EA Forums - Banner

Add Jitter to Network Performance Meter

ca5kQQN.png

It makes life much easier when EA can help me filter out the laggers by measuring their jitter.

The data is there already.

Look at the amount of fluctuations in ping response time from 5 seconds ago to a minute ago.

A great connection would have the ping response time to be as flat line and consistent as possible.

I have had to quit the game on so many people after looking at their network performance graph.

Replies

  • Here are three more graphs
    All of them stay within the "GOOD" response time metric as designed by EA.

    xFYoDZ2.png

    TNzQ4ZA.png

    eYdaNhM.png

    What the Good/Fair/Poor metric doesn't tell you, is how much fluctuation is in your ping response time
  • Ziostilon wrote: »
    Here are three more graphs
    All of them stay within the "GOOD" response time metric as designed by EA.

    xFYoDZ2.png

    TNzQ4ZA.png

    eYdaNhM.png

    What the Good/Fair/Poor metric doesn't tell you, is how much fluctuation is in your ping response time

    "Minimum" and "maximum".

    Or "highest" and "lowest" in this case.

    Not sufficient?
  • If you had one huge spike out of nowhere, but the rest of the 50 second sample size... your response time was steady.

    The high and low meter would not be able to show that.

    My main gripe is that i gotta go into the network performance menu to get this information everytime.
    There is already a GOOD/FAIR/POOR bar graph on the top right corner when you match up with your opponent.
    Just build in a threshold that if the jitter in your connection goes beyond a certain amount. Have some indicator go off, letting the user know that the game you're about to play is going to lag.
  • You're talking versus and HUT, right?
  • Packet loss should also be included if we are on this subject. I had a green bar against some guy and yet faced the worst lag because the packet loss was at 10%, that isn't really an indicator of a great connection.
  • Bmh245
    905 posts Member
    Ziostilon wrote: »
    Here are three more graphs
    All of them stay within the "GOOD" response time metric as designed by EA.

    xFYoDZ2.png

    TNzQ4ZA.png

    eYdaNhM.png

    What the Good/Fair/Poor metric doesn't tell you, is how much fluctuation is in your ping response time

    Yeah, the problem is that Good/Fair/Poor tells you what the ping is at that moment. It doesn't reflect the variability of the ping at all. So in your 2nd graf, when the ping spiked to 63ms, the connection meter would have gone to yellow at that moment. But once it passed, looking at the connection meter would never tell you it happened.

    The logical fix is that if the ping rate is "fair" at any point, the connection meter should always say "fair," since what I'm concerned about as a player is not just what the ping is at the moment, but how variable it'll be. Don't tell me the connection is "Good" because the ping is 25ms right now, if 30 seconds earlier the ping was 60ms.
  • You're talking versus and HUT, right?

    Yes you are correct
  • Here's a couple more:

    gQVZvYb.png

    nOLIexl.png

    jy4Rent.png

    WvMSfCP.png

    yRYBzAA.png

    Sometimes it levels out by the end of the game. It could be acceptable one period, and unplayable in another. That's an unbalanced playing field.
  • BoboFloggins
    2170 posts Member
    edited March 2017
    What is that from..... vs hut or eashl

    Sorry yes you already answered that.

    Well, that's going to happen but jitter is still a variance in packet delivery. The chart is still showing your highest and lowest times. If anything, perhaps the sample size is too small, no?

    I mean, either way, in peer to peer there is no guarantee that you are going to get consistency. There are too many variables.
  • Bmh245
    905 posts Member
    What is that from..... vs hut or eashl

    Sorry yes you already answered that.

    Well, that's going to happen but jitter is still a variance in packet delivery. The chart is still showing your highest and lowest times. If anything, perhaps the sample size is too small, no?

    I mean, either way, in peer to peer there is no guarantee that you are going to get consistency. There are too many variables.

    The problem with the ping charts is that the highest and lowest ping response is just the highest and lowest in the last two minutes, not the highest and lowest over the course of the game. I've played games where ping spiked as high as 500ms (which I was able to see by looking at network performance during intermissions), but at the end of the game it said my highest ping was 50ms or whatever. At the very least, EA should record "highest" and "lowest" pings for the game as a whole.

    EA should also rate any connection that at any point spikes above 50ms (or whatever the cutoff line between Green and Yellow is) as Fair. I don't care if my ping is currently low. What I care about is how high it might get, and any connection that has spiked high should always be rated as Fair, even if the ping drops to a reasonable rate.

  • Hi....I had a tech come out, he came inside and afterward said he checked the signs and it wasn't from my structure or association. Said he settled on certain decisions and a few people would grow a hub that was at high volume. Guaranteed me subsequent to running a few additional tests and all that I wouldn't be charged yakkity yak it wasn't me or my wiring and so on in light of the fact that I had him check it, while he was here. I traded out everything and redesigned, he advised me to stand by up to 14 days. I am at my absolute limit here, I will share some information on what my association resembles and what I can imitate for certain tests.
Sign In or Register to comment.

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!