Yasha wrote: »

frozywozy wrote: »

As some people mentionned in this thread, I though that the first objective of Cyrodiil was to introduce large scale PvP. I understand that this could be only a temporary solution to try to split people appart while you guys figure out a better way to solve performance issues, but I don't think this is going in the right direction anyway.

I think that engineers spent more than enough time to try to figure out the way to fix the performance issues in Cyrodiil. It is now time to accept that it isn't going to get better and try to focus on different aspects of PvP that could bring alot of customers / subscribers back.

Yes, I am indeed reffering to Arenas and Battleground. Those small scales instances of PvP would be perfect to suite the desire of battle of many folks and would be lag free.

Before you lose even more players, I think it is time to start considering small scale instances of PvP and spend more time developping those than you spend to fix performance issues in Cyrodiil. I think most will agree that we waited long enough.

I think you have hit the nail on the head here Frozy. I basically left ESO a few months after launch because of continued terrible lag in pvp. If it hasn't been fixed in all this time, and by some accounts is actually getting worse, surely it is time to consider other forms of pvp- ones that we can actually reliably participate in.

Of course the goal should be to try and fix the lag in massive pvp, but I would like to play something functional while that is being looked at.

This thread here, started by @heroofnoneb14_ESO lays out some amazing concepts for some PvP alternatives, much of it in line with stuff that's already been developed. It's been linked in this thread before, but some of us are trying to keep it prevalent so as to gain more momentum and attention.

Squeaky wheel and all.

WebBull wrote: »

Get rid of spell effects. They are way over the top and while nice, they are not needed.

Well we do need some sort of visual aid. Not everyone swings a sword.

glak wrote: »

Sacadon wrote: »

stefan.gustavsonb16_ESO wrote: »

Sacadon wrote: »

If not already, please consider using designs other non-gaming organizations have used successfully for a long time now. I'm thinking that budget is your constraint though versus talent or technology.

Think of how many calcs/sec apps like Soundhound and Shazaam have to do on HPC systems in order to recognize a song in 2-5 secs. There are many companies that already use massively parallel high-performance cloud computing systems across the globe to do these things.

Another extremely parallel system design that has been in place for a long time is high-frequency electronic trading. Yes there's arbitrage to take advantage of delays but a gaming system could just write off such delays. I bet TradeStation would give you some ideas without giving you their IP. A friend of mine is a developer on their core system and I could make a connection for ZOS, but expect that you'd not need one too.

One thing is for sure, achieving a highly distributed client-to-server design that performs well is within reach and has been for the past 5 years. So this is ZOS's opportunity otherwise someone else is going to eventually.

This is perfectly sound advice, but consider the fact that the game client has serious load balancing issues, and it sounds reasonable to suspect that massively parallel high performance solutions may not be on their roadmap. Their current server load balancing is based on spatial partitioning (zones) and grouping (phases). Redesigning it to use parallel threads to share the workload for a particularly busy area is very likely beyond their reach, if not in skill so at least in time and money.

Agreed, sadly. Edit: I was surprised they only wrote the client in 32-bit but thought they'd port it eventually. Sure wish they had more $.

At least the console clients are 64-bit.

Why would they made consoles 64 bit and not PC/Mac. Is anyone still using a 32 bit processor that could actually run this game anyway?

Rune_Relic wrote: »

glak wrote: »

Bashev wrote: »

Fizzlewizzle wrote: »

helltiger wrote: »

Remove AoE from game(pvp) and we get that battle, what we see on this video https://youtu.be/MQFxE4AESn4?t=50s. AoE can be present in pvp only a: siege machines, healing, disabling(without any damage: mass stun with AoE cap, negate, etc). Also meele>>range, single>aoe.

This isn't the worst idea, it would give a whole different feel when going to PvP.
Lets all be honest, the battles against a group of people that are "disorganized" and try to take out 1 person at the time are much more fun (and lasting) then the ones where some zerk group rolls over the field and just wipes everything.

If you're in a 1 X Many fight it wont matter if you can or can't aoe, since most will just bombard the single player with spells and skills until he dies. Group gameplay would become a lot more fun and interesting.

I proposed this idea back in June --> http://forums.elderscrollsonline.com/en/discussion/114541/can-we-have-a-campaign-where-all-aoe-dps-skills-are-disabled
Most players didn't like it.

Worth a try on the PTS. All they have to do is make a debuff in the open areas of Cyrodiil that sets the AoE cap to 0 and disables caltrops. Lore perspective: above ground outside of the city, Molag Bal removed AoE damage to make us suffer.

I would 2nd aoe disabled on the PTS and encourage a 100v100+ fight to see how much of a difference it makes to performance. I am sure a few guilds can stump up the bodies to hit the same place at the same time with 3 factions if ZOS gave them time and place.

We can't honestly play PvP without AoE's though.

The_Death_Princess wrote: »

Hmmmmm,

Why is it that no one is talking about the fact that large scale battles got WORSE with 1.6.
(We believe they built CP on an old client and didnt patch months of fixes)

Why is it that when Blue team on Yellow team fight the population can be maxed out, certainly much greater, in the area and the lag is tolerable?
(We believe red team is exploiting this)

Why is it that if large scale battle happen on other servers than Thornblade the lag is virtually nonexistant.
(We believe that it is two major guilds that use a Denial of Service attack during these battles)
(Which also correlates to lag is not bad when these guilds are not running).

Mr. Sage, are you really not seeing these things?

Sounds like a conspiracy theory to me.

helltiger wrote: »

blackweb wrote: »

2. All of my suggestions are based on experience and have been APPLIED SUCCESSFULLY IN OTHER GAMES.

keyword - "other games" this is ESO and friendly fire contradicts to RvR. Even if your raid will be in the same alliance with me, I can just wipe you from harm, running past.

blackweb wrote: »

3. NOTHING WILL STOP AE SPAM AND ASSOCIATED LAG FASTER THAN FRIENDLY FIRE.

Disabling aoe damage in pvp and switching on Collisions for preventing stacking players in one point.

"Bashev wrote: »

I proposed this idea back in June --> http://forums.elderscrollsonline.com/en/discussion/114541/can-we-have-a-campaign-where-all-aoe-dps-skills-are-disabled
Most players didn't like it.

We must also propose solution for PvE without AoE.

Collision detection will only increase lag. It will only add to the calculations that the server has to make.

stefan.gustavsonb16_ESO wrote: »

eventide03b14a_ESO wrote: »

Why would they made consoles 64 bit and not PC/Mac. Is anyone still using a 32 bit processor that could actually run this game anyway?

I would say hat almost nobody is running a 32-bit CPU, certainly not if they expect to play games on their computer, but quite a few people are still running a 32-bit OS, namely the 32-bit version of Windows 7. Somewhat depending on who you ask, about 10% of Windows installations worldwide are still 32-bit. It would be quite reasonable to require a 64-bit OS to play a modern game, but apparently ZOS did not want to exclude the 32-bit Windows 7 crowd.

There is still a lot of 32 bit architecture out there. Lots of people playing this game aren't computer savvy. The forum participants aren't an accurate representation of this game's demographic. While those of us here have a clearer picture of 64 bit advantages, most people don't understand what it entails.

Rewind back to this game's announcement. Part of it revolved around targeting more than just high end gaming machines, that includes your school/work issued laptops that tend to get purchased in bulk. Then lets not forget most user's refusal to ween themselves off Windows XP. There's still a significant number of 32 bit processors and/or OS installations out there. And people tend to hang onto a PC/Laptop for 3 to 6 years.

It's rare that 64 bit is necessarry, it's just preferable. ZOS made the business decision to do as broad a stroke as possible when targeting customers. (nothing wrong with that.) What is refreshing is this comment here from earlier in this thread. What you can glean from it is that there is a 64 bit light at the end of the tunnel, it's just a long tunnel.

eventide03b14a_ESO wrote: »

Rune_Relic wrote: »

on the PTS and encourage a 100v100+ fight to see how much of a difference it makes to performance. I am sure a few guilds can stump up the bodies to hit the same place at the same time with 3 factions if ZOS gave them time and place.

We can't honestly play PvP without AoE's though.

Never said we could. The idea is to confirm once and for all the the size of the damage queue is causing the problem.
If it is then we need to manage the damage by dropping them if the load gets too high.

Damage queue length = 0
Everyone fires skills...no damage is done.
No damage calcs are done.

Damage queue length = 5
5 attacks are stacked and no more can be calculated until one space becomes available.
Every fires skills as usual with no delay, 5+ incoming damage calcs is dropped.

So with 20 people there can never be any more than 100 damage calcs to be performed at any time on the server. Compared to now where there is no management and no limits on the size of the damage that has to be calculated.

ZOS_PaulSage wrote: »

Just to clarify. There seemed to be a misunderstanding that I was saying we are "giving up" or "blaming this on players." Absolutely not. We are looking at the issue from many angles. I certainly didn't mean to imply we couldn't or wouldn't change code to improve performance, but rather that the answer wasn't more hardware. (It is often suggested this is the best way to fix problems.) I also wasn't implying you shouldn't play naturally, but unfortunately it is true that more players right now in a smaller area causes the performance problems. I say this not to discourage behavior or try push off the issue, but rather to give completely frank information about the root cause.

Howdy @ZOS_PaulSage and @ZOS_GinaBruno,

By chance can we get some more precision on what the major issue(s) are? While it may seem like some insider baseball, I think some of the solutions that are being thought up might be a bit better if we knew just a bit more. What is a lot to a server? What types of calculations are difficult? (ex: people have mentioned the aoe, the lighting, and 64-bit clients). While some of these make sense to me some seem more like client problems as opposed to server.

Muizer wrote: »

PvP is not set up to support hundreds in one locality at a time. I do not mean the technical capacity to do that without lag. I mean the game rules. You want it to take ten times more players to take a keep? It would probably take a nerf of corresponding magnitude to player stats to justify such numbers.
Perhaps if cyrodiil were one gigantic keep. That might make our individual contributions insignificant enough to justify needing hundreds.

I'm just saying discuss lag......sure, but please in ways relevant to the game we've had from day one, and not some utterly vague marketing phrase. That's nice for ranting purposes but that's as far as it goes.

Before release and after, prior to about 1.2ish, I was in numerous 3 way battles with 200-300+ players at a keep. Siege flying everywhere, AoEs, heals, the works - all with no lag -

@Muizer it was set up to support hundreds, and was a selling point that was emphasized. A dream was reality. Vague marketing phrase? Were you in Cyrodiil for any of those fights?

Anyone fraps any of them?

NadiusMaximus wrote: »

Only real way of fixing it just dawned on me.

Reduce the skills we take into Cyrodil to two per bar and no ultimate.
Make battles be fought with weapons , light and heavy attacks.
One damage ability and one heal, or two damage, or two heals. That's it.


I'll lol myself for that one, but might be how they handle it vs. Fixing it for real before next year.

How about taking out all skills and we just emote what we do.

TheBull wrote: »

Arenas will help ease the server lag in Cyrodiil.

Tamriel Unlimited then should undergo a name change to "Tamriel 16 Players at a Time".

Also, how will arenas ease server lag?

Sylvyr wrote: »

@Muizer it was set up to support hundreds, and was a selling point that was emphasized. A dream was reality. Vague marketing phrase? Were you in Cyrodiil for any of those fights?

Not sure when I first went to Cyro, but I'd stick by the point that it's not just the lag that makes sieges poorly suited for large numbers. Perhaps back in the days you refer to people just didn't yet know how to build for and organise a proper zerg. I say, many the sieges I've been in were a foregone conclusion due to people organising simply to maximise AoE damage output concentration. Lag? Yes. Without lag? Result would be the same, but without lag to blame it on.

The problem with asking players to spread out is that cyrodiil encourages and rewards group play.

The main objective is to capture keeps, and scrolls. This cannot be achieved without a group, even undefended.

The alliance ranks are the only measurable progression system for individual players in pvp. Obtaining ranks requires large amounts of AP, which are primarily attained at large battles (sieges, choke points, between keeps).

Hence, players will always head towards battles. Asking them not to is asking them to forego AP and therefore stall progression (which is imperative for longevity and a feeling of accomplishment/attachment to characters).

So there are a few options to achieve the desired result (reducing lag by splitting up populations);

1: Add alternate objectives which are viable for earning AP. These could be instanced areas, like dungeons, battlegrounds, arena etc. Or open world, like dolmens, towns, ayelid wells.

There are numerous ways to make these rewarding, but each has their pros and cons. The most effective might be a 'king of the hill' type scenario, where the faction who holds an objective earns AP on a timer, the less players they have, the greater the reward (and risk). It could have a set limit per day/hour to prevent farming, but may require a lot of balancing.

This is just one example, cyrodiil currently has vast areas of space which could be utilised for these types of events.

2: Discourage 'zerging'. Again, there are multiple ways to achieve this. Boost siege damage, make detonation increase damage per enemies hit, reduce AP earned based on allies in proximity, increase guard damage based on enemies near keep, cap number of allies that can receive cleanse, are a few of the ideas that have been presented.

While these solutions may not be ideal, one or more could be very effective temporarily until a more permanent solution is implemented.

3: Encourage small group play. Boost AP gains for groups of 4-8 players if they are the only ones to deal damage to enemies. Give them buffs to help them stand a chance against large groups.

Again, tricky to balance, but would help enormously if done right.

4: Side quests with AP rewards. Examples: bounty hunting, dolmen challenges, dungeon challenges, exploration, scavenger hunts. Perhaps a chance at an epic set item would add more incentive, can be multiple types for solo/small group.

5: Rework PvP code. Left this til last for obvious reasons, would likely require huge amounts of time, money and manpower to achieve. However, could also result in clearing up latency issues once and for all, leading to massively improved player satisfaction (which = profit)



This is by no means an exhaustive list, but rather a compilation of previous suggestions and a few of my own.

Above all, communication is KEY. If you keep people informed of your intentions and progress, they won't feel cheated or ignored. This thread is a good step in the right direction, but for many it is too little, too late.

If you want to retain as many as possible, KEEP POSTING. Even if there's no news, acknowledgement of the issues goes a long way.

Fin.

@ZOS_PaulSage @ZOS_BrianWheeler @ZOS_JessicaFolsom

Haven't had time to read through this thread,

some initial thoughts.

- There used to be more action across the map with FCs. Now you can only port as far as the front-line and this creates more zerging.
- With champ points I am now zerging a lot more because I get more points
- Lag seems to get worse when zergs collide and meteor spamming starts.

thanks

Septimus_Magna wrote: »

stefan.gustavsonb16_ESO wrote: »

Sylvyr wrote: »

Muizer wrote: »

...
I'm just saying discuss lag......sure, but please in ways relevant to the game we've had from day one, and not some utterly vague marketing phrase. That's nice for ranting purposes but that's as far as it goes.

Before release and after, prior to about 1.2ish, I was in numerous 3 way battles with 200-300+ players at a keep. Siege flying everywhere, AoEs, heals, the works - all with no lag -

@Muizer it was set up to support hundreds, and was a selling point that was emphasized. A dream was reality. Vague marketing phrase? Were you in Cyrodiil for any of those fights?

Anyone fraps any of them?

The game prior to 1.2.3 (I think) had almost no safety checks against cheating, and a lot of the heavy computations (all collision detection and movement, perhaps some damage calculations as well) were done on the client side. That was not a realistic solution for the long haul, and once they added checks and measures to stop teleporting and invulnerability hacks, the game stopped being able to handle large, crowded battles.

But I do remember the early days, right after launch. Cyrodiil was a lot of fun until the hackers ruined it and the anti-hack countermeasures brought the server code to its knees.

If this is correct the solution is worse than the problem imo. Cheaters should be reported by players and banned by moderators, if there are systems to prevent cheating which cause these laggy situations in Cyrodiil I would vote for removing them. When someone kills you by cheating players could simply report them with a screenshot of the death recap to get the cheater banned. This might not prevent cheating but it also doesnt prevent all normal players from experiencing large scale battles.

That's probably what they thought too. But it's an imperfect solution. The only real solution is to prevent the activity- the optimistic mode of banning as caught just is not efficient enough, can catch people with false positives, and can let flagrant abuses go for longer than they should.

DHale wrote: »

I would name the players, guilds and the streamers but I want the streamers to keep their TS channel open. Because they are not just exploiters they are not very bright. I would tape them to get them banned but I don’t care.

Don't name them... just report them.

Septimus_Magna wrote: »

If this is correct the solution is worse than the problem imo. Cheaters should be reported by players and banned by moderators, if there are systems to prevent cheating which cause these laggy situations in Cyrodiil I would vote for removing them. When someone kills you by cheating players could simply report them with a screenshot of the death recap to get the cheater banned. This might not prevent cheating but it also doesnt prevent all normal players from experiencing large scale battles.

@Septimus_Magna : Reporting did not work well in the begging, there is no reason to think that it would now... that said I think that distributed, randomized, and reactive reactive mechanics (ex: torrenting) would help spread it out. Not everyones computer could do it as well but my system could probably handle some of that load and over time more and more others will as well.

Today art 13:00 - 13:15, in NA thornblade, reds were pushing blue to get their scroll back.

Only EP was pop locked

Blues had 3 bars pop

Yellow had 2 bars pop

The lag went do bad around there that we could not use abilities, share quest, walk around... NADA

Even in Chillrend during primetime and all 3 factions pop locked no such nonsense happen.

Open your eyes ZoS, there is a group of EP NA who is using some kind of video stream to overload the server and cause lag intentionally. All the time. Look at when they push for emperor or for scrolls it happends all the time