alternatelder wrote: »MincMincMinc wrote: »Quoted post has been removed
Feedback as simple as a thumbs up or thumbs down is just as legitimate as pages long explanations. They may not help the devs tailor their responses as accurately as needed, but a simple like or don't like is still legitimate feedback as a stand alone comment. Besides that, ZOS pretty much does whatever they've decided their going to do regardless of feedback. That's why their investment in this template version of Cyrodiil is so concerning for so many of us primarily PvP players.
Why is ZOS even investing time and money into a whole new mode of PvP before exhausting their options with the system they already have. As many other posters have pointed out, ZOS hasn't even tried limiting shield stacking and heal stacking in groups under Battle Spirit. We've been asking for this for years. ZOS appears to be on the road to scrapping the PvP that put ESO on the map in the first place, and that is very sad.
Feedback that is purposefully misconstrued or biased for the purpose of being spiteful, I cannot subscribe to being helpful."I do not like template pvp, it is limiting" This is fine and good feedback
"I can't believe zos is doing this to the game, how could they remove builds? Im going to quit if they continue with this" yeah these comments are not only purposefully incorrect, but also not helpful in anyway other than baiting random hate commenters to pop up.
The only investment theyve made is the instance tools that allow them to have a different skill book between pve and pvp.....Which is extremely helpful and healthy for the game. Everyone for years has been asking for pve and pvp to be balanced seperately. Practically since the dots and hots stacking change almost a decade ago.
The skills you see probably took them a week of work for one person. They are almost all copy and renamed versions of the current skills with parts disabled. It probably took them just as long to write the tooltips as it was to make the skills function.
Not to mention they stated multiple times that template pvp was never intended as the final product.....again not sure why people thought this would be the case. From nearly every aspect it makes no sense.
Please answer the question.
Why is ZOS even investing time and money into a whole new mode of PvP before exhausting their options with the system they already have?
Do you even know what this mode is? It is them investing in their current system, by testing here...🤦
I do know. I've played some on the PTS, so I've seen it first hand. It's not good in any capacity in my judgement. Have you played the game mode on PTS yet?
No. It's creating a whole new game mode. Where is the theory crafting in this new game mode. Where is the linking this game mode to PvE in any capacity?
It's not a game mode for one, it's a test. There is a difference.Please answer the question.
Why is ZOS even investing time and money into a whole new mode of PvP before exhausting their options with the system they already have?
It's not a game mode for one, it's a test. There is a difference.Please answer the question.
Why is ZOS even investing time and money into a whole new mode of PvP before exhausting their options with the system they already have?
As to why, what do you think sounds more efficient? Removing and tweaking indivual things hoping ONE of those things is the magic bullet, or stripping something down to its bare bones to see if problems still persist, and if not adding features back one at a time until they find things that cause problems?
People complaining need to realize that just randomly testing X or Y is the least efficient way of testing. It's like trying to figure out what's causing a computer not to work right. You can fiddle with X program or Y component, but in the end taking everything down to the absolute necessities and then adding things back one at a time is often the best way of finding out what's causing issues.
It's not a game mode for one, it's a test. There is a difference.Please answer the question.
Why is ZOS even investing time and money into a whole new mode of PvP before exhausting their options with the system they already have?
As to why, what do you think sounds more efficient? Removing and tweaking indivual things hoping ONE of those things is the magic bullet, or stripping something down to its bare bones to see if problems still persist, and if not adding features back one at a time until they find things that cause problems?
People complaining need to realize that just randomly testing X or Y is the least efficient way of testing. It's like trying to figure out what's causing a computer not to work right. You can fiddle with X program or Y component, but in the end taking everything down to the absolute necessities and then adding things back one at a time is often the best way of finding out what's causing issues.
I get your point, not trying to argue with it, but I do think your comparison needs some finetuning.
The current situation is like having a computer running poorly for over 10 years, bringing it to the repair guy every six months, and getting a bill and a "should be fine now" every time. Then after 10 years the repair guy says "I'll have to strip everything down to the bare minimum to run some tests first".
It's possible to agree with the narrative while still doubting the people who provide it to you.
It's not a game mode for one, it's a test. There is a difference.Please answer the question.
Why is ZOS even investing time and money into a whole new mode of PvP before exhausting their options with the system they already have?
As to why, what do you think sounds more efficient? Removing and tweaking indivual things hoping ONE of those things is the magic bullet, or stripping something down to its bare bones to see if problems still persist, and if not adding features back one at a time until they find things that cause problems?
People complaining need to realize that just randomly testing X or Y is the least efficient way of testing. It's like trying to figure out what's causing a computer not to work right. You can fiddle with X program or Y component, but in the end taking everything down to the absolute necessities and then adding things back one at a time is often the best way of finding out what's causing issues.
MincMincMinc wrote: »It's not a game mode for one, it's a test. There is a difference.Please answer the question.
Why is ZOS even investing time and money into a whole new mode of PvP before exhausting their options with the system they already have?
As to why, what do you think sounds more efficient? Removing and tweaking indivual things hoping ONE of those things is the magic bullet, or stripping something down to its bare bones to see if problems still persist, and if not adding features back one at a time until they find things that cause problems?
People complaining need to realize that just randomly testing X or Y is the least efficient way of testing. It's like trying to figure out what's causing a computer not to work right. You can fiddle with X program or Y component, but in the end taking everything down to the absolute necessities and then adding things back one at a time is often the best way of finding out what's causing issues.
I get your point, not trying to argue with it, but I do think your comparison needs some finetuning.
The current situation is like having a computer running poorly for over 10 years, bringing it to the repair guy every six months, and getting a bill and a "should be fine now" every time. Then after 10 years the repair guy says "I'll have to strip everything down to the bare minimum to run some tests first".
It's possible to agree with the narrative while still doubting the people who provide it to you.
Except in that analogy you forgot the person who brought the computer in spends his time downloading and installing random malware demanding more malware each quarter. (procs and effects running rampant are the malware)
The repair guy is faced with either getting rid of the malware or throwing out the computer entirely. First rational thing you'd do in this scenario is uninstall all of the malware and test it. Just be happy that the repair man is also the malware creator. It is much more lucrative to just keep selling malware to people than it is to repair the computer.
It's not a game mode for one, it's a test. There is a difference.Please answer the question.
Why is ZOS even investing time and money into a whole new mode of PvP before exhausting their options with the system they already have?
As to why, what do you think sounds more efficient? Removing and tweaking indivual things hoping ONE of those things is the magic bullet, or stripping something down to its bare bones to see if problems still persist, and if not adding features back one at a time until they find things that cause problems?
People complaining need to realize that just randomly testing X or Y is the least efficient way of testing. It's like trying to figure out what's causing a computer not to work right. You can fiddle with X program or Y component, but in the end taking everything down to the absolute necessities and then adding things back one at a time is often the best way of finding out what's causing issues.
Actually, the way to test is to change only one variable at a time. All of us with a hard science background and have designed lab experiments know this is how it's done. When you change more than one variable at a time you can't easily discern which variable that was changed that impacted the results of the test.
It's not a game mode for one, it's a test. There is a difference.Please answer the question.
Why is ZOS even investing time and money into a whole new mode of PvP before exhausting their options with the system they already have?
As to why, what do you think sounds more efficient? Removing and tweaking indivual things hoping ONE of those things is the magic bullet, or stripping something down to its bare bones to see if problems still persist, and if not adding features back one at a time until they find things that cause problems?
People complaining need to realize that just randomly testing X or Y is the least efficient way of testing. It's like trying to figure out what's causing a computer not to work right. You can fiddle with X program or Y component, but in the end taking everything down to the absolute necessities and then adding things back one at a time is often the best way of finding out what's causing issues.
Actually, the way to test is to change only one variable at a time. All of us with a hard science background and have designed lab experiments know this is how it's done. When you change more than one variable at a time you can't easily discern which variable that was changed that impacted the results of the test.
It's not a game mode for one, it's a test. There is a difference.Please answer the question.
Why is ZOS even investing time and money into a whole new mode of PvP before exhausting their options with the system they already have?
As to why, what do you think sounds more efficient? Removing and tweaking indivual things hoping ONE of those things is the magic bullet, or stripping something down to its bare bones to see if problems still persist, and if not adding features back one at a time until they find things that cause problems?
People complaining need to realize that just randomly testing X or Y is the least efficient way of testing. It's like trying to figure out what's causing a computer not to work right. You can fiddle with X program or Y component, but in the end taking everything down to the absolute necessities and then adding things back one at a time is often the best way of finding out what's causing issues.
Actually, the way to test is to change only one variable at a time. All of us with a hard science background and have designed lab experiments know this is how it's done. When you change more than one variable at a time you can't easily discern which variable that was changed that impacted the results of the test.
I think that's still what they're doing, they're just removing as many variables as possible and getting a baseline before they start adding.
Actually, the way to test is to change only one variable at a time. All of us with a hard science background and have designed lab experiments know this is how it's done. When you change more than one variable at a time you can't easily discern which variable that was changed that impacted the results of the test.
Actually, the way to test is to change only one variable at a time. All of us with a hard science background and have designed lab experiments know this is how it's done. When you change more than one variable at a time you can't easily discern which variable that was changed that impacted the results of the test.
I mean, if we want to get pedantic about it, the best way to test this would be purely in-house with a replicate of the live server while running thousands of bots and some profiling tools. That would tell you exactly what is slowing down your system with absolutely zero ambiguity.
MincMincMinc wrote: »Actually, the way to test is to change only one variable at a time. All of us with a hard science background and have designed lab experiments know this is how it's done. When you change more than one variable at a time you can't easily discern which variable that was changed that impacted the results of the test.
I mean, if we want to get pedantic about it, the best way to test this would be purely in-house with a replicate of the live server while running thousands of bots and some profiling tools. That would tell you exactly what is slowing down your system with absolutely zero ambiguity.
They did say in the livestream that they already ran bots, but for some reason that was not enough. Probably a higher up argument that is requiring the team to go live with the test.
MincMincMinc wrote: »Actually, the way to test is to change only one variable at a time. All of us with a hard science background and have designed lab experiments know this is how it's done. When you change more than one variable at a time you can't easily discern which variable that was changed that impacted the results of the test.
I mean, if we want to get pedantic about it, the best way to test this would be purely in-house with a replicate of the live server while running thousands of bots and some profiling tools. That would tell you exactly what is slowing down your system with absolutely zero ambiguity.
They did say in the livestream that they already ran bots, but for some reason that was not enough. Probably a higher up argument that is requiring the team to go live with the test.
Bots will also never be as purposefully erratic as an actual human tester or properly simulate what connecting to people all over the world with a myriad of different potential connection issues will be like
DDOS'ing a private server with a, most likely self hosted, bots will only get you so far
Actually, the way to test is to change only one variable at a time. All of us with a hard science background and have designed lab experiments know this is how it's done. When you change more than one variable at a time you can't easily discern which variable that was changed that impacted the results of the test.
MincMincMinc wrote: »Actually, the way to test is to change only one variable at a time. All of us with a hard science background and have designed lab experiments know this is how it's done. When you change more than one variable at a time you can't easily discern which variable that was changed that impacted the results of the test.
I mean, if we want to get pedantic about it, the best way to test this would be purely in-house with a replicate of the live server while running thousands of bots and some profiling tools. That would tell you exactly what is slowing down your system with absolutely zero ambiguity.
They did say in the livestream that they already ran bots, but for some reason that was not enough. Probably a higher up argument that is requiring the team to go live with the test.
Bots will also never be as purposefully erratic as an actual human tester or properly simulate what connecting to people all over the world with a myriad of different potential connection issues will be like
DDOS'ing a private server with a, most likely self hosted, bots will only get you so far
Well I wasn't talking about network DoSing. I was talking about the supposed issues with an overtaxed server than can't keep up with calculating combat in realtime.
Players are not really erratic though, are they? To the point where a lot of actions can be processed and even predicted by heuristics. There's also telemetry that can be recorded and used as the basis for inputs. And if you want the ultimate stress test, there's no need to even simulate real user inputs. You can just pick out the actions that are likely to be the most heavy to calculate and run them through a partial system. You don't even need a full game simulation for that.
It's not unreasonable to expect the devs to know how players interact with their product. Whatever comes out of Vengeance is still going to have a meta of good and bad strats, and the devs will need to know how to make this fun.MincMincMinc wrote: »You really think a zos employee knows the ins and outs of ball group meta?
xylena_lazarow wrote: »It's not unreasonable to expect the devs to know how players interact with their product. Whatever comes out of Vengeance is still going to have a meta of good and bad strats, and the devs will need to know how to make this fun.MincMincMinc wrote: »You really think a zos employee knows the ins and outs of ball group meta?
MincMincMinc wrote: »xylena_lazarow wrote: »It's not unreasonable to expect the devs to know how players interact with their product. Whatever comes out of Vengeance is still going to have a meta of good and bad strats, and the devs will need to know how to make this fun.MincMincMinc wrote: »You really think a zos employee knows the ins and outs of ball group meta?
Its a nice thought, but realistically the devs just wont be able to compete with people that have years worth of gameplay optimizing and finding abusable mechanics that don't get reported. It took zos years to figure out the bunny hop bug the top end of players were doing. Not to mention it still exists in the game even after they "fixed" it. At one point my guildies had a running list of bugs and exploits that they would test every pts to see if they were still in the game. I couldn't use overload for years because I was worried of being banned because it was fundamentally broken.
There's a reason why alpha and beta tests exist (well actual tests, not the AAA pre-order hype marketing scam) Just to rehash the point you quoted above is me explaining why zos needs player involvement and can't rely on just bots.
People see vengeance and think it will not be fun because all their skills nolonger have random powercreep buffs tied to them. Where in reality it may be far more fun once combat is more transparent, clear, and concise. I don't care if my hp is one number or another, neither of these is more or less fun. By some people's arguments we should just make every skill give 10k weapon damage or hp because more buffs = more fun. Functional and clear combat will increase the fun though.
MincMincMinc wrote: »MincMincMinc wrote: »Actually, the way to test is to change only one variable at a time. All of us with a hard science background and have designed lab experiments know this is how it's done. When you change more than one variable at a time you can't easily discern which variable that was changed that impacted the results of the test.
I mean, if we want to get pedantic about it, the best way to test this would be purely in-house with a replicate of the live server while running thousands of bots and some profiling tools. That would tell you exactly what is slowing down your system with absolutely zero ambiguity.
They did say in the livestream that they already ran bots, but for some reason that was not enough. Probably a higher up argument that is requiring the team to go live with the test.
Bots will also never be as purposefully erratic as an actual human tester or properly simulate what connecting to people all over the world with a myriad of different potential connection issues will be like
DDOS'ing a private server with a, most likely self hosted, bots will only get you so far
Well I wasn't talking about network DoSing. I was talking about the supposed issues with an overtaxed server than can't keep up with calculating combat in realtime.
Players are not really erratic though, are they? To the point where a lot of actions can be processed and even predicted by heuristics. There's also telemetry that can be recorded and used as the basis for inputs. And if you want the ultimate stress test, there's no need to even simulate real user inputs. You can just pick out the actions that are likely to be the most heavy to calculate and run them through a partial system. You don't even need a full game simulation for that.
Players are far more diverse than what one dev setting up a test would do. You really think a zos employee knows the ins and outs of ball group meta? You need players that know what to abuse. Otherwise the difference is like people light attacking each other with hundings vs a 4-5 proc set build that runs sets with 5x effects, timers, and stacks spamming every possible AoE timed buff skill in the game.
We already see this difference during MyM events. Where the dilution of pve players that are just light attacking each other unbuffed drastically cuts down on the strain of the server. Compared to a normal pop locked server of 40 man pvp zergs and ball groups.
Joy_Division wrote: »MincMincMinc wrote: »xylena_lazarow wrote: »It's not unreasonable to expect the devs to know how players interact with their product. Whatever comes out of Vengeance is still going to have a meta of good and bad strats, and the devs will need to know how to make this fun.MincMincMinc wrote: »You really think a zos employee knows the ins and outs of ball group meta?
Its a nice thought, but realistically the devs just wont be able to compete with people that have years worth of gameplay optimizing and finding abusable mechanics that don't get reported. It took zos years to figure out the bunny hop bug the top end of players were doing. Not to mention it still exists in the game even after they "fixed" it. At one point my guildies had a running list of bugs and exploits that they would test every pts to see if they were still in the game. I couldn't use overload for years because I was worried of being banned because it was fundamentally broken.
There's a reason why alpha and beta tests exist (well actual tests, not the AAA pre-order hype marketing scam) Just to rehash the point you quoted above is me explaining why zos needs player involvement and can't rely on just bots.
People see vengeance and think it will not be fun because all their skills nolonger have random powercreep buffs tied to them. Where in reality it may be far more fun once combat is more transparent, clear, and concise. I don't care if my hp is one number or another, neither of these is more or less fun. By some people's arguments we should just make every skill give 10k weapon damage or hp because more buffs = more fun. Functional and clear combat will increase the fun though.
The thing is they don't have too.
As was said many, many times in the BG stream thread, we don;t expect the devs to be good players. We do expect them to be knowledgeable about the very product they are in charge of balancing.
It's not that different from being a general manager or a coach of a professional sports team. Yeah coach cant play, but they do need to be on top of the sport's "meta" and coach does need to be able to correctly identify a good player from a mediocre one, and how the game's rules dictate what are effective plays and strategies.
The devs don;t know how ball groups run because they have actively avoided any conversation or the mountains of available data that would tell them exactly what's going on. I've played with an active organized group and for years we asked, begged, pleaded any dev to anonymously run with us just for a few nights to see what it is we go through. On our discord, we have years of combat data that make it glaringly obvious what mechanics are exploited. I know other groups who have made similar requests
Their customers know much of this information when it's not even their job. That makes it a hard swallow because it's not an unreasonable expectation that their clocking in for 40 hours a week precisely to collect the data that will inform them of what's going on in Cyrodiil, rather than actually playing at a competitive level. It was plainly obvious for all of us to see in that BG stream that the combination of the dev's build, hesitant and questionable remarks about PvP mechanics/strategies, and actual gameplay that this was someone who was not comfortable or familiar with what I would say are basic fundamentals of PvP gameplay.
Not that I have a degree in business management or anything like that, but if I were ZOS putting on a public showcase, the employee I would be putting up for an already skeptical public to see and answer questions would be the most talented and knowledgeable person on my payroll. And if I happened to be that ZOS employee, I would most certainly be looking up data, asking good people for builds, and honing my game before that judgmental audience. Could nobody at ZOS have a quick discreet conversation with anyone in the PvP community to tell them that, yeah, Mara's Balm and Troll King got nerfed a long time ago so it might be a good idea to run something a little more current?
So there's a lot of reason for skepticism. More than enough that the benefit of the doubt is no longer there. My hesitations about Vengeance stem not from what I see on the PTS (although, it's hard not to be concerned). Rather, because it is coming about using the same process that I have seen time and again from ZOS':
- With very minimal outside input (i.e, communication from the PvP community who have the data and the high level experience that ZOS lacks), the devs just come up with a radical change to the existing system and spring it on us to test on the PTS
- While on the PTS, immediate and obvious concerns are raised.
- The devs assure us not to worry about those concerns because they aren;t the priority, it's really something else that is getting tested, or future fixes and updates are in the works regarding our concerns
We all know steps 4 and 5:
- What was on the PTS basically becomes what we have to play with on Live, our feedback not being addressed
- The supposed fixes/updates either never see the light of day or aren't seen until numerous patches later (typically with unsatisfactory results)
So, it's not the loss of power creep.
I still remember that promised line and additional forums/outlets of better communication that came in the wake of the BG stream. What ever happened to that? Nothing but a single generic post that didn't even address what people's main concerns were about in the first place (heals and excessive defense). So despite the obvious public display that the current combat team isn't familiar with the very product that they are in charge of overseeing, it's the same old same old, nothing has been learned or changed.
MincMincMinc wrote: »MincMincMinc wrote: »Actually, the way to test is to change only one variable at a time. All of us with a hard science background and have designed lab experiments know this is how it's done. When you change more than one variable at a time you can't easily discern which variable that was changed that impacted the results of the test.
I mean, if we want to get pedantic about it, the best way to test this would be purely in-house with a replicate of the live server while running thousands of bots and some profiling tools. That would tell you exactly what is slowing down your system with absolutely zero ambiguity.
They did say in the livestream that they already ran bots, but for some reason that was not enough. Probably a higher up argument that is requiring the team to go live with the test.
Bots will also never be as purposefully erratic as an actual human tester or properly simulate what connecting to people all over the world with a myriad of different potential connection issues will be like
DDOS'ing a private server with a, most likely self hosted, bots will only get you so far
Well I wasn't talking about network DoSing. I was talking about the supposed issues with an overtaxed server than can't keep up with calculating combat in realtime.
Players are not really erratic though, are they? To the point where a lot of actions can be processed and even predicted by heuristics. There's also telemetry that can be recorded and used as the basis for inputs. And if you want the ultimate stress test, there's no need to even simulate real user inputs. You can just pick out the actions that are likely to be the most heavy to calculate and run them through a partial system. You don't even need a full game simulation for that.
Players are far more diverse than what one dev setting up a test would do. You really think a zos employee knows the ins and outs of ball group meta? You need players that know what to abuse. Otherwise the difference is like people light attacking each other with hundings vs a 4-5 proc set build that runs sets with 5x effects, timers, and stacks spamming every possible AoE timed buff skill in the game.
We already see this difference during MyM events. Where the dilution of pve players that are just light attacking each other unbuffed drastically cuts down on the strain of the server. Compared to a normal pop locked server of 40 man pvp zergs and ball groups.
Your post makes it sound like ZOS could easily solve the Cyrodiil performance problems with more/better hardware on the server end. I'm pretty sure we've had that discussion before and been punished for it.
Joy_Division wrote: »We all know steps 4 and 5:
- What was on the PTS basically becomes what we have to play with on Live, our feedback not being addressed
- The supposed fixes/updates either never see the light of day or aren't seen until numerous patches later (typically with unsatisfactory results)
Exactly this^
Personally I'm far less concerned about these tests than I am about, at some point, something even remotely similar to what's on the PTS right now becoming our only option for Cyrodiil play. I fear the primary decision on whether or not this game mode will be implemented will be based upon capitalism and not upon player satisfaction.
If the goal was to fix Cyrodiil ZOS would at least try limiting heal and shield stacking WITHIN groups and tweaking a few of the most calculation intensive sets before starting from scratch with a whole new game system.
I fear this test is more to see what the customer base will endure and still play more than it is anything else.
MincMincMinc wrote: »Joy_Division wrote: »We all know steps 4 and 5:
- What was on the PTS basically becomes what we have to play with on Live, our feedback not being addressed
- The supposed fixes/updates either never see the light of day or aren't seen until numerous patches later (typically with unsatisfactory results)
Exactly this^
Personally I'm far less concerned about these tests than I am about, at some point, something even remotely similar to what's on the PTS right now becoming our only option for Cyrodiil play. I fear the primary decision on whether or not this game mode will be implemented will be based upon capitalism and not upon player satisfaction.
If the goal was to fix Cyrodiil ZOS would at least try limiting heal and shield stacking WITHIN groups and tweaking a few of the most calculation intensive sets before starting from scratch with a whole new game system.
I fear this test is more to see what the customer base will endure and still play more than it is anything else.
Well, does it make sense to just implement vengeance? Half the endgame playerbase pvps on a consistent schedule. PvE players tend to come and go with dlc patches. Does it make sense to make half your playerbase nolonger need eso+ or dlc? Your main source of income.
Does it make sense to remove potions and glyphs and traits and....... No, if you implemented vengeance as it stands or even any template preset pvp you would see half your playerbase not participate anywhere else in your game. The consumable economy would collapse, half your endgame daily dungeon players, etc. A template pvp system would not ever hold salt.
Not to mention they already stated that gear would return, traits, enchants, etc. The more plausible changes you could expect would be:
- more streamlined skills like on vengeance. A damage skill......is a damage skill. Instead of having 8x passives or procs tied to it
- the same skills as live, but with rule changes like dots/hots not stacking. Or certain skills that wont crossheal. etc.
- Set changes or bans to prevent complicated spaghetti code procs like RoA for example
- Status effect reworks to cut down on unnecessary ticks.
- Battlespirit being removed, effectively getting rid of a whole layer of calculations.
- No more stuck in combat..... Seriously you leave combat within 2-3s after attacking.
JustLovely wrote: »like at the moment it isn't at its best but doing nothing at all is substantially more preferable to this vengeance nonsense like im concerned that this won't be totally scrapped despite the extreme amount of negative feedback. There is nothing good in vengeance you're literally a companion made playable. It is horrid.
Couldn't agree more.
I don't like anything about this new play mode for Cyrodiil.
Why isn't ZOS working on balancing the Cyrodiil we already have? They could remove a few of the most calculation intensive sets and limit cross healing and shield stacking for groups, which is something we've been asking for for years and they have yet to so much as even try it out and see if that would be enough to improve performance.
Implementing this special play mode will isolate PvP from PvE entirely so PvP players will have no reason to PvE at all, ever. Some of the PvE community thinks this is a good idea, but they have no idea how integral PvP is to ESO.
Here's an idea from another game that fixed the unkillable people issue, which is a plague in ESO and make most not want to play aside from few "pros":
- Make shields mitigate 30% of incoming damage, instead of completely stopping it, and reduce your damage done by 30%.
MincMincMinc wrote: »MincMincMinc wrote: »MincMincMinc wrote: »Actually, the way to test is to change only one variable at a time. All of us with a hard science background and have designed lab experiments know this is how it's done. When you change more than one variable at a time you can't easily discern which variable that was changed that impacted the results of the test.
I mean, if we want to get pedantic about it, the best way to test this would be purely in-house with a replicate of the live server while running thousands of bots and some profiling tools. That would tell you exactly what is slowing down your system with absolutely zero ambiguity.
They did say in the livestream that they already ran bots, but for some reason that was not enough. Probably a higher up argument that is requiring the team to go live with the test.
Bots will also never be as purposefully erratic as an actual human tester or properly simulate what connecting to people all over the world with a myriad of different potential connection issues will be like
DDOS'ing a private server with a, most likely self hosted, bots will only get you so far
Well I wasn't talking about network DoSing. I was talking about the supposed issues with an overtaxed server than can't keep up with calculating combat in realtime.
Players are not really erratic though, are they? To the point where a lot of actions can be processed and even predicted by heuristics. There's also telemetry that can be recorded and used as the basis for inputs. And if you want the ultimate stress test, there's no need to even simulate real user inputs. You can just pick out the actions that are likely to be the most heavy to calculate and run them through a partial system. You don't even need a full game simulation for that.
Players are far more diverse than what one dev setting up a test would do. You really think a zos employee knows the ins and outs of ball group meta? You need players that know what to abuse. Otherwise the difference is like people light attacking each other with hundings vs a 4-5 proc set build that runs sets with 5x effects, timers, and stacks spamming every possible AoE timed buff skill in the game.
We already see this difference during MyM events. Where the dilution of pve players that are just light attacking each other unbuffed drastically cuts down on the strain of the server. Compared to a normal pop locked server of 40 man pvp zergs and ball groups.
Your post makes it sound like ZOS could easily solve the Cyrodiil performance problems with more/better hardware on the server end. I'm pretty sure we've had that discussion before and been punished for it.
Did you mean to quote a different post? Physical hardware wise, they already did an upgrade 2.5 years ago. I point out that this test will prove whether this is a physical hardware or software issue. 99% sure if zos determines it to be a hardware issue, we will not see new servers again because they already made that investment planning for another 7-10 years. We'd sooner see zos give up on cyro and focus only on bg game modes if anything. So better hope that zos determines it to be a combat code optimization issue.