EA Forums - Banner

EASHL Connection Questions

Replies

  • Q: Is game traffic marked with IPCoS (class of service / diff serve code point) and if so, what value are the packets assigned? Are both flows marked with CoS... i.e. from client / console and server side?

    (EASHL)
  • Unfortunately not all questions can be answered at this time as we may not have enough information on the subject, so i'll do my best to answer what I can.

    QUESTION: My ping usually averages 11ms, + or - , 0 packet loss, and no huge fluctuations in ping, and yet I still have games that lag terribly. Is there a reason that while the network performance monitor looks great, my game quality is not? I've had VS games in the 30s-70s that perform better than many EASHL games at 11ms

    ANSWER: While ping response time is typically the best predictor for network performance, it is not the sole determining factor for the performance in a single game, and a good ping time does not guarantee a flawless experience. If people are experiencing good ping times and poor responsiveness there could be another underlying issues we need to track down further.



    QUESTION: As mentioned, my ping is usually low. If it's not 11ms, then it is 48 ms. I've long suspected, and have recently confirmed that means I am connected to an East Coast server. Why am I connecting to an East Coast server? (Our club is a West Coast club and we always match region).

    ANSWER: The region that is selected when creating or editing a club has no impact on matchmaking; when a group initiates matchmaking from the Dressing Room we collect the ping times from each user to each of the possible server locations and use an algorithm to determine the “best” region for the group of users. So if I have a club for which the region is set to “NA West”, but the users that I am currently matchmaking with from the Dressing Room have better connections to the “NA East” region, then for the purposes of matchmaking this club will be preferring matchups that will result in a connection to the “NA East” server. I suspect that in some cases this may be a source of confusion for users due to the disconnect between the Club Region setting and the matchmaking, as if I am located on the west coast and my club is set to the “NA West” location I may be unaware that other users that I am matchmaking with might actually have better connections to the “NA East” server, and so even though the best option for me personally would be to connect to the “NA West” server, a better option for my group as a whole might actually be to connect to “NA East”.
    If users are playing a Drop-In game or are playing Club games with the “Match Club Region” rule set to “Relaxed” then there is no guarantee that they will be connected to their “best” server. For example, if we correctly determine that the “NA West” server is the ideal location for a group of users who are matchmaking together we will attempt to match them up with another group of users for whom “NA West” is also the ideal location. If no suitable match can be found then matchmaking will eventually expand to allow them to matchup with a group for whom “NA East” is the ideal location, and in this scenario one of the groups would be forced to play on a server that is not the ideal location for their group. The only way to avoid this would be to implement a hard matchmaking rule that would prohibit users from ever matchmaking with a group that is searching in a different region, which may then make it more difficult for users to find games.
    Note that this does not apply if the user is playing an EASHL Clubs game and has selected the “Exact” option for the “Match Club Region” rule, as we will not expand the matchmaking search to include groups searching in other regions.



    QUESTION: Does the captain matter when searching for a game?Does their network affect everyone who connects in that game, or everyone on that team? Will it affect game quality or ability to find a game?

    ANSWER: The captain is irrelevant when searching for a game. All users who are participating in the matchmaking will be considered equally when evaluation the “best” region to use for that group of users, and no extra weighting is given to the captain user.



    QUESTION: Does playing home give a better connection? I think most people believe so, as lots of people back out if they are away. We've noticed better connections at home most games.

    ANSWER: The Home team would only potentially have a better connection when groups with differing “best” regions are matched together. This certainly shouldn’t be the case most of the time, as we always prefer to matchup two groups that share the same “best” region.



    QUESTION: Does playing home/away affect game experience in VS? I've had nights where every game of VS I played was away, and it was terrible, until I get home and it's like playing offline. Then I have nights where home games are terrible, and I finally play an away game, and it's buttery smooth. Is this all coincidence?

    ANSWER: Whether the user is home or away in VS/HUT should have no impact on gameplay due to the p2p topology of the connection. From game to game there could be a variance of latency since your connection is directly to your opponent. In a small percentage of cases where restrictive NAT settings would have normally blocked a connection between 2 users, we have the ability to move the 2 users to a dedicated game server and depending on the region of both users the game could be hosted on the NA East or NA West servers.



    QUESTION: Playing in a game with a bad connection, even East Coast with an ok feeling connection has lots of issues. Is there anything we can do to change the below issues?
    a. Pucks harder to pickup
    b. Pucks always go back to the other team when dislodged
    c. hits/pokes don't work
    d. AI plays much worse than nomral
    e. Goalies, even humans, give up weaker goals than usual.

    ANSWER: The gameplay, physics, and AI behaves no differently for either team when lagging compared to a game without lag. Added latency and input delay will naturally cause user actions and input to be delayed which in itself will hinder the user’s performance on the controller.



    QUESTION: Does one player having a bad connection, or being from another region affect everyone, some of the players, just one team, or just the player with a bad connection?

    ANSWER: A user with a poor connection in an EASHL game should not impact the connection of other users in the session, regardless of whether they are on his team or the opposing team. The server will broadcast inputs to all users at a fixed interval, so the rate at which I send and receive inputs to and from the server is determined only by my connection with the server, not by the connections of other users in the session. However, a user with a poor connection will potentially be unable to play the game up to their full potential due to the delay in sending and receiving inputs, so that user lagging may affect our team chemistry or going offsides for example.



    QUESTION: I'd like to know at a very basic level how they have implemented lag compensation in EASHL and VS.
    a. Is there lag compensation in VS/EASHL (I'm assuming the answer is yes)?
    b. What is the maximum added latency on a good connection to compensate for those with bad connections?
    c. Are the lag compensation values reevaluated on a regular basis in a game, and how often (roughly), if so?

    ANSWER: I don’t believe that our title utilizes any functionality that would fall under the banner of “lag compensation”. In EASHL games, the game server will broadcast inputs to all users at a fixed interval, so if one person is lagging, only they will experience the delay. For Online Versus games, the connection is peer to peer, so any lag you experience your opponent is also experiencing.

    So your saying drop in games and eashl don't run on p2p ? Does eashl run on IPv4 or 6
  • B_Bunny
    893 posts Moderator
    edited January 2017
    In a 6v6 eashl game, assuming 9 people are east coast and 3 are west - how likely is it that the server chosen is a west server? (Have had enough occurences with this already)

    And if a west server is indeed chosen, dont you find it counterintuitive to make 9 other players play on a rather large delay (30 to 50 ping in some cases) rather than have the 3 players from west play on an east server on such a delay?

    I know youve said its taking averages and choosing the best server, but with the above scenario that seems unlikely at times. As a matter of fact there are times where I feel it chooses the minority.
    PSN: B-Bunny
  • VeNOM2099
    3178 posts Member
    edited January 2017
    B-Bunny wrote: »
    In a 6v6 eashl game, assuming 9 people are east coast and 3 are west - how likely is it that the server chosen is a west server? (Have had enough occurences with this already)

    And if a west server is indeed chosen, dont you find it counterintuitive to make 9 other players play on a rather large delay (30 to 50 ping in some cases) rather than have the 3 players from west play on an east server on such a delay?

    Well here's the thing, from what he said I gathered that the game is basing its decision on what is best for the group based on every players' connection to each. So in your example, if you would connect to the East Servers, then:

    Player 1 -9 on West Coast = 45ms ping
    Players 10-12 on East Coast = 20ms ping

    But if you would connect to the West Servers, then:

    Players 1-9 on West Coast = 15ms ping
    Players 10-12 on East Cost = 35ms ping

    Then the game would select that your group connects to the West server, because overall the connection would be better for everyone as a group.
  • B_Bunny
    893 posts Moderator
    edited January 2017
    VeNOM2099 wrote: »
    B-Bunny wrote: »
    In a 6v6 eashl game, assuming 9 people are east coast and 3 are west - how likely is it that the server chosen is a west server? (Have had enough occurences with this already)

    And if a west server is indeed chosen, dont you find it counterintuitive to make 9 other players play on a rather large delay (30 to 50 ping in some cases) rather than have the 3 players from west play on an east server on such a delay?

    Well here's the thing, from what he said I gathered that the game is basing its decision on what is best for the group based on every players' connection to each. So in your example, if you would connect to the East Servers, then:

    Player 1 -9 on West Coast = 45ms ping
    Players 10-12 on East Coast = 20ms ping


    But if you would connect to the West Servers, then:

    Players 1-9 on West Coast = 15ms ping
    Players 10-12 on East Cost = 35ms ping

    Then the game would select that your group connects to the West server, because overall the connection would be better for everyone as a group.

    Ideally, but too many times ive had the bolded occur, so it doesnt seem to make sense to me.
    PSN: B-Bunny
  • B-Bunny wrote: »
    In a 6v6 eashl game, assuming 9 people are east coast and 3 are west - how likely is it that the server chosen is a west server? (Have had enough occurences with this already)

    And if a west server is indeed chosen, dont you find it counterintuitive to make 9 other players play on a rather large delay (30 to 50 ping in some cases) rather than have the 3 players from west play on an east server on such a delay?

    I know youve said its taking averages and choosing the best server, but with the above scenario that seems unlikely at times. As a matter of fact there are times where I feel it chooses the minority.

    When choosing your group’s best game location it is only looking at your own team’s pings, so in this case assuming you had 3 west and 3 East, there in theory is an even split. We look at everyone’s ping response to each game server and use the mean and variance to determine which server is best for the team as a whole. It might be that the 3 West users on your team have a worse ping time to the East server while your East guys may have a better ping to West, thus your game could be sent to West. If you have your matchmaking settings set to “match my region” and you are connecting to West, your opponent will also be on West, so this shouldn’t be affecting their connection.
  • BoboFloggins
    2170 posts Member
    edited January 2017
    B-Bunny wrote: »
    In a 6v6 eashl game, assuming 9 people are east coast and 3 are west - how likely is it that the server chosen is a west server? (Have had enough occurences with this already)

    And if a west server is indeed chosen, dont you find it counterintuitive to make 9 other players play on a rather large delay (30 to 50 ping in some cases) rather than have the 3 players from west play on an east server on such a delay?

    I know youve said its taking averages and choosing the best server, but with the above scenario that seems unlikely at times. As a matter of fact there are times where I feel it chooses the minority.

    When choosing your group’s best game location it is only looking at your own team’s pings, so in this case assuming you had 3 west and 3 East, there in theory is an even split. We look at everyone’s ping response to each game server and use the mean and variance to determine which server is best for the team as a whole. It might be that the 3 West users on your team have a worse ping time to the East server while your East guys may have a better ping to West, thus your game could be sent to West. If you have your matchmaking settings set to “match my region” and you are connecting to West, your opponent will also be on West, so this shouldn’t be affecting their connection.

    with that and:

    "In EASHL games, the game server will broadcast inputs to all users at a fixed interval"


    I would have to venture to say that, yes, there is a larger more complex issue going on as we see the slower / input lag even with my constant ping of 12 ms.

    Based on what you are saying here, when folks see a layer 3 ping with ~ 15 ms response times, they should see no issue, regardless of 2 v2 or 6 v 6 if I am reading you correctly.
  • Sgt_Kelso
    1325 posts Member
    edited January 2017
    How come my player sometimes does 'involuntary' movements, i.e. dashes or turns in completely different direction that I am guiding it, or even doing that when there's no input? What causes it, because the assumption was it was caused by some sort of lag correction? But now you say there isn't one?

    I did a test by deliberately choking down my bandwith to see if it had any effect. The end result seemed to be that the ping stayed the same, but the player felt a bit more sluggish with slower Mbps rates. How come? If the ping stays the same, one would think I'd receive the game data at same speed, but judging by the responsiveness, this wasn't the case?

    Would it be possible to make it so, that we could see the ping of all our team's players in the dressing room?

    Oh, and welcome to the forum, and thanks for taking the time to asnwer our questions.


  • Why give users the ability to choose their region at all if you never actually use that information? All it would take is one brief network hiccup for the game to think you should be on the West instead of the East, even though your connection to the East is far more reliable. Can't you give users the ability to decide what server is best for themselves?

    Thanks for answering those questions!
  • Why give users the ability to choose their region at all if you never actually use that information? All it would take is one brief network hiccup for the game to think you should be on the West instead of the East, even though your connection to the East is far more reliable. Can't you give users the ability to decide what server is best for themselves?

    Thanks for answering those questions!

    In previous years this was the case and your Club Region was where your preferred game server would have been. In many cases, especially for those that live in Central US or Canada, they may not know which region is best for themselves. Giving back control is a possibility we could look at, but the goal was to automatically decide which server would be best for the group based on metrics and not necessarily by speculation.
  • Sgt_Kelso wrote: »
    How come my player sometimes does 'involuntary' movements, i.e. dashes or turns in completely different direction that I am guiding it, or even doing that when there's no input? What causes it, because the assumption was it was caused by some sort of lag correction? But now you say there isn't one?

    I did a test by deliberately choking down my bandwith to see if it had any effect. The end result seemed to be that the ping stayed the same, but the player felt a bit more sluggish with slower Mbps rates. How come? If the ping stays the same, one would think I'd receive the game data at same speed, but judging by the responsiveness, this wasn't the case?

    Would it be possible to make it so, that we could see the ping of all our team's players in the dressing room?

    Oh, and welcome to the forum, and thanks for taking the time to asnwer our questions.


    Your test and the result sound expected. Your bandwidth likely won't change your ping time too much, but the net result as you have said will be your sluggish/delayed gameplay. I have experience this myself while playing and while someone else in the house is watching Netflix for example.
  • Sgt_Kelso wrote: »
    How come my player sometimes does 'involuntary' movements, i.e. dashes or turns in completely different direction that I am guiding it, or even doing that when there's no input? What causes it, because the assumption was it was caused by some sort of lag correction? But now you say there isn't one?

    I did a test by deliberately choking down my bandwith to see if it had any effect. The end result seemed to be that the ping stayed the same, but the player felt a bit more sluggish with slower Mbps rates. How come? If the ping stays the same, one would think I'd receive the game data at same speed, but judging by the responsiveness, this wasn't the case?

    Would it be possible to make it so, that we could see the ping of all our team's players in the dressing room?

    Oh, and welcome to the forum, and thanks for taking the time to asnwer our questions.


    Your test and the result sound expected. Your bandwidth likely won't change your ping time too much, but the net result as you have said will be your sluggish/delayed gameplay. I have experience this myself while playing and while someone else in the house is watching Netflix for example.

    OK.

    What's the expected upload / download requirements for the game.
  • Lol. I have 220mbps down and 20 up. I never play on WiFi and am usually the only one using my bandwidth. My ethernet cable is cat6 and there is no damage to it.

    Still lag. Whatever piece of **** formula you guys are using , just would like to say thank you. Lag in 7/10 games is very enjoyable. Keep up the good work.

  • Q: Is game traffic marked with IPCoS (class of service / diff serve code point) and if so, what value are the packets assigned? Are both flows marked with CoS... i.e. from client / console and server side?

    (EASHL)

    in general most providers of internet services will ignore any QOS tagging of packets. So it would be a waste of time.
  • Q: Is game traffic marked with IPCoS (class of service / diff serve code point) and if so, what value are the packets assigned? Are both flows marked with CoS... i.e. from client / console and server side?

    (EASHL)

    in general most providers of internet services will ignore any QOS tagging of packets. So it would be a waste of time.

    Not sure we are talking apples to apples. DSCP /CoS is not the exactly the same as QOS.

    But either way, you may be correct when talking about consumer grade internet services. I didn't think about that. Most ISP's do likely ignore the TOS field for good reason.

    I work on enterprise grade networks over MPLS and it is a necessity with converged services.

    Brain fart.
  • Why give users the ability to choose their region at all if you never actually use that information? All it would take is one brief network hiccup for the game to think you should be on the West instead of the East, even though your connection to the East is far more reliable. Can't you give users the ability to decide what server is best for themselves?

    Thanks for answering those questions!

    Most games offer you a choice from 3 servers or so, and sometimes 1 will be recommended. In the end though, you decide which one you want. This game should be the same.
  • Sgt_Kelso wrote: »
    How come my player sometimes does 'involuntary' movements, i.e. dashes or turns in completely different direction that I am guiding it, or even doing that when there's no input? What causes it, because the assumption was it was caused by some sort of lag correction? But now you say there isn't one?

    I did a test by deliberately choking down my bandwith to see if it had any effect. The end result seemed to be that the ping stayed the same, but the player felt a bit more sluggish with slower Mbps rates. How come? If the ping stays the same, one would think I'd receive the game data at same speed, but judging by the responsiveness, this wasn't the case?

    Would it be possible to make it so, that we could see the ping of all our team's players in the dressing room?

    Oh, and welcome to the forum, and thanks for taking the time to asnwer our questions.


    Your test and the result sound expected. Your bandwidth likely won't change your ping time too much, but the net result as you have said will be your sluggish/delayed gameplay. I have experience this myself while playing and while someone else in the house is watching Netflix for example.

    And yet my daughter will stream youtube videos while my son plays his online game and then I play NHL. 98% of the time, I will be lag free despite all this. All on a 50/5 internet connection, hardwired. Only one wireless is my son.
Sign In or Register to comment.

Howdy, Stranger!

It looks like you're new here. Sign in or register to get started.