4v4 casted replay for zhl

would you like 4v4 casted replay? (please read the thread before voting)

Yes, immediately cease live casting and cast from replays
14
42%
Maybe, try Crush's "find the xapos" solution for a week first
3
9%
No, games should always be cast live
14
42%
Frankly, my dear, I don't give a damn
2
6%
 
Total votes: 33
User avatar
wawmoose
Posts: 590
Joined: Sun Oct 12, 2014 11:27 pm
Battle.net Name: WaWMoose
Battle.net Char Code: 1878
Battle.net Server: NA
Location: Canada ofc

Re: 4v4 casted replay for zhl

Post by wawmoose » Thu Jan 08, 2015 8:20 pm

This may be pointless, but would someone please enlighten me on how this didn't come up in season 2? All the games were live casted correct? Including the playoffs, yet not alot of complaining about casters lagging games happened iirc.
Also one more thing, I think an advantage on every team 12+ members is flexibility, and I think it is a bit ridiculous if owners expect to be starting their top 4 every game, at least not until playoffs. You have lots of players, use them.

Also, daggius's last comment has a great plan imo. +1 on that.
RIP Flyers
User avatar
daggius
Posts: 442
Joined: Tue Jul 29, 2014 12:53 am
Battle.net Name: daggius
Battle.net Char Code: 0
Battle.net Server: NA

Re: 4v4 casted replay for zhl

Post by daggius » Thu Jan 08, 2015 8:56 pm

wawmoose wrote:This may be pointless, but would someone please enlighten me on how this didn't come up in season 2? All the games were live casted correct? Including the playoffs, yet not alot of complaining about casters lagging games happened iirc.
Also one more thing, I think an advantage on every team 12+ members is flexibility, and I think it is a bit ridiculous if owners expect to be starting their top 4 every game, at least not until playoffs. You have lots of players, use them.

Also, daggius's last comment has a great plan imo. +1 on that.
-Season 2 was almost exclusively cast by only 3 people: Kyle, Teroh, and Kurt. By comparison, Season 3 has a lot more different casters who as a whole have more variation in their broadcast/sc2 setting optimizations, rigs, internet, experience, etc.

-Season 2 did not have benches. Season 3 does. More players in the game means more chance for lag problems.

-Season 2 was mostly played on west since sc2 closed the east and central servers at beginning of season. So there was no arguing about server because there was only one server available-- west.

-Season 2 actually did have substantial lag problems, especially in playoffs. Every single game in the finals series was remade at least once because Vapour would try to play , then remake, try again, then remake and finally use ooter. Kyle and some random guy were casting and lagging him. If you check stream section of forum you see a few posts of people asking why is delay and lag always worse in the casted games??
User avatar
krazymen
Posts: 1170
Joined: Wed Feb 26, 2014 6:10 pm
Battle.net Name: krazymen
Battle.net Char Code: 249
Battle.net Server: NA

Re: 4v4 casted replay for zhl

Post by krazymen » Thu Jan 08, 2015 9:00 pm

wawmoose wrote:This may be pointless, but would someone please enlighten me on how this didn't come up in season 2? All the games were live casted correct? Including the playoffs, yet not alot of complaining about casters lagging games happened iirc.
Also one more thing, I think an advantage on every team 12+ members is flexibility, and I think it is a bit ridiculous if owners expect to be starting their top 4 every game, at least not until playoffs. You have lots of players, use them.

Also, daggius's last comment has a great plan imo. +1 on that.
the problem about the players is that theres a huge skill gap between lets say a 1st round and a 6-7th round so when you cant play 1 starter sure manageable when you cant play 2 it starts to be a problem and when you cant have 3 or even any of them on the ice its almost the same as a forfeit. Its for that reason that we created the probe league in order to let people with the same skills play together. Also oddly enough i never experienced the lag that i am experiencing in season 2. Sure i lagged some games in season i remember clearly lagging in a playoff game however I didnt feel like it was necessary for me to talk about it since it wasnt frequent and i actually had spikes in those games.
Bowling trophy (birthday trophy)
Achieved level 105 in starcraft 2
Undefeated in 5v5
self proclaimed director of ironic trophies
User avatar
l)arkangel
Posts: 1106
Joined: Tue Apr 08, 2014 10:50 am
Battle.net Name: Darkangel
Battle.net Char Code: 1457
Battle.net Server: NA
Location: Troll City

Re: 4v4 casted replay for zhl

Post by l)arkangel » Thu Jan 08, 2015 9:04 pm

Definitely lagged super hard about halfway through a game vs freeze in playoffs last year. The game where vapour didn't play and ooter did....
*S5 Champion*

Player Profile
User avatar
Watermelon
Posts: 1374
Joined: Sun Feb 02, 2014 2:34 pm
Battle.net Name: Watermelon
Battle.net Char Code: 1918
Battle.net Server: NA

Re: 4v4 casted replay for zhl

Post by Watermelon » Thu Jan 08, 2015 9:10 pm

daggius wrote:
wawmoose wrote:This may be pointless, but would someone please enlighten me on how this didn't come up in season 2? All the games were live casted correct? Including the playoffs, yet not alot of complaining about casters lagging games happened iirc.
Also one more thing, I think an advantage on every team 12+ members is flexibility, and I think it is a bit ridiculous if owners expect to be starting their top 4 every game, at least not until playoffs. You have lots of players, use them.

Also, daggius's last comment has a great plan imo. +1 on that.
-Season 2 was almost exclusively cast by only 3 people: Kyle, Teroh, and Kurt. By comparison, Season 3 has a lot more different casters who as a whole have more variation in their broadcast/sc2 setting optimizations, rigs, internet, experience, etc.

-Season 2 did not have benches. Season 3 does. More players in the game means more chance for lag problems.

-Season 2 was mostly played on west since sc2 closed the east and central servers at beginning of season. So there was no arguing about server because there was only one server available-- west.

-Season 2 actually did have substantial lag problems, especially in playoffs. Every single game in the finals series was remade at least once because Vapour would try to play , then remake, try again, then remake and finally use ooter. Kyle and some random guy were casting and lagging him. If you check stream section of forum you see a few posts of people asking why is delay and lag always worse in the casted games??
^this
S4 Champ | Playoff MVP | "Best Owner"
EUDL7 Champ
BaseTradeTV ZH3K Winner
record for most goals in a ZHL game: http://www.zealothockey.net/event/450/
User avatar
wawmoose
Posts: 590
Joined: Sun Oct 12, 2014 11:27 pm
Battle.net Name: WaWMoose
Battle.net Char Code: 1878
Battle.net Server: NA
Location: Canada ofc

Re: 4v4 casted replay for zhl

Post by wawmoose » Thu Jan 08, 2015 9:55 pm

daggius wrote:
wawmoose wrote:This may be pointless, but would someone please enlighten me on how this didn't come up in season 2? All the games were live casted correct? Including the playoffs, yet not alot of complaining about casters lagging games happened iirc.
Also one more thing, I think an advantage on every team 12+ members is flexibility, and I think it is a bit ridiculous if owners expect to be starting their top 4 every game, at least not until playoffs. You have lots of players, use them.

Also, daggius's last comment has a great plan imo. +1 on that.
-Season 2 was almost exclusively cast by only 3 people: Kyle, Teroh, and Kurt. By comparison, Season 3 has a lot more different casters who as a whole have more variation in their broadcast/sc2 setting optimizations, rigs, internet, experience, etc.

-Season 2 did not have benches. Season 3 does. More players in the game means more chance for lag problems.

-Season 2 was mostly played on west since sc2 closed the east and central servers at beginning of season. So there was no arguing about server because there was only one server available-- west.

-Season 2 actually did have substantial lag problems, especially in playoffs. Every single game in the finals series was remade at least once because Vapour would try to play , then remake, try again, then remake and finally use ooter. Kyle and some random guy were casting and lagging him. If you check stream section of forum you see a few posts of people asking why is delay and lag always worse in the casted games??

Thank you for the explanation. Knowing this, I agree with the plan you mentioned earlier even more. +1
RIP Flyers
User avatar
Teroh
Legend
Posts: 479
Joined: Wed Mar 12, 2014 7:47 pm
Battle.net Name: Teroh
Battle.net Char Code: 773
Battle.net Server: NA

Re: 4v4 casted replay for zhl

Post by Teroh » Fri Jan 09, 2015 6:02 am

Based on some testing I've done with my own stream in IH, with exaggerated settings (90 FPS) to push my upload up to averaging ~1.8 MB/S upload, occasionally spiking up to 2.2, but never above. I can lower my stream settings to average 1.6 MB/S, never going above 2.0 MB/S. I would say setting a limit of 2.5 MB/S upload would be enough of a buffer for good streaming, but even 2 would suffice with low enough settings.

Players (Watermelon, Vapour, PeterDLai, htcp, Eric, CountryKen, moose, Rush, I don't remember the others) reported no significant difference in subjective lag on either East or West when I was:
not streaming at all
streaming at 30 FPS
60 FPS
90 FPS
me (90 FPS, 540p), moose (60 FPS, 720p), and htcp (unknown settings) were all streaming at the same time.

The biggest reported difference occurred when a player also had the stream open.

With that data and the split in the poll, I am making a ruling for now regarding this issue:

Games will continue to be streamed live.
In the case of severe lag, as agreed upon by both team captains within a game, the caster can be forced to show they are not the one lagging AND the benches will be required to leave the game. This occurs as follows:
Step 1: Caster will show they are at less than 3 spikes.
Step 2: The caster will show that they are not actively dropping frames in OBS/Xsplit.
Step 3: Game will be paused, caster stops stream, uses speedtest.net to test upload speed.
If the caster is above 2.5 MB/S upload on test: benches leave and live stream starts again, game is unpaused and continues with empty benches.
If the caster is below 2.5 MB/S upload on test: caster and benches leave and game is casted from replay.

This is sufficient evidence for me to rule out that the caster is the cause of the lag. Removing the benches at that time should help as well.

It is now the burden of the caster to test their own stream settings to not get kicked out of games.
Run OBS. Start your stream. Open task manager, go to performance, go to your ethernet or wifi tab and see what your Send/Upload rate is. Run speedtest and ensure that your upload speed is higher by at least 0.5 MB/S. If not, lower your stream settings.

As for players: Do not watch the twitch while playing in your game.
stealthmeh
Posts: 250
Joined: Sat Apr 26, 2014 2:38 pm
Battle.net Name: stealth
Battle.net Char Code: 522
Battle.net Server: NA

Re: 4v4 casted replay for zhl

Post by stealthmeh » Fri Jan 09, 2015 1:33 pm

Teroh's test was done during low internet traffic hours and does not contain a significant xapos present. Keep in mind that streamers them selves are not usually the cause of the lag and just amplify the lag if there is a xapos player present. Should maybe redo this test between the hours of 7PM and 9PM EST (when most zhl games usually happen and as water mentioned is peak internet traffic) with at least 2 of the following players in the lobby.

Xapos Suspects: (keep in mind most of these suspects do not cause lag under all conditions and sometimes need a catalyst like streaming, off server, someone in their house downloadig/streaming, using their AC or other possible triggers)
Squid (both servers with frequent occurrence but unknown trigger)
Twista (unknown conditions but frequent occurrence)
ZergJR (unknown conditions but frequent occurrence)
Helios (on central with other unknown conditions)
OneGun (unknown conditions, but frequency is high)
Whirlwiind (seems to be internet issue cause he always has high spikes when it occurs, but seems random)
DoorMan (unknown conditions)
CountryKen (on central under other unknown conditions)
FourBolt (unknown conditions)
ArbiTeR (on east)
Guerrila (on west coast under unknown other conditions)
DrPorkchop (unknown conditions)
Bulbaasur (on west when using his AC maybe?)

With that being said it is nice to see action is being done regarding this issue and I believe your initial procedure to minimize lag should reduce lag in probably 80% of ZHL games (4v4 with only non xapos list casters, preferably not off server casters). As for the last 20% we would need to find ways to isolate the player that is lagging everyone and force them to leave the specific match. This is difficult to do objectively with out some tab where we can see everyone in lobbies ping/spikes. Or some other method I cannot think of.
Last edited by stealthmeh on Mon Jan 12, 2015 10:27 am, edited 1 time in total.
User avatar
Bulbasaaur
Posts: 277
Joined: Fri May 30, 2014 7:22 pm
Battle.net Name: Bulbasaaur
Battle.net Char Code: 615
Battle.net Server: NA

Re: 4v4 casted replay for zhl

Post by Bulbasaaur » Fri Jan 09, 2015 2:01 pm

lol… people still think I use ac………
User avatar
Kurt_58
Legend
Posts: 615
Joined: Sun Jan 26, 2014 2:39 pm
Battle.net Name: Kurt
Battle.net Char Code: 638
Battle.net Server: NA
Location: Hell. It's cold here.

Re: 4v4 casted replay for zhl

Post by Kurt_58 » Fri Jan 09, 2015 2:01 pm

Stealth you're forgetting yourself and vapour. The 2 most frequent players that lag out of games.
Locked