12 person group limit? Whaaaaaat?

Agalloch

This is not a bashing message..but how are other MMOs can handle groups of more than 24 members? All MMOs have server calculations...but probably most of them have a better coding and better servers.

DerAlleinTiger

Agalloch wrote: »

This is not a bashing message..but how are other MMOs can handle groups of more than 24 members? All MMOs have server calculations...but probably most of them have a better coding and better servers.

This is what I don't get. They've already told us that this year they're updating their servers and getting new ones. Why the hell are we making these kinds of minor performance "improvements" before that's done? Why are we jumping the gun on something like this when we should be waiting to see how the performance is once the new servers are fully installed and running? One would think THAT will be the biggest performance bump possible. Wait until then, and then take a look at the performance issues that still exist and measure up: Is the performance boost worth the cost of whatever changes need to be made? Are the performance issues, after the server update, serious enough for most players to even notice them?

Perhaps the biggest one to ask: Are the performance issues under circumstances where performance issues should be expected? I get concerned if I start seeing performance issues in small group content like 4-man dungeons and battlegrounds, or solo questing and mat farming. What I don't get concerned about is seeing a performance issue while riding a 50+ man dolmen train in the Alik'r. I EXPECT less performance there and am NOT willing to give up functionality to "fix" those issues.

JoeCapricorn

I've been in 24-person groups many times and outside of Cyrodiil I have not experienced noticeable skill delay.

The thing is, it happens in Cyrodiil even when I am not in a group. Lately mostly in Grey Host.

Jaraal

DerAlleinTiger wrote: »

Why are we jumping the gun on something like this when we should be waiting to see how the performance is once the new servers are fully installed and running? One would think THAT will be the biggest performance bump possible.

ZOS stated in the live stream that the new servers would not improve performance.

They did not elaborate.

TequilaFire

Of course new servers won't fix everything, but how can next gen silicon not improve things?
Certainly more than removing torch bugs from Cyrodiil. lol
Edit - It also depends if they are switching providers and new across the board or just replacing end of life components with equivalent new hardware.

rprice1819_ESO

I don't get why they would spend money on servers that are not going to make anything better for their customers. I guess we aren't worth spending money on. Also, limiting groups to 12 instead of 24 outside Cyrodiil isn't going to help anything except other game companies that will be picking up new customers. Logic seems to not be a factor in their decision making process.

sean.plackerb14_ESO

Jaraal wrote: »

DerAlleinTiger wrote: »

Why are we jumping the gun on something like this when we should be waiting to see how the performance is once the new servers are fully installed and running? One would think THAT will be the biggest performance bump possible.

ZOS stated in the live stream that the new servers would not improve performance.

They did not elaborate.

Dose not surprise me. I don't think the server hardware specs are the problem. I remember when they moved the EU server actually over to the EU they posted pics of the servers. They were high end, expensive HP Enterprise servers.

Take a FAR less popular MMO like Planetside 2 (MMOFPS in this case), I doubt they have made any significant investment in server hardware in a long time. Yet that game can handle HUGE group sizes /w voice chat, and huge battles with players on foot, in land and air vehicles. Yet the server has no lag issues / latency spikes. And Planetside 2 is a FPS so I can only imagine how many server side calculations are done during large battles. I'd say big battles are 150-200+ players, yet ESO's PVP often suffers massive latency spikes during FAR smaller battles.

It's def improved. I remember when it was common for the entire PVP server to crash and rollback 15-30 mins multiple times per night. I think the issue is the server side CODE not the hardware. I'd bet money all or almost everyone that wrote the server side code is no longer at ZOS. I think the server side code needs major optimization / work. New hardware def can't hurt, but I don't expect to see any meaningful difference from it (whenever it actually happens).

Other than server side code optimization, I know there was a large spike in connection problem posts after ZOS implemented DDOS protection. I understand why they did, ESO was getting hit with DDOS attacks on a regular basis for a while. So they implemented a DDOS protection solution by Akamai that requires all traffic to pass through one of their "scrubbing centers" to constantly monitor for DDOS attacks and instantly start trying to filter the DDOS traffic from the legit traffic if/when a DDOS attack occurs.

Now that ZOS is part of Microsoft, I think it would be more interesting to see them switch from using Akamai to moving the ESO servers into Microsoft Azure data centers and be hosted on their network instead. Microsoft operates one of the biggest fiber networks on the planet and has data centers all over (including of course where the US and EU ESO servers are now, Texas, and Germany).




Azure is what Xbox Live, OneDrive, Exchange, Office 365, Skype, Teams, etc all run on. As well as Microsoft's game streaming service.

Kelinmiriel

Limiting groups to 12 in Cyro didn't impact performance.

So perhaps the concern is that having Companions in larger groups would make a difference.

If this is the case - then why not just restrict Companions to groups of 12 or less? Reserve larger groups for players only.

That way, we could still have larger groups for rp, for house tours, for chat, for guilds, for social interactions - for all the things we actually need/want them for. Companions are for solo play, and should be able to help fill out groups for a 12-person trial, or a random dungeon - but there's no reason they're required in a group larger than that.

Especially if it would make the difference between being able to have groups > 12, or not.

It's probably too late for Blackwood, but maybe it could be changed still after? PLEASE?

JoeCapricorn

All they have to do is perform a DON'T on this proposed change and voila, it does not go live on June 1st!

It would show that they do listen to our feedback and care about us enough to know that this change would seriously harm social functions for guilds.

If not, it will be the case that instead we'll be forced to live with 12-man groups universally until six months down the road a different update either changes groups back to 24, or adds expanded groups that are simplified on the back-end (perhaps they don't give out health details and such, and cannot be used in Cyrodiil) that allow for 50 people.

The latter is what happened with the controversial Rapid Maneuver swap with Vigor in August (causing a lot of players who used it in PvE to lose it). Update 29 turned Major Gallop into a passive and made Rapid Maneuver outside of PVP situations obsolete. The passive is also obtainable at rank 3 of the Alliance War skill lines. Basically, ZOS did listen to our feedback, but needed time to implement an alternative as they work from home.

But this time, I really hope that @ZOS_GinaBruno (et al.) passes on the feedback to the dev-team and try to convince them NOT to shorten groups to 12. I don't want my social guilds to feel the negative effects of this decision for six months, potentially driving away players from ESO entirely and causing membership numbers to drop and guilds to die off or adapt.

I'd rather have 1 less frame per second than 12-player group limits.

Jazraena

I'd just like to note that we had a lovely crossguild RP event the other night on PC-EU - a talent show. Open World in the Rift. Within minutes, we filled up the first group and had to form a second; peak participation was somewhere around 35 to 40.

I expect the same to happen next week during the Night of Lights in Solitude. Even more groups is going to make any OOC communication a pain, and this is not that rare an event.

Jaraal

Jazraena wrote: »

I'd just like to note that we had a lovely crossguild RP event the other night on PC-EU - a talent show. Open World in the Rift. Within minutes, we filled up the first group and had to form a second; peak participation was somewhere around 35 to 40.

I expect the same to happen next week during the Night of Lights in Solitude. Even more groups is going to make any OOC communication a pain, and this is not that rare an event.

Unfortunate that things like these are not important to them. But they would rather sacrifice player fun and convenience in the name of "performance" changes that have proven to have minimal impact.

It makes you wonder about priorities, doesn't it?

LoneStar2911

I hope we’re all somehow wrong about this, and limiting groups to 12 will give performance an incredible boost. And we’ll all just be so thankful that we no longer complain about our groups being limited to 12.
I mean, we can dream.

BlueRaven

Is it possible to have addions that can in some way replicate larger groups? Tracking people, one common text box, etc? Is that a possible fix?

RedMuse

BlueRaven wrote: »

Is it possible to have addions that can in some way replicate larger groups? Tracking people, one common text box, etc? Is that a possible fix?

I doubt it or they'd already exist, if only for RP purposes. RP groups often need more than the current 24 people, so there'd have been "a marked" for them already and they'd have been developed.

If they refuse to at least keep the current group size then they absolutely need to make it possible for larger groups of people to talk to each other, such as through customized chat channels which isn't zone or guild.

My main guild will have our "farewell" WB run on Friday and it'll be a sad day. We're usually about 20 people, give or take. After Friday about 8 people will be left out of the socializing. Because the achievement hunting and lead farming is just an excuse for the socializing. Two groups aren't a viable way to do this, since we're text chat based discord isn't an option and even if we could do two groups we'd all be left out of half the social stuff.

It makes me worry about the future of not just my guilds but in general. Because without the socializing aspect then there's no reason to be in a guild unless you're a hard core pvp'er, trial runner or want to trade. This will do more damage to the social aspect of the game and thus to the game community than I think ZOS is aware of. As it doesn't look like ZOS has any immediate plans to alleviate the damage it will likely be a long time, as in a year or more, before anything is done, and by then I fear it'll be far too late to fix what was broken.

And considering that this change gave zero performance improvements when tested in Cyrodiil, I don't foresee it offering enough of an improvement on performance that it could be in any way said to offset the damage caused.

JoeCapricorn

ZOS is turning ESO from an MMORPG to just an ORPG

LoneStar2911

I guess we (PC) will find out in 7 days how much this actually sucks. I hope the devs have a gameplay stream soon and tell us why they made this change.

Varana

They already said it was "for performance reasons".
So scratching the bottom of the barrel for a few single FPS in circumstances where performance doesn't actually matter all that much, just to say "they've improved performance".

With the abysmal state of communications with ZOS, you won't get anything more.

Elsonso

Varana wrote: »

They already said it was "for performance reasons".
So scratching the bottom of the barrel for a few single FPS in circumstances where performance doesn't actually matter all that much, just to say "they've improved performance".

I am pretty sure that "performance" in this case is not "FPS". It is probably "latency" more than anything, and probably when the companions are being used. I am convinced that the group reduction is really just so that companions can be used when the player is grouped. Performance related to server-side calculations for companion AI.

Jazraena

It's not about FPS or any other performance on your screen, as I understand, but server-side calculations.

Which, while harder to gauge, is still miserable, and absolutely unacceptable without some sort of replacement communication option across more than 1 guild and 12 people.

agegarton

Of all the changes coming with this Chapter, the continuation of group sizes is the most frustrating.

I am convinced it most zero difference in Cyrodiil in terms of server performance and I’m equally convinced it will have no impact in PvE zones.

I have to be honest, this Chapter is leaving me cold. It just doesn’t seem as feature packed as the others - it feels like half a job.

JanTanhide

Game use to be able to handle groups of 100. As time went on player base increased, proc sets dominate and the server hardware got older and older.

Servers need to be upgraded/replaced or at least change the game engine to an engine that can handle everything.

Changing to a limit of 12 just means there will be more 12 person groups. Won't change a thing as far as server loading goes.

VaranisArano

JanTanhide wrote: »

Game use to be able to handle groups of 100. As time went on player base increased, proc sets dominate and the server hardware got older and older.

Servers need to be upgraded/replaced or at least change the game engine to an engine that can handle everything.

Changing to a limit of 12 just means there will be more 12 person groups. Won't change a thing as far as server loading goes.

I got into this in more detail above, but it does lower the amount of in-group communication. The in-group data going from 12 player to 12 players IS less than that of 24 players going to 24 players.

The caveat to that is that the Cyrodiil has been at 12 player groups since Markarth, which ZOS said was for behavioral reasons not for performance, and while the in-group communication may be smaller, this really doesn't do anything to the most visible source of performance issues: having lots of players in the same place.

Jaraal

VaranisArano wrote: »

JanTanhide wrote: »

Game use to be able to handle groups of 100. As time went on player base increased, proc sets dominate and the server hardware got older and older.

Servers need to be upgraded/replaced or at least change the game engine to an engine that can handle everything.

Changing to a limit of 12 just means there will be more 12 person groups. Won't change a thing as far as server loading goes.

I got into this in more detail above, but it does lower the amount of in-group communication. The in-group data going from 12 player to 12 players IS less than that of 24 players going to 24 players.

But the result will be instead of in-group communication among 24, now they will be using zone chat to get the information out, and increasing the communication to 100-200 instead of 24.

I'm not following the reasoning here.

VaranisArano

Jaraal wrote: »

VaranisArano wrote: »

JanTanhide wrote: »

Game use to be able to handle groups of 100. As time went on player base increased, proc sets dominate and the server hardware got older and older.

Servers need to be upgraded/replaced or at least change the game engine to an engine that can handle everything.

Changing to a limit of 12 just means there will be more 12 person groups. Won't change a thing as far as server loading goes.

I got into this in more detail above, but it does lower the amount of in-group communication. The in-group data going from 12 player to 12 players IS less than that of 24 players going to 24 players.

But the result will be instead of in-group communication among 24, now they will be using zone chat to get the information out, and increasing the communication to 100-200 instead of 24.

I'm not following the reasoning here.

ZOS isn't concerned about group chat, solely. When I say "communication" I'm talking about client - server - client communication, which Gina referenced, saying: "Additionally, there’s a fair amount of data that has to be exchanged on the backend for every person in your group." That's stuff like positioning, health bars, and - if you've looked at some of the grouping addons for PC, the API actually tracks and makes a whole lot more info available than the base game makes visible.

For the purposes of theoretical math, let's say that the client - server - client pipeline only has to handle two extra pieces of info per second when grouped up: health bar and chevron position on the map.

24 players sending that info to 24 players in their group: 1,152 server messages per second

12 players sending that same info to 12 players in their group: 288 server messages per second

On the face of it, ZOS is making a pretty serious reduction in their data backend.

The caveat with that is, as I said, that this backend change didn't make a big difference in Cyrodiil, and it really doesn't touch the performance problems that happen when lots and lots of players show up in the same placed without being grouped up.

Exquisition

Elsonso wrote: »

Varana wrote: »

They already said it was "for performance reasons".
So scratching the bottom of the barrel for a few single FPS in circumstances where performance doesn't actually matter all that much, just to say "they've improved performance".

I am pretty sure that "performance" in this case is not "FPS". It is probably "latency" more than anything, and probably when the companions are being used. I am convinced that the group reduction is really just so that companions can be used when the player is grouped. Performance related to server-side calculations for companion AI.

I rather have no companions and being able to group up with 24 ppl then having companions and not be able to group up anymore. The stupid groupsize reduction is absolutely horrible and will make guild events super hard to organise and this will result in having less fun.
of course, people having fun while playing ESO, doesn't pay the bill...

Earthewen

Rykoth wrote: »

Are we really going with a cap on groups outside of Cyrodiil? Bad idea. Bad bad BAD idea. I get it might help some performance but there's a lot of folks who make use of 24 player groups. Yes, roleplayers, but others as well.

Loving most of the notes but this would be a yuge mistake to keep this change. Yuge mistake.

It was a bad idea inside cyro as well. Nothing changed in the amount of lag, which was how the reduction was pitched. They said it changed our behavior, but they never said what behavior it was. People faction zerg more than ever now, making lag even worse ... shall I go on?

Elsonso

JanTanhide wrote: »

Game use to be able to handle groups of 100. As time went on player base increased, proc sets dominate and the server hardware got older and older.

Servers need to be upgraded/replaced or at least change the game engine to an engine that can handle everything.

Bolded. I doubt server hardware will fix this. They really need to figure out how to do the server software better.

Jaraal

VaranisArano wrote: »

On the face of it, ZOS is making a pretty serious reduction in their data backend.

The caveat with that is, as I said, that this backend change didn't make a big difference in Cyrodiil, and it really doesn't touch the performance problems that happen when lots and lots of players show up in the same placed without being grouped up.

That's what I don't understand. Since they stated that the reduction in group size in laggy Cyrodiil had no significant impact on performance, then how is it magically going to improve perfomance in a relatively smooth running PvE situation?

VaranisArano

Jaraal wrote: »

VaranisArano wrote: »

On the face of it, ZOS is making a pretty serious reduction in their data backend.

The caveat with that is, as I said, that this backend change didn't make a big difference in Cyrodiil, and it really doesn't touch the performance problems that happen when lots and lots of players show up in the same placed without being grouped up.

That's what I don't understand. Since they stated that the reduction in group size in laggy Cyrodiil had no significant impact on performance, then how is it magically going to improve perfomance in a relatively smooth running PvE situation?

To be honest, I'll be pleasantly surprised if we see any performance impact.

Otherwise, it feels a little like those old Pink Panther cartoons where he's chasing a fly and winds up doing way more damage to himself and his stuff than to the fly.

Elsonso

Jaraal wrote: »

VaranisArano wrote: »

On the face of it, ZOS is making a pretty serious reduction in their data backend.

The caveat with that is, as I said, that this backend change didn't make a big difference in Cyrodiil, and it really doesn't touch the performance problems that happen when lots and lots of players show up in the same placed without being grouped up.

That's what I don't understand. Since they stated that the reduction in group size in laggy Cyrodiil had no significant impact on performance, then how is it magically going to improve perfomance in a relatively smooth running PvE situation?

Simply because the two environments/scenarios are not the same. There is more to the group size than what is apparent from our perspective, and what happened with PVP simply isn't the same as with PVE.