Feb 15 Cyrodiil Test Details

weathersong

As laggy as ever for me in PC/NA Gray Host last night. Around 6PM Pacific. Queue of 20-something to get in.

From the moment I arrived it was pretty obvious - mounting up and the animation hitches in the middle.

A bigger fight at Alessia kicked me. While I remained connected to Discord seamlessly.

A fight I wouldn't even call very big at Nikel, maybe just ~40 people total, seemed to be going fine until the last 5 seconds of damage all caught up all at once.

I called it quits after that and went back to Blackreach. Zero lag, as always. Same in BG's. Maybe I'm just spoiled by doing most of my PvP in these areas. My playstyle is oriented around instant reaction to skills, dodge roll, timing when an enemy is likely to charge or DB. Even a "little" lag in GH makes none of that feasible. It isn't even the same game.

Tigor

Changes leading to changes, leading to changes. Just play this test, enjoy the moment and have some cookies.
Day 1 was good, only some minor issues that are out of scope, but still a threat.

cyberjanet

ZOS_GinaBruno wrote: »

Update: We will be activating double AP during this upcoming test. To be clear, the only actions/activities that will result in double AP are capturing and defending Keeps, resources, and outposts, and killing/healing other player characters.

Thank you @ZOS_GinaBruno for this info. May I ask you to include it in your original post? Thank you.

Izanagi.Xiiib16_ESO

@ZOS_GinaBruno

When should we expect to hear if any changes will be made to the test conditions for next week?
Perhaps disabling cross healing again to see if performance improves?

LightYagami

Somehow I'm not focusing on server performance, but the No-Proc Cyrodiil makes the game way more fun to play and I enjoy the no-proc PvP.

Finedaible

"That reduction of cost is done via an ability, which has to check anytime you cast an ability if it’s cost should be reduced."

Why is cost reduction from sets coded as a proc ability and not an added character stat? Does this mean Breton's cost reduction is also a proc? If the server has to check Seducer set for each 'ability' cast - or Seducer must proc with each ability cast - then how is that efficient? Seems to me cost efficiency should be an added stat like Hunding's Rage as you described.

Ringod123

This test already proved a failure within 6 hours of it starting (at least on Gray Host EU), performance felt slightly (about 5-10%) better when no enemies were around, the recent sprint lag which has generally been around 2-4 seconds during busy times was down to around 0.5 seconds, BUT, as soon as more than just a couple of enemies were around, ALL of the issues came straight back, being hit with 3-4GCD's of abilities all at once, only able to fire off 1 ability every 3 seconds if im lucky, sprint having a 2-4 second lag, arrow abilities leaving me standing with my arms in the air for up to 8 seconds before deciding it just wasn't going to fire while the same enemy i'm targeting has no issues hitting me with 3 times as many abilities per second than the game allows him to cast. Then there was a defence at Brin where ALL AD sieges stopped working for about 10 seconds while DC's all worked fine, a defence at Alessia where as soon as 15-20 EP showed up, EVERY SINGLE AD in the area got the slo mo bug while EP were still running around normally.

Within 1 day the test has proven useless, but we have to endure it for 3 weeks?

Jaraal

Ringod123 wrote: »

Within 1 day the test has proven useless, but we have to endure it for 3 weeks?

Agreed, there is zero improvement in lag. Same old desynchs and disconnects.

It's good to know that group size, cross healing, and proc sets are not the source of the poor performance. But my guess is they are doing another "behavioral" study.... which apparently takes weeks to complete.

Ringod123

Jaraal wrote: »

Ringod123 wrote: »

Within 1 day the test has proven useless, but we have to endure it for 3 weeks?

Agreed, there is zero improvement in lag. Same old desynchs and disconnects.

It's good to know that group size, cross healing, and proc sets are not the source of the poor performance. But my guess is they are doing another "behavioral" study.... which apparently takes weeks to complete.

At this point the only logical culprit is years and years of adding quickly-made but botched code on top of the underlying system to get new things to work instead of actually rewriting them to support these new features and/or large scale battles. Quite evident we are now paying the price for ZOS constantly taking the easy/quick/cheap option for 6-8 years.

relentless_turnip

Ringod123 wrote: »

Jaraal wrote: »

Ringod123 wrote: »

Within 1 day the test has proven useless, but we have to endure it for 3 weeks?

Agreed, there is zero improvement in lag. Same old desynchs and disconnects.

It's good to know that group size, cross healing, and proc sets are not the source of the poor performance. But my guess is they are doing another "behavioral" study.... which apparently takes weeks to complete.

At this point the only logical culprit is years and years of adding quickly-made but botched code on top of the underlying system to get new things to work instead of actually rewriting them to support these new features and/or large scale battles. Quite evident we are now paying the price for ZOS constantly taking the easy/quick/cheap option for 6-8 years.

This and stadia 😂
Prior to the changes made for stadia it was stuff moved server side because of cheating. I think the quick coding has caused a lot of bugs, but the latency issues were caused by moving too many calculations server side.

Syrpynt

relentless_turnip wrote: »

Ringod123 wrote: »

At this point the only logical culprit is years and years of adding quickly-made but botched code on top of the underlying system to get new things to work instead of actually rewriting them to support these new features and/or large scale battles. Quite evident we are now paying the price for ZOS constantly taking the easy/quick/cheap option for 6-8 years.

This and stadia 😂
Prior to the changes made for stadia it was stuff moved server side because of cheating. I think the quick coding has caused a lot of bugs, but the latency issues were caused by moving too many calculations server side.

So which is more tedious for ZOS?
A.) Everyone gets client-side calcs again? & Banning cheaters that crop up once and a while?
B.) Limiting D-O-T's, limiting group sizes, disabling 85% of their set creations, and Performing server load tests via a low-ley forum post and drastically changing the PvP mechanics for weeks at a time multiple times a year.

lol

hummm...

RajinPVP

PC NA grayhost = we played about an hour on primetime and we did not noticed any improvement on server performance.

PC NA ravenwatch( no cp) = very very noticeable improvement. The server is packed ,all pop locked ,big fights all over the place.. it gotten to the point where theres too much going on on my screen that fps drops significantly low.. but i can use my abilities perfectly fine with little to no latency delays.

This just my personal experience last night playing with my guildmates and they experienced the same.

Dark_Aether

This test is the solution to 1 problem in Cyrodiil, but unfortunately it is not the lagg. I'm enjoying PVP more than I have in the past 2 years

Gilvoth

for me the lagg is now "Worse"
and my Ping "was" (before the test) normaly stayed around 210 constantly
my ping "now" during this test is about 375 constantly

DUTCH_REAPER

I bet if you leave no proc AND revert the healing to Only group performance will increase even more.

techyeshic

RajinPVP wrote: »

PC NA grayhost = we played about an hour on primetime and we did not noticed any improvement on server performance.

PC NA ravenwatch( no cp) = very very noticeable improvement. The server is packed ,all pop locked ,big fights all over the place.. it gotten to the point where theres too much going on on my screen that fps drops significantly low.. but i can use my abilities perfectly fine with little to no latency delays.

This just my personal experience last night playing with my guildmates and they experienced the same.

I am now hoping that no-procs is still a thing when the next patch hits so I can avoid the new CP at the same time as not go where procs are even more effective.

Nordic__Knights

PucMudDofuss wrote: »

Pink_Pixie wrote: »

Worst time to do tests, a few days before the campaign ends to get trolled by pve door, kinda kills any motivation to work to win a campaign for all that work to go down the drain. They should of waited for a new campaign to start before slapping us in the face with such tests.

I couldn't agree more. In fact a lot of grief and unhappiness could have been avoided entirely if they would have just made a couple of separate servers one CP enabled one No-CP enabled server where they could run these test. Then people who are in favor of these changes and/or want to test them could do so while those of us that don't care to participate could keep playing as per usual. We have just seen them do just that during MYM so that is surely a possibility.

These additional test servers could have their own campaign with reward tier with double AP and maybe other bonus goodies as an encouragement for people to participate.

I keep hearing people say "its just three weeks come on just do something else if you don't want to join". It wouldn't have to be a take it or leave it situation if they had thought to do that.

Personally Cyro PvP even laggy as it is, is what I enjoy the most about this game and PvP really don't interest me much. Imagine how people would react if they ran a similar test on PvE and how that community would respond...

you want something done are you just want to talk all day i myself want something fixed and zos knows players in game NEVER test *** if not made to

Galarthor

LarsS wrote: »

Cyro EU/PC gray host was great, havent been this good for years. We had a raid up of some 30 people and were in a number of huge fights on a locked server. Still skills went of as they should.

I was there. Lags started when you guys showed up at Brindle with those 30 players. We had a good time before that!

BigBearPaw

10 pm and PC-eu Grayhost is unplayable just like before the test.. Maybe it's time to assume servers ARE the MAIN problem?

Starshadw

I feel like we need to do this same test again after the new changes to CP are in place.

Nordic__Knights

Starshadw wrote: »

I feel like we need to do this same test again after the new changes to CP are in place.

zos way of things has always been to take load off, by putting more load back on, while trying to lighten the load . ive never gotten it

if you was testing how much of an load your truck can pull and what it does to motor you wouldnt take one load out just to put in another , - procs / + healing
way to do this would of been
week 1 take off all load cp / proc sets / healing
week 2 add healing
week 3 add sets
week 4 add cp
week 5 add healing / cp
week 6 add sets / cp
week 7 add cp / sets / healing

i get this would be an long test but only true way to see what load is overloading the system

Gilvoth

BigBearPaw wrote: »

10 pm and PC-eu Grayhost is unplayable just like before the test.. Maybe it's time to assume servers ARE the MAIN problem?

or it could be animation canceling that causing it.
its worth looking into.
and yes i am being serious about that, not joking, it very well could be the cause.

Galarthor

Gilvoth wrote: »

BigBearPaw wrote: »

10 pm and PC-eu Grayhost is unplayable just like before the test.. Maybe it's time to assume servers ARE the MAIN problem?

or it could be animation canceling that causing it.
its worth looking into.

More likely the tons of AoE spamming that occurs in and around ball groups / zergs.

Just think about it. Hitting 10 times as many players means 10 times more calculations. Imagine you had to do 10 times more work in the same time frame.

And AoE spamming is basically all these groups do. And since they stack up nicely to, they also get hit my more AoEs and at the same time every single AoE hitting them hits a lot of players at once.

Nordic__Knights

Gilvoth wrote: »

BigBearPaw wrote: »

10 pm and PC-eu Grayhost is unplayable just like before the test.. Maybe it's time to assume servers ARE the MAIN problem?

or it could be animation canceling that causing it.
its worth looking into.

AC is an big factor but no one well say it because then they have to fix the longest running bug in the game that became skill lol

Finedaible

I will say that even if the lag is not entirely mitigated, combat in Cyro feels better without procs so far. However, I am really annoyed that sets with an efficiency bonus/penalty are considered a proc, making builds using New Moon (which by all accounts should not be a 'proc' since it is entirely stats) useless for the duration of this test.

Gilvoth

Nordic__Knights wrote: »

Gilvoth wrote: »

BigBearPaw wrote: »

10 pm and PC-eu Grayhost is unplayable just like before the test.. Maybe it's time to assume servers ARE the MAIN problem?

or it could be animation canceling that causing it.
its worth looking into.

AC is an big factor but no one well say it because then they have to fix the longest running bug in the game that became skill lol

well said.
thank you for the honesty.
it has been my belief all along that it is the servers and also the animation canceling that causes lagg.
(or maybe the 2 combined)
but that does not mean im right, is just my theory, and i could very well be wrong.
it would have to be tested and examined to tell if true or not.

Gilvoth

Galarthor wrote: »

Gilvoth wrote: »

BigBearPaw wrote: »

10 pm and PC-eu Grayhost is unplayable just like before the test.. Maybe it's time to assume servers ARE the MAIN problem?

or it could be animation canceling that causing it.
its worth looking into.

More likely the tons of AoE spamming that occurs in and around ball groups / zergs.

Just think about it. Hitting 10 times as many players means 10 times more calculations. Imagine you had to do 10 times more work in the same time frame.

And AoE spamming is basically all these groups do. And since they stack up nicely to, they also get hit my more AoEs and at the same time every single AoE hitting them hits a lot of players at once.

that as well sounds fair and believable.
but again, that means recoding even the skills themselves.
basicly a new game?!
right?

Crom_CCCXVI

PEOPLE DON'T DIE

What game are you playing?
90% of the people are 35k health, 30k resist Werewolves or tight little bomb groups..

Maybe someone actually dying other than to a mob, might increase performance?

You have added so many ways to either run away or stay alive... CC Immunity, stun immunity, mist form, ring of the hunt, massive healing sets, endless recovery options--- and now you want to get rid of what might be left to "burst damage"?

any set that has any damage identity is gone--- Elegant, Sheildbreaker, and now what? Turog's Pact, Winterborn, Ice Furnace, ... what some Nightlblade killed the wrong person and now PROC sets are OP?


omg
ANOTHER STEP IN 100% THE WRONG DIRECTION FOR PVP. I have been here since 2015.... this right now, is the worst PvP yet, and you making it more vanilla, is going to lose what few players you have left All sets can't be 1 Resource, 1 damage, 1 crit or Pen, and uh.. how 350 in an element... 200 sets and every update they look more and more alike

JusticeSouldier

@ZOS_BrianWheeler it can be very interesting to test eso without animation canceling

Crom_CCCXVI

I played all weekend in Cyrodil..

I don't think I died once to a Proc Set. I died to 12 man groups in a little ball all spamming Sap Essence or Impulse... that can't be slowed down or broken up because they are all permanently immune to any CC's, Stuns or Immobilizes.

and I died a couple times to groups of werewolves, because despite running "different proc sets" even with a good combo you aren't breaking 25k damage, they do a one howl full health heal and then there are 5 of them....

again, what has it been 5 years since Viper/Tremorscale came out--- and now, you claim Proc sets are the issue....? nonsense