Aces High Bulletin Board
General Forums => Aces High General Discussion => Topic started by: Anaxogoras on May 13, 2009, 04:33:08 PM
-
A Zscore is a fighter's performance number, minus the average (for all fighters), and then divided by the standard deviation. For example, the formula for speed would be:
Zscore = (speed - average speed) / standard deviation
Why do this instead of comparing raw numbers?
The Z-Score curve is a bell curve. -2.00 = a score that is just over 2% better than all the rest of the POSSIBLE scores. Possible not necessarily meaning exisiting. -1 = better than just under 16%, 0.00 = right at 50%, 1.00 = better than just over 84%, 2.00 = better than just over 97%. So you see (or maybe not) that the better or worse a score is, the further up the curve it is and in reality the more impressive it is.
Because of the nature of Z-Scores and the bell curve, a 2.00 in one category is the SAME as a 2.00 in another, or any other like number. Z-Scores turn apples and oranges comparisons into Apples to apples.
In other words, this way we can meaningfully compare top speed to firepower, relative to the whole planeset. Instead of using Zscores to rank aircraft over all, I am simply going to share the individual categories with you because some of you might like it or find it interesting.
Speed at sea level:
(http://farm4.static.flickr.com/3634/3529482404_31dd2553ec_o.png)
Best speed at 10k ft (or less):
(http://farm3.static.flickr.com/2285/3528669263_d881751f0c_o.png)
Best True Air Speed:
(http://farm4.static.flickr.com/3369/3529482466_d328c080cf_o.png)
Climb at Sea Level:
(http://farm3.static.flickr.com/2319/3529482484_299106fb53_o.png)
Climb at 10k ft:
(http://farm3.static.flickr.com/2190/3529482502_5d329514f6_o.png)
Climb at 15k ft:
(http://farm4.static.flickr.com/3337/3529482520_01130404fd_o.png)
Sustained Turn Rate:
(http://farm4.static.flickr.com/3608/3528669325_8379c19fd6_o.png)
Turn Radius No Flaps:
(http://farm3.static.flickr.com/2413/3528669341_d2b52bdc44_o.png)
Turn Radius Full Flaps:
(http://farm3.static.flickr.com/2291/3529482564_33022a6c51_o.png)
Roll Rate 200mph:
(http://farm3.static.flickr.com/2237/3528669375_8edb8dda34_o.png)
Roll Rate 300mph:
(http://farm3.static.flickr.com/2204/3529482596_d1d10886c9_o.png)
Roll Rate 400mph:
(http://farm4.static.flickr.com/3409/3529482606_d66f1c57cf_o.png)
Level Acceleration 150-250mph:
(http://farm3.static.flickr.com/2264/3528706027_de7103ec7b_o.png)
Dive Acceleration 150-400mph:
(http://farm3.static.flickr.com/2368/3529518776_2dfba61bdc_o.png)
Energy Retention 400-150mph:
(http://farm4.static.flickr.com/3414/3529482620_ed715cab18_o.png)
Lethality (1 second burst):
(http://farm3.static.flickr.com/2065/3528669361_db865edb5c_o.png)
Primary Weapon Muzzle Velocity:
(http://farm3.static.flickr.com/2219/3529482630_490c805b74_o.png)
Secondary Weapon Muzzle Velocity:
(http://farm3.static.flickr.com/2302/3528669455_4dd98e972b_o.png)
Gun Destruction Potential; ammo duration time * lethality:
(http://farm4.static.flickr.com/3402/3528637219_72f5991475_o.png)
-
Gun Destruction Potential; ammo duration time * lethality:
(http://farm4.static.flickr.com/3402/3528637219_72f5991475_o.png)
Any chance of getting a score chart that factors muzzle velocity, ballistics and range in with fire duration and lethality?
-
Any chance of getting a score chart that factors muzzle velocity, ballistics and range in with fire duration and lethality?
What's "range?" I could do a chart that has different colored bars for lethality, ballistics, and duration, but it would still have to be sorted by a single category. Do you mean a single number that factors in those 3 categories?
-
Yes. I'd like to see a chart that rates the weaponry based on a combination of ALL factors, because the ballistics, muzzle velocity and rate of fire are just as important to the weapon's capability as how big of a hole it makes.
-
Bell curves ought not be used in classes, and IMO don't really help that much when cold hard numbers (not subjective curves) tell you what you need in an easier to understand manner.
-
WOW! My P38G is a piece of crap :) :rock
WTG Gav.
-
WOW! My P38G is a piece of crap :) :rock
WTG Gav.
I thought you only flew the la7?
-
Gun Destruction Potential; ammo duration time * lethality:
(http://farm4.static.flickr.com/3402/3528637219_72f5991475_o.png)
I think this chart shows pretty well that it's difficult to weigh different factors.
"lethality*ammo duration" sounds good at first. But if you examine the chart you will see the Ta-152 classed as "average", even behind the 110C with its somewhat anaemic MG/FF and its battery of 4 MG-17.
I think most of us take the 152 into the air for several reasons, the immense firepower being not the least one ;)
-
nice job and Thanks for all the work you put into this! :salute
-
What is 'ammo duration time'. Total number of seconds of full fire? If I'm not mistaken, firepower (or lethality?) is usually determined by the 'throw weight' for an individual round times the number of guns/cannons times the number of rounds fired (often at different rates) in one second. Throw weight establishes both the kinetic destructive power (using size, physical weight, muzzle velocity and trajectory info) plus chemical destruction (if any) imparted by explosives, for a single round.
While ammo duration or size of an aircraft's gun loadout is important, IMO it should be an additional criteria added after pure firepower is determined.
-
What is 'ammo duration time'. Total number of seconds of full fire? If I'm not mistaken, firepower (or lethality?) is usually determined by the 'throw weight' for an individual round times the number of guns/cannons times the number of rounds fired (often at different rates) in one second. Throw weight establishes both the kinetic destructive power (using size, physical weight, muzzle velocity and trajectory info) plus chemical destruction (if any) imparted by explosives, for a single round.
While ammo duration or size of an aircraft's gun loadout is important, IMO it should be an additional criteria added after pure firepower is determined.
Yup, what you're describing is under the heading "Lethality."
The last chart does tend to throw people for a loop, but after previous debate on this board, it was recommended that instead of simple ammo duration, the total destructive potential of an aircraft's weapons was a better thing to measure.
-
Perhaps you should take rifle caliber machine guns out of the equation when dealing with that. No matter how much firepower they have, when someone sees that they're out of cannon, they will most likely go home. It's just not practical to sit and spray bb's at things for 10 minutes before they exlpode.
-
If I called it ammunition capacity would it be better?
-
Where did you get the turn-performance data and roll data?
-
I tested it with a stopwatch and averaged 3 trials for each aircraft per category. Other categories I tested myself were dive acceleration, and energy retention, not to mention testing everything for the P-39.
Fwiw, regarding Lethality, it came up in another thread that some aircraft with the same weapon types have different lethality depending on synchronization with the prop: this is accounted for in the data, as you'll see that the Ki-84 has greater lethality than the Ki-61.
-
Is that E retention data from the off-throttle trials? If it is, you need to make it clear.. It's not a real world figure. No one zooms up with their throttle off.
-
Energy retention was tested by diving the aircraft to ~5k ft, using auto level, and when speed dropped to 400mph, I cut the engine (rpm's already pulled back so prop is feathered for glide).
Fwiw, for lethality we should also disclose that it was tested by taxiing to a hangar and firing at it nearly point blank, because that's also something impractical for combat. ;)
Moral of the story: the practicality of a test method doesn't always have bearing on the practicality of what it tests.
-
I thought you only flew the la7?
:rofl :rofl Hush frog, I'll send piggy home when I'm finished :aok
-
Btw, standard zoom climb is something I would like to test, but last time it came up we argued about how to conduct the test. For instance, Saxman wanted to let auto-climb level the plane before it would stall, whereas I wanted to hold it nearly straight up, let it stall, and record that altitude. Then Hitech came in and said that using auto-climb at all would skew the results because its behavior depends on the aircraft. So, if anyone can tell me how to test zoom climb in a way that wouldn't lead to another argument, I'm all ears! :D
-
Anax...use Widewings method's.
Set arena time to noon. Dive past 400mph to sea level, when the speed bleeds back to 400, use a 3g pull up into the straight vertical, nose at the sun, hold the plane up until you can't anymore and it falls off. Note the highest altitude obtained before this happens. Straight vertical is the most practical test for a plane's zoom ability. Note to those who will nitpick by saying this is not an unloaded zoom: An airplane going straight up *is* unloaded. The G-meter will be at zero, the wing is *not* generating lift. If the wing was generating any positive/negative lift, the plane would be looping over or nosing back down respectively...a pure vertical zoom is quite nessecarily an unloaded one.
This gives the most meaningful results for the usefulness of an aircraft's zoom in combat.
Btw, standard zoom climb is something I would like to test, but last time it came up we argued about how to conduct the test. For instance, Saxman wanted to let auto-climb level the plane before it would stall, whereas I wanted to hold it nearly straight up, let it stall, and record that altitude. Then Hitech came in and said that using auto-climb at all would skew the results because its behavior depends on the aircraft. So, if anyone can tell me how to test zoom climb in a way that wouldn't lead to another argument, I'm all ears! :D
-
Alright, I'll give that a try sometime.
Btw, one of the interesting differences between Snailman's lethality ratings and mine is that I account for ROF. Thus the Ki-84's 20mm shells are less lethal than Hispanos, but the Ki-84 fires them at ~27 rounds/sec while the Spitfire only fires ~22 rounds/sec. So for a 1 second burst the Ki-84 is actually more lethal.
I don't know if this is historically correct or not. I'm just taking the total number of rounds and dividing it by the time it takes to fire them.
-
Alright, I'll give that a try sometime.
Btw, one of the interesting differences between Snailman's lethality ratings and mine is that I account for ROF.
So do I. (Respectively Mr. Williams)
May I quote myself:
In a nutshell, firepower of a weapon is defined as (kinetic energy + chemical energy)*ROF, measured at the muzzle.
Also it says on my chart "one second burst". :)
Thus the Ki-84's 20mm shells are less lethal than Hispanos, but the Ki-84 fires them at ~27 rounds/sec while the Spitfire only fires ~22 rounds/sec. So for a 1 second burst the Ki-84 is actually more lethal.
Which would be true if the individual shell of the Hispano weren't way more powerful than the Ho-5's.
BTW, latest version can be found here: http://bbs.hitechcreations.com/smf/index.php/topic,264782.0.html
-
I should have said "difference in ROF between synchronized and unsynchronized weapons." My bad.
-
Awesome work, thanks.
Has anyone ever done a study on bullet dispersion? That might be interesting to see.
-
Energy retention was tested by diving the aircraft to ~5k ft, using auto level, and when speed dropped to 400mph, I cut the engine (rpm's already pulled back so prop is feathered for glide).
Fwiw, for lethality we should also disclose that it was tested by taxiing to a hangar and firing at it nearly point blank, because that's also something impractical for combat. ;)
Moral of the story: the practicality of a test method doesn't always have bearing on the practicality of what it tests.
So you're putting that chart up for people to see which plane is the best, deadstick? The other charts are good but that one's pretty useless compared to a shallow and vertical powered zoom comparison.
-
So you're putting that chart up for people to see which plane is the best, deadstick? The other charts are good but that one's pretty useless compared to a shallow and vertical powered zoom comparison.
Let's agree to disagree. :)
-
I tested it with a stopwatch and averaged 3 trials for each aircraft per category.
Excellent!
-
Let's agree to disagree. :)
Agree to disagree that no one has more use for deadstick figures than powered ones? Either way, you should tell people that that's what it is. The ranks don't look like that in powered flight.
-
Agree to disagree that no one has more use for deadstick figures than powered ones? Either way, you should tell people that that's what it is. The ranks don't look like that in powered flight.
One of the counter suggestions I've heard is that "energy retention" should be a test of deceleration from 500mph to 400mph in powered, level flight. The only problem is that some aircraft break apart long before they reach 500mph ias. ;)
What the glide test tests is manifest in actual combat. You've showed me as much yourself with the way you make use of the Ta-152's energy retention. It's probably the most difficult performance trait to make use of in the whole game.
-
One of the counter suggestions I've heard is that "energy retention" should be a test of deceleration from 500mph to 400mph in powered, level flight. The only problem is that some aircraft break apart long before they reach 500mph ias. ;)
What the glide test tests is manifest in actual combat. You've showed me as much yourself with the way you make use of the Ta-152's energy retention. It's probably the most difficult performance trait to make use of in the whole game.
Yeah, I realize difficulty getting to a top dive speed testably higher than the top deck speed of the fastest is a problem for some airplanes.
Just a thought:
In a practical sense I suppose the question we are testing is whether a an airplane with a lower top speed following another in a dive, lets say to the deck, can hold position actually gain for long enough after leveling out to perhaps gain a firing position. Say an F4U vs. a P-51.
Or, another possibility, can the airplane with lower top speed but better E-retention gain significant separation after leveling out from the dive? Say a P-51 vs. a D9 with the drop-tank rail.
That idea should perhaps shape the tests.
EDIT: Another thing to keep in mind: Halve the available thrust for Moot's Ta-152 and the power-off efficiency of the airframe wouldn't mean much in combat. Double the thrust for a Hurricane and its draggy airframe wouldn't mean near as much either. So.......
-
I don't know how you tested the Mossie's climb rate, but if it was from the AH charts, that was a Mossie with 100% fuel and 2,000lbs of bombs. It climbs better than quite a few aircraft at lower alts.
-
I get sad when I see this. . .
-
I don't know how you tested the Mossie's climb rate, but if it was from the AH charts, that was a Mossie with 100% fuel and 2,000lbs of bombs. It climbs better than quite a few aircraft at lower alts.
Thanks for the tip. I will retest its climb myself.
-
All I'm saying is that the "E retention" graphic needs a less misleading title. I don't fly around on the 152 with the power off. The mossie is pretty awful at E retention in light maneuvers or extremely conservative BNZ alike.
-
Nice work. Have you made a master chart that accounts for each one you made and give overall ranking?
-
All I'm saying is that the "E retention" graphic needs a less misleading title. I don't fly around on the 152 with the power off. The mossie is pretty awful at E retention in light maneuvers or extremely conservative BNZ alike.
It would be misleading if kinetic energy retention were not the precise property that was tested. I remember having this discussion with you before. You were of the opinion that because the AH community mostly uses the phrase "e-retention" to mean a bundle of properties like induced-drag, power-loading, and others along with true kinetic energy, we should only use the term the way the unwashed masses use it, imprecise though it may be. The trouble is that my method is to tease out these individual properties and test them independently. You can't look at any data set in isolation and say "this tells me what to expect." Rather, you have to look at numerous data sets and see what their combination means for performance in game. Nothing shows this so well as comparing turn rate data to turn radius data. Both tell you something important about how the aircraft performs in a turn-fight, but neither is sufficient alone. In the same way, looking at the data for energy retention in isolation will also tell you nothing. It only becomes meaningful when you also look at dive acceleration and top speed, among other categories.
Ultimately, if I were to adopt your suggestion and use the term "energy retention" to mean a bundle of things, then energy retention would be untestable because it would mean different things in different circumstances. What I give you is the abstract; but it is universal and unequivocal.
-
So you plot up some charts showing various performances for the plane set, and for energy retention you pick its one manifestation that people rarely use, over one that's by far got the most practical importance? Just add "dead stick" to the chart title. That's less misleading than a chart that shows the mossie as the one best model at "energy retention", when in fact it's one of the worst, or at least nowhere near contention for top spot.
-
So you plot up some charts showing various performances for the plane set, and for energy retention you pick its one manifestation that people rarely use, over one that's by far got the most practical importance?
Which is that?
-
Zooming, man. What else? Full power zooms (vertical and shallower, how shallow's an arbitrary left to the investigator) and retention of speed during maneuvering are the two things that "unwashed" players actually need to know (as far as energy retention goes) for air combat. What exactly is the use of dead stick slipperiness? Nothing that I can think of makes my point more evident than the fact that in practice the Mossie is a mediocre boom and zoomer.
In other words, this way we can meaningfully compare top speed to firepower, relative to the whole planeset.
And here anyone can see that there's little sense in choosing dead stick "energy retention" before performance in zooming and maneuvering, for this insight. Comparing planes based on their paired dead stick energy retention and firepower? Is the point to show how the planes rank for the purpose of shooting things down while dead stick?
I'm arguing this the same way I'd expect to be corrected if I ranked the planes for zoomability and never specified that the excellent P51 and Ta152 zoomers are dead meat to the crappy zero if the zoom starts at low enough speed.
-
Zooming, man. What else? Full power zooms (vertical and shallower, how shallow's an arbitrary left to the investigator) and retention of speed during maneuvering are the two things that "unwashed" players actually need to know (as far as energy retention goes) for air combat.
Zoom climb ability is a combination of energy retention and power loading. I do intend to test that at some point if Widewing's method is acceptable.
What exactly is the use of dead stick slipperiness? Nothing that I can think of makes my point more evident than the fact that in practice the Mossie is a mediocre boom and zoomer.
Did you notice that the Mossie has one of the worst sustained climb rates? That's a big part of the equation for boom 'n zoom. So I don't agree that the Mossie's mediocre boom 'n zoom ability invalidates the data.
And here anyone can see that there's little sense in choosing dead stick "energy retention" before performance in zooming and maneuvering, for this insight. Comparing planes based on their paired dead stick energy retention and firepower? Is the point to show how the planes rank for the purpose of shooting things down while dead stick?
Ok, I get the point that you don't like that testing for energy retention in isolation has to be done power off. ;)
-
Ok, I get the point that you don't like that testing for energy retention in isolation has to be done power off. ;)
That's not the point. You picked the one aspect of energy retention that's barely of any real, practical use. Anyone without a clue will look at that collection of charts and figure that the mossie is a competitive model as far as energy retention goes during combat (which is like, 95% of what the planes are used in). It's wrong and I'm telling you you ought to at least clue-in the reader about that subtlety. "These figures are for off-power energy retention."
-
I found these graphs really interesting and decided to tally up the plus and minuses of a few planes. The number I came up with is the result,where zero was the middle of the road on each catagory. Some planes may have a +2 on lethality but a -1 on sustained turn. When finished I saw where they lay compared to zero.
The 110G ended a -2 plane. Which in my computations (Im no math whiz bare with me) is just a tad tougher to fly in the LW MA than the most average plane (zero, which I have yet to find). So if in the LW MA the 110G has an ENY of 10, I decided to call the average ENY, the middle of the road plane a 12 ENY. This would adjust the -2 I figured for it from these graphs. So then since the average zero plane is a 12 ENY in the LWMA, a -2 rating would assign it a 14 adjusted ENY.
The figured the following planes. I put the number I figured from the graphs, followed by the actual LW MA
number for that plane, and lastly the number I feel th LW MA should actually be for the plane.
............................. ............................. ............................. ....
Plane--------Graph Tally----LWMA ENY----Adjusted LWMA ENY
............................. ............................. ............................. ....
190D9-_______ +7 ___________15_____________ 5
P47D-40-______+4___________ 20_____________ 8
A6M5-_________ -9__________ 20_____________ 20
P40E-__________-6 __________30______________18
P51D-__________+5___________8______________7
P38L-__________+7____________20____________5
Ki61-___________-7___________25_____________19
109K-___________+7__________20_____________5
110G2-__________-2__________10_____________14
F6F-____________+3__________15_____________9
SpitXVI-_________+15_________5______________-3
La7-____________+6__________5_______________6
Tempest-________+10_________5_______________2
SpitIX-__________+3__________20______________9
F4U1D-__________+9__________15______________3
Hurri1-__________-16__________40_____________28
Some of the planes are real close in the end result, like the A6m5 and P51, where some are way off. The Spit16 as guessed came out totally whacked, and in my opinion should be perked. Just too many plusses.
Perhaps with this setup, rather than denying planes with low enys, simply perk any ride less than 5 ENY, and for score, use the eny as a mutiplier.
Say a Tempest shot down 5 planes, he would get a landed score of 10 points since his ENY is 2.
If a Hurri1 shot down 5 planes his landed score would be 140.
Of course the cost of perk planes would have to be cheapened a bit to make it worthwhile.
I like what Gav did with these and it kinda creates an overall value when you tally them up.
Whether or not its science or not..ok not...it does show the Spit16 is way too much plane not to be perked,among others.
-
Whether or not its science or not..ok not...it does show the Spit16 is way too much plane not to be perked,among others.
Not all performance aspects are equal. Weighting them equally may produce an rating that is too low or two high depending on the strengths of a given aircraft.
The ultimate ratings review, usage, says the Spitfire Mk XVI is not overpowering.
-
Not all performance aspects are equal. Weighting them equally may produce an rating that is too low or two high depending on the strengths of a given aircraft.
I wonder how the 262 would rate when being included in those tables. Perhaps we would see that, being outperformed in all but 2 categories, it should be unperked and get a midrange ENY value? ;)
-
I wonder how the 262 would rate when being included in those tables. Perhaps we would see that, being outperformed in all but 2 categories, it should be unperked and get a midrange ENY value? ;)
If it could not cruise around at speeds that were literally above the critical mach number of the prop planes around it, you would actually be right!!! And I do think its perk price should be lowered. Just not worth it for what you get, especially since forwards view was ruined.
-
If it could not cruise around at speeds that were literally above the critical mach number of the prop planes around it, you would actually be right!!! And I do think its perk price should be lowered. Just not worth it for what you get, especially since forwards view was ruined.
Sure, unperk it. But only make it available at the 163 base.
-
Whether or not its science or not..ok not...it does show the Spit16 is way too much plane not to be perked,among others.
Doing this kind of number stuff is what convinced me that some corrections are needed to what is perked and what is not. Since most are very resistant to perking something that is free today, and the same goal of consistency can be achieved by unperking aircraft instead, you might want to try that perspective, too.
If you want the original spreadsheet that provides a cumulative ranking system that you can tinker with, just pm me your email address.
-
Alright, it looks like I'm totally wrong on the mossie. I am not making up that for the couple of months where I flogged it for all I could, it simply couldn't match the P51/38/152 for zooming/e-fighting. So I don't know what the deal is.. Probably maneuvering e-retention. I just ran a couple of trial dives/zooms in real time between a 152 and mossie, and the 152 only outzoomed the mossie by maybe 200 yards at most. In a dive, the 152 does walk away from the mossie around 500mph, though. The quick trials I ran were with 3G and 5G pulls into wep'd vertical zooms, from >500mph dives.
-
I understand what you are saying Karnak, but Im looking at this a real general yardstick for each plane. Not so much how it does in a given situation, rather, in every situation vs all planes. A mix of offensive +'s and defensive +'s thrown into a blender, and see what spice you taste the strongest.
-
Whether or not its science or not..ok not...it does show the Spit16 is way too much plane not to be perked,among others.
Hmmm...I notice no crowd forming to stone YOU. Yet. ;)
-
WOW! My P38G is a piece of crap :) :rock
WTG Gav.
Your P38G! Hah! Just cause you finally figured out what fun it is. That's my P38G. It's got my name on it too! :)
-
I understand what you are saying Karnak, but Im looking at this a real general yardstick for each plane. Not so much how it does in a given situation, rather, in every situation vs all planes. A mix of offensive +'s and defensive +'s thrown into a blender, and see what spice you taste the strongest.
But the yardstick you will end up with is almost useless.
-
Its workin good at stirrin so far :P
-
Sure, unperk it. But only make it available at the 163 base.
That would actually be a good compromise to go along with lowering its perk price some. :aok
-
I wonder how the 262 would rate when being included in those tables. Perhaps we would see that, being outperformed in all but 2 categories, it should be unperked and get a midrange ENY value? ;)
It wouldn't be good for the tables. I also don't believe you can make a meaningful statistical comparison with Zscores between aircraft like the A6M2 and the 262; the Tempest and F4U-4 already stretch the limit.
Secondly, if I did add the 262, speed and firepower are two of the most weighted categories. Without a doubt, the 262 would rank above the Tempest over all.
-
It wouldn't be good for the tables. I also don't believe you can make a meaningful statistical comparison with Zscores between aircraft like the A6M2 and the 262; the Tempest and F4U-4 already stretch the limit.
Secondly, if I did add the 262, speed and firepower are two of the most weighted categories. Without a doubt, the 262 would rank above the Tempest over all.
Sometimes you have to throw in a little well-reasoned qualitative judgement...
The 262 is not only faster than all the prop-planes...It can also *cruise* at beyond the dive limit of prop planes. This is a crucial difference, the 262's performance goes beyond merely being the fastest plane except for the 163.
-
A rough estimate by looking at the chart and how I feel about how the 262 handles:
262- Tally +31
MA adjusted at a -19 ENY remembering a 12 ENY was my middle of the road plane
-
<<< Currently perking coffee