How should lag be handled in tournaments?

[Charlight]

(shameless spinoff off @Citrus 's thread)

Since Calamari Cup 2, TO's have been trying to figure out how to handle lag in tournaments. The problem was present before but got more attention with how in CC2 there were a few teams with severe lag issues, making the game more or less unplayable.

First, a little bit of theory. So we don't run into any misconceptions about latency and lag, the gist of it is this:

Latency (or ping) is the time it takes for your Wii U sending a packet of information to another Wii U. (one-way delay time. There's also round-trip delay time, but I'm just trying to explain the concept of it).

Lag is a large, noticeable delay between you sending a packet of information and the other Wii U receiving it. This can be caused either by a high ping or anything else the data bumps into along the way (e.g. data loss/corruption, wifi interference, heck, even a ****ty ethernet cable).

Even though you might have good ping and no packet loss on an internet connection test you do on your computer, your Wii U's internet connection might still be affected.

I'm sorry for using you as an example @Ryanator , but this is one of the craziest things I've ever seen lag do:

While the Wii U doesn't support Java or Flash, it's possible to do an HTML5 connection test in your Wii U internet browser. Using www.ping.eu/ping, connecting to www.nintendo.co.uk I get the following results:

While I'm not sure how reliable these results are, I'd really appreciate it if somebody a little bit more tech savvy could elaborate on whether or not an internet connection like this could be used as proof of lag.

It's obvious we need some sort of lag rule, but how? As some teams don't own a capture card, it'd be unfair to demand footage of the other team lagging. Even so, if they did, it would take too long for a team to upload the footage and the TO to review it in an actual tournament.

What do you guys think?

 

[pixelatedcody]

It also factors into account, if you're streaming that would create issues as well, Honestly the really tedious way of doing a lag test, is both teams hosts or the whole team provide an IP Address to ping, this would show you your latency (ms) to each teammate (or an ip of an ISP in that area maybe? as that's a bit more secure than giving out your ip address. limit the possibility of a DDOS.), however that's extremely tedious- and doesn't even factor in if you can't even ping the ip due to firewalls or NAT.


So It's a real mess.

 

[Aweshucks]

what are you suggesting we do? test the ping for each player in the tournament using a system that sonicwaffle has already explained to you doesn't adress the problem all that well? That would take absolutely forever.

There's not too much we can do about it right now, unfortunately. we can punish them for DCing, but up to that I don't know what we could do. If someone can come up with a decent way to adress lag that isn't too restricting for players and doesn't take forever, then I'd love to hear it and implement it

 

[SupersonicWaffle]

It also factors into account, if you're streaming that would create issues as well, Honestly the really tedious way of doing a lag test, is both teams hosts or the whole team provide an IP Address to ping, this would show you your latency (ms) to each teammate (or an ip of an ISP in that area maybe? as that's a bit more secure than giving out your ip address. limit the possibility of a DDOS.), however that's extremely tedious- and doesn't even factor in if you can't even ping the ip due to firewalls or NAT.


So It's a real mess.

Well it would be really stupid to try to ping a private IP address and expect it to reach something that is not in your own network so NAT doesn't play a role. Most consumer grade routers won't have ICMP disabled so pinging another player wouldn't be that big of a problem.
I agree though that it's not exactly a good idea to give everyone your IP-address assuming your ISP's DHCP lease is longer than a few hours which it probably is.

The only reason your Wii U's latency should be worse than your PC's is if you have problems with your wifi which in all honesty any "serious" player shouldn't be using anyway (I know they still do though so this test may be worthwhile to some).
While running the test you would also have to simulate the load of a Splatoon game on your connection, I might look at the exact bandwidth usage of Splatoon this weekend but I can't promise.

Regarding streaming while playing, yes this can be problematic but only if you're maxing out your upload speed because then the packets will enter a queue and will have to wait to be sent, i.e. increasing latency. When that happens though your stream is also likely to drop a lot of frames, meaning it will buffer a lot for your viewers. At that point you would probably be best of to just stop streaming

 

[Citrus]

RIPOFF ARTIST

 

[Agosta44]

Last tournament a team we played was streaming and had massive lag. We told them and they didn't care. It made the game bull**** and I lost interest in playing the set. If someone is lagging and you have proof TOs should hold the other team accountable if they choose not to do anything to remedy the situation.

 

[Static]

Maybe everyone should be hard wired with an ethernet adapter.

 

[applepie]

I find it funny how some teams will accuse others of lagging even after they have a game where someone on their team disconnects (and is teleporting around). Lag is just always going to be a thin line that you have to tread lightly on, until some kind of LAN thing happens, it's always going to be an issue. Also I cannot stress this enough but "MOST" cases of lag... goes both ways. As in, say if there is a single person on a team that seems laggy, there is a rare case where this is only person at an advantage. Yes lag only adds more RNG to the game, as it can be RNG based in the sense of who gets the "upper" hand from said lag.. but on a statistical bases it should be close to the same? Lag is always going to come down to some from of a gentlemen's agreement, outside of very extreme cases.. until/if we can ever lan.

Maybe everyone should be hard wired with an ethernet adapter.

Yea I agree, I don't even know how people who are not hard wired can even complain about lag. No matter how good your internet is the wii u's built in wireless adapter is very limited let alone for competitive play.

 

[Saturn]

We really need to build the community to the point where we can support intra-continental teams and tournaments.

It's silly how many teams consist of players halfway across the world from each other. Those distances make ping around a third of a second (20 goddamn frames).

 

[SupersonicWaffle]

Last tournament a team we played was streaming and had massive lag. We told them and they didn't care. It made the game bull**** and I lost interest in playing the set. If someone is lagging and you have proof TOs should hold the other team accountable if they choose not to do anything to remedy the situation.

see here's the problem we would have as TOs
really the most accurate measurement we could have is getting video of how a laggy player is killed and calculate the latency between the last hitsound and the kill confirmation then put it in relation to the locations the players are in and gauge if it's X% more than it should be. This measurement would be extremely inaccurate though because we're talking about milliseconds here and a ton of gamer-grade capture cards won't capture well enough to provide accurate video not to mention splatoons framerate dipping especialy for hosts. Next thing is that ryanator's case gives us pretty decent information that hit-information is directly sent to the affected player so we would need multiple POVs because it could be a problem with the accuser instead.

So this would end up requiring multiple guys to upload a bunch of videos, this is just not doable in a tournament setting, it simply takes too long.

 

[Charlight]

Do you guys think it would be fair to disqualify a player based on their download/upload speed? As in, a bad internet connection?

How this would work for example is, if a team complains about a player lagging, the player would do a speedtest and send a screenshot of the results to a TO. If their connection was e.g. below 1mbps they would be told to sit it out for the rest of the tourney.

But if internet connection isn't the cause of lag (lag can be caused by all sorts of things, but connection is a common culprit), what would we do then?

 

[Sendou]

We really need to build the community to the point where we can support intra-continental teams and tournaments.

It's silly how many teams consist of players halfway across the world from each other. Those distances make ping around a third of a second (20 goddamn frames).

Looking forward to Ocenian Splatoon tournaments myself!

 

[jsilva]

Yea I agree, I don't even know how people who are not hard wired can even complain about lag. No matter how good your internet is the wii u's built in wireless adapter is very limited let alone for competitive play.

I got the UTechSmart ethernet adapter. The first time I used it I had the most crazy lag I've ever seen—this included a disembodied inkbrush swinging for a few seconds after getting splatted and the tower suddenly advancing something like 40 points in just a few seconds moving ridiculously fast. So I switched back to wireless (a really nice Asus router). Recently I connected the ethernet adapter again wondering if it was a fluke with Nintendo's servers and it has been mostly fine, but I've experienced moments of lag that I've never experienced over wireless. Certainly no better than wireless, at least from my end.

Just personal experience...