Ramming A/C

[Devildrivr]

Please fix the damage model of the Spitfire and the Hurricanes, all models.  A fair number of Spitfire pilots have now resorted to deliberately ramming axis a/c knowing they will survive the collision 9 out of 10 times. In refence to fighter a/c. I would guess this is largely due to the flight model of the allied a/c in this game being far to robust.   Please Audit head on ramming for a/c

Edited January 13, 2022 by devildrivr
spelling

[Kidd27]

This is not true. Plane type or side has no bearing on collisions. Internet latency is the culprit.

You need to adjust your timing from what you see.

If you fly your plane into another plane, u die.

While YOU see the collision, (during the time it takes for your info to go through the internet and to the server and back) the other pilot has pulled up/out/moved and has avoided the collision.

Its has nothing to do with the model strength being able to withstand a collision.

Fact

[Devildrivr]

I disagree, have seen the ea deliberately Ram other axis ac without pulling out of the head on.  Will test with another player during intermission to confirm. 

[Kidd27]

1 minute ago, devildrivr said:

I disagree, have seen the ea deliberately Ram other axis ac without pulling out of the head on.  Will test with another player during intermission to confirm. 

this has been a topic for 20 years. Wrapping your head around internet latency is not an easy task.

The only real option for CRS would be to have both plane blow up when it got he message that one plane detected a collision. That would not be very fair to the guy who avoided the collision on his end.

I'd give you a hand with testing it if im on. PM me

[vanapo]

1 hour ago, Kidd27 said:

That would not be very fair to the guy who avoided the collision on his end.

Avoiding collision doesn't include "being the guy who pulled out of a head-on 50 milliseconds earlier".

I'd say the normal outcome should be for both planes to blow up and no kills awarded. The only exception should be someone ramming your 6. So if a plane is struck on a shallow angle from the rear, ok, that's an exception to deter low skill griefers.

BTW: I said many times that I think there IS a latency dependent bug with the detection of who pulled out first. If someone runs into your belly (so you are not able to see the guy running into you), he should always be the guy blowing up. If you can record instances of such cases in which you blow up with a bunring wreck stationary in the air afterwards, this would be great to prove my case.

Edited January 13, 2022 by vanapo

[Kidd27]

52 minutes ago, vanapo said:

Avoiding collision doesn't include "being the guy who pulled out of a head-on 50 milliseconds earlier".

I'd say the normal outcome should be for both planes to blow up and no kills awarded. The only exception should be someone ramming your 6. So if a plane is struck on a shallow angle from the rear, ok, that's an exception to deter low skill griefers.

BTW: I said many times that I think there IS a latency dependent bug with the detection of who pulled out first. If someone runs into your belly (so you are not able to see the guy running into you), he should always be the guy blowing up. If you can record instances of such cases in which you blow up with a bunring wreck stationary in the air afterwards, this would be great to prove my case.

avoiding does include "being the guy pulled out out of a head on"   if 50 milliseconds made the difference between collision or no collision.

The guy ramming from the 6 always blows up- this is latency working its magic.

Its got nothing to do with beating the other guy in avoidance by milliseconds per se;  its all internet and the time it takes for information to makes it way from computer to server and back.

 Im not connecting how a stationary wreck  (ive been them) is evidence of something.

[jester]

4 hours ago, devildrivr said:

Please fix the damage model of the Spitfire and the Hurricanes, all models.  A fair number of Spitfire pilots have now resorted to deliberately ramming axis a/c knowing they will survive the collision 9 out of 10 times. In refence to fighter a/c. I would guess this is largely due to the flight model of the allied a/c in this game being far to robust.   Please Audit head on ramming for a/c

You've had this account for like 12 years and really think this? 

[vanapo]

29 minutes ago, Kidd27 said:

 Im not connecting how a stationary wreck  (ive been them) is evidence of something.

Then you probably didn't experience aerial collisions in game very often because this is exactly what you become if you end up exploding due to one, apart from getting killed any other way in which case you don't. So in order to report on those cases, at least one of the involved planes has to end up as such.

And I don't agree: "Avoiding collisions" is the avoidance of colliding with another plane. "Avoiding collisions" is not "pulling out early enough for the other plane to hit your mid or aft section". That is named "pulling out of a collision a few milliseconds earlier than your opponent". Which is, as many people report over and over and over and over again, not working correctly or not working in a consistent way. Probably also because of latency issues. Who knows. But it's probably not a good feature if people complain about it all the time.

 

[madrebel]

28 minutes ago, jester said:

You've had this account for like 12 years and really think this? 

seriously how many times does this need to be explained? this conversation was dead and buried over a decade ago.

[Devildrivr]

2 hours ago, madrebel said:

seriously how many times does this need to be explained? this conversation was dead and buried over a decade ago.

Then there should be nothing wrong with testing just to see which sides planes fair the best in a head on

[TR6AL]

From what I've been told it has to do with internet lag and predicter coding. Info gets moved between clients and if the client pc thinks it hit another plane then it causes damage etc. If it doesnt think it did.. then no damage. 

It doesn't care what side you are on.. 

[greyman]

2 hours ago, madrebel said:

seriously how many times does this need to be explained? this conversation was dead and buried over a decade ago.

It's been happening since the late 80s with Air Warrior. It will continue until the eggheads eliminate internet lag.

 

21 minutes ago, devildrivr said:

Then there should be nothing wrong with testing just to see which sides planes fair the best in a head on

I'll save you some trouble:

 - if you collide with another player on your screen, you die.
- if you don't collide with another player on your screen, you live.

[jester]

1 hour ago, devildrivr said:

Then there should be nothing wrong with testing just to see which sides planes fair the best in a head on

There's a way to game the mechanic which some people do. Considering the population of the air game lately.. you're probably just running into the same people frequently who are abusing it. Even then, abusing it is not 100% sucessful, it's very high risk, and also it takes two to tango so look in the mirror if you're frustrated with it happening often

I'm still mind boggled that you think this has anything to do with DM or faction when your account has been around half the time this game has.

[vanapo]

3 hours ago, greyman said:

I'll save you some trouble:

 - if you collide with another player on your screen, you die.
- if you don't collide with another player on your screen, you live.

and that, my friend, just doesn't work correctly. As said many times. And is probably due to:

3 hours ago, TR6AL said:

From what I've been told it has to do with internet lag and predicter coding. Info gets moved between clients and if the client pc thinks it hit another plane then it causes damage etc. If it doesnt think it did.. then no damage. 

But it's not a side thing, it's rather a latency thing. That's why I advice every player outside of North America to avoid collision at all cost.

3 hours ago, TR6AL said:

It doesn't care what side you are on.. 

It only cares what side of the atlantic/pacific you are on.

[Sudden]

While you are at it fix the damage model between a truck having fun running through the hangars full of airplanes. It only happened once.

[jwilly]

2 hours ago, vanapo said:

It only cares what side of the atlantic/pacific you are on.

Not correct, because the mechanics do not work such that greater-packet-latency-to/from-the-server-increases-likelihood-of-dying. It only matters that the total latency between any two clients is non-zero. It doesn't matter whose latency is lower.

The essential mechanic to understand is that you see me at a position in game-space that's one latency trip (my client --> server --> your client) older than where I see me, and I see you at a position in game-space that's one latency trip (your client --> server --> my client) older than where you see you. No one in game that's moving is ever seen by any other player at the same location where they see themselves.

So if your client thinks we collided, your client notifies the server that you're dead; but if in the same interaction my client thinks we didn't collide, my client doesn't notify the server that I'm dead.

My client always manages death notifications about me. Your client always manages death notifications about you.

Here's an old-but-previously-CRS-used illustration of how it works. Remember, where I see your plane and where you see your plane cannot ever be the same location, unless either your plane is stationary or latency has been magically reduced to zero.

[Devildrivr]

12 hours ago, jester said:

There's a way to game the mechanic which some people do. Considering the population of the air game lately.. you're probably just running into the same people frequently who are abusing it. Even then, abusing it is not 100% sucessful, it's very high risk, and also it takes two to tango so look in the mirror if you're frustrated with it happening often

I'm still mind boggled that you think this has anything to do with DM or faction when your account has been around half the time this game has.

If a pilot knows his chances of surviving are 9 out of 10 then he might feel compelled to exploit. Will test with multiple peeps during intermission.

[vanapo]

9 hours ago, jwilly said:

Not correct,

Again. As said many times, this apparently doesn't work correctly, as I am experiencing collision deaths when someone hits my underbelly, so apparently in cases where the OPPOSITE is true. We had this discussion several times now @jwilly - I understand how it is supposed to work, I always tell you that it doesn't work as it is supposed to work, and then you answer by telling how it is supposed to work.

Again, this is what I think I encountered a lot of times and which I would ask everybody to record if they have the means to do so. Because if recorded, this would be the proof it is not working as intended:

 

I can see why you should not be able to score a kill by just running into the dead 6 of another plane, especially a bomber. But if you collide with your enemy in a dogfight, you shouldn't be rewarded with a score, you should be rewarded with ... well, let this guy speak for me:

 

Edited January 14, 2022 by vanapo

[jwilly]

Quote

I can see why you should not be able to score a kill by just running into the dead 6 of another plane, especially a bomber. But if you collide with your enemy in a dogfight, you shouldn't be rewarded with a score,

Discussions like this one are never about how the game should work, game-philosophically. They instead are about how a wide-area-network, strong-client-weak-server game has to work unless it either magically has zero latency to all players, or a magical ability to determine each player's motivations and intentions.

That'd be a fine alternate discussion to have, except there's no way to write code that'll operate as you want. Game-players from the beginning of online air-combat gaming have argued "you should not be able to score a kill by just running into the dead 6 of another plane". The game however has no way to know intent. The game has no way to know if the fast plane behind intended to collide, or the slow plane in front decelerated and veered into the path of the fast plane making a firing pass. The only way to code some-collisions-are-ignored is to code all-collisions-are-ignored...and that has been tried, and makes a very goofy air combat game that's nothing like what it's trying to simulate.

Quote

Again. As said many times, this apparently doesn't work correctly, as I am experiencing collision deaths when someone hits my underbelly,

It seems you don't understand the explanation. 

It's not relevant what you as the pilot can see happening. It's also not relevant what you intend to happen, or what the other pilot intends to happen. All that's relevant is what your client perceives has happened, from a simulated-physics perspective.

Your client can "see" in all directions simultaneously, even though for realism the client only shows you-the-pilot what you can see in a single direction where you're looking.

If your client is told by the server that another aircraft is now at the same location in game-space as your aircraft...as your client perceives, i.e. your true location and the other aircraft's reported location, which necessarily is one server-trip delayed from what that pilot knows to be his true location...then your client determines that a mid-air collision has occurred. 

It doesn't matter if you-the-pilot visually saw it coming. It doesn't matter what you intended, or what the other pilot intended.

And it can't be resolved using two true locations, because...as long as there's non-zero latency...it's impossible to know another (moving) player's true location in real time.

Your edited version:

Clearly enough, you intended to miss a collision with the delayed position of the enemy plane, but did not succeed. Maybe you pulled up, but he pulled up too.

The other pilot missed a collision with the delayed position of your plane. His intent is unknowable and makes no difference.  

I'm not making this stuff up. It's not based on my perceptions of how game-code like this works. This was sorted out by the game-designing-and-coding community way back at the beginning of online air combat games, i.e. Air Warrior --> WarBirds --> Aces High. What you want is if-I-die-the-other-guy-should-die. There's no way to code that without either magic or a worse gameplay result than how it works now.

Edited January 14, 2022 by jwilly

[Kidd27]

9 minutes ago, jwilly said:

Discussions like this one are never about how the game should work, game-philosophically. They instead are about how a wide-area-network, strong-client-weak-server game has to work unless it either magically has zero latency to all players, or a magical ability to determine each player's motivations and intentions.

That'd be a fine alternate discussion to have, except there's no way to write code that'll operate as you want. Game-players from the beginning of online air-combat gaming have argued "you should not be able to score a kill by just running into the dead 6 of another plane". The game however has no way to know intent. The game has no way to know if the fast plane behind intended to collide, or the slow plane in front decelerated and veered into the path of the fast plane making a firing pass. The only way to code some-collisions-are-ignored is to code all-collisions-are-ignored...and that has been tried, and makes a very goofy air combat game that's nothing like what it's trying to simulate.

It seems you don't understand the explanation. 

It's not relevant what you as the pilot can see happening. It's also not relevant what you intend to happen, or what the other pilot intends to happen. All that's relevant is what your client perceives has happened, from a simulated-physics perspective.

Your client can "see" in all directions simultaneously, even though for realism the client only shows you-the-pilot what you can see in a single direction where you're looking.

If your client is told by the server that another aircraft is now at the same location in game-space as your aircraft...as your client perceives, i.e. your true location and the other aircraft's reported location, which necessarily is one server-trip delayed from what that pilot knows to be his true location...then your client determines that a mid-air collision has occurred. 

It doesn't matter if you-the-pilot visually saw it coming. It doesn't matter what you intended, or what the other pilot intended.

And it can't be resolved using two true locations, because...as long as there's non-zero latency...it's impossible to know another (moving) player's true location in real time.

I'm not making this stuff up. It's not based on my perceptions of how game-code like this works. This was sorted out by the game-designing-and-coding community way back at the beginning of online air combat games, i.e. Air Warrior --> WarBirds --> Aces High.

I think "see" has been misinterpreted as an actual "visual" -  You cant simply look away to avoid a collision. The fact you dont have a "visual" on the belly of your plane, doesnt mean you avoid a collision.

If your place in space occupies the last KNOWN place in space as the other guy, you die. His last KNOWN place in space is gonna lag behind where he really is.

[OLDZEKE]

The microsoft combat flight sims had collision bubbles where both died if they touched. We, as players, tried to adjust or change that as with internet lag the bubble seemed to grow. We did find a way to reduce it, which helped but still it was always a gripe.

This game is if you collide, with enough force, you explode. I've had many collisions where both I and the enemy I collided with went in to spins rather than either explode. Some I recovered from some I didn't.

As to "seeing" the collision, it's what your game "sees" or tracks as being in the same zyx space as you are in at that moment.

[vanapo]

2 hours ago, jwilly said:

Clearly enough, you intended to miss a collision with the delayed position of the enemy plane, but did not succeed. Maybe you pulled up, but he pulled up too.

I don't know how it is so hard to understand that I am just telling you it is not working in a consistent manner. Why would you ever think I'd presume you could avoid collision or exploding by looking the other way? Seriously. That's  misunderstanding my argument on purpose in a way it's almost offensive.

I literally can't see a plane colliding with me when we both pull up hard and he is the second one to pull up. When I pull away from a headon, the enemy disappears under my nose. The guy who pulls up last would be the one  who is pulling into the underbelly of the guy who pulled up first - in order TO RAM HIM, on purpose or by accident. He would be able to see the collision coming up, the guy pulling up first won't be able to see it only if he had a window between his feet like a stuka pilot. So if I pull up in front of him and the kill cam shows that he obviously pulled up as well and right into me, from all the information I got, I pulled out first and was the first to avoid collision.So either we both blow up or I shouldn't blow up at all.

So, coming back to your example, wich is exactly depicting the scenario above, the plane that SEES itself colliding with the other plane is going to explode. The other plane which SEES itself pulling away before a collision, won't explode. And that is - given that scenario - completely right.  That's how it supposed to work. Like you explained yourself, this is to not punish the guy who avoided - from all he was able to see - to collide with another plane.

All I am saying, is - again - that sometimes it's sometimes exactly the other way around.

When I pull away and the enemy collides - from my perspective - with my underbelly, this can't be the result of me being the guy(from my perspective) pulling away second. It does mean that (from my perspective) he pulled out after I did. Yet my plane reports "been hit" and explodes. Yet apparently, his plane doesn't report that and flies on through my wreckage. So this is randomly not working or it is not working for the guy with the bad latency because he lags behind so much, that he is not able to avoid collisions based on the data he SEES. In both cases, I don't see the benefit in keeping one of the planes alive, as acting upon what you are seeing and pulling away according to what you are seeing might still get yourself killed in a collision while the other side flies away.

 

3 hours ago, jwilly said:

The game however has no way to know intent. The game has no way to know if the fast plane behind intended to collide, or the slow plane in front decelerated and veered into the path of the fast plane making a firing pass. The only way to code some-collisions-are-ignored is to code all-collisions-are-ignored.

Are you sure on that? Strangely enough I had a lot of occasions in which a fighter plane, coming  at me in a straight line from my 6,  collided quite visually with my client's level flying Heinkel or even sometimes my fighter and they blew up with me flying away from those collisions. Seems hard to grasp how it should occur to my client which just depicted to me that this plane ran into my 6, touched, and blew up, might not have registered that a plane ran into my 6, touched and thus collided with me.

2 hours ago, OLDZEKE said:

he microsoft combat flight sims had collision bubbles where both died if they touched. We, as players, tried to adjust or change that as with internet lag the bubble seemed to grow. We did find a way to reduce it, which helped but still it was always a gripe.

Very interesting. So what you are telling here is that when both planes by layout have the same size of a hitbox which, when collided with, leads to explosion, the hitbox of the plane with the higher latency seem to grow. Now we have a game where in case of one-sided collision, only the plane that is struck on said hitbox explodes, not the striking plane. Let's just think one second about what this means for the plane with the bigger hitbox due to higher latency. Yeah, it means exactly what I am telling you guys for years now: It is at a disadvantage regarding collision outcomes. And a growing one with growing latency.

Now, there probably is no good solution for that. But I doubt putting high latency players at even a greater disadvantage than they already are facing is the best one.

[OLDZEKE]

My microsoft combat flight sim example does not cross over to wwiiol. 2 completely different collision systems.

Ms had a very simple damage model

[vanapo]

8 minutes ago, OLDZEKE said:

My microsoft combat flight sim example does not cross over to wwiiol. 2 completely different collision systems.

Serious question if answering is possible: How are the differences concerning the outcome of high velocity mid air collisions between 2 planes one of which doesn't report being struck?

Edited January 14, 2022 by vanapo

[hillstorm]

3 hours ago, jwilly said:

What you want is if-I-die-the-other-guy-should-die. There's no way to code that without either magic or a worse gameplay result than how it works now.

I'm the kid in the dunce cap reading that and saying, um, yes - that IS what I want. In the case of a midair collision, both planes should at least be catastrophically damaged ... I mean, right?

Here's a crazy thought. Get rid of collision effects altogether. Literally allow planes to fly through one another, the way same-side infantry can run through each other or vehicles can run through them, like magic. 

Would that be so terrible? Not as terrible as the feeling when an enemy plane is bearing down on me to collide with me on purpose, and I can't get out of the way because I'm in an old, slow bomber that can't maneuver to save its (my) life. It's a perversion of the game.