Thanks to Barambas for this one. I don’t think it needs any comments.
43 thoughts on ““Ghost Shells Are Fixed””
Hmm. It seems to have been on a path through the mantlet.
Accuracy of (visual) shell termination is ‘iffy’ at range, on pentrating hits you can often see the tracer extend beyond the tank.
It could be a visual glitch coupled with a bounce?
I have a nice replay of bouncing on one scout, and the shell hitting and penetrating another about 25m from the original tank and about 30° from the original path of the shell.
The visual shell carried on straightforward until it buried itseld in a hill.
Type have quite thin mantlet, so i don’t think so.
WZ-120: 249 maximum penetration (AP fired), actual penetration can be as low as 187mm. at 100m.
Type 61: 90 – 120mm of spaced armour + 70mm of actual armour on mantlet (checked the armour model).
On 500m distance, it’s well bouncable.
Still, it’s an interesting glitch even “if” it’s purely visual and it’d better get fixed, because it confuses players.
Lag lamp was red at the moment of impact so it could be some package lost?
It could be (the shell going few cm over the turret, likely, judging by arc). That’s not really fixable, yeah…
package loss as in the client did not receive the type 61 taking damage or bouncing the shot?
I need name of that HUD mod :/
Zayaz Damage Panel IIRC.
They must mean “Ghost shells are fixed. For a given value of “fixed” Hands behind the back, fingers crossed.
When will you guys realize, that WGN is just bullshitting you.. Ghost-Shells don’t exist.. yeah, ofc… Ok then, but Ghost-Shells are fixed now.. aha
You’re right.
WG dont make any sense in what they say:
“Bots are impossible / you can report bots, we’re fightig them”
“Ghost shells dont exist / ghost shells fixed”
“E25 is balanced fine / E25 removed: too good”
It makes me think about minecraft’s “removed herobrine” in each patchnote… But WG jerks about something real.
Exactly that ;)
I often see artillery shells explode in air nowadays in 9.5, is this some bug, too? Or is it only graphical?
They explode if they go out of range of the gun.
Happens a lot with low tiers.
one day even noobs like ss will realize that tracers are client side while real shell traction is calculated on server. On 500+ m its no surprise if they dont match 100%
You are making a big deal about nothing. Tracers have never been that accurate anyway and at that range the client and server only have to be slightly out of synch to get this kind of effect.
He missed, shit happens.
OP here.
Well, that I didn’t know. Thanks for the info lets move along (:
but bashing on wg is way easier than accepting that you missed.
just take the wot forum:
” I got killed, must be hacks!!!”
“I missed/bounced, wg is manipulation rng to make me lose”
and so on
I am playing this game for like four years now and I have NEVER EVER heard anyone claim that ghost shells are normal. Sure, they can be an effect of de-sync, but definitely not this often (not the first mail I got).
A “ghost shell” is an effect, where the shell seemingly hits the targets but goes through or disappears, regardless of its cause. If this doesn’t pose a problem for you, then it sure does for me.
So take the tracer away and it’s fine.
What you seem to want is impossible with today’s technology. When we have infinite bandwidth with zero latency then you can have it. Of course that would be defying the laws of physics but lets not let a small technicality like that get in the way.
See above. Ghost shells were never “normal”, not in this amount anyway. I got more mails like this. Why do I have the feeling you just came to bash me and are looking for a stick?
Try to use server side aim.
Been using that since 0.6.1.
If you have plenty, please show some at shorter ranges where it isn’t just client server desync
This is such a bullshit. Server-client would have to be WAY out of sync for this to happen.
if this IS actually case of server-client sync, then its definetly BUG. Server should send EXACT direction of newly appeared shell and client should be able to calculate shell treacer precisely.
Lastly, if there was a problem with synchronization when the shell was fired, it wouldn’t appear at all, why do you think, that sync fail would cause shell to fly in slightly different direction? BULLSHIT.
“Working as intended”
+1
I was driving my Tiger 2 in 3rd person view and saw a shell go through my turret side and do no damage. I saw it hit the hill next to me. The ghost shell thing is not fixed at all. (This was last week.)
Very “out of place” -sensation to hear finnish in a video found on this blog. =)
Well just few days ago I shot through enemy M103 with my VK 45.02 B.
He was front of me like 50 meters and I fully aimed and shoot towards the hull weakspot. Then I see my shot just going right trough the tank and impacted on the wall behind the M103.
My friend was watching next to me and he confirmed that he saw the same as I did: my shot just went through middle of the enemy M103.
Replay?
There are less of em, but they are not gone. The past 2 patches the few videos of ghost shells I’ve seen are shots that had a good chance of bouncing. Before ghost shells was side shots that would have done damage. Now it seems like they occur at times instead of bounces. Could be a packet loss issue or anything as stuff that happens in the game when the replay is recorded, be it high pick, rubber banding or even that spin bug will still be there as it was during the fight. Even “glitches” like this. Could even be a Type 61 model bug or hit box issue “anyone looked at that around the mantle” Or like I said, could be that the client was to lazy to animate the bounce and just did a ghost shell instead. Or it was shown like that due to loss of info. Even if the server did register it as a shot even tho on his client it was not shown due to loss of info. The damage would still be there.
Did the guy say “stronk” at the very end? :D
To all you “boo birds”….what would qualify as a genuine ghost shell to you? The shot must be point blank on a stationary Maus with under 10ms of lag with a gold round on Bill Gate’s personal computer before you believe it?
Before we can see server-side replay we can’t be sure what is the problem. Personally I think there are two reasons for ghost shells:
- de-sync – especially visual desynchronisation of the client. Have you ever noticed that tracers are drawn after the shell path and impact is calculated? Many times I saw enemy tank’s HP going down before the tracer got rendered. A split second difference but it is there. Also in multi-tank clashes with many shot, exposions etc. the visuals are delayed that everyone must notice. Maybe it’s only the GPU who “gets late” but the tracers are also rendered by GPU.
- botched collision models – as proven many times before, collision models are carelessly assembled with a gap here, gap there, something missing, something in the wrong place… you all know, if you read FTR. As the shell have no mass, no diameter (it’s a mass-less flying point) it can slip through any, even one-pixel wide, gap.
BTW, Silent, if you have so many replays with this phenomenon as you claim, it would be interesting to make a sort of stats of it – like if a certain tank model is more prone to let shells through, or certain locations/connection providers (also wifi vs. wire), if it happens on certain maps or even certain places on maps etc. Also the places where tanks are hit would be interesting, I personally suppose that most of the botched areas are around the tracks and on hull-turret boundary.
I personally think that connection quality has non-negligible influence on this phenomenon. I’ve experienced ghost shells very rarely, let’s say ten times over past four years, most of them in the beginning. But I almost never have the ping indicator red
(well, I dont play right after patches or during the daytime when x5 are out) and I have quite constant 45-70 ms ping. Also my framerate is not jumping up and down, steadily hovering around 35-40 fps. I think these parameters are factors too when considering ghost-shells.
All the proof in the world means nothing. If WG says Ghost shells do not exist then (to them) they don’t.
Why? Because it’s their game so everything they say is right.
Now go to gulag.
Wargaming should advertise their game with this 1 min video :D
It’s a packet loss side effect. The server only tells whether it damaged anything after the packet loss. Ricochets or misses are not given retroactively.
WG: We were going to fix this but something even more important came up. Check out the HD detailing of the rails in the garage. It will be fixed when it’s fixed.
Before 9.5 it was rare that I would get a ghost shell, I usually thought it was due to lag but since 9.5, I have been getting 1 or more during a match, to the point, I have shot the same tank 3 times, in a row, over a period of 30 seconds, I would have killed them but instead died because they were able to kill me, only because I could not hurt them. I am getting sick of the bullshit in this game.
That’s how Wg does it to do what they can to make sure u get a loss and that that person gets to live…. RNG…. Rigged Numbers Game
Damn that dude in the background sounds retarded ;)
Hmm. It seems to have been on a path through the mantlet.
Accuracy of (visual) shell termination is ‘iffy’ at range, on pentrating hits you can often see the tracer extend beyond the tank.
It could be a visual glitch coupled with a bounce?
I have a nice replay of bouncing on one scout, and the shell hitting and penetrating another about 25m from the original tank and about 30° from the original path of the shell.
The visual shell carried on straightforward until it buried itseld in a hill.
Type have quite thin mantlet, so i don’t think so.
WZ-120: 249 maximum penetration (AP fired), actual penetration can be as low as 187mm. at 100m.
Type 61: 90 – 120mm of spaced armour + 70mm of actual armour on mantlet (checked the armour model).
On 500m distance, it’s well bouncable.
Still, it’s an interesting glitch even “if” it’s purely visual and it’d better get fixed, because it confuses players.
Lag lamp was red at the moment of impact so it could be some package lost?
It could be (the shell going few cm over the turret, likely, judging by arc). That’s not really fixable, yeah…
package loss as in the client did not receive the type 61 taking damage or bouncing the shot?
I need name of that HUD mod :/
Zayaz Damage Panel IIRC.
They must mean “Ghost shells are fixed. For a given value of “fixed” Hands behind the back, fingers crossed.
When will you guys realize, that WGN is just bullshitting you.. Ghost-Shells don’t exist.. yeah, ofc… Ok then, but Ghost-Shells are fixed now.. aha
You’re right.
WG dont make any sense in what they say:
“Bots are impossible / you can report bots, we’re fightig them”
“Ghost shells dont exist / ghost shells fixed”
“E25 is balanced fine / E25 removed: too good”
It makes me think about minecraft’s “removed herobrine” in each patchnote… But WG jerks about something real.
Exactly that ;)
I often see artillery shells explode in air nowadays in 9.5, is this some bug, too? Or is it only graphical?
They explode if they go out of range of the gun.
Happens a lot with low tiers.
one day even noobs like ss will realize that tracers are client side while real shell traction is calculated on server. On 500+ m its no surprise if they dont match 100%
You are making a big deal about nothing. Tracers have never been that accurate anyway and at that range the client and server only have to be slightly out of synch to get this kind of effect.
He missed, shit happens.
OP here.
Well, that I didn’t know. Thanks for the info lets move along (:
but bashing on wg is way easier than accepting that you missed.
just take the wot forum:
” I got killed, must be hacks!!!”
“I missed/bounced, wg is manipulation rng to make me lose”
and so on
I am playing this game for like four years now and I have NEVER EVER heard anyone claim that ghost shells are normal. Sure, they can be an effect of de-sync, but definitely not this often (not the first mail I got).
A “ghost shell” is an effect, where the shell seemingly hits the targets but goes through or disappears, regardless of its cause. If this doesn’t pose a problem for you, then it sure does for me.
So take the tracer away and it’s fine.
What you seem to want is impossible with today’s technology. When we have infinite bandwidth with zero latency then you can have it. Of course that would be defying the laws of physics but lets not let a small technicality like that get in the way.
See above. Ghost shells were never “normal”, not in this amount anyway. I got more mails like this. Why do I have the feeling you just came to bash me and are looking for a stick?
Try to use server side aim.
Been using that since 0.6.1.
If you have plenty, please show some at shorter ranges where it isn’t just client server desync
This is such a bullshit. Server-client would have to be WAY out of sync for this to happen.
if this IS actually case of server-client sync, then its definetly BUG. Server should send EXACT direction of newly appeared shell and client should be able to calculate shell treacer precisely.
Lastly, if there was a problem with synchronization when the shell was fired, it wouldn’t appear at all, why do you think, that sync fail would cause shell to fly in slightly different direction? BULLSHIT.
“Working as intended”
+1
I was driving my Tiger 2 in 3rd person view and saw a shell go through my turret side and do no damage. I saw it hit the hill next to me. The ghost shell thing is not fixed at all. (This was last week.)
Very “out of place” -sensation to hear finnish in a video found on this blog. =)
Well just few days ago I shot through enemy M103 with my VK 45.02 B.
He was front of me like 50 meters and I fully aimed and shoot towards the hull weakspot. Then I see my shot just going right trough the tank and impacted on the wall behind the M103.
My friend was watching next to me and he confirmed that he saw the same as I did: my shot just went through middle of the enemy M103.
Replay?
There are less of em, but they are not gone. The past 2 patches the few videos of ghost shells I’ve seen are shots that had a good chance of bouncing. Before ghost shells was side shots that would have done damage. Now it seems like they occur at times instead of bounces. Could be a packet loss issue or anything as stuff that happens in the game when the replay is recorded, be it high pick, rubber banding or even that spin bug will still be there as it was during the fight. Even “glitches” like this. Could even be a Type 61 model bug or hit box issue “anyone looked at that around the mantle” Or like I said, could be that the client was to lazy to animate the bounce and just did a ghost shell instead. Or it was shown like that due to loss of info. Even if the server did register it as a shot even tho on his client it was not shown due to loss of info. The damage would still be there.
Did the guy say “stronk” at the very end? :D
To all you “boo birds”….what would qualify as a genuine ghost shell to you? The shot must be point blank on a stationary Maus with under 10ms of lag with a gold round on Bill Gate’s personal computer before you believe it?
Before we can see server-side replay we can’t be sure what is the problem. Personally I think there are two reasons for ghost shells:
- de-sync – especially visual desynchronisation of the client. Have you ever noticed that tracers are drawn after the shell path and impact is calculated? Many times I saw enemy tank’s HP going down before the tracer got rendered. A split second difference but it is there. Also in multi-tank clashes with many shot, exposions etc. the visuals are delayed that everyone must notice. Maybe it’s only the GPU who “gets late” but the tracers are also rendered by GPU.
- botched collision models – as proven many times before, collision models are carelessly assembled with a gap here, gap there, something missing, something in the wrong place… you all know, if you read FTR. As the shell have no mass, no diameter (it’s a mass-less flying point) it can slip through any, even one-pixel wide, gap.
BTW, Silent, if you have so many replays with this phenomenon as you claim, it would be interesting to make a sort of stats of it – like if a certain tank model is more prone to let shells through, or certain locations/connection providers (also wifi vs. wire), if it happens on certain maps or even certain places on maps etc. Also the places where tanks are hit would be interesting, I personally suppose that most of the botched areas are around the tracks and on hull-turret boundary.
I personally think that connection quality has non-negligible influence on this phenomenon. I’ve experienced ghost shells very rarely, let’s say ten times over past four years, most of them in the beginning. But I almost never have the ping indicator red
(well, I dont play right after patches or during the daytime when x5 are out) and I have quite constant 45-70 ms ping. Also my framerate is not jumping up and down, steadily hovering around 35-40 fps. I think these parameters are factors too when considering ghost-shells.
All the proof in the world means nothing. If WG says Ghost shells do not exist then (to them) they don’t.
Why? Because it’s their game so everything they say is right.
Now go to gulag.
Wargaming should advertise their game with this 1 min video :D
It’s a packet loss side effect. The server only tells whether it damaged anything after the packet loss. Ricochets or misses are not given retroactively.
WG: We were going to fix this but something even more important came up. Check out the HD detailing of the rails in the garage. It will be fixed when it’s fixed.
Before 9.5 it was rare that I would get a ghost shell, I usually thought it was due to lag but since 9.5, I have been getting 1 or more during a match, to the point, I have shot the same tank 3 times, in a row, over a period of 30 seconds, I would have killed them but instead died because they were able to kill me, only because I could not hurt them. I am getting sick of the bullshit in this game.
That’s how Wg does it to do what they can to make sure u get a loss and that that person gets to live…. RNG…. Rigged Numbers Game
Damn that dude in the background sounds retarded ;)
WOW, is that really you? HE MAN??? (:
http://www.comparestoreprices.co.uk/images/fa/fame-and-fortune-ladies-my-boyfriend-is-a-real-he-man-t-shirt.jpg