Way Too Long Discourse on Flight-Modeling Disputes

[MackStones]

I've been in some discussions with other flight-sim pilots about Il-2: Battle of Stalingrad, and we got talking about the inevitable disputes that will blow up the BOS forum about flight-models once the game is released, even (or especially) in the beta version.  One of the pilots reminded me about a long post (using the above title) on the Warbirds of Prey forum back in the late 2000s (when the in-development Cliffs of Dover was still known as Storm of War), and suggested that I repeat it on this forum as well in any effort to give the "bitchers and moaners" something more to consider before posting about flight-model "problems" with BOS planes.  

 

We have to realize that BOS is never going to get all flight models "correct", in large part because (as noted in the post below) that is an almost impossible task historically, because there may be computer/technical limitations within any game engine, and also there might need to be some game-balancing to consider.  We also do expect the BOS developers to do a good job in trying to get the flight-modeling correct, but we can't and shouldn't expect perfection either.

 

With that in mind, here's my old post (Oleg, of course, refers to Oleg Maddox of Il-2 fame):

 

Introduction.

It is not uncommon on this forum, and especially on the other Il-2 forums over the years, for people to complain that Oleg’s flight model for a particular aircraft model is “porked” (worse than real-life) or “too uber” (better than real-life). In fact, there is an active thread on the Spits Maps and Missions sub-forum which has raised flight modeling at times.

I have at times in various posts in this forum alluded to problems in determining the “real-life” flight model. I wanted to go ahead and discuss those problems in more detail (“oh no, not another lengthy post from Mack, and this one with subtitles at that” exclaims the assembled populace - ha ha). Many, but not all, of my comments in this post are based in part upon an article I read several years ago, but which I unfortunately cannot now locate online despite repeated and extensive efforts. Other comments are simply a result of my “real-life”, occupational habit of examining, and often attacking, the credibility and probative value of evidence submitted for a particular proposition (I'm a lawyer).

My goal is to encourage people to think long and hard about their position before complaining about flight models in Il-2 (if only by thinking, “If I bring up this flight model criticism in a post, Mack will probably do a 20-page reply post” - ha ha). Of course, I’m as guilty as anyone of making blanket critiques of flight modeling, so this post is partially directed at me.

Oleg’s flight model for each plane (as modified by patches, and as constrained by computer coding issues) is presumably based upon his team’s view of that particular plane real-life flight model. Furthermore, many of the complaints about Oleg’s flight model for a particular plane are presumably based upon the complainer’s view of the real-life flight model for that plane. The question is therefore, with respect to each aircraft: “What was the plane’s real-life flight model?” There are several different sources that can be reviewed to answer that question, but each source suffers from deficiencies.

Manufacturer Specifications.

For instance, as a source for real-life flight models, one can review specifications published by the aircraft’s manufacturer. By specifications, I am referring to the data published by the manufacturer on aircraft dimensions, capabilities, and flight characteristics. This is probably one of the primary sources used. Those manufacturers specifications are useful, but aren’t necessarily correct and complete, either.

In the first place, I think that most will agree that private, for-profit manufacturers at times exaggerate the capabilities of their planes to varying degrees, especially when trying to sell it to governments and/or to meet procurement requirements. For instance, if the government says it wants the plane to go 450 kph at 1,500 meters, bidders will likely show that in their claimed specifications (which claim might be “truthful” in the sense of doing it with a 10%, high-octane fuel load, removal of guns, pilot armor, and radio, and in a dive). We all come to expect exaggeration whenever undergoing a sales call; it wasn’t any different with respect to aircraft sales in the 1930s and 1940s, particularly since military sales were the primary source of revenue for most aircraft companies at the time. In addition, in totalitarian countries during the late 1930s and early 1940s, I suspect that there would have been an incentive to exaggerate aircraft specifications in order to satisfy the (at times unreasonable) demands of the leaders (e.g.- “Yes, Our Dear Leader, the X aircraft satisfies each and every one of the fifty-eight requirements that you ordered our design team to meet upon pain of death”). Of course, even in “free countries”, it is not unknown for subordinates to lie or exaggerate achievements or capabilities to their superiors either, for a variety of reasons.
 

Moreover, at times, the manufacturer in question may have published several different specifications for the same model of aircraft. Some specifications may actually represent “goals” and not “objectives met”. Other specifications may be from earlier in the design and manufacturing process, and do not accurately represent the flight characteristics of the final production version of the aircraft. Additionally, different departments of the manufacturer may utilize different specifications; for instance, the marketing department might use different specifications when trying to make a sale than the engineering department when studying flight-test failures. On other occasions, one design team of a manufacturer might understate capabilities of an aircraft created by another design team or overstate capabilities of aircraft created by its team, as part of a competition over limited resources for their competing aircraft projects. In many cases, there might be multiple sets of differing specifications, without clear descriptions of the date, stage, etc. identified. Furthermore, where records have been lost or destroyed during or after WWII (e.g.- bombing raid or seizure by invading army), perhaps only a single set of specifications remains and it is unclear whether or not it is the “official”, authoritative version. Accordingly, there may be some question as to whether, or which, particular specifications are those for the final, production version of a particular model of aircraft.

It is also conceivable that, on occasion, a manufacturer might understate the capabilities of its aircraft, in an effort to obtain additional funding for “needed improvements”. Please remember that we’re talking about “war profiteers” in the old vernacular (ha ha).

Furthermore, depending upon governmental requirements for secrecy or for political purposes, the actual specifications of a particular aircraft might have been modified, for better (e.g.- “look how great our new plane is, so enemies should beware”) or for worse (e.g.- “don’t let our enemies know how good [or bad] our new plane is”). Additionally, some specifications or test records, etc. might still be designated “top-secret” and might never have yet made publicly available [maybe more from oversight at this point, than anything else].

There is another problem with any sort of specifications insofar as they are supposed to reflect the ideal aircraft, or, in some cases, the typical aircraft, if properly manufactured with proper components using proper processes and techniques. During wartime, many aircraft would not meet those “ideal-plane” or “typical-plane” standards for a variety of reasons (e.g.- different factories, different workforces in a particular factory, sabotage, laziness, use of slave labor, poor oversight and supervision, bombing raids, poor assembly, poor or inconsistent components, etc.). Some countries in WWII also still used artisanal or small-workshop production (as opposed to more-standardized factory production) for at least some components of aircraft. Importantly, quality-control standards were not the same during WWII (especially in war-torn countries) as you might find today. Depending upon who was working with what components that week, an aircraft constructed that week might have some minor, yet appreciable differences from one constructed the following week or the prior week. Even today using modern quality control methods, different production runs of the same product will often have slight differences (e.g.- one run of a particular make and model of carpet or tile will usually have a slightly different color than another run of the same make and model). I also seem to remember old jokes about not buying cars constructed during December, since the workers may not have the same level of focus on their craft.

Along the same lines, many pilot accounts from WWII and other wars describe individual aircraft of the same make and model being slightly different from another similar aircraft in the same squadron (e.g.- “The Major always takes plane #2 because it is - or seems - faster than plane #4"). Sometimes those differences could be due to the manufacturing process (e.g.- Bob filled in to assemble the right elevator because Jim was out sick), or instead simply due to in-theatre maintenance (e.g.- Roger the crew chief takes better care of his plane’s cooling system than Fred the crew chief).

 

Governmental Records.

Another commonly-utilized source for flight models is government test bureaus or agencies. This is another useful source, but likewise has its drawbacks.

In some cases, maybe there wasn’t much testing of a particular plane, and the bureau for whatever reason simply accepted much, if not all, of the manufacturer’s claimed specifications. Additionally, it is not always clear as to what testing criteria was used or what developmental stage of the particular model was tested. Testing wasn’t necessarily always done on the final production model (but perhaps a similar one, for which the production go-ahead was given, with some final suggested modifications). In addition, manufacturers would commonly make sure that a “perfect” plane was sent for testing, and the “perfect” plane was probably better-built [with more quality control efforts] than the “typical” run-of-the-mill factory-built plane. Testing also was usually done at the beginning of, or right before, production, as opposed to during the middle of production, and there might be further differences between the tested plane and the typical mid-production plane. Testing criteria could differ, as could testing techniques and testing personnel. Depending upon what group tested a particular plane under what circumstances, different results might follow. As we all know, different people usually grade or judge things differently, whether through personal preference, “how I was taught or raised to do it”, bias, prejudice, self-interest, improper promotion [e.g.- I like the X plane more than the Y plane, and will do anything to make the X plane “win” the competition]. or even corruption. There is also a problem of what test records remain, and the nature and completeness of the existing test records [similar to problems with manufacturer specifications]. To the extent the government testing involves in-flight testing, additional problems are present as described below.

Flight Test Records.

A third major source of flight model data is comprised of records of flight tests. These are very useful, but like the others, not perfect.

Indeed, there are many variables that can affect the results of flight tests. Was a special, “perfect” plane used? What was the fuel load-out? What was the weapons load-out? Were particular components taken out [or added] for testing purposes? Was special or extra-high octane fuel used? How were the instrument calibrated? What were the weather conditions? What was the skill level of the pilot? What were the personal preferences or prejudices of the pilot? What was the pilot told to do, or not do, in the test? How “controlled” was the test? In addition, what records of flight tests remain? Are those complete records? Are they necessarily accurate ones? For instance, does the only extant flight test records for a particular model represent the typical flight test, or is it instead the aberrational flight test?

In a similar vein, flight testing of captured opposition aircraft provide some advantages (e.g.- avoid pro-enemy propaganda claims about aircraft capabilities), but also has its own disadvantages. The same flight-test concerns raised above apply here, but, added to it, are problems relating to the condition of the aircraft [e.g.- crash-landed but rebuilt; found in water and drained; abandoned in “working condition” at airfield [perhaps because it was a “bad” plane], the usual lack of design manuals or training manuals for enemy plane, the lack of test pilot experience in such an enemy plane, perhaps the inability of the test pilot to fully read or understand the gauges, differences in fuel, oil, or other materials, differences in maintenance and repairs, differences in test pilot preferences and experiences (e.g.- if I only usually flew, and loved Hellcats, would I be the best pilot to test out a captured Oscar?), and the simple variables in particular production-runs or individual aircraft.

Pilot Accounts.

Accounts by pilots of aircraft they flew in, or against, are yet another source of flight model information. Although these descriptions are often interesting, and provide some utility, they have significant drawbacks. At minimum, pilot accounts also subject to basic problems with any remembrances of things past (i.e.- faulty memory, bravado, intentional lies, “spin”, etc.), especially if written years after the fact.

Furthermore, pilot accounts are usually non-quantitative in nature (e.g.- “My P-51 could outrun a Fw-190"), and therefore are more difficult to translate into game terms (i.e.- taking the same example, “how much faster was your P-51D-NA in level flight at 3,000m than a Fw-190A-9").

Unfortunately, in most combat settings, I’m not sure that most pilots would be able to identify the exact model of opposing planes (e.g.- was that a Fw-190A8 or a Fw-190A9 that passed in front of me for my split-second deflection shot?) Even in-game, you often can’t tell the difference between models of opponents, unless you know the restricted plane-set for that mission. Please remember that, in real life, different models did not necessarily have uniform, “default” color schemes as they do in the Il-2 game (e.g.- it’s a dark green Y plane, so it must be a Mk II model). If your real-life Bf-109 outran a Spitfire in level flight, it makes a big difference as to which Spitfire you were facing from a quantitative perspective, since, according to specs, some Spits would be faster than your plane and others would be slower. Indeed, how many of you could recite the exact make, year, and model of each car that you pass on the freeway today, or that you were involved with in an auto accident five years ago? In any event, as we are all aware, misidentification of opposing aircraft is a common feature in all aerial warfare (e.g.- at times, any British fighter was later reported by Germans to be a Spitfire, or any Japanese fighter was reported by Americans to be a Zero). Taking the earlier example, maybe it was actually a Hurricane that you outran (you only could see him in your long six o’clock and you had only brief glances at him) instead of a Spitfire. We all see this sort of misidentification online all the time, which leads to many friendly kills as well as to mistaken calls about the type of enemy plane in the area.

Real-life also throws many variables into the mix when trying to base a flight model on pilot accounts, such as the type of mission flown and loadout of the planes (e.g.- did one plane have 15% fuel, not 80% fuel, or was one carrying bombs or rockets), respective pilot skills (e.g. - an Experten in a Fw190 may out-turn a rookie in a Spitfire even at low speeds), the speed and/or altitude characteristics of the fight (e.g.- at the right speed, a Wildcat will out-turn a Zero), the ever-important factors of luck and surprise, the fuel type and octane rating, maintenance standards, availability of parts and supplies for repair, need for jury-rigged repairs, the variable characteristics of individual planes, weather conditions, etc. Much of the time, however, at least some of these important variables are left out of the description by the pilot. There is the question of how well your anecdotal evidence is statistically representative of the planes in question. For instance, if you easily shot down two Georges in mid-1945 over Japan, you might think that they were bad planes, whereas it might have more to do with lesser-trained or lesser-experienced pilots who happened to be flying them at the time. Alternatively, if you were able to kill a Corsair with a single burst from your Zero, you might think they were easy to kill, even though you actually just got in a lucky shot. Maybe your Hurricane outran the Tony in Burma only because he lost sight of you over the jungle and gave you a good head-start, not because your plane was faster. Anecdotes can be representative of typical behavior and characteristics, but don’t have to be so representative.

Another important factor in accounts by pilots of aircraft is simply bias or prejudice in favor of, or against, whatever plane you flew. There are many examples of pilots refusing to fly, or at least resisting efforts to make them fly, “better” planes since they were familiar with the current ones, had been through “thick and thin” with the current ones, knew the proper tactics to use with the current planes, etc. If I had survived dozens of missions in a P-40E, and the P-40E had been able to get me home even with battle damage, I might think it was the best plane around, and be reluctant to transfer to the “better” P-51s. Of course, some pilots were probably “bitchers and moaners” and were never satisfied with the type of plane assigned to them, and their accounts describe the bad features of their plane.

Test Records for Flyable Historic Aircraft.

 

There are many WWII-era aircraft that remain in existence and in flyable condition, although alas fewer than we would prefer and fewer as time passes. Efforts have been made to qualitatively or quantitatively determine flight models of such planes based upon the existing flyable aircraft. That is useful, but again not necessarily accurate.

In the first place, most of these flyable historic aircraft probably receive more love and attention from the owners than do the family members of the owners (ha ha). They are generally “pampered” planes, with extraordinary amounts of care, maintenance, and attention. As such, they are not necessarily emblematic of a typical aircraft of that type in the field during WWII. They often have had upgrades in particular components, lubricants, and/or fuels over wartime versions. They also are probably not flown as “hard” as they might have been in real-life WWII (e.g.- a 54-yr old owner of a historic P-47D-22 worth $1 million, who has a wife and three kids, isn’t going to necessarily push the plane as hard as might a 20-yr. old “hot-shot” bachelor USAAF second lieutenant in combat with Bf-109G-14s over the Ardennes in January 1945), and modern test results may thus not necessarily reflect actual WWII capabilities.

Secondary Sources.

Secondary sources such as books, magazine articles, online articles, and even forum chat can provide useful information about particular WWII aircraft and their flight models. Unfortunately, the question then becomes “on what source(s) did this secondary source base its opinion of a particular plane’s flight model”, which then brings up all of the issues discussed above.

General Impressions among Aviation Enthusiasts.

Frequently, a blanket statement about the real-life flight model of a particular aircraft based upon a general understanding of the community, such as “Everyone knows that a Spitfire can overturn a Bf-109". At times, there is some usefulness under a “wisdom of the crowd” theory (e.g.- if everyone thinks something, maybe there is something to it). On the other hand, as we all know, such a statement is frequently subject to bias, prejudices, and ignorance. Even if the statement is based upon a more educated understanding, we still have to address what sources were reviewed in coming up with that basis, and thus encounter the problems with sources discussed above.

Limitation of Sources.

It is also important to remember that not all of the above-described sources are available for each model of each Il-2-game aircraft. Some models of planes only have some of these sources, but not others. Depending upon what sources are available, or not available, there may be errors in our understanding of the flight models of particular aircraft.

Conclusion (to the cheers of all readers - ha).

In conclusion, in my opinion, there is NO single perfect source for the real-life “flight model” of aircraft in the Il-2 game. Each source has advantages and disadvantages, and none of them has the whole story. The true “flight model” (if one can exist) probably takes into account several of those sources. Consequently, before spouting off about how a particular flight model is unrealistic, consider your sources and the pros/cons of each of those sources and the fact that the “other side” might have its own legitimate [though also partially flawed] sources supporting its position. In short, each side is probably “right” to some degree, and each side is probably “wrong” to some degree.

We must all recognize that there is a relatively large “grey area” concerning real-life flight models of the Il-2-game aircraft. In fact, as many of you recognize, there are also dramatic differences of opinion on what constitutes the “in-game” flight models of particular Il-2 aircraft (e.g.- Hardball says X plane has a maximum speed of 550 kph at 1,000 m, Il-2 Compare says 565 kph, Official Manual says 545 kph, and my own experiments on Quick Mission Builder show 580 kph). In theory, the in-game flight models should be more readily capable of accurate measurement and determination than 60+ year old “real life” flight models, yet we still constantly argue about the in-game flight models.

Quite simply, we are doomed to argue and argue over these issues, and no one will likely ever have the “absolutely correct” answer. To be clear, I’m not saying that everyone’s argument is just as good as everyone else’s. Instead, some arguments are definitely wrong, some arguments are partly correct, and some arguments are mostly correct, but no arguments are absolutely and completely correct. When discussing in-game flight models, people ought to simply avoid blanket statements, and recognize that there can be reasonable, good faith opinions and arguments on the other side.

Just remember that when Storm of War finally comes out, we’ll all get to re-visit longstanding arguments over flight models of each and every aircraft (plus of the newly-added aircraft). I, for one, am studying up on the specifications of the S.25 Short Sunderland Mk. I Flying Boat so I can be prepared to argue about its flight model “ad nauseam” when SOW comes out (ha ha).

Edited August 30, 2013 by MackStones

[leitmotiv]

Since December 2012 they talk about BoS FM, DM http://forum.il2sturmovik.ru/topic/122-fizika-i-aerodinamika-v-bzs/?p=3786

Who is this Oleg guy, theirs new guy in town hope he comes back soon from vacation and joines guys in comments on stream

[Skoshi_Tiger]

A very interesting read. Which basically can be summed up with that depending on the sources used by the developers there will ALLWAYS be arguments about flight models if they don't match a particular customers sources, anecdotes or gut feeling of what the flight model should be.  

 

Luckily the developers seem to fairly thick skinned and will be able to deal with it. They have their sources and I sure there will have to be a fairly well thought argument containing appropriate verified sources of information, presented in an eloquent and appropriate manner for the information to be accepted. Then it would be up to time and availability of resources and priority to see if the changes (ever) get implemented!

 

Sometimes it comes down to (as I often have to tell my kid while I'm preparing their school lunches) "You get what you get and your don't get upset!". In my opinion this (or for that matter any) flight sim is a work of art. The developers as the artist have the right to implement their vision of what it should be. The sims popularity or whether it is a failure  or ( hopefully ) a success will depend on how enjoyable and immersive experience flying in the sim will be. Flight models are only one part of that.

 

Cheers!

[LukeFF]

TL:DR

[6./ZG26_Emil]

TL:DR

[Heywooood]

Mack - that was the most detailed, lucid, articulate post I have read in a while....(not to mention the longest)

 

butt - it won't matter more than a fart in a hurricane when the time comes, even as I appreciate your heartfelt concern (dread?) over the coming...errmm hurricane anyways.

 

once the BoS is released, you may have to dig through a great many gripe threads to find anything readable..

 

its just the nature of competitive, online combat flight simulation - not that the offliners wont be doing it too - its just that they arent as strident because there are no 'points' to lose

 

its inevitable...let it happen..its entertainment

[FG28Artist]

TL:DR

A valid point of view: Your volontary and self-imposed total abstinence from any FM-discussions hereafter has been noted in the annals. ;-)

[6./ZG26_Emil]

A valid point of view: Your volontary and self-imposed total abstinence from any FM-discussions hereafter has been noted in the annals. ;-)

 

[MackStones]

TL:DR

I guess the title of the post didn't clue him in.   (ha ha)

 

Shoski_Tiger, you are correct.  Due to a multiplicity of sources, each of which might have their own errors, biases, or incompleteness, it is difficult to ever get a completely authoritative flight model (not to attention the difficult task of trying to recreate it in a flight sim).  In fact, you'd be hard-pressed to get a completely authoritative flight model for a modern aircraft.  Are you going to rely on the specifications that a big aerospace company uses in its marketing materials?  Are you going to rely upon test pilot data?  There is little aerial combat these days, and mock battles have self-imposed limits, so it's hard to say what the modern aircraft can do when pilots are pushing them to the extreme in combat.  Even if you study the limited modern aerial combat data, there usually has been such a disparity in skills, and/or logistical/command & control back-up, to make a fair evaluation difficult.  Plus, much of the data is secret anyway now.  Now, try to go back 70 years, with the understanding that much of the data might have been destroyed in battle, bombing, etc. or lost in the meantime.  That's my first point.  The second point, as you note, is that there remain sources of flight modeling data from WW2, but they often differ, and it's easy for anyone of us to claim that the developers "porked" or "ubered" an aircraft based upon the source we've got in hand (even if not cherry-picked) - we just need to realize that the developers are relying upon another source (and presumably on a rough amalgamation of a number of sources).

 

I also recognize that there will be flight modeling disputes.  Like most pilots, I've never lost an aerial battle in my flight sim career in which my opponent didn't have some unfair flight-modeling advantage (ha ha).  The funny thing is that, once I switch to the "uber" plane that beats me, the flight-modeling advantage seems to go away ... I've never figured that one out (ha ha).

[6./ZG26_Emil]

This might be a revelation to you but I would be very surprised if any WW1/WW2 flight simmer worth his salt isn't already aware of this.

 

Also you could have summed that whole document up in a couple of sentences instead of writing an entire book, hopefully you enjoyed doing it anyway

[arjisme]

Mack - that was the most detailed, lucid, articulate post I have read in a while....(not to mention the longest)

It would be significantly improved and potentially draw more people into reading the entire thing if it opened with his conclusions first. Start with a summary that makes the point(s), then follow with the elaboration that justifies them.

[MackStones]

This might be a revelation to you but I would be very surprised if any WW1/WW2 flight simmer worth his salt isn't already aware of this.

 

Also you could have summed that whole document up in a couple of sentences instead of writing an entire book, hopefully you enjoyed doing it anyway 

 

We'd all love to think that, but ... forums for flight sims are nevertheless filled with thousands of pages arguing over flight modeling.  The same thing will happen here.  At Warbirds of Prey, though, the disputes slackened somewhat after I posted that, and people would be referred back to that from time to time.  They might still raise an issue, but at least be prepared to better argue for their position.

 

Yes, I could summarize that post, but a summary isn't going to convince many people, especially ones that are passionate that they (and they alone) have the actual, "true" authoritative source that all others have ignored.  Once pressed, it ends up that they are relying, for example, upon sales literature of a manufacturer (I've got an old early WW2 magazine showing Westland Lysanders shooting scores of Bf-109s out of the sky) or some "secret" report that was given to Stalin himself by an aircraft designer (yes, I'm sure that any report given to Stalin by an aircraft designer - even some famous ones were imprisoned by Stalin from time to time - would not tell him of all of the problems with the aircraft).

 

All (and I mean all) that being said, I like to learn of new sources.  The new sources can add to our understanding, even considering their imperfections.  

 

"Book"?  It's only a few pages long.  I'm a lawyer; we write thank-you notes longer than that (ha ha).

 

P.S.-  It was a re-print of an original post. I'll add a little summary upfront since I guess people are too busy watching "cat videos" to read anything (ha ha). - Apparently, it won't let me Edit the original post now; I corrected a spacing error already, and can't do it.

 

 

 

Edited August 31, 2013 by MackStones

[nynek]

Since we are 13% into the 21 century how about we start using wind tunnels and computers to model that. Then we can subtract some 10 to 30 percent from "ideal" state and we are done.

EVERY pilot is loving his her plane in which they stop sucking.

[FlatSpinMan]

MackStones - thanks for posting that, to say nothing of actually thinking it all. I'd never seriously considered al the factors that could impact upon a plane's performance, let alone when that performance itself is then filtered through a game.

I wish the people who would benefit the most from reading this would actually read it.

[Heywooood]

It would be significantly improved and potentially draw more people into reading the entire thing if it opened with his conclusions first. Start with a summary that makes the point(s), then follow with the elaboration that justifies them.

the university's editing board has noted your critique and filed it accordingly - thank you

[JaMz]

As Heywood said, It will happen anyway....Its part of the process and your never irradicate it