All players in the match should have a ping of 200ms or better to the host, but the individual connections between players could be greater. Let's take a worst-case scenario: the host is in Los Angeles and one player is in Toronto and has a ping of 150ms to the host. Another player is in Recife, Brasil and has a ping of 190ms to the host. The connection between the player in Toronto and Brasil is going to be >200ms. This is the worst-case scenario however - matchmaking will always be first looking for hosts that are geographically close to you, and this expands in stages as the search continues.
The further worst-case scenario is if a player is behind a strict NAT (type 3). In that case all the traffic from that player will have to be routed through the host and then to all the other players, which will increase the connection time further.
newuxtreme: So to answer your question, yes it would still be useful to get some bandwidth tests to players with >200ms ping, although I'm most interested in these tests from players who have a severely limited upstream that may be the cause of those players to getting the network disconnect frequently. This is a separate case from the lag you showed in your video.
And thank you for the video, those are some really bad cases. The later ones look like the more typical lag - a second or so after you down a player you are receiving that player's damage events to you. Which is still not good - it means that even if a player is not dropping packets severely enough to cause the lag detect to kick in, their baseline latency is still high enough to cause these double-down situations with over a second between the down events.
The first case is much more strange since the player doesn't die when they should but also doesn't appear to be frozen from the lag detect, which should be the only case that would prevent them from going down immediately. Then when you shoot them again the downed event doesn't come until 11 seconds later. On Monday I'll talk to our network programmers about these cases. The delayed-damage / double-down case is the baseline lag we've been struggling with for a long time and isn't something we expected would be solved with the new lag detect threshold that was added in 1.06, but it's obviously still a big issue. The 11-second delay to down a player is one of the extreme cases that I would have expected the lag detect threshold should be catching but obviously it isn't.
BTW did you happen to check the scoreboard during the match and see what the ping bars were on that player?
I just wanted to second and agree that we truly appreciate you being so vocal about the way things work and the efforts that are being taken in order to help those still might have issues. I play with a lot of folks and they really are from all over the world, and still randomly keep having the white plug errors. India, Germany, US, Canada, Australia , each one having different speeds and decent DL/UL as well. Though I must agree with most that the recent patch has improved things by a signiicant amount say 75% compared to the pure 1.06 patch.
I will go ahead and try to use the bandwidth tests with players from all of the above players if possible. The question is which connection would you prefer I test, the 2/.5 one (this connection has been used for years, it's stable and reliable, for all games ,BF3, GOW, UC2,3 and every other, even TLOU until 1.06) or the current 20/20 where upload speed coulldn't possibly be an issue. The pings for both connections are here and there the same, and since we play with the above mentioned players as opposed to those in say India/Pak/Asia a lot more our average pings in gaming are always on the higher end. Not that this was an issue pre 1.06. Also matchmaking in ND games since U2 has never for some reason been able to find players in Asia fast enough, so inevitably we use someone in the US/UK/Can to host and keep playing. Again, the lag would be felt only say once in 5-10 games and 1 or 2 players max, and these players would lag for everyone not just one of us.
I'm sorry but I don't think that I saved that high school video's scroreboard, I'll try to do so the next time we ever face such weird lag.
Ndi edaly thanks for the question . No no one dropped and it was towards the end of the match . The person who was invisible is from Canada I believe not sure exactly what part I'm from long island Ny . A person who had the black screen on my team nick a 30 is from a European country I think . Would be hard to recreate because not all of the peopleexpeople are on my friends list . Good explanation though.
i'm from Canada lol.
And yeah it was weird. it was during a last man standing for the enemy in SR. i got killed by the invisible player and then i got a black screen and couldn't respawn (had to wait until the game was over).
newuxtreme: if you could do some bandwidth tests with your 2 / .5 connection that would be useful. It would also be good to know, just as a reference point, what the speedtest measures on the player(s) you're testing with (in addition to the ping).
I still find the most annoying thing that came with patch 1.06 is the invisibility bug.. It happens almost every game now. If it's not me it's my teammate or the enemy team. Either way it ruins the experience big time. Invisibility bug never happened to me prior to 1.06 so there must be something that has been introduced that causes it.
Last night in a game a very first time that the invisibiity came back to normal.. At a sudden time all enemies and teammates except the ones in my party were gone and 3 min later I could see them again. This hasn't happened before that invisible people became visible again. Although the enemy team began to leave and I think it's because they had some game breaker bug too, since we played in the same lobby for 3 games already.
If I make the network test on the ps3 somehow I get only about 1/3 of my usual dl and upload speed at the moment (with wired connection! and NAT2).
I dont know if it's that what caused the lag, but why is the ps3 not using my full bandwidth every other device in the network can (there are also 3 pcs connected o0, that's strange).
Some weeks ago I got always the full speed at the ps3 network test and The Last of US was lag free too.
It seems that the router cuts down the speed if other devices are connected in my case, some sort of traffic splitting even if there are no downloads, so i advice you to check if your speed test results are ok.
Things I've done for this speed test.
Used the 2/.5 connection.
Used a wired connection completely dedicated to the laptop that was testing.
NOT DMZ though, which is what I always do on the PS3. Simply because it's not 'safe' on the PC.
I could repeat the ping tests with DMZ Nat if you require, if it makes any difference to the ping test, because the speeds should remain the same.
I hope these results are useful, took forever to get them!
1. Speed test results
2. File transfer speeds over Skype, highest, lowest and what we saw was a consistent average transfer speed. Both ways.
3. Ping Tests.
My location and Speedtest result.
Test 1 - Germany
1.Transferred a song file from me to him first.
First time :- 52-62kbps Range of speeds with it just varying between these two numbers
Second time :- Range of Speeds seen 39-47-52-62kbps 47-52kbps avg speed.
Transferring from him to me
1st time Range :- 20-62 Avg:- 38
2nd time Range :- 19-74 61 Avg :- 50
2. Ping test was attempted from CMD but couldn't even connect to each other. So went to ping.eu/ping and got this result:-
My result pinging him.
--- PING 126.96.36.199 (188.8.131.52) 56(84) bytes of data. ---
--- 184.108.40.206 ping statistics ---
packets transmitted 9
packet loss 100 %
time 8058 ms
His result pinging me.
3. His speedtest results