4.006 DM Discussion

May 26, 2020

7 hours ago, catchov said:

I agree. C'mon It's absurd to suggest the current DM is (kind of) correct based on aircraft failure during training. That is akin to politicians and their rhetorical cherry picking. I understand some aircraft builders cut corners in the manufacturing process, used inferior cheaper material etc to increase profits at the (ultimate) cost of trainees and combat pilots alike. That is why some kites just broke up. If I'm flying in a game such as this, I don't expect to be given a poorly manufactured, defective kite. That's taking reality a bit too far.

First, I said nowhere in that post (or anywhere else) that the current DM is correct.

Second, how do you know those failures were due to 'defective' aircraft? They all are caused by dives. They could be (and likely were) caused by pilots exceeding the limits of the aircraft structure - and that is relevant to the current discussion.

But the main point was really just to point out survivor bias.

May 26, 2020

I miss X-Wing vs. TIE Fighter.

Don't get me wrong, I like WWI and all that, but if ~~LucasArts~~ Disney ever gets off its ass, stops squandering the Star Wars license by making mediocre sequels and Forced micro-transaction video games, and finally gives us a proper remaster of XvT, I'm outta here. Good riddance, right? Anyway, I remember one of the last expansion/updates (ca. 2000) where they reduced the maneuverability of the A-Wing. It made sense: it was a tad faster than a TIE Interceptor, could turn with a TIE Fighter, and it had some shields (typically you'd disable these to have faster laser recharge). As an A-Wing pilot, I was pissed off. There were some people using footage from Return of the Jedi Special Edition to convince the developers that A-Wings were, in fact, able to stay with TIE Fighters in a turn. Obviously the change was made for game balance, still to this day I think it was one of the contributing factors that led to the XvT community becoming a toxic wasteland, and the end of its popularity in multiplayer. I maintain they shouldn't have touched the A-Wing: Alliance was meant to be outnumbered by Empire, but have better fighters overall. I don't think the devs truly understood their own game.

If you want an in-depth analysis (oh God, please no):

But I digress, back to FC: strong reactions regarding changes to someone's favourite plane (@SeaW0lf, @US93_Larner) are going to be met with resistance, especially this late in the development cycle. This game/sim isn't being sold to robots, we need passionate people with interest in the subject, time and money to dedicate to it, for it to be profitable in the long run. If people truly hate a change, then there's a disconnect with the developers happening. In this case, though, I think the devs do understand their own game. They want a simulation, and everything in terms of balance and multiplayer fun is a nice coincidence.

There's really nothing more to say, and I don't think it's right to insult people like @emely for wanting a different experience, because at least at some point before (and even after) release, it was marketed to them as such. I'm all for more options: more engine variants (110hp Camel, 200hp Albatros, 235hp SPAD), gunnery spread to simulate turbulence without actually killing server performance, and even stronger wings as a multiplayer server setting.

Edited May 26, 2020 by J5_Hellbender

May 26, 2020

I was late to the party, Rogue Squadron was my 1st Sim

But hell what a game !

I agree with engine variants. There's little point in replacing an engine when it's just as easy to add value to the game with a variant.

Mission builders are big and ugly enough to apply planes historically, if they wish.

May 26, 2020

I have to say emely don't self destruct yourself over this, yes it's irritating that the patch didn't change the bugged damage and even more irritating that a lot here won't admit it but getting yourself banned won't change anything.

3 minutes ago, Zooropa_Fly said:

I was late to the party, Rogue Squadron was my 1st Sim

But hell what a game !

Yep, one of my favorite games of all time.

May 26, 2020

11 minutes ago, kendo said:

First, I said nowhere in that post (or anywhere else) that the current DM is correct.

We are in agreement then kendo. I'm sorry if I misinterpreted the thrust of your arguments.

11 minutes ago, kendo said:

Second, how do you know those failures were due to 'defective' aircraft? They all are caused by dives. They could be (and likely were) caused by pilots exceeding the limits of the aircraft structure - and that is relevant to the current discussion.

It's all speculation innit. Neither of us really know. But I'm willing to bet defective aircraft and/or poor rigging, poor maintenance was a greater factor in "catastrophic failure" than poor piloting. And trainees were not in combat.

So I think using the trainee stats in this discussion is irrelevant and should be ignored. That's pretty much all I'm saying.

This tread is about apparent minor damage in combat resulting in wings falling off too easily when performing reasonably gentle manoeuvres.

May 26, 2020

I agree with a lot of what you say.

And on the DM, I really don't know? I have no problems with wing stength being adjusted a little more if it seems right and can be justified.

But a big theme that has come up in last few days is about the wing folding (in MP at least) being caused not so much by the hits, as by the high-G maneuvering afterwards - so then we get into another debate about how often that actually happened in real life and whether the sim is wrong there too. And then Bidu posts what he found about wing-folding accidents (and yes, no combat damage at all) and frankly I was surprised at the number in such a short time.

So, I think it is all relevant here.

Edited May 26, 2020 by kendo

May 26, 2020

Some interesting references posted by @SeaW0lf and @US93_Talbot.

TL;DR: While I think that hit damage currently is correct (enough), I think the hitboxes as we have them are not a good way to depict damage. The DM we have systemically disourages good aim while rewarding distance shooting. I explain why this is the case systemically. I suggest we either factor in the shooting distance in hit lethality, or there should be two kinds of hit boxes. One that only does cosmetic damage and one, representing much smaller sections of the aircraft, that makes the structure collapse this latter kind does not require any probabilties, but just hit counts vs structural strenght. What we have now is in fact something correct that does the wrong thing in the induvidual case.

I think I start this by posting some artwork, from A G Lee.

I suspect anyone drawing such would use topics that are both alarming as well as common.

Regarding wing strenght as such, we have Cecil Lewis telling us about getting shot at and presenting us with the poster boy of all survivorship bias. Here:

Of course we had no business to be taking on Huns like this. It was our own fault. And now we were going to be shot down. I tipped the machine up on one wing, diving steeply, engine full on. The Parasol was not built for this sort of flying, and I knew it. The next moment the wings would drop off. I cut off the engine and dived steeper, turning in a close spiral. By now the Hun could not get his sights on us (for the Parasol dropped like a brick), so, having secured a moral victory (and damn nearly a material one), he sheered off. I pulled the machine out gently, opened up, and we headed for home, feeling a bit sheepish. I was worried about the machine. The right-hand plane seemed loose. It shook in a peculiar way. Throttled right down, we came back at stalling speed, nursing her on to the ground.

Once down, we examined the machine. There were hundreds of holes through the planes, of which about fifty formed a group, almost severing the mainspar, about six feet to the right of my head. No wonder the wing had felt sloppy coming home ! That Hun must have held his machine wonderfully steady to put a close group through the spar like that. Said Pip : “I think he was sighting on your head, and his sights were a fraction out. He could never have kept his gun steady on a point in the plane.”

“ Maybe it was your head.”

“Maybe. It’s a lovely group anyway.”

(Lewis, Cecil. Sagittarius Rising . Casemate Publishers. Kindle-Version.)

From this we know.

- He used a spiral dive to get away.

- He is just extraordinarily lucky to be alive. He should be dead by all means. The next 500 pilots experiencing this ordeal would end up dead.

- It is possible that a plane recieves 100+ hits while not falling out of the sky.

- The spar had almost 50 hits and it held, but barely so.

How to read that? First: the spiral dive. That is where our wings come off most frequently. His didn't. Why? First, he knew that the Morane was not supposed to be flown that way, else catastrophe. In extremis, he did. How so? He cut the engine and went down straight, rolling. Why didn't the wings come off even though they "shook"? (Wing shake!) Simple, he didn't do any pulling on the very light and overbalanced elevator. It is evident that he didn't, as he specifically mentiones pulling out "gently". Then, he returned "almost at stalling speed" due to fear it would come apart. I doubt he ever went beyond 1.5 g in that whole maneuver.

In a nutshell, we have what corresponds to 1.5-g-wing-comes-off-damage in our sim. How do we have that implemented in the sim? @unreasonable posted @AnPetrovich's DM data above:

We see that 50 rounds would get us where the wing would come off as long as it hit the right kind of wood. So if the graph above is from the Camel wing being shot from dead astern, where the Spar would likely be hit, these numbers are rather perfectly in line with Lewis' account. Both chances and number of hits are consistent with "literature".

The third point, that planes *can* be perforated to an obscene degree while still remain (somewhat) flyable is now a problem that cannot be reconciled with the DM mechanics that give every bulled a lethal potential, regardless of where it hits on the aircraft.

What we have here is the classic case that statistics of big numbers are utter meaningless to the individual case. Same as it is of utter irrelevance to you individually whether your chance of developing cancer is 10% or 25%, yet for your insurance it is of great importance. For you, it is still only a question of either healthy or not. You cannot be 20% dead. What the DM as we have is doing, it negates specific aim and the individual case. Unless you hit meat or metal, it does not matter where on the aircraft you hit. And the wings are about the largest target in a dogfight, hence it is mathematically certain that the wings are the first to go. This makes shooting from distance so much more effective, much more that it would be in real life. In Lewis' account, it mattered that the one hit group was such narrow, as only like that, it *almost* did catastrophic damage to the wing. In our game, I have no requirement to precisely hit the spar, I can hit anywhere for the same result, meaning if my target is now four times the size, I can shoot from 200 m with the same difficulty as in real life would require a shot from 50 m.

This is what applying the statistics of big numbers to the individual case does: It makes shooting FAR easier. In RoF, the added dispersion was the crutch to make you go to 50 m again instead to 200 m.

Now why is our damage model systemicaly wrong, while it is in fact tuned to plausible values? The statistics of large numbers vs. the individual case. How does that work? this is how:

This is a target:

I will shoot at it ten times from 300 m now.

I get 10, 8, 7, 7, 6, 5, 5, 5, 3, 2. (Note to self: one notch left and two up). A socre of 58 out of 100. I can say now my hit probability is 58%. If i took the whole target as a single DM box, then I'd bona fide give it a "58% damage probabilty" to my if ten rounds from my SIG-510 at the center is what it took for *kaboom* or whatever. In essence, the DM sees this:

As soon as I realize that I utterly suck at shooting, but my compulsive behavior keeps me doing it and I want better results, I just move closer.

That is what I'm dong now. I just move to 150 m, instead of 300 m.

This is what i get:

It is the same group of hits, but half the distance. But now my score is: 10, 9, 9, 9, 8, 8, 8, 8, 7, 7. My score is now 83! This means, the adjusted lethality for hits has become 83%!! But If I tuned the DM to 58%, Thus, by getting closer, in real life, I was rewarded with added 25% lethality of my hits. I didn't get better at shooting at all, I just moved closer and I got tremendously better at hitting. But the DM being toned to 58% still sees this:

instead of now this, with my lethality upped to 83%:

Thus, in the game, I don't get rewarded for being closer, as long as I am able to hit ANYTHING within the target. Hitting "1" does absolutely the same as hitting dead center. I have no benefit at all for moving in closer, hence the spray at distance is by a great marging my best option. The increased exposure to danger by merging closer does by far not outweight the benefit for hitting meat and metal that I could have if I chose to expose myself to lethal distance shooting.

Taken together, the DM we have can only be correct at a given shooting distance. You go closer, the same volley of shots should potentially more lethal or conversely less lethal at distance. Yet, our DM does not discriminate as long as I hit anywhere, hence it is systemically off in reprocucing actual damage.

How could one handle that?

One way is to simply factor in the shooting distance in bullet lethality. The numbers we have now basically reflect you being VERY close. At 300 m distance, you never, EVER get 50 rounds in a 10 cm group seering your wing spar as described by Lewis. It is unlikely to a degree of making it impossible. Yet at 300 m I can easily hit something as large as an aircraft, hence in the game it is a great option.

Another way would be just discarding the system as we have it and replace it by more detauiled DM boxes and no probability factoring. As it is clear that every plane is way too large of a target now (in terms of DM box that can create terminal damage) and we can't have hitboxes consistent to the static elements of the aircraft, why not just silghtly decouple the visual model from the DM boxes for simple geometry if that hitbox geometry is an issue? One could have a DM box approximating wing and fuselage shape, and this DM box would only cause cosmetic damage, round holes to ever-increasing extent. Inside that box, you can have the DM boxes for engine and pilot as well one representing a spar in a wing. Only hits there would cause structural damage and weakened wings. Lewis mentioned in his account how the wing would feel lose; hence you could introduce those rigging changes as a cue for structural damage before catastrophe happens. Also (like in RoF) at some further point in getting structural damage, you can make the bracing wires go, hence it should be clear to any pilot to stop working the stick too much.

I don't think just using our current hitbox layout and making the aircraft flying tanks by upping hit requirements significantly helps much for getting more "realistic" results. All it would do is make it impossible to do any damage other than engine and pilot, even from close range where you can in fact hit vital structural parts. And this is clearly NOT what is supposed to be. I'd say dump all those probabilities and give us proper hitboxes and just give us a defined reduction in load tolerance per hit. Just use a table. It's compute-friendly. The freed ressources should be good for another vertex on that hitbox. Or more.

(Note to self: this is too long of a post.)

Edited May 26, 2020 by ZachariasX

May 26, 2020

Damn! @ZachariasXbeat me to it while I was busy typing! Anyway: Agree with the lethality index based on distance: That would be a good add-on to the DM I'm suggesting below which does not take this factor into account.

Some input on the DM model from a structural engineering perspective (I have worked as a structural engineer on both gliders and fighter jet aircraft). Caveat: In the reasoning below I’m assuming most hits are from deflection shots from multiple passes and not on a non-manouvering target soaking up hits in the same spot!

First divide battle damage into two categories: Damage to single point of failure (SPF) structures or to structures with some form of alternative load path (ALP) capability:

SPF damage on a WW1 scout would be damage to some of the bracing wires, lugs and some strut joints etc.

Damage to ALP structure is easiest to visualize on semi- monocoque structure like a WW2 fighter wing and includes wing skin, spars (to some extent), stringers etc.

In case of the former, one bullet could be enough. OTOH the chance of hitting this SPF point is of course extremely small. In case on the latter, A WW2 wing has a very large capacity for ALP: even if the spar is damaged at some point, there is usually an ALP in the adjoin wing skin. In fact you actually count on this calculate when you structurally dimension a wing: That a part of the wing skin close to the spar or stringer helps carrying the load. However, in this case the damage is of course additive and the more structure is damaged, the less g-capability you will have left.

So a suitable DM would IMHO have two parts: Every time you get a hit in a hitbox you roll the dice and risk a SPF hit. This damage is not accumulative and the risk is extremely low given how small part of the structure is SPF. How high the failure percentage should be is of course dependent on the size of the hitbox but since we have graphic damage displayed as in punctured fabric assuming two slab hitboxes representing the wings, multiple tens to hundreds of hits would be required to trigger this IMHO. Say each round hitting the box gets a 0.995 chance for a SPF hit: This means 100 hits in the hitbox gives you about a 40% chance of a failure. The 0.995 figure is of course pulled out of a hat (But highly tunable and adjustable for each aircraft depending on ruggedness!) but you get the idea.

However, if you have gotten a hundred rounds through the wing, some of those have probably punched through ALP structure like spars (ALP in the sense that the load can “flow” around the hole but that the ultimate load capability is reduced) so here as a starting point assume that the residual g-load capability is reduced by the same amount: 0.995**100 =0.6. Or, if you prefer being more conservative (1- 0.995**100)= 0.4. (Ribs, fabric etc. should IMHO as a first order of approximation simply be left out of the analysis.)

So to sum up: True, I do not have any comprehensive statistics from WW1 on how many structural failures occurred due to battle damage (Note: overstressing due to excessive control surface deflection when shot at is a completely separate issue!) but from a structural engineering perspective adding common sense statistics analysis on top my estimate is that shedding wings due to battle damage should be an extremely unusual event as opposed to pilot/engine/flaming kills.

Finally, on the subject on banning, this should IMHO be reserved for extreme repeat offenders only and not for someone venting frustration in the heat of the moment. Maybe a cooling off period but a permanent ban? No. However, the “banning DM” should be accumulative and also include those who instigate conflict by goading.

Edited May 26, 2020 by Holtzauge

May 26, 2020

Beautifully written @ZachariasX but i afaik that is not how the DM is working. For the diagram below, imagine the whole target is the wing, and the black area is the spar. Then the DM doesn't register 58. Everything on 6 or farther out hits the wing but doesn't register as weakening the spar. DM may not distinguish between a hit dead centre (9 or 10) V a glancing hit (two of the 7s in this diagram), or they may have a way to score that too. I don't know.

So long as the spar dimensions V overall wing dimensions are correct, then for big numbers there will be no appreciable difference between the hits registered using this probabilty method and one using a more detailed hit box model. And remember they have said this would be too heavy for current PCs. That's the whole reason they went this probablity route

Also, who deliberately aims at spars from 300m, or even 30m? They can't be seen. From 6 0'clock you almost can't avoid hitting a spar if you hit a wing. And from other angles their method should work, if their dimension numbers are right.

Edited May 26, 2020 by kendo

May 26, 2020

45 minutes ago, ZachariasX said:

(Note to self: this is too long of a post.)

Long, interesting and accurate post, but I disagree with shot lethality being wrongly modeled. We tend to open fire much earlier, because we know that hitting anything even non-critical such as (most of the) wing is still going to secure us a probably victory... because people overstress their planes like mad when they get shot at.

Back in Darling's glory days, it was easy to distinguish an AI from a human player: their flying was more coordinated, more predictable, and in general they would survive longer. AI Fokker Dr.I especially were annoying as hell, no amount of damage would ever cause them to catastrophically overstress their wings and I'd have to move in close for Darling to score a pilot kill. Human pilots who looked like they were done for and fled home with a critically damaged plane we always let go. Sometimes we'd even escort them home if there were friendlies around who might have finished the job, we made a point of that. It was very few of them indeed, most people choose to fight regardless of combat damage. For my part, as soon as the Bristol had any wing damage whatsoever, we were on our way home. Typically we were not shown the same kind of mercy for running away, nor did we expect it.

It's true, I do feel some vindication for Darling shooting at wings for our entire RoF career. It was never with the intention of shooting them off, but with every intention of the other guy to pull them off for us. Case in point: when facing Gothas we always went straight for the under-the-belly pilot kill, and if at all possible, we'd do the same for scouts too if they hadn't spotted us creeping up on their low 6.

Again I have no problem with a multiplayer option to decrease shot lethality, since we are far more accurate with our 2D displays than real life pilots, and yanking on the stick is a natural reaction when you get shot at, which I don't think should be punished this catastrophically. Even though @Chill31 says he only needs to pull 40% of his stick back to reach 3-4g, I still think that the sheer strength required to pull more than that might deter you from doing so and there is some kind of force/audio/visual feedback missing.

May 26, 2020

5 hours ago, unreasonable said:

The DM is modeling the effects of compound and/or cumulative damage. The probability of hit # 10 causing a breakage is higher than that of bullet #9, and so on. Meanwhile the visual damage and aerodynamic penalty is, I think, mounting up in a more "Hit Points" sort of way.

I think this is where my want for Spar Hitboxes (oh no, not that again!) comes in - despite all my rather strong opinions here, I'm not trying to bash the Devs' work, but I keep envisioning a circumstance where an aircraft shoots 60 rounds at another, catches the trailing edge of a wing with 20 of those rounds, and should be hitting nought but empty canvas and a couple ribs - but instead the DM calculates one of the rounds as a Spar Hit - ergo, severe structural damage where there might not be any in the 'real thing'. Perhaps I'm focusing on this a little too much.

I guess I'd try to summarise it as: "Yes, the chance to hit a Spar from certain angles is LOW, but you CAN miss the spot where the physical Spar is and register a hit on it. With a hitbox, you only hit the Spar if you hit the Spar. I hope that's reasonable enough.

But, like you say, if it has the desired effect in the eyes of the Devs - then that's just that! ?

5 hours ago, unreasonable said:

1) What are the maximum permitted G-loads of undamaged aircraft

2) How do these limits change with combat damage

3) Does the DM offline give a reasonable representation of the historic data

4) Does the DM online produce the same results as offline.

Great questions. I'm certainly not qualified in any way to try to answer 1 through 3, but I could give 4 a shot for sure. My initial answer would be 'No', and it seems very apparent that this is due to Human flying aggression vs. AI aggression. I've posted this graph earlier, but I'll do so again to show just how large that divide is:

Now, I wonder how similarly these would match up if I got one of my buddies to 'Fly like an AI' ?? Or even if I shot up some AI on our training server? I'll definitely look at giving that a go.

5 hours ago, unreasonable said:

I would not be at all surprised if the complaints here of Vpilots losing wings after very few shots, even if factually correct, show loser bias.

This crossed my mind...I've started recording the type of loss for all our pilots during our 'ops nights' for this reason (although this is a little harder to do as we don't report on these in the same way we report on kills)

@ZachariasX That's a really well-written post!

Edited May 26, 2020 by US93_Larner

May 26, 2020

3 hours ago, Zooropa_Fly said:

What I find sad here - is people who don't play a game, trolling it's forum at every opportunity for the sole reason of : winding up certain individuals ; hoping to get people banned by provoking reactions ; getting personal kicks from it. If that's not the psychology of a 'bully' then I don't know what is ? But if you're happy exuding that, fine.

You left ww1 behind some time ago, perhaps it's time to leave the ww1 forums behind too ?

I'm not denying your right, just pontificating that's it's not a particularly profitable use of your time. Unless the superiority complex is integral to your survival ?

The only missing feature on this forum is automatically hiding people's quotes of other people one has blocked (my, do they still exist?!). But in the case of this post of yours, the profit far exceeds the loss. Thank you! ?

May 26, 2020

2 hours ago, ZachariasX said:

Some interesting references posted by @SeaW0lf and @US93_Talbot.

TL;DR: While I think that hit damage currently is correct (enough), I think the hitboxes as we have them are not a good way to depict damage. The DM we have systemically disourages good aim while rewarding distance shooting. I explain why this is the case systemically. I suggest we either factor in the shooting distance in hit lethality, or there should be two kinds of hit boxes. One that only does cosmetic damage and one, representing much smaller sections of the aircraft, that makes the structure collapse this latter kind does not require any probabilties, but just hit counts vs structural strenght. What we have now is in fact something correct that does the wrong thing in the induvidual case.

Ietc

(Note to self: this is too long of a post.)

I agree with your discussion of the specific account, but your general comments about the DM later, I fear, miss the ~~hitbox~~ mark.

Range still matters a great deal because:

a) At closer ranges you get more hits on the wing hitboxes. That should be obvious. Hits on the fuselage have no effect on wings, AFAIK, it certainly matters where on the target you hit. You also get more hits on the pilot and engine hit boxes, which are still an order of magnitude more effective at ending a fight than a wing hit even after 4.006

b) Assuming the DM works as described in earlier developer posts about BoS, the energy of hitting projectiles matters. Energy is a function of mass and speed, speed is a function of range. You will note that in AnP's graphs posted earlier, the range is specified,as is the precise gun used. I expect that if he performed the same experiment at a much longer range the results might be shifted to the right a little - ie more hits needed to break a spar. Over a normal shooting range it is true that bullets have enough energy to break canvas and wood, so it might make little practical difference, both in the DM and in RL.

It is true that you cannot aim for a specific wing spar, since a hit on the wing box is determined as a spar hit probabalistically, but you have to ask whether anyone has ever actually tried to aim at a wing spar in real life - even in the account you give the protagonists think that the shooter was aiming at the pilot - which is a hit box of it's own. This was true before the DM change and is true afterwards.

As for having loads of little hit boxes - however small you make them, you still have to have a probability distribution for the results on each of them. As AnP has explained, hit boxes take up a large amount of resources in the hit recognition stage, which is why their number is an issue.

People complaining about wings failing after a very few hits are not complaining because the hits did not hit where their spar happens to be. Many of them may not even know where that is - I certainly do not. They are complaining because their wings fell off (or were torn off). This has nothing to do with the mechanics of the DM, just the choice of G-limits and how much damage affects said limits. I know that this can create anomalies where you fire your own gun into an area of the wing that in RL has no structural importance - but for actual gameplay this is a complete red herring.

Edited May 26, 2020 by unreasonable

May 26, 2020

1 hour ago, Holtzauge said:

[...] So a suitable DM would IMHO have two parts: Every time you get a hit in a hitbox you roll the dice and risk a SPF hit. This damage is not accumulative and the risk is extremely low given how small part of the structure is SPF. [...]

Food for thought.

One immediate problem I have with that (as far I make sense of it now) is that me, as an attacker, am till not given any incentice for a close in fight.

I mean, my post above basically revolves around the idea that any DM is a two way road. Any definition on how the victim will fare, automatically produces a most suitable behavior for the successful assailant.

I have no doubt that in essence that with the current DM that I, on the receiving end, get what I deserve even though my Camel wings come off like all the time. Now, I'm not the only one contributing to my misery (even tough I'm far better at that than I care to admit), but it is also how an attacker would best tackle with me to get best result (for him).

As @kendo has probably a more detailed understanding of the DM than I have, his expanation of the DM intrinsics still make it ineficient to apprach a target closer than I require to land hits on the whole target, as still every hit rolls the dice to my doom again. I am still not rewarded for going super close. As it is described, it just reduces the chances of spar damage to soome percentage, but it doesn't change how I am supposed to play the game successfully.

Now, If I shoot at 300 m (as in the game it is super easy to aim, try that in a real biplane with the face in the wind, good luck) and I spray over the whole silouete of the aircraft (easy from 300 m), maybe 70% of the surface is wings, 40% is fuselage and 5% is metal and meat. (It adds to more than 100%, as surfaces are stacked, but enter your own numbers here.) So if I spray over that silouette, the wings take most hits, yet they are structurally weaker than the fuselage. A lot weaker. The pilot and the engine are small enough that I probably won't hit them at all with 100 rounds. But 100 rounds on the whole area give maybe 70 hits on the wings. If it takes 50 hits for them to collapse, voilà! I will get the same result from shooting from closer behind, still a volley of 100 rounds and he goes down after 50 in the wings... but at what price? During the merge I expose myself as well. I just have nothing to gain and everything to lose.

The DM is to me the single most important determinant in how we play this game. Hit numbers alone are not the full story. If I have no requirements for hitting, I won't play in a way that would cater such requirements that were clearly present in the real world back then. You don't have to be a bad person to spray and pray, not being stupid is enough. Why should I go in close if there is nothing in the game requiring me to do so to win?

I disagree that we cannot hit small SPF's in aircraft. We can shoot tremendously well on average when we gore each other on Flugpark. If it took 300 bullets somewhere on the wing to "make the difference", yet I knew those SPF on the wing, then by all means we would go in at arms lenght and and shoot those. We can hit a 10 cm structure at 40 m. If such would be rewarded, people would fly accordingly.

23 minutes ago, unreasonable said:

a) At closer ranges you get more hits on the wing hitboxes.

Ahh.. you were quicker. And no, there is a distance, from where you ALWAYS hit the aircraft. Any approach closer than that is a net negative for the attacker. We do have good aim.

23 minutes ago, unreasonable said:

b) Assuming the DM works as described in earlier developer posts about BoS, the energy of hitting projectiles matters.

I'm perfectly fine with the DM related to plain damage to the victim. It is what it translates to for the attacker that I'm concerned about.

Any machanics sets incentives. Our biggest incentive now is spray and pray. Actually, you don't have to pray. If you can land hits at 300 m, it's just "Amen".

23 minutes ago, unreasonable said:

As for having loads of little hit boxes

E.g. for the wings, you could have just one for "the spar". Andthen the whole thing as one box for cosmetic alterations. Already that would be a huge game changer.

Edited May 26, 2020 by ZachariasX

May 26, 2020

55 minutes ago, US93_Larner said:

I think this is where my want for Spar Hitboxes (oh no, not that again!) comes in - despite all my rather strong opinions here, I'm not trying to bash the Devs' work, but I keep envisioning a circumstance where an aircraft shoots 60 rounds at another, catches the trailing edge of a wing with 20 of those rounds, and should be hitting nought but empty canvas and a couple ribs - but instead the DM calculates one of the rounds as a Spar Hit - ergo, severe structural damage where there might not be any in the 'real thing'. Perhaps I'm focusing on this a little too much.

I would think the odds of something like that happening would be very low. It could happen, but we're probably talking lottery win probabilities, but yes, in those few cases the DM would get the spar hits wrong.

There is another issue though - and Emely would probably mention this, so as he isn't here for a bit ? I'll say it for him - it's how you calculate the amount of damage on a spar when it does get hit. The devs have i'm sure used their best research and modelling in the attempt to get this right, but as they have admitted themselves no-one really knows. It's here that there could be room for adjustment maybe.

But then the whole AI v human MP wing loss issue comes right to the fore again. If they do adjust wing strength to make the results look better for high G MP players, then that will throw the singleplayer AI to a place where wings never come off at all...!

So, maybe that suggests again the real issue is that people fly too hard on MP?

May 26, 2020

29 minutes ago, ZachariasX said:

If it takes 50 hits for them to collapse, voilà! I will get the same result from shooting from closer behind, still a volley of 100 rounds and he goes down after 50 in the wings... but at what price? During the merge I expose myself as well. I just have nothing to gain and everything to lose.

It does not take n hits for them to collapse - just the p of collapse increasing with each incremental hit on the same wing hit box. Most biplanes have 5 or 6, I think.

The Camel Behind example, from which I suspect you have got the number 50, is about hits on one hit box, taken at an angle where a high proportion of hits would roll a spar hit. I expect (do not know for sure) that the increase in probability of successive hits causing a spar break is hit box specific. ie So 50 hits spread over a number of hit boxes, even from the same angle, has a far lower probability of causing a break than all fifty in one box. (Looks like I may have to work it out - there goes the evening....)

In an actual fight where you are not firing from directly behind at close range, as in AnP's chart, the individual hit probability to break a section is far lower, as his other chart of comparing behind with 90 degrees above illustrates.

The wing hits that people are complaining take their wings off are doing nothing of the sort - the players are.

Without knowing exactly how accurate the DM is in it's allowed loadings, we have a simple choice:

1) Change the way we fly, or

2) Create a DM which we know to be unrealistic in order to offset the behavioural traits of MP gamers.

3) Very much my preferred option, have fun optional DM setting with lowered lethality. People could chose this on or offline, just as they can fly with unlimited ammo or invulnerability now.

I expect, however, that (3) would not actually be a popular decision for MP: my observation of gaming over many genres and years is that there is a strong tendency for the competitive types to want everyone to fly at their preferred settings while simultaneously claiming that these are the most realistic.

Edited May 26, 2020 by unreasonable

May 26, 2020

The problem with adding more hitboxes or hitbox behaviour is that it adds complexity, which then requires the netcode to be reviewed.

We know from historical reports that a single well-placed bullet could bring down an entire wing of a two-seater flying straight and level. The odds of that happening are astronomically small, but then again a lot of fighting happened during WWI, so it was meant to happen at least once.

In FC this is currently impossible. As far as I know there is no single wing hitbox which can be hit once and cause an entire unloaded (1g) wing to fail. The devs could add a randomizer to make that happen anyway, say a 0.01% chance, and a 50% chance of wing hits count as "duds", as they are passing through the wing harmlessly. In the same vein we could have random engine failures, random complete gun stoppage, random structural failure. None of this is implemented in the WWII portion of the sim either, and I'm not sure if it would be welcomed or not.

May 26, 2020

2 minutes ago, J5_Hellbender said:

The problem with adding more hitboxes or hitbox behaviour is that it adds complexity, which then requires the netcode to be reviewed.

We know from historical reports that a single well-placed bullet could bring down an entire wing of a two-seater flying straight and level. The odds of that happening are astronomically small, but then again a lot of fighting happened during WWI, so it was meant to happen at least once.

In FC this is currently impossible. As far as I know there is no single wing hitbox which can be hit once and cause an entire unloaded (1g) wing to fail. The devs could add a randomizer to make that happen anyway, say a 0.01% chance, and a 50% chance of wing hits count as "duds", as they are passing through the wing harmlessly. In the same vein we could have random engine failures, random complete gun stoppage, random structural failure. None of this is implemented in the WWII portion of the sim either, and I'm not sure if it would be welcomed or not.

Agreed , except that it might be possible with one hit, although it may be that the first hit can only "prime" the "spar" for a later hit. I do not recall any explicit developer comment on that. It is possible that his graph showing 0 cases at 1 hit is just an artifact or a random result, possible with such a small number.

With 3 hits certainly. Camel behind ~5 cases out of 10,000 after 3 hits, (0.05%) so 0.01% would be about right as the starting p.

Data off this graph, just in case anyone has not noticed it in the original DM thread.

May 26, 2020

1 hour ago, ZachariasX said:

E.g. for the wings, you could have just one for "the spar". Andthen the whole thing as one box for cosmetic alterations. Already that would be a huge game changer.

I was considering the modelling of Spars a little more today.

Just out of pure interest, let's quickly take a look at this. AnP very briefly explained the positioning of the hitboxes regarding the SPAD XIII to me in an earlier thread:

Quote

That's correct! Moreover, not only for the visual DM, but for the skin DM (and aerodynamics destortion) as well. I guess you guys would like to see all of your hits at enemy's wings, even if bullets "hit the air".

The SPAD XIII.C1 in FC has 10 hit-boxes for the top wing, including 7 wing's sections, two ailerons, and one top fuel tank. If we want to put additional hit-boxes for the spars, we have to add 14 new hit-boxes: two for each section of the wing. Otherwise, if we add only two hit-boxes (one per spar), we won't be able to detect which exactly section of the spar (wing) should be broken, and also there will be a possibility to wrongly "hit" the spar in the area where the section of the wing has already been lost (this is not a critical issue since the wing is broken and the airplane goes down, but... still weird).

By the way, to break a spar, firing, for example, from the 'dead six', you need to get a few hits at the same place of the spar, lengthwise. The minimum number of hits, placed like this, which are needed to break the unloaded spar simply depends on the spar's 'height'. What is the chance, that you put all of your hits at the same place of the spar, lengthwise? It depends on its length, obviously. Therefore, the average number of hits for breaking some section of the spar, also depends on its length. If the section is long, you need to make a good number of shots, untill at least few of them will reach the same spot. But sometimes it might happen occasionally faster. Or, instead, longer. This is how the probablity theory works.

Certainly, if the spar is loaded (because of aerodynamic forces) then it needs less number of hits at the same place of the spar for breaking it. The most damaged place of the spar breaks the faster the higher G-load is.

Let's see if we can find the positions for those top wing hitboxes. My guess would be something roughly along these lines:

If we number those out we can see roughly what the hitboxes cover; Namely, Central section, Central fuel tank, inner 1/3, middle 1/3 and outer 1/3 of wings, and the ailerons.

Now let's add some spar hitboxes. I would imagine that the Devs would want to model the spars divided up between each 1/3 of the wings, like so:

...So we've just jumped from 10 hitboxes to 22. That's quite a significant leap, and as you can see we now have a highly complex layout of hitboxes for the upper wing. We haven't even done the lower wings yet! If the Devs modelled each wing spar as just one long hitbox (which could potentially be more, or less, realistic, I don't know), that would be 14 hitboxes - which is a little more reasonable perhaps from a workload perspective.

But, let's try to put this into the context it deserves - say AnP decides "Okay. We'll model wing spars for Flying Circus". He then has to add the extra, what, 12 hitboxes for the spars in each SPAD wing? Then the 6 for the Albatros, then the ones for the Camel, the S.E, the Dolphin.....etc etc.

THEN, because FC is only a module of IL2, the feature would have to be brought to the other modules. Let's say, just for talking's sake, that all aircraft in IL2 only have one spar per wing. Even if the Spar was modelled as its own hitbox, that would be 142 new hitboxes required.

...142 hitboxes that would need to be individually applied, tested for bugs, coded to interact with bullets of different calibers, coded to detect and react accordingly to collisions, coded to detect and react accordingly to over-stress, coded to represent the material the Spar is made from....etc. etc. etc.

That is a MAMMOTH task for the devs, and would constitute a MAJOR update. One I doubt we'll see for quite a long time, if ever.

Don't get me wrong - I'm in the 'model the spars!' camp - but I also have to appreciate that the task of doing that, simply put, is absolutely unrealistic in the context of a side-project FM tweak.

However, if @AnPetrovich ever wants to make that huge leap into the next level for an IL:2 damage model, perhaps he'd consider starting with the Albatros or SPAD as his first 'test subject'

Two questions in regard to Spars, which may be a little more realistic in terms of a DM update, are:

1) If an aircraft with two sets of spars in each wing receives damage to only one spar, will that wing be more resilient to G-forces with its remaining intact spar?
2) If so, is this represented in AnP's current system?

Edited May 26, 2020 by US93_Larner

May 26, 2020

2 minutes ago, US93_Larner said:

THEN, because FC is only a module of IL2, the feature would have to be brought to the other modules. Let's say, just for talking's sake, that all aircraft in IL2 only have one spar per wing. Even if the Spar was modelled as its own hitbox, that would be 142 new hitboxes required.

...142 hitboxes that would need to be individually applied, tested for bugs, coded to interact with bullets of different calibers, coded to detect and react accordingly to collisions, coded to detect and react accordingly to over-stress, coded to represent the material the Spar is made from....etc. etc. etc.

That is a MAMMOTH task for the devs, and would constitute a MAJOR update. One I doubt we'll see for quite a long time, if ever.

It is. I hope it never happens, and we go straigt to soft-bodies (*cue saxophone music*).

On top of that, even though the DM is calculated per wing and the lift of both wings is added up with some complexity, we are actually flying monoplanes. At least in the eyes of the engine we are. Unless this was changed from RoF, though I don't believe so as many of the FMs were ported straight over.

May 26, 2020

47 minutes ago, unreasonable said:

Without knowing exactly how accurate the DM is in it's allowed loadings, we have a simple choice:

1) Change the way we fly, or

2) Create a DM which we know to be unrealistic in order to offset the behavioural traits of MP gamers.

3) Very much my preferred option, have fun optional DM setting with lowered lethality. People could chose this on or offline, just as they can fly with unlimited ammo or invulnerability now.

I did a couple tests this morning in SP to try and determine how easy this would be. I haven't sorted the footage I gathered into hard data yet (Which is going to be hard - lots and lots of variables), but one thing I have determined is that, depending on which aircraft you're flying, there's a big difference in how feasible this might be when trying to put it in the context of a MP dogfight.

For example - the Albatros in a full turn from roughly 160 km/h, with the nose dipping slightly below the horizon, was pulling around 4 G. Trying to keep the nose above the horizon, that would go down to about 2.5 - 3 G. Split-S's were right out, coming in at 5 G with a tight Split-S, and roughly 4 G with a loose Split-S.

I think that it's reasonably do-able to fly that aircraft in a more conservative way (although I might not agree with myself the next time a Camel settles in on my six).

The SPAD was a little more worrying. Trying some dives like I would normally do - nothing too extreme, maybe about 300m down as if I was BnZ'ing someone then pulling out of the dive at what I would deem a 'safe' angle - was pulling over 5 G consistently - and often over 6! Trying to pull very gently out of the same dives was still over 4 G. A 'breaking turn' in a SPAD cruising along at its usual speeds would be pulling over 4 G every time.

In general, once I started trying to fly 'safely', I was still finding it hard to stay below 4 G with any kind of manoeuvre in the SPAD, simply due to its speed.

I want to fly all the aircraft a little more and firmly establish a 'baseline' for how many Gs you can typically be expecting to pull in certain situations / manoeuvres, and then test them for how many Gs they can then take after certain amounts of bullet hits, to get a more firm idea of how plausible 'Option 1' is

Edited May 26, 2020 by US93_Larner

May 26, 2020

Where did this assumption that what we have now is accurate come from?

Its different from before, it might even have better science behind it. But more accurate in terms of how it plays out and is experienced in-game? Yeah nah.

May 26, 2020

2 hours ago, J5_Hellbender said:

Again I have no problem with a multiplayer option to decrease shot lethality, since we are far more accurate with our 2D displays than real life pilots, and yanking on the stick is a natural reaction when you get shot at, which I don't think should be punished this catastrophically. Even though @Chill31 says he only needs to pull 40% of his stick back to reach 3-4g, I still think that the sheer strength required to pull more than that might deter you from doing so and there is some kind of force/audio/visual feedback missing.

I hope I didn't post that. If I did, it was an accident for sure. I won't try to correct myself here as it wouldonly be my unsupportedestimate. I have made all preparations for flying the Dr1 with a G meter mounted in the cockpit. I need a dry day now to fly it and post actual results.

Edited May 26, 2020 by Chill31

May 26, 2020

3 hours ago, Holtzauge said:

[snip]

So to sum up: True, I do not have any comprehensive statistics from WW1 on how many structural failures occurred due to battle damage (Note: overstressing due to excessive control surface deflection when shot at is a completely separate issue!) but from a structural engineering perspective adding common sense statistics analysis on top my estimate is that shedding wings due to battle damage should be an extremely unusual event as opposed to pilot/engine/flaming kills.

We have historic data which suggests that shedding wings due to battle damage (obviously in real conditions a combination of damage and pilot action) was very far from extremely unusual.

Eg MvR's combat reports, which explicitly mention victims losing their wings on many occasions. As posted earlier.

May 26, 2020

9 minutes ago, unreasonable said:

We have historic data which suggests that shedding wings due to battle damage (obviously in real conditions a combination of damage and pilot action) was very far from extremely unusual.

Hopefully soon I'll be supplementing this with Franks' other analysed claims for the other German Aces, as well as historical combat reports from the 93d / 103d Aero Squadrons. I'll need to analyse these myself - I don't expect to do even half of as good a job as Mr. Franks - but, I'll do my best to present my findings in a way that reflects their plausibility to what 'really happened'.

10 minutes ago, Chill31 said:

I have made all preparations for flying the Dr1 with a G meter mounted in the cockpit. I need a dry day now to fly it and post actual results.

Absolutely cannot wait! Will there be any filming involved? ?

Edited May 26, 2020 by US93_Larner

May 26, 2020

2 minutes ago, US93_Larner said:

Absolutely cannot wait! Will there be any filming involved? ?

No doubt! Planning a head mounted go pro with the G meter in view and a rear facing go pro to see the tail surfaces.

May 26, 2020

9 minutes ago, Chill31 said:

No doubt! Planning a head mounted go pro with the G meter in view and a rear facing go pro to see the tail surfaces.

Brilliant. Thanks kindly for taking the time!

---------------------------------

Re: Additional analysed claims, here's just a little taste. I still have to analyse the claims of von Tutschek, Wolff, Immelmann, Voss, Göring and LvR, as well as sorting the data into greater specifics before tackling the USAS claims. 'Not Specified' can refer to a number of things - either the claim was difficult to historically verify, the details were not mentioned / are unknown, or in some cases the victory might not even have been scored! Therefore, I felt it was important to leave it in.

Edited May 26, 2020 by US93_Larner

May 26, 2020

Interesting - I suppose the best explanation of so many crashed/FTL is that these were a combination of pilot wounds, engine damage or loss of control for other reasons. I would generally be inclined to include them in the "meat and metal" section if we simplify and have only three, which seems reasonable since we are interested here mostly in the % of structural kills:

1) Structural collapse

2) Pilot/engine damage

3) Flamers

Having only one identified flamer is interesting, although both of these aces were killed fairly early on (Nov 1917 and Jan 1918) so they missed much of the incendiary era. If you add their own deaths to the sample, you will get two more certain flamers.

Edited May 26, 2020 by unreasonable

May 26, 2020

5 minutes ago, unreasonable said:

Interesting - I suppose the best explanation of so many crashed/FTL is that these were a combination of pilot wounds, engine damage or loss of control for other reasons.

Right on the money. Typically (but not always) a wounded pilot would immediately land. Most other Forced to Land cases were engine out / fuel leak, and several cases of 'crashed' were attempted FTLs with shot out controls.

One very interesting case (I'd have to find it again) had a badly wounded pilot black out and go into an extreme speed spiral dive (in a Pup, I believe) before coming to, righting himself, and crash-landing.

EDIT: I just finished von Tutschek's - he had five flamers in his tally (four in 1917, one in 1918)

Edited May 26, 2020 by US93_Larner

May 26, 2020

I still think many of the wing failures recorded in combat were not due to battle damage but simply overloading of the airframe by a pilot reacting to being under fire: Take the SE5 for example: IIRC correct then it had an ultimate load factor (as in the factor it breaks at with no safety factor) as low as 5.5g. Now check out the g-loading history below from a mock combat taken from RAE R&M 469 from June 1918: In some manouvers the pilot hits 4g even in mock combat. Now imagine you dive away from a foe and suddenly tracer wizzes by your crate: Maybe, just maybe you would jerk the stick? If you did its absolutely possible to exceed 5.5 g or higher momentarily which is all it takes. Of course some aircraft were significantly stronger and IIRC then the Fokker D7 held up to somewhere around 10g before it failed in a post war test. Most scouts would probably fall in the range between those or even lower and provided you are not limited by stick forces, causing a 10g load momentarily is certainly possible even in a WW1 scout provided you had your speed up.

BTW: here is the link to the report: Has many nice figures of other aircraft, including Camel and Bristol Fighter etc.

May 26, 2020

1 hour ago, US93_Larner said:

Let's see if we can find the positions for those top wing hitboxes.

You know, hitboxes that have no relation to internal structure I find terminally flawed. No matter what super system you tack on it, it will always have a flawed result in the individual case.

Looking at your wings, I would think that with only three (3) hitboxes, I get a far more plausible result than what we have. I need:

- leading edge (a lot of hits required for significant structural damage)

- spar (60 hits and kaboom)

- trailing edge (only does cosmetic damage)

You know, compared to the wing spar, the ailerons are utter redundant hitboxes. It just doesn't matter if they get hit compared to what a broken spar does. Somehow, people realize that structure should be represented in hitboxes, hence we have the silly aileron hitboxes. Now, who ever lost an aileron in flight? Has that been seen, ever? I could take away those two hitboxes and nobody would notice. Yet we have them, the ONLY hitboxes representing individual structure components.

May 26, 2020

I mean, technically losing the entire wing means they lost the aileron too... ?

Edited May 26, 2020 by J28w-Broccoli

May 26, 2020

I don't think there is much problem using a statistical model. We can't see each individual bullet strike anyway AFAIK so you won't know that something different occurred than should have. Whether the model is any cop is a different question though. ?

Edited May 26, 2020 by slug_yuugen

May 26, 2020

2 minutes ago, slug_yuugen said:

We can't see each individual bullet strike anyway

But how would you make statistics if you don't even know the hit numbers? What Am I missing here?

Generally speaking, I find it not an intelligent procedure to dischard info you have at hand and replace it with info you make op by means of a great algorithm.

May 26, 2020

You don't necessarily need to take a monte carlo approach to generate a model although that's definitely appealing, take a much fuller simulation of the wing/part and run that offline to generate a model. Of course that implies making a much more complex simulation in order to generate the simpler model which is a lot more work to achieve.

For example if you know the direction the shot came from and the relative dimensions of the wing and spar at that angle you can generate a crude %chance to hit a spar using (width of spar in that projection/width of wing in that projection).

Edited May 26, 2020 by slug_yuugen

May 26, 2020

31 minutes ago, Holtzauge said:

I still think many of the wing failures recorded in combat were not due to battle damage but simply overloading of the airframe by a pilot reacting to being under fire:

[snip]

BTW: here is the link to the report: Has many nice figures of other aircraft, including Camel and Bristol Fighter etc.

I actually agree: I can well imagine a novice pilot with 10 hours in type, finding himself under fire for the first time, exceeding his limits before taking a single hit.

Unfortunately there is no way to tell in the vast majority of reports of structural failure in a fight how much the damage has reduced the maximum allowed, or even if the target has been damaged at all. Typically what we get is something like "under my guns, his wings came off". (Thanks for link!)

27 minutes ago, ZachariasX said:

You know, hitboxes that have no relation to internal structure I find terminally flawed. No matter what super system you tack on it, it will always have a flawed result in the individual case.

[snip]

You mean like the pilot hitbox? Do we really need a spine, brain, and heart hitbox so that we do not mix up disabling hits with just a scratch?

7 minutes ago, slug_yuugen said:

You don't necessarily need to take a monte carlo approach to generate a model although that's definitely appealing, take a much fuller simulation of the wing/part and run that offline to generate a model. Of course that implies making a much more complex simulation in order to generate the simpler model which is a lot more work to achieve.

For example if you know the direction the shot came from and the relative dimensions of the wing and spar at that angle you can generate a crude %chance to hit a spar using (width of spar in that projection/width of wing in that projection).

Quite so. Which latter part is exactly what the new DM does.

Edited May 26, 2020 by unreasonable

May 26, 2020

32 minutes ago, Holtzauge said:

Maybe, just maybe you would jerk the stick? If you did its absolutely possible to exceed 5.5 g or higher momentarily

I find this hard to believe, as jerking a control colum takes far more travel than a flightstick on the PC plus the control forces can be higher (dpending on the type of aircraft). I think pilots were painfully aware of the flimsy nature of their rides and treated them with care. Probably most of the lads have seen one go down after ripping off the wings. Ripping the column back would translate into knowingly attempting suicide for not getting killed.

Your chart shows that 3 g is actually a lot regarding how those lads fought close in combat. The chart shows one thing clear, namely that the common combat speed of the aircraft was far lower than we fly them in the game. I'm sure most of the time during maneuvers, they hardly exceeded 150 km/h, making it impossible to pull much more g than 3 without flicking out of the turn.

Also in order to pull 5+ g, the crate had to go maybe 250 km/h, which is well over maximum flight speed and far above maneuvering speed. We routinely fly that fast, as we attack in a way the common pilot would flee for his life.

2 minutes ago, unreasonable said:

You mean like the pilot hitbox? Do we really need a spine, brain, and heart hitbox so that we do not mix up disabling hits with just a scratch?

Yes. One box has always been good enough for one pilot.

Edited May 26, 2020 by ZachariasX

May 26, 2020

35 minutes ago, ZachariasX said:

You know, hitboxes that have no relation to internal structure I find terminally flawed. No matter what super system you tack on it, it will always have a flawed result in the individual case.

Without 'empty' hitboxes for those sections you'd never see the 'puffs' of the aircraft being it - nor would the visual damage texture update itself.

35 minutes ago, ZachariasX said:

Now, who ever lost an aileron in flight? Has that been seen, ever?

Many times...I've pulled them off in dives, I've shot them off EA.

39 minutes ago, Holtzauge said:

I still think many of the wing failures recorded in combat were not due to battle damage but simply overloading of the airframe by a pilot reacting to being under fire:

In some manouvers the pilot hits 4g even in mock combat.

Oh, absolutely, I bet there were cases where that happened. Distinguishing between those cases and 'Shot to Pieces' cases are very difficult, though.

I find it exceptionally easy to hit 4G in a dogfight as it stands! In a SPAD, you'll be hitting 5G pretty frequently as well, and even 6G if you're a more aggressive BnZer - and as Zach just said, "Your chart shows that 3 g is actually a lot regarding how those lads fought close in combat". Which is really interesting given the direction the discussion has been taking!

39 minutes ago, Holtzauge said:

Take the SE5 for example: IIRC correct then it had an ultimate load factor (as in the factor it breaks at with no safety factor) as low as 5.5g.

In that case, according to the G-meter in game, S.E's should be falling apart after every attempted BnZ!

Great post by the way. That R&M is a sweet find.

Edited May 26, 2020 by US93_Larner

May 26, 2020

3 minutes ago, unreasonable said:

Which latter part is exactly what the new DM does.

The devil is in the details though as it would be pretty expensive to calculate those widths with precision for arbitrary angles so presumably there is some sensible fudging going on. For example having pre-calculated %chances defined for vertical and horizontal shots and then interpolating between them based on the dot product of the shot direction and the direction the wing is generating lift. Then it matters how you interpolate as the width of the wing increases very quickly as you increase angle. Etc. etc. etc.

May 26, 2020

1 hour ago, Holtzauge said:

BTW: here is the link to the report: Has many nice figures of other aircraft, including Camel and Bristol Fighter etc.

The link goes to report about the calculated loads on a high speed aeroplane in flight (BE2 ) no sign of the chart or anything about Camels etc.

Edit - solved. The link goes to report 496, the report you refer to is 469 496 also worth having a look at.

Edited May 26, 2020 by unreasonable

4.006 DM Discussion

Recommended Posts

kendo

BMA_Hellbender

Zooropa_Fly

Tycoon

ST_Catchov

kendo

ZachariasX

Holtzauge

kendo

BMA_Hellbender

No.23_Triggers

J2_Bidu

unreasonable

ZachariasX

kendo

unreasonable

BMA_Hellbender

unreasonable

No.23_Triggers

BMA_Hellbender

No.23_Triggers

US103_Baer

Chill31

unreasonable

No.23_Triggers

Chill31

No.23_Triggers

unreasonable

No.23_Triggers

Holtzauge

ZachariasX

US63_SpadLivesMatter

slug_yuugen

ZachariasX

slug_yuugen

unreasonable

ZachariasX

No.23_Triggers

slug_yuugen

unreasonable

Browse All

Activity