000 meters). I estimated based on A6M7 data, and came up with 3341 fpm. Now MAYBE the rate was 4500 fpm - if the Gunston A6M2 data is not a misprint it must be for A6M8. Using 3341 we get a maneuverability rating of 33. This aircraft also adds 2 x 50 gallon drop tanks, loses 100 minutes of endurance (made up by them), and increases its load by 660 pounds (increasing its logistic cost). IF we assume the 4500 fpm value is right - that would add 6 more to the maneuverability value - which sounds excessive. But we will do that if it can be confirmed. Note that the A6M8 is one of those aircraft present in standard RHS scenarios but NOT in EOS (or AIO) - so EOS people - don't be looking for it.
A7M2 has initial ROC of 3621 fpm - previously we had to estimate from a 10,000 meter base - and average ROC to 10,000 m is never close to initial ROC. This changes the maneuverability rating to 36 - and shows why JNAF wanted and needed it. It could carry 2 x 77 gallong drop tanks - but these are too big for our code to handle - so I fitted 2 x 50 gallon tanks - dropping endurance by 100 minutes (which gets added back by the tanks - but it makes extended range missions cost more supply points - which is right).
-
el cid again
- Posts: 16983
- Joined: Mon Oct 10, 2005 4:40 pm
RE: Aircraft ROC Review
G4M1 Initial ROC was 1800 fpm. This is enough to change maneuverability to 10 (up from 9) in spite of the effect of two engines on the formula.
G4M2m22 initial ROC was 1380 fpm. This is also a dramatic change from using average ROC to 8000 m (26,245 ft) - so it is also enough to increase maneuverability to 9 (up from 8).
G4M3m34 (found ONLY in EOS and AIO) initial ROC is also 1380 fpm. Again, this is sufficiently greater than using average ROC to 7000 meters (22,965 feet) that the maneuverability increases to 10 (up from 9). Production was set back 5 months from 5/44 to 10/44.
G4M2e (The Okha missile Betty) is estimated from G4M2 and weight data to have an initial ROC of 1279 fpm. That increases maneuverability from 7 to 8. It is still a dog.
G3M2. Gunston does not give any ROC for this aircraft. But I was able to find data for climb to 3000 meters and climb to 4000 meters. From this, and the G4M1 ratio of 1.04 between average ROC and initial ROC, I estimated the value at 1255 fpm - yielding a maneuverability value of 8 (up from 7).
A5M4 initial ROC is 2790 fpm. This is not enough to change the maneuverability rating (being only 45 fpm greater than the value we were using).
Ki-30 initial ROC is 1640 fpm. This should not be enough to change the maneuverability rating, but that was apparently wrongly calculated initially - it increases to 19 (up from 14, but it should have been 18). This aircraft also represents the Ki-32 of almost identical performance, for which no ROC data is given (the Ki-32 could carry 50 kg more bombs, but this is not represented). The aircraft is given 8 x 100 pound clusters each representing 3 x 15 kg bombs. This is also right: the bomber could carry 24 x 15 kg bombs. It might alternatively carry 4 x 100 kg bombs.
Gunston gives the Ki-15 initial ROC at "about 1640 fpm" - but we use the C5M2 - which if nominally similar - had a much better ROC (3000 meters in 3 min 58 sec vice 6 min 49 sec). I estimate initial ROC from this at 2380 fpm, which leaves maneuverability at 26. [This aircraft is not in EOS or AIO]
J2M2 initial ROC is 3610. This is enough to increase maneuverability from 34 to 35.
A6M2-N initial ROC is not given by Gunston. But using climb to 3000 meters and 6000 meters data from Mikesch (Zero) and Francillon, it is possible to estimate initial ROC at 2592 fpm. This is not enough greater than average ROC data (2440 fpm) to change the maneuverability, which remains 24.
Ki-21 II initial ROC is 1640 fpm. This raises maneuverability from 10 to 11.
G4M2m22 initial ROC was 1380 fpm. This is also a dramatic change from using average ROC to 8000 m (26,245 ft) - so it is also enough to increase maneuverability to 9 (up from 8).
G4M3m34 (found ONLY in EOS and AIO) initial ROC is also 1380 fpm. Again, this is sufficiently greater than using average ROC to 7000 meters (22,965 feet) that the maneuverability increases to 10 (up from 9). Production was set back 5 months from 5/44 to 10/44.
G4M2e (The Okha missile Betty) is estimated from G4M2 and weight data to have an initial ROC of 1279 fpm. That increases maneuverability from 7 to 8. It is still a dog.
G3M2. Gunston does not give any ROC for this aircraft. But I was able to find data for climb to 3000 meters and climb to 4000 meters. From this, and the G4M1 ratio of 1.04 between average ROC and initial ROC, I estimated the value at 1255 fpm - yielding a maneuverability value of 8 (up from 7).
A5M4 initial ROC is 2790 fpm. This is not enough to change the maneuverability rating (being only 45 fpm greater than the value we were using).
Ki-30 initial ROC is 1640 fpm. This should not be enough to change the maneuverability rating, but that was apparently wrongly calculated initially - it increases to 19 (up from 14, but it should have been 18). This aircraft also represents the Ki-32 of almost identical performance, for which no ROC data is given (the Ki-32 could carry 50 kg more bombs, but this is not represented). The aircraft is given 8 x 100 pound clusters each representing 3 x 15 kg bombs. This is also right: the bomber could carry 24 x 15 kg bombs. It might alternatively carry 4 x 100 kg bombs.
Gunston gives the Ki-15 initial ROC at "about 1640 fpm" - but we use the C5M2 - which if nominally similar - had a much better ROC (3000 meters in 3 min 58 sec vice 6 min 49 sec). I estimate initial ROC from this at 2380 fpm, which leaves maneuverability at 26. [This aircraft is not in EOS or AIO]
J2M2 initial ROC is 3610. This is enough to increase maneuverability from 34 to 35.
A6M2-N initial ROC is not given by Gunston. But using climb to 3000 meters and 6000 meters data from Mikesch (Zero) and Francillon, it is possible to estimate initial ROC at 2592 fpm. This is not enough greater than average ROC data (2440 fpm) to change the maneuverability, which remains 24.
Ki-21 II initial ROC is 1640 fpm. This raises maneuverability from 10 to 11.
-
el cid again
- Posts: 16983
- Joined: Mon Oct 10, 2005 4:40 pm
RE: Aircraft ROC Review
Ki-46 II initial ROC is "about 1970 fpm" - enough to increase maneuverability to 13 (up from 12).
Ki-46 III initial ROC is given as the same, but it had a slower average ROC by 12.7%, so I estimate it at 1748 fpm.
That is still enough to increase maneuverability to 15 (up from 14), in spite of the impact of 2 engines on the formula.
Ki-67 initial ROC is 1476. This is not much higher than the average ROC we were using of 1361, but for some reason maneuverability now calculates at 11 vice 10. [The old value turns out to be right for that data, and the new one for this data]
All the above aircraft in this thread so far were Mitsubishi machines.
Ki-46 III initial ROC is given as the same, but it had a slower average ROC by 12.7%, so I estimate it at 1748 fpm.
That is still enough to increase maneuverability to 15 (up from 14), in spite of the impact of 2 engines on the formula.
Ki-67 initial ROC is 1476. This is not much higher than the average ROC we were using of 1361, but for some reason maneuverability now calculates at 11 vice 10. [The old value turns out to be right for that data, and the new one for this data]
All the above aircraft in this thread so far were Mitsubishi machines.
RE: Aircraft ROC Review
Wonder why you want a more imperfect way to rate ROC.
ROC is a wrong way to rate planes. Time tTO (at least to 5000-6000m) is much better due to the fact that are planes that arent good at altitude. ROC dont shows that since is measured at sea level. Time To shows.
ROC is a wrong way to rate planes. Time tTO (at least to 5000-6000m) is much better due to the fact that are planes that arent good at altitude. ROC dont shows that since is measured at sea level. Time To shows.
-
el cid again
- Posts: 16983
- Joined: Mon Oct 10, 2005 4:40 pm
RE: Aircraft ROC Review
Nakajima aircraft in RHS in this box:
B5N2 (representing also B5M1 and B5N1) had an inital ROC of 1378 fpm. This does not change maneuverability, which remains 16. B5N2-Q is similar - slightly heaviar - initial ROC estimated at 1375 fpm - and again this does not change maneuerability (nor would any value down to the value we were using - since it was also 16). The 2-Q is only found in EOS and AIO.
B6N2 initial ROC is 1885 fpm. This is enough to increase maneuverability to 25 (up from 22). If only this machine had good forward armament (it does not) it would not be a horrible secondary fighter (as many attack aircraft in history have been).
C6N1 (recon) and C6N1-S (night fighter) have initial ROC of 2512 fpm. This does not change the outstanding maneuverability for these machines of 29. C6N1 is the only specialized carrier recon aircraft in the world war II era.
Ki-27 is listed as having an initial ROC of 2953 fpm. That must be the Ki-27a, because the Ki-27b average ROC is higher, at 3057 fpm. We have learned how to estimate IROC, so we can now determine it was about 3179 fpm for the Ki-27b. That is sufficient to increase manuverability rating to 29. Gunston says (incorrectly) the Ki-27 may be the most maneuverable of all Japanese fighters - and speculates the greatest of any country in any age. This must mean he is only familiar with WWII and later aircraft - and never looked at the Ki-10 - itself the reason the Ki-27 was preferred by JAAF. The Ki-27 is ALMOST as maneuverable as a Ki-10! It was just too old for the war, serving only in secondary roles like training and home defense when the war began, and we lack the slots to include it.
KI-43 II is listed with an initial ROC of 3250 - significantly higher than the 2821 average ROC we were using. That bumps maneuverability by 2 to 31, and helps model the dramatic value of the Ki-43 II (which I think was the most produced IJA fighter - or plane of any kind). But the Ki-43 I is not listed, is unlikely to have an initial ROC 15% higher than what we are using, and will take some effort to figure out.
EDIT: Wow! The Ki-43 I actually climbs almost 6% faster than the Ki-43 II does! [Less weight I guess] The II reached 5000 meters in 5 minutes and 49 seconds, but the I could reach that altitude in 5 minutes and 30 seconds. From this, and the ratio between average ROC and initial ROC for the II, I can estimate the initial ROC for the I very closely: I get 3437 fpm. That means its maneuverability value also increases to 31 - also up from 29. So they remain as they were - relatively the same maneuverability rating - but both jumped 2 points - which is unusual. This helps explain in simulation how IRL "the Oscar was almost as great a technical surprise as the Zero." Our A6M2 Zero has a maneuverability value of 28 - similar but not quite as large - but better armament - and more range. And the Oscar now indeed seems to continue the JAAF preference for extream mineauverability (begun with the Ki-10, continued with the Ki-27, at the expense of everything else - structure - weapons - range). Gunston does say "it could outmaneuver every aircraft ever ranged against it" - and its unique flaps once caused a proposal we give it a special maneuverability bonus. At the moment - this does not appear entirely necessary.
Ki-44 II initial ROC was 3832 fpm. That causes maneuverability to increase to 36, up from 33. This helps explain why the Japanese preferred it to the Me-109E (against which it flew in competition) - and why Francillon thinks it was a better interceptor. Speed, ROC and firepower make it an outstanding interceptor. But the III is a problem: not many were built and there isn't much data to work with. Gunston cites an instance (19 Feb 1945) on which a "small number" of IIc intercepted "120 B-29s" - and destroyed ten. [EDIT: I was able to confirm the tale - using a different source: B-29 reports admit a force of 119 B-29s on that date - over that target (Musashino plant of Nakajima) - lost six to a small force of Shoki's. Just over a 5% loss rate to interceptors alone is severe by US wartime standards.]
After determining the Ki-44 III had a new wing area of 204.504 sq ft and a new power plant of 2000 hp, I was able to estimate an initial ROC of 4134 fpm - and to calculate from that a maneuverability rating of 38 (up from 34).
Ki-49 II (found only in EOS & AIO) has an initial ROC of 1312 fpm. This does not change the maneuverability rating of 10.
Using the ratio of average ROC to initial ROC from the II model for the I, we get 1249 fpm for the I. This corresponds to a maneuverability value of 9 (down from 10) - and since ROC went up - we must have had a wrong value before. Since the maneuverability of the Ki-21 is 11 - you can see why the Ki-49 was a mixed bag in comparison. It had marginally better range and defensive armament, and armor - but it was not as fast nor as maneuverable. Since that was all true IRL, it is nice to see it in the data.
Ki-84 I had an initial ROC of 3600 fpm, dramatically greater than the average ROC of 2780 fpm we were using. This results in a maneuverability rating of 36 - a gigantic improvement - as you would expect. This puts it in the league of the Ki-44 II and most of the other first rate Japanese fighters - but it is also a serious bomb carrier - a nice plane. Gunston says a captured one easily "outclimbing and outmaneuvering a P-51H and a P-47N!" [So much for the Americans were always better - later in the war - school of thought] "First batches went to China, where the 22nd Sentai flew rings around Gen Chennault's 14th Air Force."
B5N2 (representing also B5M1 and B5N1) had an inital ROC of 1378 fpm. This does not change maneuverability, which remains 16. B5N2-Q is similar - slightly heaviar - initial ROC estimated at 1375 fpm - and again this does not change maneuerability (nor would any value down to the value we were using - since it was also 16). The 2-Q is only found in EOS and AIO.
B6N2 initial ROC is 1885 fpm. This is enough to increase maneuverability to 25 (up from 22). If only this machine had good forward armament (it does not) it would not be a horrible secondary fighter (as many attack aircraft in history have been).
C6N1 (recon) and C6N1-S (night fighter) have initial ROC of 2512 fpm. This does not change the outstanding maneuverability for these machines of 29. C6N1 is the only specialized carrier recon aircraft in the world war II era.
Ki-27 is listed as having an initial ROC of 2953 fpm. That must be the Ki-27a, because the Ki-27b average ROC is higher, at 3057 fpm. We have learned how to estimate IROC, so we can now determine it was about 3179 fpm for the Ki-27b. That is sufficient to increase manuverability rating to 29. Gunston says (incorrectly) the Ki-27 may be the most maneuverable of all Japanese fighters - and speculates the greatest of any country in any age. This must mean he is only familiar with WWII and later aircraft - and never looked at the Ki-10 - itself the reason the Ki-27 was preferred by JAAF. The Ki-27 is ALMOST as maneuverable as a Ki-10! It was just too old for the war, serving only in secondary roles like training and home defense when the war began, and we lack the slots to include it.
KI-43 II is listed with an initial ROC of 3250 - significantly higher than the 2821 average ROC we were using. That bumps maneuverability by 2 to 31, and helps model the dramatic value of the Ki-43 II (which I think was the most produced IJA fighter - or plane of any kind). But the Ki-43 I is not listed, is unlikely to have an initial ROC 15% higher than what we are using, and will take some effort to figure out.
EDIT: Wow! The Ki-43 I actually climbs almost 6% faster than the Ki-43 II does! [Less weight I guess] The II reached 5000 meters in 5 minutes and 49 seconds, but the I could reach that altitude in 5 minutes and 30 seconds. From this, and the ratio between average ROC and initial ROC for the II, I can estimate the initial ROC for the I very closely: I get 3437 fpm. That means its maneuverability value also increases to 31 - also up from 29. So they remain as they were - relatively the same maneuverability rating - but both jumped 2 points - which is unusual. This helps explain in simulation how IRL "the Oscar was almost as great a technical surprise as the Zero." Our A6M2 Zero has a maneuverability value of 28 - similar but not quite as large - but better armament - and more range. And the Oscar now indeed seems to continue the JAAF preference for extream mineauverability (begun with the Ki-10, continued with the Ki-27, at the expense of everything else - structure - weapons - range). Gunston does say "it could outmaneuver every aircraft ever ranged against it" - and its unique flaps once caused a proposal we give it a special maneuverability bonus. At the moment - this does not appear entirely necessary.
Ki-44 II initial ROC was 3832 fpm. That causes maneuverability to increase to 36, up from 33. This helps explain why the Japanese preferred it to the Me-109E (against which it flew in competition) - and why Francillon thinks it was a better interceptor. Speed, ROC and firepower make it an outstanding interceptor. But the III is a problem: not many were built and there isn't much data to work with. Gunston cites an instance (19 Feb 1945) on which a "small number" of IIc intercepted "120 B-29s" - and destroyed ten. [EDIT: I was able to confirm the tale - using a different source: B-29 reports admit a force of 119 B-29s on that date - over that target (Musashino plant of Nakajima) - lost six to a small force of Shoki's. Just over a 5% loss rate to interceptors alone is severe by US wartime standards.]
After determining the Ki-44 III had a new wing area of 204.504 sq ft and a new power plant of 2000 hp, I was able to estimate an initial ROC of 4134 fpm - and to calculate from that a maneuverability rating of 38 (up from 34).
Ki-49 II (found only in EOS & AIO) has an initial ROC of 1312 fpm. This does not change the maneuverability rating of 10.
Using the ratio of average ROC to initial ROC from the II model for the I, we get 1249 fpm for the I. This corresponds to a maneuverability value of 9 (down from 10) - and since ROC went up - we must have had a wrong value before. Since the maneuverability of the Ki-21 is 11 - you can see why the Ki-49 was a mixed bag in comparison. It had marginally better range and defensive armament, and armor - but it was not as fast nor as maneuverable. Since that was all true IRL, it is nice to see it in the data.
Ki-84 I had an initial ROC of 3600 fpm, dramatically greater than the average ROC of 2780 fpm we were using. This results in a maneuverability rating of 36 - a gigantic improvement - as you would expect. This puts it in the league of the Ki-44 II and most of the other first rate Japanese fighters - but it is also a serious bomb carrier - a nice plane. Gunston says a captured one easily "outclimbing and outmaneuvering a P-51H and a P-47N!" [So much for the Americans were always better - later in the war - school of thought] "First batches went to China, where the 22nd Sentai flew rings around Gen Chennault's 14th Air Force."
-
el cid again
- Posts: 16983
- Joined: Mon Oct 10, 2005 4:40 pm
RE: Aircraft ROC Review
ORIGINAL: Dili
Wonder why you want a more imperfect way to rate ROC.
ROC is a wrong way to rate planes. Time tTO (at least to 5000-6000m) is much better due to the fact that are planes that arent good at altitude. ROC dont shows that since is measured at sea level. Time To shows.
Well - actually - we don't want "a more imperfect way to rate ROC" - and I posted a week ago a query if we might not better choose average ROC (I suggested to 10,000 feet, which is widely available for Allied planes) vice initial ROC. Not one person (including you) said you preferred that. We agreed that more consistent ROC data would be better - and IF there was a CONSENSUS in the Forum (vice two voices, yours and mine) saying average ROC to a medium altitude was better - I would adjust all the initial ROC data we have now to that standard - and as well all the average ROC to other altitudes to the selected medium altitude (or maximum in the rare case the plane does not reach the selected standard).
But I don't quite agree that "ROC is the wrong way to rate planes." Which I am translating to mean "initial ROC is the wrong way to rate planes." First, it seems to be the WITP standard. Second, it is the right way to rate them in two cases, critical to WITP air modeling:
1) The pure interceptor case: a plane taking off from the runway to intercept an incoming raid naturally is going to use its initial ROC - and surely that is the one case we can be sure the field will be used; We don't actually know if it is used in any other instance;
2) In a dogfight, if it is extended and if neither aircraft has won, the natural course of events is to lose altitude, until the ground or the sea is reached. As a dogfight extends in time, both planes will tend to get closer and closer to the altitude at which initial ROC is indeed a significant factor.
I had forgotten (2) when I made my query - but thinking about it - I think the system as it is may be better than the other option. What I had in mind (and possibly what you have in mind) is that the average rate of climb to operational altitudes (which seem to be 6000 to 10000 feet in terms of stock settings) may more generally be the case most of our aircraft are in most of the time. But - thinking it through - we don't actually know that ROC is EVER used for a plane not actually taking off from the runway. And - probably - it is only used INDIRECTLY - in the case of air combat - IF it is a part of the maneuverability value. [It appears that maneuverability in stock may have been based - in the single engine case - on speed and ROC - but it might be only speed. For 2 engines the value was divided by 2. For four engines it was divided by 8 in stock and CHS - but only by 4 in RHS. RHS added - well subtracted - wing loading and power loading as well.]
There is a third suggestion - I have no clue why? - that ROC might be used to see if an aircraft targeted by AAA can evade it? Anyway - it does not appear to matter what average ROC is - it may apply most of the time - but when is it used in our model?
Now this is as close to difinitive as I can be: I hope it is clear I do not know exactly what is used when - and that neither does anyone else know. Even if we were reading code, until we read every routine related to air combat, and maybe also AAA combat, we would not know where else it was used? The epistomology of this (how I know what I know) is mostly related to testing and to what programmers have said (i.e. that air combat as such is dominated by maneuverability field) - and to common sense. That means that actual information (as opposed to pure assertion) which would clarify these matters would always be welcome and, if germane, used. Further - even if you completely accept my understanding of what is used in what way where - but you have a different view of why we might prefer to use average ROC data vice initial ROC data - I would consider why you thought so? I am proceeding as a technician: do the best we can for now. And that by definition must mean "in the context of my present understanding." I am very open to modifying my understanding for any reason other than someone has an emotional attraction for this or that. An actual rationale is of interest to me. It should by now be clear I have no emotional stake - no matter how many hours are invested - in any product I have worked on (or adopted or adapted from others). I will shamelessly throw out my past database - as I am now doing - for anything that is even just a slight bit better. I am ending RHS as I began it: with aircraft data. It has been reviewed extensively in between at least ten times. And while I think this is the "end" - until we get more slots - I will instantly reverse that opinion if you (or anyone else) give me any sound reason to do so. I do not "own" this data - and I constitutionally seek the best possible modeling of it in the context of the technical limitations of our model. I have tried (and perhaps sometimes succeeded) to even improve on the model as designed - where that is possibly via data manipulation. [I also have suggested possible code changes, but I cannot implement those.] Here my intent was to explain "why" - but I also am attempting to say if there is a technical reason you don't like my "why" - you should feel free to go into the grounds for that. This is an unnecessary public thread for two reasons:
1) To generate better data (or in this case possibly a better approach to using data)
2) To build what Joe Wilkerson calls "credibility" for the data set - and I suppose to show off how rich it is - with all the planes and even plane types we have added.
Constructive comments are welcome - as they always are. In my world, there is no ego associated with data - it is just data. All that matters is to get the best data possible, and to model it in the best way we can dream up. If you dream better than I do, quoting Jim Dunnigan, "good game designers shamelessly steal ideas from others." I only object to his use of the term "steal" - I think one should actually give credit to the source - and I hope everyone I borrowed ideas from feels recognized.
-
el cid again
- Posts: 16983
- Joined: Mon Oct 10, 2005 4:40 pm
RE: Aircraft ROC Review
As an afterthought, I think one reason this very simple air combat model may work so well is that, while ROC essentially declines with altitude, potential energy due to altitude starts and zero and increases with altitude. So the ROC value to air combat - if it is thought of as combined with potential energy - does not decline nearly as sharply as ROC alone does. Rather in some cases the sum may be almost constant, and in others a whole lot closer to a constant value than one approaching zero would be. For this reason, it may be that representing air combat at different altitudes is better done this way than ever one might guess. [I would prefer, of course, to track everything in much more detail, and figure it out at every altitude.]
Another aspect of this is that ROC itself is somewhat related to speed and wing area. That is, if a given aircraft had more power than it does, it would also have a greater ROC. Similarly, if a given aircraft had more wing area than it does, it would also have a greater ROC. So when the ROC field is greater, and to the extent this is used directly or indirectly, it is telling the code "this is the better combat aircraft" in proportion to how much greater the value is. It may be that initial ROC data - which maximizes the difference between aircraft - is a better indicator of which is better than average ROC data would be?
Another aspect of this is that ROC itself is somewhat related to speed and wing area. That is, if a given aircraft had more power than it does, it would also have a greater ROC. Similarly, if a given aircraft had more wing area than it does, it would also have a greater ROC. So when the ROC field is greater, and to the extent this is used directly or indirectly, it is telling the code "this is the better combat aircraft" in proportion to how much greater the value is. It may be that initial ROC data - which maximizes the difference between aircraft - is a better indicator of which is better than average ROC data would be?
RE: Aircraft ROC Review
While not every plane can turn on a dime, and not every plane is fast enough to "shoot and skoot", ALL planes must lose altitude to make any turns whatsoever. This is a principal of aerial physics.
The plane that loses altitude less by regaining altitude in quicker climbs has a major advantage, enough so that the over-all combat survivability is affected.
As Sid stated, ROC also affects an interceptors' ability to get up with oncoming enemy bombers.
The plane that loses altitude less by regaining altitude in quicker climbs has a major advantage, enough so that the over-all combat survivability is affected.
As Sid stated, ROC also affects an interceptors' ability to get up with oncoming enemy bombers.
-
el cid again
- Posts: 16983
- Joined: Mon Oct 10, 2005 4:40 pm
RE: Aircraft ROC Review
This section contains Kawasaki aircraft:
Ki-45 (all versions) are given as 1300 fpm by Gunston. He says source material "widely differs" and I concur. I show average ROC at 2610.8 fpm to 5000 meters (16,405 feet). IF that is correct, initial ROC for the Ia should be about 2715 fpm. That would not change our maneuverability rating of 14. It should be about 25 fpm less for the Ib, which would leave the maneuverability rating unchanged. From average ROC data, we know the Ic could climb about 11.4% slower than the Ib, so we estimate its initial ROC at about 2406 fpm, less than the value we had been using. Since the average ROC was 2345 fpm, and that is in the right range for initial ROC, we will go with it. That yields a maneuverability rating of only 13 - not good enough for a night fighter - but since the aircraft was marginal - it may be a good rating in this case. It is down from 14 before. [This aircraft is not found in EOS or AIO]
Ki-45 II (found only in EOS and AIO) - a two seat precursor to the Ki-96 - probably had an initial ROC around 3380 fpm, an empty equipped weight of 10,471 pounds and a gross weight of 13,668 pounds. If the maximum speed was about 370 mph, that yields a maneuverability rating of 17, up from 15. This aircraft may be thought of as a high performance Ki-45, and it is not a night fighter. The higher performance comes from using larger engines. The aircraft was added to fulfill a percieved need for a high performance attack aircraft without waiting far into the war to get it - and players who don't want to spend the HI points for a 2 engine attack plane should not build it. The Ki-96 - the variant officially approved - deleted the rear seat and machine gun - and in the event took too long to develop - but weighed slightly less and had slightly better performance. The Ki-96 in turn was cancelled - because the JAAF decided to reinstate the rear seat - and so the project went on to become the Ki-102 - which is in all forms of WITP.
Gunston gives no data for either Ki-48 ROC. And it is clear we are using average ROC data to 5000 m (16405 feet). We know from this the Ki-48 II took 30 seconds less to reach that altitude, so we know their relative values. We can estimate initial ROC at 1895 fpm for the I and 2007 for the II. This does not change the maneuverability ratings of 11 and 12, respectively.
Ki-61 I and 2 are both credited with 2200 fpm by Gunston. Something is very wrong here: this is well below the average ROC - and they aren't the same as each other. The II could climb to 5000 meters in exactly the same time as the Ki-100 I could (6 minutes) - so I gave it the same initial ROC. The I could reach that altitude 30 seconds faster - so I give it 3578 fpm. This results in a maneuverability rating of 34 for the I and 33 for the II. [Both aircraft are not present in EOS or AIO] You can see why these aircraft were highly regarded, by both sides, in a dog fight.
Ki-100 I initial ROC was 3280 fpm. That causes a maneuverability rating of 32 (up from 30). Note the progression downward for the family aircraft Ki-61 I (34), Ki-61 II (33), Ki-100 I (32). As aircraft gain speed and weight, but do not increase wing area, wing loading increases, and sometimes maneuverability suffers slightly.
Gunston also gives no data for Ki-102b ROC. Here NO ONE gives initial ROC (normal for Japanese aircraft) and almost no one for average ROC. But Francillon gives us quite good data for the b and c. He lists average ROC at 2377 fpm, and that should correspond to about 2472 fpm. But RHS - inherited from CHS - inherited from stock - is using 2750! Now that is awfully round - and sounds like a guess. Unless there is some indication it is correct, I think we should go with a value closer to the average ROC to 5000 meters. Using my estimate, we must reduce the maneuverability factor to 14 (from 15). There also was a night fighter variant that nearly saw service - and if we get a slot we might want to think about adding it. There was also a high altitude version which saw limited service. Suspecting we might overrate the altitude, I checked, and found it slightly underrated for the RHS system (80% for non-turbosupercharged engines). The high altitude model would have a considerably higher operational ceiling, and a completely revised armament. Again - if we only had a slot for it.
These four are Kawanishi aircraft:
H6K4 initial ROC was 1222 fpm. This value does not change the maneuverability factor of 4 - which occurs because we divide 14 by the number of engines for a 4 engine plane. It isn't likely ANY 4 engine aircraft will change its durability value due to this review, but I wanted to check a few cases to make sure.
H6K2-L (a variant of H6K4 in spite of its name) had an initial ROC of about 1197 fpm. The maneuverability factor should have been 4 (but was 5) and it still should be 4 with this value.
H8K2 initial ROC was 1673 fpm. This does not change maneuverability = 5.
H8K2-L initial ROC was 1286 fpm. This does not change maneuverabilty = 4. The transports run heavier than the patrol planes do, and in particular that applies to this one, which had a normal load about 3 times greater than the patrol plane's normal weapons weight.
Ki-45 (all versions) are given as 1300 fpm by Gunston. He says source material "widely differs" and I concur. I show average ROC at 2610.8 fpm to 5000 meters (16,405 feet). IF that is correct, initial ROC for the Ia should be about 2715 fpm. That would not change our maneuverability rating of 14. It should be about 25 fpm less for the Ib, which would leave the maneuverability rating unchanged. From average ROC data, we know the Ic could climb about 11.4% slower than the Ib, so we estimate its initial ROC at about 2406 fpm, less than the value we had been using. Since the average ROC was 2345 fpm, and that is in the right range for initial ROC, we will go with it. That yields a maneuverability rating of only 13 - not good enough for a night fighter - but since the aircraft was marginal - it may be a good rating in this case. It is down from 14 before. [This aircraft is not found in EOS or AIO]
Ki-45 II (found only in EOS and AIO) - a two seat precursor to the Ki-96 - probably had an initial ROC around 3380 fpm, an empty equipped weight of 10,471 pounds and a gross weight of 13,668 pounds. If the maximum speed was about 370 mph, that yields a maneuverability rating of 17, up from 15. This aircraft may be thought of as a high performance Ki-45, and it is not a night fighter. The higher performance comes from using larger engines. The aircraft was added to fulfill a percieved need for a high performance attack aircraft without waiting far into the war to get it - and players who don't want to spend the HI points for a 2 engine attack plane should not build it. The Ki-96 - the variant officially approved - deleted the rear seat and machine gun - and in the event took too long to develop - but weighed slightly less and had slightly better performance. The Ki-96 in turn was cancelled - because the JAAF decided to reinstate the rear seat - and so the project went on to become the Ki-102 - which is in all forms of WITP.
Gunston gives no data for either Ki-48 ROC. And it is clear we are using average ROC data to 5000 m (16405 feet). We know from this the Ki-48 II took 30 seconds less to reach that altitude, so we know their relative values. We can estimate initial ROC at 1895 fpm for the I and 2007 for the II. This does not change the maneuverability ratings of 11 and 12, respectively.
Ki-61 I and 2 are both credited with 2200 fpm by Gunston. Something is very wrong here: this is well below the average ROC - and they aren't the same as each other. The II could climb to 5000 meters in exactly the same time as the Ki-100 I could (6 minutes) - so I gave it the same initial ROC. The I could reach that altitude 30 seconds faster - so I give it 3578 fpm. This results in a maneuverability rating of 34 for the I and 33 for the II. [Both aircraft are not present in EOS or AIO] You can see why these aircraft were highly regarded, by both sides, in a dog fight.
Ki-100 I initial ROC was 3280 fpm. That causes a maneuverability rating of 32 (up from 30). Note the progression downward for the family aircraft Ki-61 I (34), Ki-61 II (33), Ki-100 I (32). As aircraft gain speed and weight, but do not increase wing area, wing loading increases, and sometimes maneuverability suffers slightly.
Gunston also gives no data for Ki-102b ROC. Here NO ONE gives initial ROC (normal for Japanese aircraft) and almost no one for average ROC. But Francillon gives us quite good data for the b and c. He lists average ROC at 2377 fpm, and that should correspond to about 2472 fpm. But RHS - inherited from CHS - inherited from stock - is using 2750! Now that is awfully round - and sounds like a guess. Unless there is some indication it is correct, I think we should go with a value closer to the average ROC to 5000 meters. Using my estimate, we must reduce the maneuverability factor to 14 (from 15). There also was a night fighter variant that nearly saw service - and if we get a slot we might want to think about adding it. There was also a high altitude version which saw limited service. Suspecting we might overrate the altitude, I checked, and found it slightly underrated for the RHS system (80% for non-turbosupercharged engines). The high altitude model would have a considerably higher operational ceiling, and a completely revised armament. Again - if we only had a slot for it.
These four are Kawanishi aircraft:
H6K4 initial ROC was 1222 fpm. This value does not change the maneuverability factor of 4 - which occurs because we divide 14 by the number of engines for a 4 engine plane. It isn't likely ANY 4 engine aircraft will change its durability value due to this review, but I wanted to check a few cases to make sure.
H6K2-L (a variant of H6K4 in spite of its name) had an initial ROC of about 1197 fpm. The maneuverability factor should have been 4 (but was 5) and it still should be 4 with this value.
H8K2 initial ROC was 1673 fpm. This does not change maneuverability = 5.
H8K2-L initial ROC was 1286 fpm. This does not change maneuverabilty = 4. The transports run heavier than the patrol planes do, and in particular that applies to this one, which had a normal load about 3 times greater than the patrol plane's normal weapons weight.
RE: Aircraft ROC Review
There was a suggestion that aircraft ROC data might not be perfect
-
el cid again
- Posts: 16983
- Joined: Mon Oct 10, 2005 4:40 pm
RE: Aircraft ROC Review
If there was I probably made it. If I did not make it, I certainly agree! That does not mean it isn't a vital bit of data, that we can afford not to get it at least consistently to one standard, or that were we making the model from scratch that we would not also elect to use it, after all. [It is safe to say you and I would use it differently from this model, but we would probably use it more often rather than less often than this system does.] It is hard to get data of this sort, and initial ROC is a statistical concept - because it changes at every altitude - how do you measure initial ROC? And not in this age with digital instrumentation, but in that one? We have almost no initial ROC data for Japanese aircraft - because apparently they didn't record it - and perhaps didn't use it. That means - to the extent the game is supposed to use it - we have consistently underrated Japanese aircraft - to the extent we were able to get initial ROC data for the Allies. [Yet another case where there may be a built in bias, but not an intentional one by anyone. Note that, as usual, it is a bias against Japan - most such things are - and the perception that things are unbiased in favor of the Japanese is sub-rational. In this case - if this theory is correct - the culprit was Japanese cultural or industrial practice - inherited by us from the data from most sources.] Yet I admit to using both average ROC data and initial ROC data when I could get it - and that is at least inconsistent. And I saw both standards in the data as I inherited it. I think we need to get (or estimate) to a single standard to be where we want to be - to be fair to all types in a relative sense. Being consistent is at least better than being inconsistent. And in spite of my suggestion we adopt as standard average ROC to a medium altitude, the more I think about it, the more I am inclined to go with the original system (if, indeed, the designer intended to use initial ROC). Once again, I must say that this model, which is "too simple to be any good" (my thoughts being quoted), turns out to be remarkably good in terms of its simplicity. That is the mark of a thoroughbred concept, one that is likely to become a classic: like Fletcher Pratt mineature naval games did. It may not be obvious why estimating ranges with the human eye models optical fire control well, but it does.
RE: Aircraft ROC Review
I think have made clear my opinion: Initial ROC presents the best case and makes poor altitude planes look better. I have no wish to continue this discussion. If you think ROC is still the best(or less worse) go for it. It's your Mod.
-
el cid again
- Posts: 16983
- Joined: Mon Oct 10, 2005 4:40 pm
RE: Aircraft ROC Review
I do not understand your point. This last is too cryptic - and previously I saw no indication of what your analysis is - other than "it was not right."
It is absolutely clear that initial ROC is the highest possible value for ROC - for virtually all aircraft. It has nothing whatever to do with the altitude capability of the aircraft (as far as I know) - it just is what it is. The opposite extreme is absolute ceiling where ROC is (in still air, by definition) zero. Service ceiling is very nearly the same thing: all aircraft are equal at their cervice ceiling (in still air, by definition) - at 100 fpm.
What is NOT absolutely clear is why we should adopt some value between initial ROC and these upper extremes? The nearer we get to service ceiling, the more nearly equal we make the aircraft. Which is to agree absolutely with what you said - putting in the opposite way - it is to minimize the difference between planes. WHY is that a good idea? And IF it is a good idea, at what point between maximum difference and absolutely all are equal is the right point to be using?
Is your understanding of when ROC data is used different from mine? When does code use it? It is ever used directly in any instance other than a plane attempting to intercept from the runway? If not, what is wrong with using initial ROC then? When, surely, the interceptors diserve the advantage. It IS used in air combat in RHS (and possibly in CHS and stock IF they are including ROC as a componant of maneuverability): but only indirectly. Still - all such cases either begin - and at least potentially end - near the ground/sea. And relative climbing ability is an absolutely vital factor in determining when they reach the ground, or indeed any other position involving turning. Why shoud we want to minimize the advantage of the planes that can both out climb and out turn (because of better ROC)?
No question here is meant to say you are on the wrong track. They are pure questions - because the answers are NOT clear to me.
It is absolutely clear that initial ROC is the highest possible value for ROC - for virtually all aircraft. It has nothing whatever to do with the altitude capability of the aircraft (as far as I know) - it just is what it is. The opposite extreme is absolute ceiling where ROC is (in still air, by definition) zero. Service ceiling is very nearly the same thing: all aircraft are equal at their cervice ceiling (in still air, by definition) - at 100 fpm.
What is NOT absolutely clear is why we should adopt some value between initial ROC and these upper extremes? The nearer we get to service ceiling, the more nearly equal we make the aircraft. Which is to agree absolutely with what you said - putting in the opposite way - it is to minimize the difference between planes. WHY is that a good idea? And IF it is a good idea, at what point between maximum difference and absolutely all are equal is the right point to be using?
Is your understanding of when ROC data is used different from mine? When does code use it? It is ever used directly in any instance other than a plane attempting to intercept from the runway? If not, what is wrong with using initial ROC then? When, surely, the interceptors diserve the advantage. It IS used in air combat in RHS (and possibly in CHS and stock IF they are including ROC as a componant of maneuverability): but only indirectly. Still - all such cases either begin - and at least potentially end - near the ground/sea. And relative climbing ability is an absolutely vital factor in determining when they reach the ground, or indeed any other position involving turning. Why shoud we want to minimize the advantage of the planes that can both out climb and out turn (because of better ROC)?
No question here is meant to say you are on the wrong track. They are pure questions - because the answers are NOT clear to me.
RE: Aircraft ROC Review
Sid,
You touched on the real issue: because we do not know exactly how the game engine code uses ROC, we do not know precisely which ROC it wants (initial, at a given altitude, etc.).
Only if we know how the code uses ROC can we make a confident choice. What if the code uses the database ROC (meaning the ROC value we entered into the scenario database) as 'initial ROC' for intercepts from the runway, and uses a modified ROC based upon altitude? For example, maybe the code takes database ROC and uses a percentage of it based on the percentage of max altitude (the maximum altitude entered into the database). [Concrete example, database ROC is 3,000 ft/min, database max altitude is 28,000 ft. The engagement takes place at 14,000 ft. In this example maybe the game engine calculates that 14,000ft is 50% of 28,000ft, so it uses 1,500 ft/min (50% of 3,000 ft/min) in it's combat formulas.]
Dili,
I understand your point, but unless we know just which ROC the game engine expects to find in the database, your assertion is just as much a guess as using initial ROC is. In fact, guessing that the game engine wants initial ROC is more likely to be true, because Grigsby would know that bit of data is more available for all aircraft.
You touched on the real issue: because we do not know exactly how the game engine code uses ROC, we do not know precisely which ROC it wants (initial, at a given altitude, etc.).
Only if we know how the code uses ROC can we make a confident choice. What if the code uses the database ROC (meaning the ROC value we entered into the scenario database) as 'initial ROC' for intercepts from the runway, and uses a modified ROC based upon altitude? For example, maybe the code takes database ROC and uses a percentage of it based on the percentage of max altitude (the maximum altitude entered into the database). [Concrete example, database ROC is 3,000 ft/min, database max altitude is 28,000 ft. The engagement takes place at 14,000 ft. In this example maybe the game engine calculates that 14,000ft is 50% of 28,000ft, so it uses 1,500 ft/min (50% of 3,000 ft/min) in it's combat formulas.]
Dili,
I understand your point, but unless we know just which ROC the game engine expects to find in the database, your assertion is just as much a guess as using initial ROC is. In fact, guessing that the game engine wants initial ROC is more likely to be true, because Grigsby would know that bit of data is more available for all aircraft.
Intel Monkey: https://sites.google.com/view/staffmonkeys/home
-
el cid again
- Posts: 16983
- Joined: Mon Oct 10, 2005 4:40 pm
RE: Aircraft ROC Review
It might be that the game uses ROC for an intercept case involving CAP - it would not surprise me a bit. And it might do that with a modifier based on altitude. I am pretty sure altitude is used in that routine - but ROC might be a factor as well.
But IF it is pro rated for altitude - we pretty much would have to feed it the initial ROC - so we would not be cutting a value that is going to be cut again in the routine. I think initial ROC is a better guess - and I think it either says that somewhere or we were told that - about 18 months ago when we first worked on aircraft. It certainly seems unlikely they wanted to "equalize" all aircraft for some altitude where the field would not be as different as it can be.
Note also that even a code specialist (for complex code not written by him - and modified a number of times over a long period) can never be sure how code uses something? While he or she can look at a line and say "I see that field used in this algorithm in this specific way" - he or she cannot know what other algorithms also use the same field? Even if you have actually seen them all - you don't know that you have done. A tester is only slightly better off. While you can show this happens under those conditions, with this range of outcomes, you cannot know if you have tested for every possible branch in the code? You might have missed a routine that is only used in some case you didn't think of. You might also have missed an outcome if you didn't run enough tests: if an outcome is very unlikely, you might never have seen it.
The philosophy of our original programmer (GG) seems to be "use a few values, throw in lots of die rolls and conditional tests, and arrange for a crudely correct outcome set - where the die rolls permit the less likely outcomes to happen - or not - sometimes." How can ROC be for any altitude at all? Some planes climb to vastly higher altitudes than others. The average ROC to 10,000 meters is very different than to 3,000 meters - and both are common values. It may be that is what was done - at the data entry stage - but it hardly seems likely it was the design intent.
But IF it is pro rated for altitude - we pretty much would have to feed it the initial ROC - so we would not be cutting a value that is going to be cut again in the routine. I think initial ROC is a better guess - and I think it either says that somewhere or we were told that - about 18 months ago when we first worked on aircraft. It certainly seems unlikely they wanted to "equalize" all aircraft for some altitude where the field would not be as different as it can be.
Note also that even a code specialist (for complex code not written by him - and modified a number of times over a long period) can never be sure how code uses something? While he or she can look at a line and say "I see that field used in this algorithm in this specific way" - he or she cannot know what other algorithms also use the same field? Even if you have actually seen them all - you don't know that you have done. A tester is only slightly better off. While you can show this happens under those conditions, with this range of outcomes, you cannot know if you have tested for every possible branch in the code? You might have missed a routine that is only used in some case you didn't think of. You might also have missed an outcome if you didn't run enough tests: if an outcome is very unlikely, you might never have seen it.
The philosophy of our original programmer (GG) seems to be "use a few values, throw in lots of die rolls and conditional tests, and arrange for a crudely correct outcome set - where the die rolls permit the less likely outcomes to happen - or not - sometimes." How can ROC be for any altitude at all? Some planes climb to vastly higher altitudes than others. The average ROC to 10,000 meters is very different than to 3,000 meters - and both are common values. It may be that is what was done - at the data entry stage - but it hardly seems likely it was the design intent.
-
el cid again
- Posts: 16983
- Joined: Mon Oct 10, 2005 4:40 pm
RE: Aircraft ROC Review
Aircraft from minor Japanese manufacturers:
Q1W1 - no ROC data given at all by Gunston. Francillon gives data to 2000 meters that indicates an average ROC of 751 fpm. This is so bloody awful - no possibly initial ROC figure that is reasonable is going to change the maneuverability of the aircraft. I estimate it at 766 fpm. [This aircraft is no in EOS or AIO]
J7W1 (by the same manufacturer - Kyushu - as the Q1W1 above) is not even listed by Gunston. But its climb to 8000 meters - 8 minutes and 40 seconds - does indicate a truly spectacular initial ROC. That is an average ROC of 2460 fpm. At these speed regimes, climb to half the altitude should result in an ROC of 2.7 times the higher ROC = 6642 fpm. The ratio between a medium altitude (in this case 4000 meters) and initial ROC is about 1.04 = 6908 fpm. Call it 6900 (so as not to imply this is a precision calculation). That would result in a maneuverabilty value of 54 - which sounds excessive. Comments - or actual data - are welcome. For lots of reasons, this would clearly be a fantastic aircraft - but does 54 overstate that? CHS and stock both use 2500 - which sounds a lot like a guess to me. For some reason unclear to me RHS was using 2693 fpm. For the moment, I will apply the principle "if it is too large, try a close order of magnitude smaller" - or 3450 fpm. That yields a maneuverability rating of 38- which at least seems more reasonable.
N1K1-J initial ROC is 3300 fpm. That increases maneuverability to 33 (up from 29).
B7A2 initial ROC is 1973 fpm. This does not change maneuverability of 25.
M6A1 initial ROC is 1747 fpm. This increases maneuverability to 21, up from 18. This is the only aircraft ever build intended to attack from a submarine, although the US did design a similar submarine in the 1950s for a never built jet bomber. At the time we introduced it, it was the first time WITP could put actual attack bombers on I-400 or I-12 class submarines - which if in the game - lacked their actual aircraft.
D3A1 initial ROC is 1640 fpm. This leaves maneuverability unchanged at 20. For some reason we were using a slightly higher ROC.
E13A1 initial ROC is only 1 fpm greater than we were using. But maneuverability decreases to 17, down 2, because of an apparent error before.
E16A1 fits the same pattern: initial ROC 1 fpm greater than we were using, and no change to maneuverability.
Ki-36 initial ROC is 1525. We had too large a value. The maneuverability decreases to 16, from 17.
D4Y2 initial ROC is 2700 fpm. This increases maneuverability to 29, up from 22.
P1Y1 initial ROC is listed as 2100 fpm by Gunston. This is unlikely, as the average ROC is more than 200 fpm higher! I estimate it at 2385. This results in a maneuverability factor of 13, a decrease of 3 (apparently a miscalculation).
L2D2 initial ROC 828 fpm leaves maneuverability value at 6.
Ki-56/LO initial ROC 803 fpm reduces maneuverability to 7 (from 9).
Ki-34/AT-2/L1N1 initial ROC 1262 fpm reduces maneuverability to 6 (from 8). This aircraft is not in EOS or AIO.
Q1W1 - no ROC data given at all by Gunston. Francillon gives data to 2000 meters that indicates an average ROC of 751 fpm. This is so bloody awful - no possibly initial ROC figure that is reasonable is going to change the maneuverability of the aircraft. I estimate it at 766 fpm. [This aircraft is no in EOS or AIO]
J7W1 (by the same manufacturer - Kyushu - as the Q1W1 above) is not even listed by Gunston. But its climb to 8000 meters - 8 minutes and 40 seconds - does indicate a truly spectacular initial ROC. That is an average ROC of 2460 fpm. At these speed regimes, climb to half the altitude should result in an ROC of 2.7 times the higher ROC = 6642 fpm. The ratio between a medium altitude (in this case 4000 meters) and initial ROC is about 1.04 = 6908 fpm. Call it 6900 (so as not to imply this is a precision calculation). That would result in a maneuverabilty value of 54 - which sounds excessive. Comments - or actual data - are welcome. For lots of reasons, this would clearly be a fantastic aircraft - but does 54 overstate that? CHS and stock both use 2500 - which sounds a lot like a guess to me. For some reason unclear to me RHS was using 2693 fpm. For the moment, I will apply the principle "if it is too large, try a close order of magnitude smaller" - or 3450 fpm. That yields a maneuverability rating of 38- which at least seems more reasonable.
N1K1-J initial ROC is 3300 fpm. That increases maneuverability to 33 (up from 29).
B7A2 initial ROC is 1973 fpm. This does not change maneuverability of 25.
M6A1 initial ROC is 1747 fpm. This increases maneuverability to 21, up from 18. This is the only aircraft ever build intended to attack from a submarine, although the US did design a similar submarine in the 1950s for a never built jet bomber. At the time we introduced it, it was the first time WITP could put actual attack bombers on I-400 or I-12 class submarines - which if in the game - lacked their actual aircraft.
D3A1 initial ROC is 1640 fpm. This leaves maneuverability unchanged at 20. For some reason we were using a slightly higher ROC.
E13A1 initial ROC is only 1 fpm greater than we were using. But maneuverability decreases to 17, down 2, because of an apparent error before.
E16A1 fits the same pattern: initial ROC 1 fpm greater than we were using, and no change to maneuverability.
Ki-36 initial ROC is 1525. We had too large a value. The maneuverability decreases to 16, from 17.
D4Y2 initial ROC is 2700 fpm. This increases maneuverability to 29, up from 22.
P1Y1 initial ROC is listed as 2100 fpm by Gunston. This is unlikely, as the average ROC is more than 200 fpm higher! I estimate it at 2385. This results in a maneuverability factor of 13, a decrease of 3 (apparently a miscalculation).
L2D2 initial ROC 828 fpm leaves maneuverability value at 6.
Ki-56/LO initial ROC 803 fpm reduces maneuverability to 7 (from 9).
Ki-34/AT-2/L1N1 initial ROC 1262 fpm reduces maneuverability to 6 (from 8). This aircraft is not in EOS or AIO.
RE: Aircraft ROC Review
Witpqs: average ROC at all altitudes including service ceiling limit will be the answer. The best way to get this is with TIME TO.
RE: Aircraft ROC Review
Dili,
Maybe, maybe not. What I am pointing out is that we can't know unless Matrix examines the code and tells us.
Maybe, maybe not. What I am pointing out is that we can't know unless Matrix examines the code and tells us.
Intel Monkey: https://sites.google.com/view/staffmonkeys/home
RE: Aircraft ROC Review
What difference what method you use, just as long as you stay consistent? Thus you will always be comparing apples to apples regardless of what the code expects. For my two cents worth the average would be the best simply because you cannot govern at what altitude air to air is occurring, it is data that is easily available, and tends to even out anomalies.