DM poll

[J5_HellCat_]

34 minutes ago, AnPetrovich said:

Sorry guys, it was a busy time!
I remember, I promised the test results for 90°, here it is:

(P.S. Actually "fix is coming" is needed to read as "now in 4.006")

 

 

Thanks!

[the_dudeWG]

Thanks, @AnPetrovich! I think this was the real culprit of the 4.005 DM. Bullet hits from these angles through the wing plane hit boxes should do little to no damage to all aircraft, and by the looks of it, that would be the case in 4.006 now. S!

[No.23_Triggers]

@AnPetrovich Thanks so much for taking the time to implement these changes! From what I see, 4.006 is an all-round improvement in EVERY way! 

I would love for the FC community to be able to provide some thoughts and feedback on the new 4.006 DM to you, once we have tested it out for a good amount of time (of course!) Is there any way we can do that? 

Thanks again! 

[US103_Baer]

Let's not we exaggerate, it's only 1450 bullets. 1200 bullets variation between planes is nothing. ?

 

More seriously though, I would like to see that chart at 3G loading, the SPAD and Dolphin I've been trying, seem to fall apart similar to 4.005.

Edited May 21, 2020 by US103_Baer

[1PL-Sahaj-1Esk]

Is the G-loading factor properly modelled for WWI planes in the first place? I doubt it that one could achieve so many G is WWI planes??

 

After flying a little with the Camel I do not find any substential difference in WING-breaking, I would even say it is worse in turns. The G-clock in WWI runs so fast IMO - I read so many WWI books/accounts and the blackouts from manouvering was never a topic. In FC it is the NEXUS of all problems. This is really not looking good.

Edited May 22, 2020 by 1PL-Sahaj-1Esk

[US63_SpadLivesMatter]

2 hours ago, 1PL-Sahaj-1Esk said:

I read so many WWI books/accounts and the blackouts from manouvering was never a topic.

 

Probably because they didn't fling around their planes the way you're all used to doing.

[US103_Baer]

Idk. With life actually on the line, I'm sure they explored the outer edges of the envelope. Many accounts seem to verify that.

Edited May 22, 2020 by US103_Baer

[No.23_Triggers]

4 hours ago, J28w-Broccoli said:

 

Probably because they didn't fling around their planes the way you're all used to doing.

I get what you're saying here - why would you chance your luck in a flimsy WW1 plane? The flip-side of that, though, is like Baer says. If you have someone behind you actively kill you, doing some 'dangerous' manoeuvres might be the lesser of two evils...

[ZachariasX]

16 minutes ago, US93_Larner said:

I get what you're saying here - why would you chance your luck in a flimsy WW1 plane? The flip-side of that, though, is like Baer says. If you have someone behind you actively kill you, doing some 'dangerous' manoeuvres might be the lesser of two evils...

 

Rickenbacker wrote that the N28 biggest downside was that pilots were affraid of flying it to the limits because of the upper wing canvas shedding, hence they would not maneuver as they would do if they had faith in the aircraft, resulting them getting shot down.

 

It seems that certain dangers are too severe to test them too much, even when getting shot at.

[unreasonable]

1 hour ago, US93_Larner said:

I get what you're saying here - why would you chance your luck in a flimsy WW1 plane? The flip-side of that, though, is like Baer says. If you have someone behind you actively kill you, doing some 'dangerous' manoeuvres might be the lesser of two evils...

 

Of course pilots might feel like that. Probably why 7 out of MvR's 20 BE kills broke up in the air.

 

That does not mean that they were right to push dangerous manoeuvres to the limits.  Take a plane past it's limits and your death (sans chute) is guaranteed.  Have some Boche trying to kill you: his guns might jam, incendiaries damage his own plane, run out of ammunition, other friendlies appear on the scene, get too far over friendly lines so the Boche chickens out....  

 

Breaking your own plane in the air is always the worst possible choice. (Unless you are are on fire.....) 

 

 

My hypothesis (and that of others too based on the threads so far) is that the root of the wing shedding problem is that many players in MP are over-stressing the planes.

 

They do this, because:

 

1) They have got used to doing it since there was almost no penalty with prior DMs

2) The disincentive to die in MP is less than the disincentive to allow your opponent a clean kill. (MP ego).

3) Sorry to have to say this, (not really) but many of them have almost no idea about the fundamental physics of flight.

4) More charitably, because it has been difficult, given the lack of seat-of-the-pants feedback and no g-meter, to judge just how much stress is inflicted on the planes.

 

Now that the team is putting together a bottom up, physics based version of the DM people can either lobby to adjust to it piece-meal based on evidence, or lobby for what they think makes for a "better game".

 

Personally I hope that the developers stick to their guns and make only changes that are evidence based.  I also happen to believe that this is also the right formula commercially, that is if FC has any future commercially, (which I doubt, but hope I am wrong).  Let's face it, the RoF MP die-hards have either already  bought the game or will not, largely because of things that have nothing to do with the DM, like invisible planes, which need solving as a matter of urgency.

 

 

Edited May 22, 2020 by unreasonable

[No.23_Triggers]

46 minutes ago, unreasonable said:

My hypothesis (and that of others too based on the threads so far) is that the root of the wing shedding problem is that many players in MP are over-stressing the planes.

 

They do this, because:

 

1) They have got used to doing it since there was almost no penalty with prior DMs

2) The disincentive to die in MP is less than the disincentive to allow your opponent a clean kill. (MP ego).

3) Sorry to have to say this, (not really) but many of them have almost no idea about the fundamental physics of flight.

4) More charitably, because it has been difficult, given the lack of seat-of-the-pants feedback and no g-meter, to judge just how much stress is inflicted on the planes.

Right you are. 

[keeno]

I'm really enjoying this new DM. I've only had a few hours playing SP and I haven't had one wing shedding incident, all victories were either enemy pilot kills or engine failures.

 

Thanks to the DEVS for looking into and addressing the issue.

 

Cheers

[No.23_Starling]

2 hours ago, unreasonable said:

Personally I hope that the developers stick to their guns and make only changes that are evidence based.  I also happen to believe that this is also the right formula commercially, that is if FC has any future commercially, (which I doubt, but hope I am wrong).  Let's face it, the RoF MP die-hards have either already  bought the game or will not, largely because of things that have nothing to do with the DM, like invisible planes, which need solving as a matter of urgency.

 

 

Yes! And once the invisibility is solved we can look into solid data on the SE5a RPM (Coming soon from Shuttleworth for a real survivor) and probably the Spad vs DviiF dive capabilities where we have data.

[unreasonable]

I would just add, in case anyone has not also been following the parallel 4.006DM thread, that my in game checks of the g-meter show no sign of it calculating incorrectly.

For instance, if you sustain a level coordinated turn at 60 degrees of bank, the G meter shows 2.0 - just as it should. 

 

I do wish the g-meter could be incorporated into the display in a more intuitive and less "GUI" way - something like a subtle colour shift of the whole screen, or a band around the edge.  More instinctive feedback and less glancing at a little 2D dial on an intrusive HUD.   This might make it easier for everyone to relearn how much they can yank the stick.

[Zooropa_Fly]

1 hour ago, keeno said:

I'm really enjoying this new DM. I've only had a few hours playing SP and I haven't had one wing shedding incident, all victories were either enemy pilot kills or engine failures.

 

Thanks to the DEVS for looking into and addressing the issue.

 

Cheers

 

I'm still shooting at meat and metal.

The only plane I'm tempted to go for via the wings now is the Camel.

 

S!

[ST_Catchov]

Just now, Zooropa_Fly said:

 

 

The only plane I'm tempted to go for via the wings now is the Camel.

 

 

 

WHY YOU LITTLE ….! Be fair, you're only supposed to shoot at my head. No ace of the month for you!

[Tycoon]

11 hours ago, unreasonable said:

 

 

 

1) They have got used to doing it since there was almost no penalty with prior DMs

2) The disincentive to die in MP is less than the disincentive to allow your opponent a clean kill. (MP ego).

3) Sorry to have to say this, (not really) but many of them have almost no idea about the fundamental physics of flight.

4) More charitably, because it has been difficult, given the lack of seat-of-the-pants feedback and no g-meter, to judge just how much stress is inflicted on the planes.

 

 

 

 

You forgot reason 5) pilots are used to and perfectly capable of flying the plane within the correct g load, then a few bullets wiz by and your plane shows no visible damage but you suddenly lose 3 Gs worth of maneuverability before the wings come off.  

Edited May 22, 2020 by Tycoon

[ST_Nooney]

16 hours ago, unreasonable said:

 

 

 

 

3) Sorry to have to say this, (not really) but many of them have almost no idea about the fundamental physics of flight.

 

What of those that do and the wings still fall off, despite the knowledge of knowing they shouldn't?

[unreasonable]

8 hours ago, Tycoon said:

You forgot reason 5) pilots are used to and perfectly capable of flying the plane within the correct g load, then a few bullets wiz by and your plane shows no visible damage but you suddenly lose 3 Gs worth of maneuverability before the wings come off.  

 

Hyperbole is not a winning debating strategy in the long run.  

 

I do think the question of how visible damage is shown in the game is legitimate, although I am not sure if real WW1 pilots would necessarily have had clear warning of imminent failure either.  After all, in my analysis of MvR's victories, before the date where I think he started to use incendiaries, excluding BEs which shed wings easily, 6/35 (17%) of the planes he shot down suffered structural collapse.

 

Perhaps they knew they were critically damaged and risked it anyway - or perhaps they did not. Unfortunately we will never know.  Perhaps some clearer indication of increased potential for structural damage would be useful.

 

3 hours ago, ST_Nooney said:

What of those that do and the wings still fall off, despite the knowledge of knowing they shouldn't?

 

 Despite the belief of believing that they shouldn't.  

[Tycoon]

39 minutes ago, unreasonable said:

 

Hyperbole is not a winning debating strategy in the long run.  

 

 

Just look at the videos I posted, it's not hyperbole.

[unreasonable]

2 minutes ago, Tycoon said:

Just look at the videos I posted, it's not hyperbole.

 

Trying to download that and look at it frame by frame, but unfortunately the part where the SPAD may or may not be hit by the D.VII is shot from the D.VII's pov, which is not helpful.

Then there is the question of what else had happened in the sortie before that.  But at least this is trying to give evidence, so I am all in favour of that. 

 

I would just remind you, however, that a single case, even if proven to be as you say, cannot discredit a stochastic model.   Plus there is the long standing developer advice that MP cannot be used to assess the DM because there are too many other factors.  For instance, see @J5_Baeumer 's post in the 4.006 DM thread about the unreliability, or difficulty of interpretation,  of the  parser stats.

 

What @US93_Larner's comparative tests of PvP and PvAI show is that the wings in MP are not being shot off, they are being ripped off by the human pilots - the AI rarely does it (possibly too rarely).  If there is some way in which visible damage or sound cues could help players realise they have been compromised in an immersive way I am all in favour: but players would still have to change their behaviour.

 

Another possible solution is just to have two DMs. One for MP, one for everyone else. The MP DM could be the same for all planes, to avoid arguments and claims of unfairness - I propose the Pfalz. 

[No.23_Triggers]

1 hour ago, unreasonable said:

What @US93_Larner's comparative tests of PvP and PvAI show is that the wings in MP are not being shot off, they are being ripped off by the human pilots - the AI rarely does it (possibly too rarely).  If there is some way in which visible damage or sound cues could help players realise they have been compromised in an immersive way I am all in favour: but players would still have to change their behaviour.

I would agree with this in all cases except the Halberstadt CL.II, which is still exhibiting cases of having its wings shot completely off in level flight on an alarmingly frequent basis. We haven't really seen this with any other aircraft, save for the Albatros and SPAD in some isolated cases. We haven't tested the Dolphin or the Bristol F2B to any meaningful degree yet.

 

I'll put out a chart detailing cases of the CL.II doing this soon (AKA whenever I stop being lazy). 

The burning question in my mind at the moment, based on our tests, is - are the amount of G forces caused by basic manoeuvres plausible? From what Chill31 has said in regards to his experiences flying his Dr.I, my interpretation is that  the G-loads are seemingly a little too exaggerated. It would stand to reason, then, that if the G-loads were less severe (aka, 3 Gs in a split-S instead of 5 Gs), aircraft wouldn't be turning to confetti quite so quickly. 

At the moment, I think the DM is stopping some aircraft from performing even the most basic combat manoeuvres after only minimal damage. Adam's SPAD example would be a good one to base that statement off of - in his video, he takes a short, inaccurate burst before winging over and losing his wings while pulling out of the short dive. On the parser for that fight, the D.VII pilot he was facing had a total recorded 5 hits for that sortie, and Adam hadn't taken any prior damage. I appreciate Baeumer's point about some of the finer points of using Parsers to determine aircraft damage, but 5 bullets is bizarrely low to shoot down any aircraft, let alone a SPAD.

I'll need to find the quote again, but one pilot of the 22nd Aero Squadron cites going into a steep spiral-dive out of a fight with D.VIIs - a fight in which he accumulated 70+ bullet holes in his aircraft. Unless every hit was in the centre fuselage, that wouldn't even be close to possible based on Adam's video. 

That being put aside for a moment, though, I thought you were on to something really interesting by looking at different reported types of historical kills from MvR and comparing to FC 'reported' kills. I suppose when looking at data like that, one question would be 'Should the DM be perfectly number-accurate to the devs' calculations, or should they reflect the historical experience more?'. One conclusion to that question is to say that we are pulling more G-forces than our historical counterparts would have done -- and that brings us back to the top of this post! 

I'm not 100% sure about having a 'one size fits all DM', although I suppose to a degree that's what we had when nobody's wings were coming off. If that was the answer, I'd agree that the Pfalz would be the best DM to base it on. 


 

Edited May 23, 2020 by US93_Larner

[unreasonable]

@US93_Larner

 

Having no wings coming off is historically completely implausible - hence my Pfalz joke.  Even after MvR was torching most victims with incendiaries, ~10% of his kills were due to structural collapse. Since the small numbers of flamers in FC currently suggest the ammunition is intended to model ball + tracer, not incendiaries, I think about 15-20% structural collapse would be the best target, if you were trying to match historic data.  Maybe looking at the other books you mentioned will change that somewhat, but I doubt very much.

 

As to what gs are incurred in basic manoeuvres  - well that depends on how you fly them. You can do a split S more or less tightly, just like a flat turn, and incur a wide range of loads.

 

One of the things that the G meter - and looking at the relevant graphs etc - has reminded me of is how the addition of Gs as you tighten a turn do not come in a straight line.

n=1/cos theta  in a flat turn, so adding more angle adds increasing G per degree added.  The same would be true in any loaded manoeuvre. You could see that in the gif I posted a while ago, also here.  So if you are pushing the envelope a tiny mistake will suddenly take you well over your limits.

 

 

 

 

 

[No.23_Triggers]

13 minutes ago, unreasonable said:

@US93_Larner

 

Having no wings coming off is historically completely implausible - hence my Pfalz joke.  Even after MvR was torching most victims with incendiaries, ~10% of his kills were due to structural collapse. Since the small numbers of flamers in FC currently suggest the ammunition is intended to model ball + tracer, not incendiaries, I think about 15-20% structural collapse would be the best target, if you were trying to match historic data.  Maybe looking at the other books you mentioned will change that somewhat, but I doubt very much.

Totally in full agreement with you on that point - ironically, I do honestly think that the Pfalz's DM is fairly in line with your MvR report (sans 10G mayhem). 

 

15 minutes ago, unreasonable said:

As to what gs are incurred in basic manoeuvres  - well that depends on how you fly them. You can do a split S more or less tightly, just like a flat turn, and incur a wide range of loads.

Another good point, and one we considered - we tested an Albatros with varying degrees of Split-S 'tightness'...full elevator deflection gave 4.6 - 5Gs, a wider Split S gave around 6 Gs. My interpretation of Chill's reports of performing loops would suggest, and I stand to be corrected here, that those numbers should be closer to 3G and 4G. We didn't, however, factor in the variance in G-loading not being 'in a straight line', so cheers for that - I'll keep that in mind. 

As a general disclaimer to all - the test results I'm posting have no 'ulterior motive' - they're simply to provide some in-game context to the discussions and are there to be used as a tool for people reaching whatever conclusions they come to. Although, of course, I've been sprinkling in my own thoughts and opinions when discussing that data. 

[Tycoon]

1 hour ago, unreasonable said:

 

 

 

I would just remind you, however, that a single case, even if proven to be as you say, cannot discredit a stochastic model.  

 

 

I understand this, I'm just saying what I've experienced from my admittedly limited testing, but then again that's all we can do.

 

21 minutes ago, unreasonable said:

1 hour ago, unreasonable said:

 

.   Plus there is the long standing developer advice that MP cannot be used to assess the DM because there are too many other factors.  For instance, see @J5_Baeumer 's post in the 4.006 DM thread about the unreliability, or difficulty of interpretation,  of the  parser stats.

 

 

 

 

 

 

  That may be, but this problem is extremely easy to replicate be it offline or online.

 

1 hour ago, unreasonable said:

 

 

 

What @US93_Larner's comparative tests of PvP and PvAI show is that the wings in MP are not being shot off, they are being ripped off by the human pilots - the AI rarely does it (possibly too rarely). 

 

 

I don't think we can use the ai results as legitimate evidence, Unless I'm wrong there is no way to see what kind of G's they are pulling. 

[unreasonable]

1 hour ago, US93_Larner said:

Totally in full agreement with you on that point - ironically, I do honestly think that the Pfalz's DM is fairly in line with your MvR report (sans 10G mayhem). 

 

Another good point, and one we considered - we tested an Albatros with varying degrees of Split-S 'tightness'...full elevator deflection gave 4.6 - 5Gs, a wider Split S gave around 6 Gs. My interpretation of Chill's reports of performing loops would suggest, and I stand to be corrected here, that those numbers should be closer to 3G and 4G. We didn't, however, factor in the variance in G-loading not being 'in a straight line', so cheers for that - I'll keep that in mind. 

As a general disclaimer to all - the test results I'm posting have no 'ulterior motive' - they're simply to provide some in-game context to the discussions and are there to be used as a tool for people reaching whatever conclusions they come to. Although, of course, I've been sprinkling in my own thoughts and opinions when discussing that data. 

 

On the last point - absolutely - you have been clear and transparent throughout, with your excellent tests and your above average data presentation, just like my significant other when she tells me how wonderful I am. I am in no doubt that you want a fun and sustainable game,  as does She.  In my case that consists of taking her shopping for ludicrously overpriced garments and  bejewelled thingamijigs, in her case it consists of long post-prandial sessions of ....  hang on a moment.  Was it the other way around - three months of lockdown separation is driving me bonkers.  Reminder to self - no more posting after lunch, especially if it involves a bottle of a rather good Côtes du Rhône. Which it usually does - except on Claret days, of course.

 

I hesitate to be critical of @Chill31's posts (not really) since he has a real Dr.1 etc and we are indeed very lucky to have him around. That does not mean that we have to pretend that his command of deductive reasoning is better than that of anyone else. 

 

His post - IIRC, went along these lines:

 

1) I pulled the stick in game Dr.1 as I do in my real one

2) This cause me to loop as I do in my real one

3) The g meter showed 3+, but in the real one I feel 2+

4) I know the difference - it is pronounced

5) Therefore I think there is something wrong with the G-meter

 

I am quite willing to stipulate (1), (3) and (4) above, accepting @Chill31's good faith reports. The problem is (2).

 

Either his game Dr.1 did not respond to the stick inputs in exactly the same way as his real plane, or the G meter is wrong - or both. I have tested the G meter and I think it is right. No-one so far has reported trying to replicate or falsify my results, so unless and until that happens I am sticking with my judgement (easily tested!) that the g-meter is correct. Therefore, he did not loop in the game exactly as he did in his actual Dr.1    ie (2) is false.

 

This could be because the FM and control surface response of the game Dr.1 =/= Chill's Dr.1    Or it could be because, lacking the normal seat-of-pants feedback, Chill misjudged the loop in the game. Or a bit of both.  

1 hour ago, Tycoon said:

I understand this, I'm just saying what I've experienced from my admittedly limited testing, but then again that's all we can do.

 

  That may be, but this problem is extremely easy to replicate be it offline or online.

 

I don't think we can use the ai results as legitimate evidence, Unless I'm wrong there is no way to see what kind of G's they are pulling. 

 

1) Yes, I agree. There is only so much we can do. On the other hand, AnP has a bot that can carry out thousands of tests to generate a probability distribution.

 

2) How do you replicate "this problem" offline? Certainly as a player offline you can rip your wings off.  The whole issue is whether this is indeed a "problem". You see it as a problem because you cannot do what you are used to doing. I see it as an improvement.

 

3) The AI obviously know how much they are pulling, and more to the point how much they can pull.  I would have no complaints if the AI made mistakes a little more often - I think they should, but they should be modeling the mistakes of a real WW1 pilot not a MP veteran who has got used to making bat-turns with impunity.   Real WW1 pilots did sometimes over stress their aircraft in a fight with dire results.  But given the developers statements, I think we have to accept that the AI has the same DM and g limits as humans.

Edited May 23, 2020 by unreasonable

[Chill31]

4 hours ago, unreasonable said:

I hesitate to be critical of @Chill31's posts (not really) since he has a real Dr.1 etc and we are indeed very lucky to have him around. That does not mean that we have to pretend that his command of deductive reasoning is better than that of anyone else. 

 

His post - IIRC, went along these lines:

 

1) I pulled the stick in game Dr.1 as I do in my real one

2) This cause me to loop as I do in my real one

3) The g meter showed 3+, but in the real one I feel 2+

4) I know the difference - it is pronounced

5) Therefore I think there is something wrong with the G-meter

 

I am quite willing to stipulate (1), (3) and (4) above, accepting @Chill31's good faith reports. The problem is (2).

 

Either his game Dr.1 did not respond to the stick inputs in exactly the same way as his real plane, or the G meter is wrong - or both. I have tested the G meter and I think it is right. No-one so far has reported trying to replicate or falsify my results, so unless and until that happens I am sticking with my judgement (easily tested!) that the g-meter is correct. Therefore, he did not loop in the game exactly as he did in his actual Dr.1    ie (2) is false.

 

This could be because the FM and control surface response of the game Dr.1 =/= Chill's Dr.1    Or it could be because, lacking the normal seat-of-pants feedback, Chill misjudged the loop in the game. Or a bit of both.  

What you've said here is perfectly accurate.  I am going to be shocked, jaw on the floor shocked, if I am pulling more than 3 Gs in the loop at any point, but right now, I don't have anything to back it up other than the seat of my pants.  On the bright side, the weather is perfect today, so if I can rig up a mounting for my G-meter, I can test this today.

[Tycoon]

11 hours ago, unreasonable said:

 

 

 

1) Yes, I agree. There is only so much we can do. On the other hand, AnP has a bot that can carry out thousands of tests to generate a probability distribution.

 

2) How do you replicate "this problem" offline? Certainly as a player offline you can rip your wings off.  The whole issue is whether this is indeed a "problem". You see it as a problem because you cannot do what you are used to doing. I see it as an improvement.

 

3) The AI obviously know how much they are pulling, and more to the point how much they can pull.  I would have no complaints if the AI made mistakes a little more often - I think they should, but they should be modeling the mistakes of a real WW1 pilot not a MP veteran who has got used to making bat-turns with impunity.   Real WW1 pilots did sometimes over stress their aircraft in a fight with dire results.  But given the developers statements, I think we have to accept that the AI has the same DM and g limits as humans.

Just to clarify all I'm complaining about is exaggerated dm, not G force durability in general. I think what you're mostly talking about is G force durability being too strong which I might agree with. Look at it this way, let's say G durability is way down to let's say 5 Gs, with no damage. Ok fine, I can learn to fly the plane like that, but then you take some bullet strikes that don't damage wires, struts, or anything and you instantly lose 3 Gs worth of stress, now that I do have a problem with. Regardless of what the G limit should or shouldn't be this is it's own problem, and a game breaking one at that.  

[unreasonable]

1 hour ago, Tycoon said:

Just to clarify all I'm complaining about is exaggerated dm, not G force durability in general. I think what you're mostly talking about is G force durability being too strong which I might agree with. Look at it this way, let's say G durability is way down to let's say 5 Gs, with no damage. Ok fine, I can learn to fly the plane like that, but then you take some bullet strikes that don't damage wires, struts, or anything and you instantly lose 3 Gs worth of stress, now that I do have a problem with. Regardless of what the G limit should or shouldn't be this is it's own problem, and a game breaking one at that.  

 

I do not know what the undamaged G durability really was for these crates, or how much it should change with damage.  I have to trust the developers to decide that - perhaps their thinking will continue to evolve if new facts emerge.  But I assume that you would agree that damage to spars should reduce G durability to some extent.

 

You might even agree that frequently repeated near black out manoeuvers are not how WW1 pilots would have flown, whether damaged or not, even in SPADS.  Certainly seeing some of the videos I am gobsmacked that people actually think that WW1 biplanes were flown like that, damaged or not.  (edit: I am talking about high G turns and pull ups, not fast dives).

 

What I do know is that a significant proportion of MvR's victims lost their wings, not subsequent to entering a terminal death spiral, but while under fire or turning in a fight.  Tests vs AI suggest that wings are not being shot off in huge numbers, if anything, too few lose wings,  so there is no evidence that the DM is exaggerated at all. 

 

We are left with player behaviour, and/or online hit recognition as possible explanations for the higher incidence in MP.

 

 

  

 

 

 

 

Edited May 23, 2020 by unreasonable

[Tycoon]

27 minutes ago, unreasonable said:

 

I do not know what the undamaged G durability really was for these crates, or how much it should change with damage.  I have to trust the developers to decide that - perhaps their thinking will continue to evolve if new facts emerge.  But I assume that you would agree that damage to spars should reduce G durability to some extent.

 

 

 

 

  

 

 

 

 

Yeah, I agree 100%, what I don't agree with is losing wing strength when there isn't damage to spars or even wires.

[1PL-Sahaj-1Esk]

 

@AnPetrovich

 

can you please explain those two graphs in conjunction with one of my parser-missions-extracts from Flugpark server? below the parser screenshot. My wings came off as I was BnZ a Pfalz (around 3-4G).

 

 

 

 

Another question: why is the scale different on both graphs?

 

 

Edited May 23, 2020 by 1PL-Sahaj-1Esk

[kendo]

1 hour ago, Tycoon said:

Yeah, I agree 100%, what I don't agree with is losing wing strength when there isn't damage to spars or even wires.

 

Not sure your assumption is right. Expect the sim is registering damage to the wing elements. It comes back again to getting one of the unlucky low-number breaks on the left hand side of the curve. Not that the bullets are passing through 'air' without hitting anything. The deficiency here may be in the visual model being too limited in depicting damage. 

 

58 minutes ago, 1PL-Sahaj-1Esk said:

Another question: why is the scale different on both graphs?

 

Top one is for shots from above, while bottom one is for shots from directly behind. Firing from above has much less chance of hitting the spar.

[Tycoon]

5 minutes ago, kendo said:

 

Not sure your assumption is right. Expect the sim is registering damage to the wing elements. It comes back again to getting one of the unlucky low-number breaks on the left hand side of the curve. 

 

 

 

Those unlucky low-number breaks happen with reliability and consistency from what I've tested.

 

8 minutes ago, kendo said:

 

 

 

 

 

 

[1PL-Sahaj-1Esk]

1 hour ago, kendo said:

Top one is for shots from above, while bottom one is for shots from directly behind. Firing from above has much less chance of hitting the spar.

 

Exactly, however in game it feels flipped, shots from 90° seems to do a greater damage than shots from 6.

 

I got two shots from 90° angle from the ground and my wings folded - this is in NO WAY conisitent with that chart.

 

 

Edited May 23, 2020 by 1PL-Sahaj-1Esk

[kendo]

I don't know? The parser reads 5 hits, so were the other 3 on the fuselage? (haven't read all he way back on this) Also, I think the %age rating in the parser for any individual hit is probably complete BS and should be ignored. No idea how that would even be calculated. All you can say from the parser is you were hit a certain number of times (and that may even be wrong because of the acknowledged issues with MP?).  

 

The most important thing is that the graphs above are both for unloaded wings (see AnP's original post):

On 5/6/2020 at 8:09 PM, AnPetrovich said:

I performed the test for all WW1 planes to compare their durability before and after I have corrected their spars size. This test was performed before I have updated the tool, so it was done under zero-gravity conditions. But it gives quite a good understanding of airplanes durability anyway:

 

So he did that test under zero gravity conditions just as a way to compare the before/after effect of the spar fix. Works ok for that one purpose to give an illustration of the change, as both tests (before/after ) are made under the same conditions. But it's totally useless for deciding how many actual hits are needed in game - even when flying level at 1G. And you were pulling ~4G.

 

So, who knows? What we would need is an updated version of this one below, but for shots from 90 degrees.

 

Edited May 24, 2020 by kendo

[US63_SpadLivesMatter]

The parser doesn't show every bullet.

[emely]

On 5/24/2020 at 2:46 AM, 1PL-Sahaj-1Esk said:

 

Exactly, however in game it feels flipped, shots from 90° seems to do a greater damage than shots from 6.

This suggests that the probability theory of getting into the spar that (possibly) works well on ww2 planes, works very poorly on our planes. But it is impossible to change this separately for the circus.  Here gentlemen, this is another advantage of RoF!  RoF had problems, but they never came from outside.

17 hours ago, kendo said:

But it's totally useless for deciding how many actual hits are needed in game - even when flying level at 1G. 

From the very beginning, this topic was designed to show the user the depth of their misconceptions.  This logically follows even from the test conditions.  Beautiful color pictures and diagrams probably convinced the community that with DM, in fact, everything is OK.  But this presentation didn’t make me believe in the reality of what is happening in the game.

10 hours ago, J28w-Broccoli said:

The parser doesn't show every bullet.

Yes, the parser doesn’t count everything, the server doesn’t count everything either, some bullets can be lost due to the network rassynchron ... Hits in the wings are considered approximately, the damage from hitting the bullet is also a convention.  However, the cross-section of the side members displayed is super accurate, down to a micron! ? This is the most important thing, isn't it?  )))

[kendo]

On 5/24/2020 at 12:46 AM, 1PL-Sahaj-1Esk said:

1. Exactly, however in game it feels flipped, shots from 90° seems to do a greater damage than shots from 6.

 

7 hours ago, emely said:

1. This suggests that the probability theory of getting into the spar that (possibly) works well on ww2 planes, works very poorly on our planes. But it is impossible to change this separately for the circus.  Here gentlemen, this is another advantage of RoF!  RoF had problems, but they never came from outside.

 

2. From the very beginning, this topic was designed to show the user the depth of their misconceptions.  This logically follows even from the test conditions.  Beautiful color pictures and diagrams probably convinced the community that with DM, in fact, everything is OK.  But this presentation didn’t make me believe in the reality of what is happening in the game.

 

3. Yes, the parser doesn’t count everything, the server doesn’t count everything either, some bullets can be lost due to the network rassynchron ... Hits in the wings are considered approximately, the damage from hitting the bullet is also a convention.  However, the cross-section of the side members displayed is super accurate, down to a micron! ? This is the most important thing, isn't it?  )))

 

I'll comment on those 3 points in sequence:

 

1. Does it suggest that? Sahaj-1Esk (and I don't mean to be disrespectful here) talks about 'feels' and 'seems'. And you (Emely) jump in behind and take that as established fact. Maybe it is the truth, but surely it needs a little more testing to be sure? And the second AnP table published above does show (based on thousands of tests again) a much greater number of hits needed when firing from 90 degrees. I would suggest that degree of evidence outweighs (at the moment anyway) your assertion.And as I said above you can't use the numbers in that graph (at zero G) as a guide to number of expected hits to break a wing when pulling 4G+.

 

2. Yes, I think the topic was designed to do that. But maybe it needed to be done? There are two different issues here, and you mix them both together in your answer.

 

The first issue - was there user misperception? Yes, I think AnP's posting of his test data showed there was. And that has led to a much deeper re-examination of all the possible contributing factors - from how the aircraft are flown, to absence of incendiary ammunition, etc.

 

The second issue. Does the fact that misperception (probably) did exist mean the DM is now all ok and correct? No. It doesn't! But it means we should now have a better idea of how far there may still be to go. (Recall that graph with our current estimate,  our ideal, and the current (4.005) reality. Surely knowing this helps and is positive for moving in the right direction? ). 

 

And maybe user perception still needs to shift some more?  Even the evidence published in last few days showing the number of pilots pulling off their wings in accidents should be enough to make people reconsider a little. Larner's testing showing the gap between wing losses for AI (limited to 5G pulls I believe) and human MP pilots might also suggest the same thing.

 

But none of this means the current 4.006 DM is correct, ot that it wouldn't benefit from further changes, but surely, please, it should make us reconsider our own positions at least a little? 

 

3. Was going to comment on this...but what's the point? 

 

Anyway, I have hopes that devs may be open to further revision of the DM if they think it is justified. AnP said the below when publishing the test data and the spar fixes. I read this as saying they are open to going further if needed.    

 

On 5/6/2020 at 8:09 PM, AnPetrovich said:

Seems like we need to change the airframe durability. Or change the DM taking into account some other factors related to the process when bullet damages a spar, to make this damage not very hard. Perhaps...

 

Edited May 25, 2020 by kendo

[emely]

There is a task to bake a cake exactly according to the technology of a hundred years ago.  Not all parts of the recipe have been preserved.  For example, you know that you need flour, but you don’t know how much it needs, you know that there is sugar in the list, but again its quantity is unknown.  Some things are known for sure - no doubt salt needs 5 tablespoons ...

And so you take the flour, (but too little) add one tablespoon of sugar (and you need 30) and at the end put the exact amount of salt (5 tablespoons)

And when they say to you “Hey man, in the memoirs this cake is called sweet, and you have it salty.” You just say that memoirs are not proof, you suggest that consumers change their perception and give evidence, which is 5 tablespoons of salt in the recipe

Edited May 25, 2020 by emely

[Panzerlang]

On 5/22/2020 at 10:57 AM, unreasonable said:

I would just add, in case anyone has not also been following the parallel 4.006DM thread, that my in game checks of the g-meter show no sign of it calculating incorrectly.

For instance, if you sustain a level coordinated turn at 60 degrees of bank, the G meter shows 2.0 - just as it should. 

 

I do wish the g-meter could be incorporated into the display in a more intuitive and less "GUI" way - something like a subtle colour shift of the whole screen, or a band around the edge.  More instinctive feedback and less glancing at a little 2D dial on an intrusive HUD.   This might make it easier for everyone to relearn how much they can yank the stick.

 A sound-cue would do the job quite nicely, something like RB2-3D had (a creaking sound iirc).