Denying that you tried to insult me is pathetic. You simply refuse to grow up. That’s even worse.
Ok have some time away.
`
`
Soo, with the in the name of the tsar patch the hit detection for me is a lot more consistent.
Something like 1 to 2 wtf moments a game down from 10 wtfs and i’m able to shoot knee sliders 9/10 times so big improvement.
I do feel tho that sometimes i’m getting hits when I miss but no where near as much, and that those hits might just be a lagged enemy moving into the bullet on the server as I see them fall in the direction of the miss..?
It also seems like sometimes the death notification is really delayed and I find myself wanting to empty a clip to make sure.
Occasionally i’ll put 3 bullets into someone with a medic rifle and the third won’t register so again i’m emptying clips instead of relying on 3 shots per kill in most situations… but I have been using the 1906 a bit…
The hit/miss performance of the 1895 trench feels better (which is what i’m basing my perception of increased consistency on).
Anyway, whatever you did has made it better, thanks!
There are acknowledged server issues that could be part of it?
No fix as of yet.
The invisible players are back
of course i know about the two protocols.
i was merely pointing out what isps are very likely to be doing to get round the problem net neutrality has brought to them.
you cant simply pick on traffic types any more.
but what about protocols and if they can hide it wey hey even better…obviously they havent hidden it well enough.
also an update.
when i have had dodgy gameplay with my new isp ive checked the icmp routes and the udp routes.
not surprisingly they arent the same …
now if they are measuring latency with icmp…can anyone tell me what the obvious flaw is with the observation in the above paragraph?
question re ping cap.
If its below 200ms for the european servers then how a 700ms ping player get on and stay on?
btw this game had no ill effects.
in fact it doesnt matter how many 100ms+ players are on it doesnt make a difference.
the only thing that makes a difference is if im getting raised ping on certain noes en route…even if the end result is good.
imo ea arent measuring udp latency.
they are using icmp.
this needs to change.
some people think they have good connections when they wont have.
THEY ARE NOT MEASURING LATENCY WITH ICMP!!!
The network graph is the only UI in game that shows latency. Latency is based off the protocol it uses. In this case it is UDP.
The 200ms threshold is not a ping kicker. The 200ms threshold stipulates when a player’s shots are handled strictly by the server.
Below threshold : (Client Reg → Server Auth) Client determines a hit and sends a hit claim to the server. The server in turn arbitrates the hit claim. If validated the server awards the hit. Only shots the client deems a hit are sent to the server.
Above threshold : (Server reg & Auth) All shots fired by a client are sent to the server for arbitration. Every shot is scrutinized.
Have you read any of this thread or the actual patch notes?
then something is up with their measurements.
when i do get iffy gameplay i check pingplotter and see a dramatic change in routes on udp.
i check my “LATENCY” on the net graph and i see very little variance.
yet on fifas if its iffy i see 3 bars instead of 4 and it might jump between red orange or green pre match.
i think the fact they dont bother showing a true packet loss figure is very suspect.
i think ea need to be transparent.
i dont believe the net graph latency figure…it simply can not be correct.
not all because im testing 2 games…plus other things.
im more interested why the “LATENCY” bear no resemblence to pingplotter or how the game plays
if it is 50ms all games id expect it to be the same every time…even more so as you kindly pointed out that its tested with udp lol
im looking at no major differences on the netgraphs but variable gameplay…how can this be?
i also see a few players with 100ms ping…most are between 20 and 100.
my guess is packet loss is a major factor and ea are hiding it.
and a question.
if someone is getting 1% udp packet loss and another is getting 5% what are ea doing about this?
Compensating for the lost packets purposefully ruining the game for everyone else.
not sure how they could compensate when they wouldnt know what instruction was in that packet as the one in front and behind arrived
The packet that is lost is resent by the client, that information when it eventually is received by the server is then applied late to the player who the player with packet loss is interacting with, being held by damage , late and stacked damage are the effects of this, it causes large amounts of de-sync in game and is basically what we have been mostly complaining about for 257 pages.
Then decompile your client and look at the code. If they’re adjusting latency values you’ll see the modification just before it is sent to the UI.
*since BF3.
Do you know anything about networking? If the routes, distances, network traffic etc are different obviously the outcomes will be different.
A player in NY will not have the same latency/loss as a player in VA for a server in DC. The same even applies to players living right next door to each other playing on the same server.
Explain what you think the compensation is and how it works. There maybe a misunderstanding.
The client holds a copy of each sent packet in memory for roughly 1 second. This is a direct copy of the ack redundancy system I noted in a previous post. If the client doesn’t receive an ack from the server within X ms (offset + UTT) for that packet it will resend.
High latency players (above threshold) do not resend data of any sort. The server extrapolates for loss. The extrapolated position is bound to the players game history as though the data was sent by him/her. Rightly, hit arbitration will be applied to the extrapolated data. HP Shots that arrive late or are lost are outright discarded and not resent (aka automatic MISS). Also the FHT is clamped, meaning it will only rewind history to a set ms. This time frame does not take into consideration your latency.
Refer to pg 62
Side note … Network Graph only shows downstream packet loss. refer to pg 69.
udp doesnt resend.
HP Shots that arrive late or are lost are outright discarded and not resent (aka automatic MISS). Also the FHT is clamped, meaning it will only rewind history to a set ms. This time frame does not take into consideration your latency.
thanks
this is what i was hoping to read.
this is why ea need to be more transparent on their netgraph.
lets say out o 100 packets yu lose 1 and i lose 5.
there is an advantage to you.
now ignoring that.
im concentrating on what can i do about my 5% loss.
its not for ea to solve as they do not dictate the routes i use to get my packets to their server.
but what they can do to help is the netgraph for packet loss.
maybe add a far more detailed section so say after a match we can go in and have alook at a graph of data.
compare whta we sent and how it was received.
this info could be used as evidence when “consulting” our isps.
from my testing ive seen some interesting things.
i test when playing a lot so i can see things in real time in the game and on the pingplotter graphs.
thanks again.