Why not break this current test up into multiple parts to get more specific data?
Maintenance for the week of March 31:
• [COMPLETE] Xbox: NA and EU megaservers for patch maintenance – April 2, 6:00AM EDT (10:00 UTC) - 12:00PM EDT (16:00 UTC)
• [COMPLETE] Playstation®: NA and EU megaservers for patch maintenance – April 2, 6:00AM EDT (10:00 UTC) - 12:00PM EDT (16:00 UTC)
• [IN PROGRESS] ESO Store and Account System for maintenance – April 2, 6:00AM EDT (10:00 UTC) - 6:00PM EDT (22:00 UTC)
• [COMPLETE] Xbox: NA and EU megaservers for patch maintenance – April 2, 6:00AM EDT (10:00 UTC) - 12:00PM EDT (16:00 UTC)
• [COMPLETE] Playstation®: NA and EU megaservers for patch maintenance – April 2, 6:00AM EDT (10:00 UTC) - 12:00PM EDT (16:00 UTC)
• [IN PROGRESS] ESO Store and Account System for maintenance – April 2, 6:00AM EDT (10:00 UTC) - 6:00PM EDT (22:00 UTC)
Lags
✭✭✭✭✭
✭
✭
Maybe i dont understand how things work behind the scenes but wouldnt it have been better break this test up into different parts? Like a week of no cp, a week of no sets, and a then a week of the simplified skills, etc? Or even like 3 days each. It just seems like that would make it easier to pin down what is causing the biggest issues. Like if performance was greatly improved while we had regular skills, regular sets, but no cp, you could assume cp is causing a bigger issue. Still in the 1 test campaign of course.
In the Q&A one of the questions and answers was;
Q. What are the next steps after the Cyrodiil Champions testing? For example, if maxing player population is the performance goal, and it's met, will the next step be adding additional skills to the character template and seeing where it breaks? How would that be done? E.g., adding a few weapon skills or adding a few sets. – kiheikat
A. The whole goal of the test is to prove out if the complexity of abilities is a core problem performance-wise. We've run a number of tests looking at various other things over the last few years on live, (CP/No-CP, Proc sets, group healing, population caps, target caps, hardware...etc.) and this is the next step in that progression.
No cp has always performed better. And the main campaign, whether it was haderus, vivec, or gray host, has always had major performance issues. Unfortunately its only gotten worse over the years, and more consistent. But i dont think you guys have ever done a test like this, unless i missed it, with NO CP or no sets, or at least no proc sets. A test where you only give people one option to play in cyrodiil, so everyone will pack into one campaign. I mean i have been playing for like 8 years so ill be the first to admit maybe i am forgetting how the tests went. But if you havent had a recent test where the only campaign was no cp, or the only campaign had no sets or proc sets, with a population like this, then wouldn't it be beneficial to do that?
Idk maybe im off base. Ofc i do not have the information or knowledge zos does. But it just seems like a given that removing almost everything combat related from cyrodiil would improve performance in a major way. And it seems like it would have been very beneficial to test these specific things separately, simple skill, no cp, no set/proc, during the same test phase (like a week or two), in a single campaign, with a similar amount of players. But like i said maybe im wrong. Maybe you guys have done similar tests before only in one campaign and i missed it.
Edited by Lags on March 26, 2025 12:55AM In the Q&A one of the questions and answers was;
Q. What are the next steps after the Cyrodiil Champions testing? For example, if maxing player population is the performance goal, and it's met, will the next step be adding additional skills to the character template and seeing where it breaks? How would that be done? E.g., adding a few weapon skills or adding a few sets. – kiheikat
A. The whole goal of the test is to prove out if the complexity of abilities is a core problem performance-wise. We've run a number of tests looking at various other things over the last few years on live, (CP/No-CP, Proc sets, group healing, population caps, target caps, hardware...etc.) and this is the next step in that progression.
No cp has always performed better. And the main campaign, whether it was haderus, vivec, or gray host, has always had major performance issues. Unfortunately its only gotten worse over the years, and more consistent. But i dont think you guys have ever done a test like this, unless i missed it, with NO CP or no sets, or at least no proc sets. A test where you only give people one option to play in cyrodiil, so everyone will pack into one campaign. I mean i have been playing for like 8 years so ill be the first to admit maybe i am forgetting how the tests went. But if you havent had a recent test where the only campaign was no cp, or the only campaign had no sets or proc sets, with a population like this, then wouldn't it be beneficial to do that?
Idk maybe im off base. Ofc i do not have the information or knowledge zos does. But it just seems like a given that removing almost everything combat related from cyrodiil would improve performance in a major way. And it seems like it would have been very beneficial to test these specific things separately, simple skill, no cp, no set/proc, during the same test phase (like a week or two), in a single campaign, with a similar amount of players. But like i said maybe im wrong. Maybe you guys have done similar tests before only in one campaign and i missed it.
1
-
Amottica✭✭✭✭✭
✭✭✭✭✭First, the test already started so it is a little late to break this test up into different parts as the title suggests.
As for the future, it may be best to see the data, actual data vs anecdotal. Zenimax is collecting information from this test. It seems logical they plan further tests and may already have ideas in mind. It also stands to reason that they may develop new ideas from the test results.
1 -
Joy_Division✭✭✭✭✭
✭✭✭✭✭You missed it.
About 3-4 years ago, ZOS ran like 4-5 consecutive tests with different permutations. I remember one was limit group size to 4. Another was no proc sets. Another was skill cooldowns. There was something else. Then they combined everything the last week.4 -
Lags✭✭✭✭✭
✭First, the test already started so it is a little late to break this test up into different parts as the title suggests.
As for the future, it may be best to see the data, actual data vs anecdotal. Zenimax is collecting information from this test. It seems logical they plan further tests and may already have ideas in mind. It also stands to reason that they may develop new ideas from the test results.
whats the point of nitpicking a post witch a question that you cant answer? And it seems like you didnt read it either. Why didnt they break it up, why wouldnt they break it up, why did they choose to do it all at once instead of testing the main specific factors separately. Its all basically the same question. Obv its too late to do it right now considering we are 2 days into the test, but that is not the point of the question.
Also i am talking about actual data. Of course. Im simply asking would it have not been more beneficial to get more specific data by having a test broken up to test the main factors that are suspected to be impacting performance? To get more precise data on whether its skills, sets, or cp. In a more controlled test thats closer together than whatever tests may or may not have been done in years past. Instead of testing all those factors at once when clearly zos wants to mainly see if its a skill issue with this test. But again, its a question for zos because someone from zos is the only one who can really answer it.
0 -
Lags✭✭✭✭✭
✭Joy_Division wrote: »You missed it.
About 3-4 years ago, ZOS ran like 4-5 consecutive tests with different permutations. I remember one was limit group size to 4. Another was no proc sets. Another was skill cooldowns. There was something else. Then they combined everything the last week.
hmm that does sound familiar. I remember group size, no proc. Cross healing in another test.. And i dont remember a skill cooldown test but maybe i missed that part. I also dont remember a no cp test. Because one of the comments in the Q&A said no cp was something they tested. But maybe they did one idk.
I also dont remember if it was in one campaign like this one, with no other campaigns available. Because i think that makes a difference for the amount of players that will show up. But If they did do similar tests with no cp and no proc then this one makes sense. 3-4 years ago is a very long time but maybe it wont really matter since the performance issues were still there. I guess this makes me feel a little more hopeful for this test though. I just really hope they have some actual good data on the impact of any kind of proc set, cross healing, and cp. And that it is still relevant.
0 -
Muizer✭✭✭✭✭
✭I can't speak for ZOS evidently, but I'd say the Vengeance setup serves to ensure progress with minimal wasted effort.
Trying to eliminate one aspect from the game at a time does not guarantee progress, because a) there may be multiple aspects interacting and b) the problem might be something that is not an obvious candidate to test. This is basically making informed guesses and hoping you get lucky. ZOS have already tried that and they're not happy with the results.
So instead they strip everything down, ensure performance is good and then add more stuff back in until they hit on something which breaks performance. Now they could indeed have 'anticipated' the success of this first step and try a variety of follow ups immediately. However if the first step failed, that would be wasted effort.
And there's another thing. Even if follow up iterations are already on the shelf, ZOS would not want to risk not hitting the player cap during the tests by spreading the population over multiple campaigns. With one pop-locked campaign, they can control population and eliminate imbalances as variables, e.g. effectively comparing performance with different population sizes.
Edited by Muizer on March 26, 2025 11:47AMPlease stop making requests for game features. ZOS have enough bad ideas as it is!0 -
the1andonlyskwex✭✭✭✭✭
✭✭
hmm that does sound familiar. I remember group size, no proc. Cross healing in another test.. And i dont remember a skill cooldown test but maybe i missed that part. I also dont remember a no cp test. Because one of the comments in the Q&A said no cp was something they tested. But maybe they did one idk.
I also dont remember if it was in one campaign like this one, with no other campaigns available. Because i think that makes a difference for the amount of players that will show up. But If they did do similar tests with no cp and no proc then this one makes sense. 3-4 years ago is a very long time but maybe it wont really matter since the performance issues were still there. I guess this makes me feel a little more hopeful for this test though. I just really hope they have some actual good data on the impact of any kind of proc set, cross healing, and cp. And that it is still relevant.
They don't need a no CP test, they've had continuously running no CP campaigns for years.0 -
MincMincMinc✭✭✭✭✭The real question is how long do we need to wait for the followup tests.
Proves:
+ Skills were too complex and bloated
+ Server hardware is fine
+ Outfits and mount skins are fine
+ Keep Ticks are fine
+ Majority of addons are fine (as far as the server cares)
Layers that need Testing:
+ Mundus
+ Enchants
+ Poisons
+ Proc sets (think about the latency issues when a VD ROA ball group enters the keep)
+ Status effects
+ Buffs/debuffs
+ Cross healing
+ Smart healing
+ Player Dilution (bring in more new and pve players through incentives)
From what we can tell just off skills it isnt likely ONE layer causing issues it is likely many layers chain reacting to cause issues. Once you add back in all of the above layers you suddenly go from 1-2 ticks per second to potentially 50 or 100 depending on combat and builds. The more proc heavy and automatic builds are the worse it gets tick wise.Edited by MincMincMinc on March 26, 2025 1:06PM0 -
Muizer✭✭✭✭✭
✭MincMincMinc wrote: »The real question is how long do we need to wait for the followup tests.
yeah, at this rate it could be a long drawn out process.MincMincMinc wrote: »Proves: + Skills were too complex and bloated
That's a bit premature. The problem could be with anything that was left out of this test.And that's rather a lot!Please stop making requests for game features. ZOS have enough bad ideas as it is!0 -
Credible_Joe✭✭✭✭Standard procedure for debugging. Filter wide, then narrow it down.
People have already mentioned that there have been smaller scope performance tests in the past. I can see why; on a dev schedule as tight as 4 dungeons 1 chapter 1 zone, committing to a full on investigation is out of the question. So they tested a couple of their suspect items one at a time, hoping lightning would strike.
With the signaling from ZOS regarding dev cycle changes, it looks like there's a lot more resources available for more thorough testing.MincMincMinc wrote: »
Layers that need Testing:
+ Mundus
+ Enchants
+ Poisons
+ Proc sets (think about the latency issues when a VD ROA ball group enters the keep)
+ Status effects
+ Buffs/debuffs
+ Cross healing
+ Smart healing
+ Player Dilution (bring in more new and pve players through incentives)
Add mail to that list. It'd be hilarious if a huge portion of the performance issues was from the mail system listening to hundreds of players together all at once.Thank you for coming to my T E D talk1 -
MincMincMinc✭✭✭✭✭
yeah, at this rate it could be a long drawn out process.
That's a bit premature. The problem could be with anything that was left out of this test.And that's rather a lot!
Correct it doesn't prove that skills were THE problem. What it proved was that gameplay is 99% the same with the toned down skills. Sorc feels like sorc, dk feels like dk, etc. The overall playstyles were retained without the tacked on extra values. Meaning they can trim the fat so to speak. It is power creep in a nutshell. If everyclass has a buff that gives 10k WD...... why not just remove it and save the tick on the server. In essence go to a clean slate for skills and balance from there.Credible_Joe wrote: »Add mail to that list. It'd be hilarious if a huge portion of the performance issues was from the mail system listening to hundreds of players together all at once.
For a while people also thought you had to leave trading guilds in pvp because the UI updating player zone locations was causing latency issues. Which doesn't seem to be the case. Or addons that talked to guild stores lagging the pvp server because the quartermaster at keeps could have a guild store tied to it.0 -
Amottica✭✭✭✭✭
✭✭✭✭✭
whats the point of nitpicking a post witch a question that you cant answer?
Creating a thread in this forum is specifically asking people, inviting them, to comment on it as they see fit.
Further, I did not nitpick anything witch or with a question I cannot answer. I did not ask any question so I could not have nitpicked anything with a question I could not answer.
BTW, I did read the entire post, and my comment was all I found necessary to test. Time will tell what Zenimax has planned or what plans they may develop form this. For now, they are testing performance with a design that eliminates a lot of variables. What they do going forward is only a guess but for this test there is nothing to split up into multiple parts.
yeah, at this rate it could be a long drawn out process.
True, but how long have we been waiting already? Yes, it would be nice if we had a roadmap on this but we have seen how well the previous roadmaps have helped.
1 -
Soraka✭✭✭✭✭I think they said they're waiting to see the results of this before developing a road map/plan moving forward1 -
Lags✭✭✭✭✭
✭I can't speak for ZOS evidently, but I'd say the Vengeance setup serves to ensure progress with minimal wasted effort.
Trying to eliminate one aspect from the game at a time does not guarantee progress, because a) there may be multiple aspects interacting and b) the problem might be something that is not an obvious candidate to test. This is basically making informed guesses and hoping you get lucky. ZOS have already tried that and they're not happy with the results.
So instead they strip everything down, ensure performance is good and then add more stuff back in until they hit on something which breaks performance. Now they could indeed have 'anticipated' the success of this first step and try a variety of follow ups immediately. However if the first step failed, that would be wasted effort.
And there's another thing. Even if follow up iterations are already on the shelf, ZOS would not want to risk not hitting the player cap during the tests by spreading the population over multiple campaigns. With one pop-locked campaign, they can control population and eliminate imbalances as variables, e.g. effectively comparing performance with different population sizes.
thats one of the main points i was talking about. Which is why i said, for example, 1 week of no cp, 1 week of simple skills, 1 week of no sets/no proc sets. Which is also why i said having it only in one campaign makes a big difference to how many people will show up. Its important.the1andonlyskwex wrote: »
They don't need a no CP test, they've had continuously running no CP campaigns for years.
That has nothing to do with an actual stress test. Especially when its compared to something like this. Like idk how else to try to explain this to you. Sure no cp always performs better but i dont think its ever been the only option with a ton of people like the vengeance campaign. But if its has then that would be good.
0 -
Lags✭✭✭✭✭
✭MincMincMinc wrote: »The real question is how long do we need to wait for the followup tests.
Proves:
+ Skills were too complex and bloated
+ Server hardware is fine
+ Outfits and mount skins are fine
+ Keep Ticks are fine
+ Majority of addons are fine (as far as the server cares)
Layers that need Testing:
+ Mundus
+ Enchants
+ Poisons
+ Proc sets (think about the latency issues when a VD ROA ball group enters the keep)
+ Status effects
+ Buffs/debuffs
+ Cross healing
+ Smart healing
+ Player Dilution (bring in more new and pve players through incentives)
From what we can tell just off skills it isnt likely ONE layer causing issues it is likely many layers chain reacting to cause issues. Once you add back in all of the above layers you suddenly go from 1-2 ticks per second to potentially 50 or 100 depending on combat and builds. The more proc heavy and automatic builds are the worse it gets tick wise.
Idk about skills being too complex. Maybe some of them. But even with almost everything removed from cyrodiil, and these super simple skills, performance is still awful. And i suspect it will continue to get worse as the week goes on. I think in general its better but at larger fights it is just as bad as always. And this isnt just me saying this, many people who are actually paying attention to skill delay have noticed it. Of course there are more players, and like i said its not as bad or often, but to me that just says the servers can barely handle large scale combat even with things this simplified.0 -
Izanagi.Xiiib16_ESO✭✭✭✭✭
✭✭✭MincMincMinc wrote: »Proves:
+ Server hardware is fine
.
This is not the case. There are a lot of unknown variables from our side. For example, was the server hardware adjusted for this test - we've seen similar things in the past with more memory dedicated during MYM and also during the free trial period on the 'upgraded' servers which coincidentally lead to the performance suddenly getting worse 6 months after being upgraded...
Secondly servers being able to process low amounts of data without issue but not handle the combat requirements of ESO doesn't mean that the solution or hardware they have picked is correct and 'fine' it just means that they are more suitable for this slow paced boring combat.
@Solar_Breeze
NA ~ Izanerys: Dracarys (Videos | Dracast)
EU ~ Izanagi: Banana Squad (AOE Rats/ Zerg Squad / Roleplay Circle)1 -
xylena_lazarow✭✭✭✭✭
✭✭✭✭✭
Way off base. They did a 3mo long no-cp no-proc test on the main Cyro server a few years ago. Zero impact on performance, it lagged the same as it always did, and Ravenwatch proved to be unpopular anyway.No cp has always performed better.
Idk maybe im off base.PC/NA || Cyro/BGs || RIP ground oils2 -
MincMincMinc✭✭✭✭✭
Idk about skills being too complex. Maybe some of them. But even with almost everything removed from cyrodiil, and these super simple skills, performance is still awful. And i suspect it will continue to get worse as the week goes on. I think in general its better but at larger fights it is just as bad as always. And this isnt just me saying this, many people who are actually paying attention to skill delay have noticed it. Of course there are more players, and like i said its not as bad or often, but to me that just says the servers can barely handle large scale combat even with things this simplified.
This is simply not the case. Ive been on for the entirety of US primetime 5pm to 1am each night. Being at multiple hundreds of player fights with next to no server side issues. Likely you are misinterpreting fps issues on your own client as others have pointed out. There have been only a handful of pause/stutter crash style events that cleared up in seconds.
Skills being easy to understand vs complex is completely different. Jabs for instance is an extremely complex skill and throughout its life has had many iterations to somewhat work. Leap is another one where it has many behind the scene spaghetti'd code patched together. Think it is a single gap closer, but also only does damage through the aoe, which tries to calculate angular direction from the landing point from the single target while also reverting and holding the target in place for it to connect.Izanagi.Xiiib16_ESO wrote: »
This is not the case. There are a lot of unknown variables from our side. For example, was the server hardware adjusted for this test - we've seen similar things in the past with more memory dedicated during MYM and also during the free trial period on the 'upgraded' servers which coincidentally lead to the performance suddenly getting worse 6 months after being upgraded...
Secondly servers being able to process low amounts of data without issue but not handle the combat requirements of ESO doesn't mean that the solution or hardware they have picked is correct and 'fine' it just means that they are more suitable for this slow paced boring combat.
The end result is the same for data collection. They will collect what is needed for what number of calculations. Considering how well performance has been proves that it is certainly possible to have this number of players. The question is how can they bring back the other elements without making each player worth 5x, 10x, or 100x the calculations like on live.Edited by MincMincMinc on March 27, 2025 1:33PM0 -
MincMincMinc✭✭✭✭✭xylena_lazarow wrote: »Way off base. They did a 3mo long no-cp no-proc test on the main Cyro server a few years ago. Zero impact on performance, it lagged the same as it always did, and Ravenwatch proved to be unpopular anyway.
The problem is that most players/humans will end up following the bandwagon to whatever the most populated server is regardless. People already didnt play nocp because the wagon was the cp30day as it has been for years. Why would established guilds in the cp30day ever swap to a no population server that other rival guilds aren't on?
How much proc sets play into performance is probably on the same level as status effects or buff timers or enchants etc. It probably matters more about how skills facilitate activating all of these effects. If all 4 of these effects trigger off any event and an aoe dot skill hits 10 people every 10 seconds that quickly spirals out of control. Compared to in vengeance where the aoe cap limits it to 3 people from the start. Don't get mistaken thinking its just aoe caps, likely this same concept applies to multiple types of skills and event callouts within the code.0 -
xylena_lazarow✭✭✭✭✭
✭✭✭✭✭
Yes it was the main server, either Gray Host or that one with the forgettable dragon name. There was no shortage of players of varying skill levels complaining that it made the build system boring, like why even have a build system at all at that point... so it's not surprising to me that Vengeance seems much better received than Ravenwatch was.MincMincMinc wrote: »The problem is that most players/humans will end up following the bandwagon to whatever the most populated server is regardless.PC/NA || Cyro/BGs || RIP ground oils1 -
MincMincMinc✭✭✭✭✭xylena_lazarow wrote: »Yes it was the main server, either Gray Host or that one with the forgettable dragon name. There was no shortage of players of varying skill levels complaining that it made the build system boring, like why even have a build system at all at that point... so it's not surprising to me that Vengeance seems much better received than Ravenwatch was.
Ravenwatch had the issue of nobody wants to be the first to step off the ladder. If zos pushed all of pvp to be re-standardized at once than it would be accepted more widely. Like any patch there would be people complaining that they didnt want to change (while probably hypocritically complaining and asking for change)
Sorc for example plays near identical to what youd expect a sorc to play like.. Skill wise I dont see a reason to not implement sorc's performative changes. The use of cooldowns is worrisome, otherwise streak cut half its code and functions perfectly fine. The stun for instance was coded such that 3 aoe+stuns is now only 1.
The scamp skill getting rid of the pet is now its own unique skill that serves the same purpose and again is probably 10x more performative.
Also there is the huge benefit of them having to start from scratch and that is avoiding the desync issues caused by poorly copy and pasted rush Frankenstein skills over the past decade. Who knows how that tolls the server.0 -
Izanagi.Xiiib16_ESO✭✭✭✭✭
✭✭✭xylena_lazarow wrote: »Yes it was the main server, either Gray Host or that one with the forgettable dragon name. There was no shortage of players of varying skill levels complaining that it made the build system boring, like why even have a build system at all at that point... so it's not surprising to me that Vengeance seems much better received than Ravenwatch was.
I think the main reasons the tests were so poor at the time were because they invalidated current setups without providing an alternative for free. Forcing players who wanted to participate to farm completely new sets.
At least this time a template is provided so players dont need to get anything in advance / spend resources on sets which will be invalidated the next week.
Thats kind of the only benefit of the current system - anyone can participate in cyrodiil as soon as they hit lvl 10 because they are given a template to use.Edited by Izanagi.Xiiib16_ESO on March 27, 2025 2:45PM@Solar_Breeze
NA ~ Izanerys: Dracarys (Videos | Dracast)
EU ~ Izanagi: Banana Squad (AOE Rats/ Zerg Squad / Roleplay Circle)0 -
Muizer✭✭✭✭✭
✭Yeah it's an interesting exercise to try to figure out what aspects of the game cause the most calculation per second to happen. A lot will be triggered by effects with timers because every time an effect is added or expires, the player's stats needs to be recalculated. Vengeance skills seem designed so there will be far fewer timed effects to keep track of.
Telemetry is another thing. Tracking who is within range of abilities for every player. Area increases with the square of the radius, so that would be a non-linear effect right there. Come to think of it, it makes AOE caps a double edged sword. You'd actually be expending more resources on calculating something like 'nearest x', and the pay-out would have to be fewer timers running?
Anyway, just some random thoughts about this test and what could be next.
I think safest bet would be to reintroduce attribute customization and sets that are purely stat based. They could also introduce more skill lines. However with the current simplicity of skills, I'm not sure they could be diversified in a meaningful way. My NB's 3 vengeance class skill lines provide little variety as it is. Really do not need more of the same under a different name.Please stop making requests for game features. ZOS have enough bad ideas as it is!0 -
YandereGirlfriend✭✭✭✭✭
✭✭✭✭✭MincMincMinc wrote: »
The problem is that most players/humans will end up following the bandwagon to whatever the most populated server is regardless. People already didnt play nocp because the wagon was the cp30day as it has been for years. Why would established guilds in the cp30day ever swap to a no population server that other rival guilds aren't on?
How much proc sets play into performance is probably on the same level as status effects or buff timers or enchants etc. It probably matters more about how skills facilitate activating all of these effects. If all 4 of these effects trigger off any event and an aoe dot skill hits 10 people every 10 seconds that quickly spirals out of control. Compared to in vengeance where the aoe cap limits it to 3 people from the start. Don't get mistaken thinking its just aoe caps, likely this same concept applies to multiple types of skills and event callouts within the code.
The market spoke very clearly when it came to the No-Proc tests and it said that, despite the hypebeasting in the early days of the ruleset, nobody actually wanted to play there long-term. And so it became an abandonware campaign (that also completely destroyed the existing Ravenwatch population because their campaign was wrecked and never fixed...).
Humans are intrinsically drawn to novelty. Vengeance is the new hotness. Make people play like this for 6-months or a year and interest will once again fall off of a cliff.
But to address the actual OP: yes, let's also have a test module where add-ons are restricted.0 -
Lags✭✭✭✭✭
✭MincMincMinc wrote: »
This is simply not the case. Ive been on for the entirety of US primetime 5pm to 1am each night. Being at multiple hundreds of player fights with next to no server side issues. Likely you are misinterpreting fps issues on your own client as others have pointed out. There have been only a handful of pause/stutter crash style events that cleared up in seconds.
Skills being easy to understand vs complex is completely different. Jabs for instance is an extremely complex skill and throughout its life has had many iterations to somewhat work. Leap is another one where it has many behind the scene spaghetti'd code patched together. Think it is a single gap closer, but also only does damage through the aoe, which tries to calculate angular direction from the landing point from the single target while also reverting and holding the target in place for it to connect.
The end result is the same for data collection. They will collect what is needed for what number of calculations. Considering how well performance has been proves that it is certainly possible to have this number of players. The question is how can they bring back the other elements without making each player worth 5x, 10x, or 100x the calculations like on live.
Telling people that the problem is on their end is so old. I remember when zos used to tell us it was on our end, back in like 2017 2018. They would tell you to "check with your isp". Ive probably spent over 20k hours in cyrodiil. I have done everything over the years to try and improve performance. Upgrading my xbox, my internet multiple times, port forwarding, upgrading my pc after i started playing on pc, and of course many random things throughout the years that people would say may help improve performance. All that being said, i know the difference between the performance issues that eso struggles with. I know the difference between fps issues and skill delay. And i always ask multiple other people if they are experiencing the same thing, although i stopped doing that after a while since its well known everyone experiences the same thing now.
But this is a new test so of course i am asking multiple people, even on different factions, if they are experiencing they same things. And they are. So please, stop with the " its not happening to me so its not happening ". Many people run around in massive zergs button mashing and dont notice skill delay. Maybe other people are just lucky. Maybe its based on where people live, who knows. All im telling you is sometimes there is still really bad skill delay. And of course fps issues but thats to be expected with so many people. And like i said before its not always but its happened multiple times after the first day of the test. Ive even seen other people talking about it on the forums. Its not a secret. Obv performance will be better with such a bare bones form of cyrodiil, but on the other hand obv there is some bigger issue here if delay like that persists even with this extremely simple server.
also, when it comes to skill complexity like i said i can agree about some skills but deff not all. I understand the point you are making but if the skills in eso are that much of an issue then zos really messed something up along the way. Because these skills are not like reinventing the wheel here. And i think the skills in the vengeance camp are way too over simplified. So hopefully they can find some kind of middle ground. People have been telling them for years to balance pve and pvp separately. I just hope they dont take it too far. Some games, like gw2 and bdo, change some skills slightly based on if you're in pvp or pve.
But again, those games seem to handle large scale pvp just fine with mainly fps issues more than anything at large fights. Not the crazy input lag/delay that eso suffers from. Among other things. So it makes me wonder how bad zos has screwed up if these skills, that are not complex at all (on the surface) are causing issues like this.1 -
Lags✭✭✭✭✭
✭xylena_lazarow wrote: »Yes it was the main server, either Gray Host or that one with the forgettable dragon name. There was no shortage of players of varying skill levels complaining that it made the build system boring, like why even have a build system at all at that point... so it's not surprising to me that Vengeance seems much better received than Ravenwatch was.
No cp(azuras star/ravenwatch) was very popular in past years. Zos ruined it, then everyone got used to playing cp. Plus the fact that the playerbase is slowly declining, and more so with the pvp playerbase, doesnt help the issue either. It used to be more popular because pvp used to be more popular, and people like variety. Many people played np cp, its just more played CP. But since the population has dwindled everyone just goes to what has more people. Because who wants to play somewhere with almost no one else to play with or fight.
Vengeance is being received well because of multiple reasons. For one its completely different. Its also a test to try and hopefully improve things. Its also simple which some people like in the short term. And it kind of kills ball groups which is the main enemy of the average zerg player. Its just zerg vs zerg and whoever has the bigger zerg wins. Its all well and good for a test, but if this was the only option cyrodiil would be dead in a month or two.
Theres also a very big difference between this test and what happened to ravenwatch. Many people probably dont even remember how azuras star was back in the day. Long before it was ravenwatch. And the under level 50 campaign, but mainly azuras star. It was very popular. Always full bars every day. Up until like maybe 2019-2020? Idk its hard to remember, and i was playing xbox and pc at the time.
No proc was not well received and killed no cp because of a few reasons. First the actual set list was hard to find and not in game. It wasnt like the game would tell you that X set wouldnt work in no cp, you had to track down a list. Second, the list made no sense and was incomplete. Something like pariah was allowed, which grants you a ton of resistance based on your health, but a set like stuhns, which grants you a ton of pen when you off balance someone, was not allowed.1 -
Lags✭✭✭✭✭
✭Izanagi.Xiiib16_ESO wrote: »
This is not the case. There are a lot of unknown variables from our side. For example, was the server hardware adjusted for this test - we've seen similar things in the past with more memory dedicated during MYM and also during the free trial period on the 'upgraded' servers which coincidentally lead to the performance suddenly getting worse 6 months after being upgraded...
Secondly servers being able to process low amounts of data without issue but not handle the combat requirements of ESO doesn't mean that the solution or hardware they have picked is correct and 'fine' it just means that they are more suitable for this slow paced boring combat.
wow i cant believe i missed that part of that post. Server hardware is fine? Ya there is no point in arguing this anymore. If anyone thinks the server hardware is fine then we are so far in opposite directions that there is no point in arguing it.
But ya I agree with you, and i would also repeat what i said in another post, if server hardware is fine then why would skills that are so simple, on the surface, have a sizeable impact on performance? When games with more complicated skills dont see it. This whole performance mess has been such a nightmare for pvpers for years.
And after all the things they've tried, and all the years, its just hard to believe they will ever fix it. At least without butchering what cyrodiil is. Thats what makes me think its a server issue. And they just keep trying to do these band aid fixes but are unwilling to deal with the actual issue. Ive spent a lot of time in many mmos and ive never seen the performance issues i see in eso. And i just find it so hard to believe that a company like zenimax online studios, now with microsoft behind them, cannot pin point exactly why we have these issues. Consistently. For years.Edited by Lags on March 28, 2025 11:44PM0