Maintenance for the week of March 25:
• [IN PROGRESS] ESO Store and Account System for maintenance – March 28, 9:00AM EDT (13:00 UTC) - 12:00PM EDT (16:00 UTC)

Simplified calculations over cooldowns

MerguezMan
MerguezMan
✭✭✭✭
Dear Zenimax,

The cooldowns implementation will for sure improve server performance. Having less players would also improve server performance, but I guess it also means you get less money, as players would not spend on crown crates if they don't play the game anymore...

Dear players,

The server performance is bad. So far, the best idea ZoS has is to add cooldowns. May we suggest something else ?

Simplified calculations over cooldowns 28 votes

Get rid of non-combat pets. (only player can see his pet, or no pets at all)
10% 3 votes
Get rid of back-and-forth calculations (skills don't heal depending on damage done)
3% 1 vote
Get rid of stackings from skills and sets (you may charge attack instead keeping button pressed for better effect)
7% 2 votes
1+2
7% 2 votes
1+3
3% 1 vote
2+3
3% 1 vote
1+2+3
3% 1 vote
I'd prefer cooldowns
7% 2 votes
Other
53% 15 votes
  • MerguezMan
    MerguezMan
    ✭✭✭✭
    Voting "other" without suggesting anything else doesn't add to the discussion.
    Please comment rather than just trolling the poll.
  • Joy_Division
    Joy_Division
    ✭✭✭✭✭
    ✭✭✭✭✭
    None. #1 is trivial and the other two are just as disruptive to gameplay as ZoS's.

    I know you're trying to help, but the reality is that ZoS needs , err needed, to improve their servers in order to provide the goal of having less lag and a game that will still be enjoyable to play.
  • BXR_Lonestar
    BXR_Lonestar
    ✭✭✭✭✭
    IMO, the elephant in the room here is that high APM players add to increasingly complicated damage calculations because of the pace of their attacks. IMO, they should be trying to resolve animation cancelling in order to slow down the pace of combat to prevent the server from getting overwhelmed by calculations. Light attack weaving is fine, but block/bash cancelling to further increase your dps rate shouldn't be a thing IMO. Addressing the cancelling issue would not only slow down the pace of combat, it would reduce the number of calculations per minute the server has to do which would also help performance.

    They should be trying to address the issue in this fashion rather than doing so with cooldowns because enemies don't stay rooted long enough for aoe abilities to be effective. At least now, you can compensate for this by moving your aoe spells around the battlefield.

    One other possibility that may help reduce server calculations is to extend the duration of abilities. ESO has so many short-lived buffs that it is ridiculous.
  • the1andonlyskwex
    the1andonlyskwex
    ✭✭✭✭✭
    I'm not convinced any of these ideas will have enough impact to matter.

    Honestly, what I was expecting from the subject line was something like using Manhattan distance instead of Euclidian distance for range calculations.
  • Gilvoth
    Gilvoth
    ✭✭✭✭✭
    ✭✭✭✭✭
    None. #1 is trivial and the other two are just as disruptive to gameplay as ZoS's.

    I know you're trying to help, but the reality is that ZoS needs , err needed, to improve their servers in order to provide the goal of having less lag and a game that will still be enjoyable to play.

    this ^
    Exactly
  • MerguezMan
    MerguezMan
    ✭✭✭✭
    Honestly, what I was expecting from the subject line was something like using Manhattan distance instead of Euclidian distance for range calculations.

    How come using Manhattan distance instead of Euclidian would REDUCE calculations in ESO's combat system ?


    For the comments about suggestions not being impactful enough on server calculations:


    1) Anyone that has run organized Trials has probably already heard about pets disabling.

    It makes the action more readable (less moving things, so you focus on actual enemies rather than that mutant frog-dog-thing jumping around), but it also improves performance (which makes sense, as the games doesn't have to render 12 frog-dog-things jumping around).

    If such little trick works for a 12-player party, I'm quite confident it would give rather good results on whole server scale.


    2) Back-and-forth effects.

    On top of regular server damage calculations, the system has to track how much damage is done (not sure if it's before or after damage mitigation, but on later case, it's even worse) to add a layer of additionnal damage or heal-back. No need to be a math genius to understand that "you get [x] Health if enemy is hit" is way easier to handle than "you heal for [x]% of damage done with this skill".

    The bow passive, recalculating damage depending on target distance, is also a non-sensic added complexity (and honestly, IRL, I doubt your arrow hits harder when being afar...).

    Thats 3 times less calculations per relevant skill per player in combat. That may not impact the core functionality of the skill (you may be healed or deal added damage on a fixed number rather than calculated number, so what ?). It's not sillier than butchering AoE effects...


    3) Effects with stack counting.

    On top of regular common buffs/debuffs, some specific skills or sets give you stacks. PC players may have improved addons to track that, but on console, you get a full line on the bottom of your screen displaying at least a dozen of blinking icons that are almost impossible to fully track for a regular human being.

    Each of these icons has a dedicated status and name. ZoS adds more with each DLC, each new set having its own stack-thing associated. Sure, the effect alone is cool, but can't we have a similar result with a simpler system ? Can't we have for instance a limited range of stacks that would add up from player actions, whatever the skill or set used ?

    Wouldn't it be easier to manage something like 3 counters of stacks per player (on top of the basic 9 types of damage rolling into target mitigation) rather than the 20 or so we have now ?


    [Edit] Improve servers may help... But that won't fix everything.
    You may call ZoS greedy, though investing on months of coding as they try to improve performance may be more expensive than just buying more server space/memory/power.

    The game was designed at the beginning to have calculations handled on player side for optimal results, cheaters have forced ZoS to handle almost everything on server side, who should be blamed for that ?
    Edited by MerguezMan on August 5, 2020 5:00PM
  • Firstmep
    Firstmep
    ✭✭✭✭✭
    ✭✭✭
    Remove stacking hots.
    Remove smart healing.
    Rework hots to completely snapshot their values from the bar they're casted and replace defile effects with healing absorbtion to reduce calculations on the server.
    Remove cp completely from pvp(this would help with balancing too).
    Extra stats may be added via battlespirit to help with the lack of cp.
    Rework cyrodiil into multiple smaller zones, as most of the map is unused anyway.
    Add more small objectives that don't require sieging, but can have an effect on score to spread people out more.
    Rework AP bonuses to discourage mass groups from stacking in 1 spot.
    Remove participation rewards from taking empty keeps with a 40 man raid.
    The list goes on..
    Edited by Firstmep on August 5, 2020 5:36PM
  • the1andonlyskwex
    the1andonlyskwex
    ✭✭✭✭✭
    Calculating manhattan distance is addition and subtraction. Calculating Euclidian distance requires multiplication (and a square root, although you can work around that) too. This would affect literally every single action requiring a range check. The downside is that your actual ability range in Euclidian distance would depend on the direction to your target, which would be weird.

    Most of the things you're concerned about either happen entirely on the client side (non-combat pets, visual stack indications), happen infrequently (most AoE abilities don't have things like healing based on their damage done), and/or don't really require significant additional calculation (the server already calculates how much damage you're doing, so a heal based on that is just one multiplication and an addition to your health pool).
    Edited by the1andonlyskwex on August 5, 2020 5:39PM
  • ZarkingFrued
    ZarkingFrued
    ✭✭✭✭✭
    IMO, the elephant in the room here is that high APM players add to increasingly complicated damage calculations because of the pace of their attacks. IMO, they should be trying to resolve animation cancelling in order to slow down the pace of combat to prevent the server from getting overwhelmed by calculations. Light attack weaving is fine, but block/bash cancelling to further increase your dps rate shouldn't be a thing IMO. Addressing the cancelling issue would not only slow down the pace of combat, it would reduce the number of calculations per minute the server has to do which would also help performance.

    They should be trying to address the issue in this fashion rather than doing so with cooldowns because enemies don't stay rooted long enough for aoe abilities to be effective. At least now, you can compensate for this by moving your aoe spells around the battlefield.

    One other possibility that may help reduce server calculations is to extend the duration of abilities. ESO has so many short-lived buffs that it is ridiculous.

    Removing swap, bash, and roll cancel would make combat so insanely clunky feeling that no one would play. They absolutely need to override skills animations in order to actually move about in the game. Could you imagine how awful it would feel? You need to be able to react quickly in order to play efficiently. Being stuck in animation would mean you die when that Executioner/ Lash comes down every time because you can't block/roll/ or bar swap until the skill you just casted finishes its unnecessarily long animation. They need to reduce animation times across the board.
    Edited by ZarkingFrued on August 5, 2020 9:18PM
  • Pinja
    Pinja
    ✭✭✭✭
    I think what people are saying about #1 is that pets already aren't allowed in Cyrodiil, which is where the focus is.
    IMO, the elephant in the room here is that high APM players add to increasingly complicated damage calculations because of the pace of their attacks. IMO, they should be trying to resolve animation cancelling in order to slow down the pace of combat to prevent the server from getting overwhelmed by calculations. Light attack weaving is fine, but block/bash cancelling to further increase your dps rate shouldn't be a thing IMO. Addressing the cancelling issue would not only slow down the pace of combat, it would reduce the number of calculations per minute the server has to do which would also help performance.

    They should be trying to address the issue in this fashion rather than doing so with cooldowns because enemies don't stay rooted long enough for aoe abilities to be effective. At least now, you can compensate for this by moving your aoe spells around the battlefield.

    One other possibility that may help reduce server calculations is to extend the duration of abilities. ESO has so many short-lived buffs that it is ridiculous.
    @BXR_Lonestar I wouldn't call that the elephant as even low skill trial instances lag. If it is a problem high preforming raids should only curse themselves, if the servers are allocated properly.

    #2 after you explain it doesn't sound that bad, but not all too many abilities work like that where they calculate off the result of the first calculation resulting in double calculations for one global. They kinda did the same with certain sets making them a flat stat instead of a percent of a max stat or other benchmark.

    #3 isn't really a calculation, it's just a counter and a another variable. Simplifying it from a processing stand point would mean just to get rid of it. Because if the code is already written in it's simplest terms, it's an if statement followed by a counter.

    My suggestions I've put out before but I'll drop them here.
    Edited by Pinja on August 5, 2020 9:40PM
    Pinja for Dual Wands.
    Pinja's three server solutions:
  • MerguezMan
    MerguezMan
    ✭✭✭✭
    Calculating manhattan distance is addition and subtraction. Calculating Euclidian distance requires multiplication (and a square root, although you can work around that) too. This would affect literally every single action requiring a range check. The downside is that your actual ability range in Euclidian distance would depend on the direction to your target, which would be weird.

    I wouldn't expect players to estimate target distance in Manhattan distance in a 3D game. That would feel weird.
    Most of the things you're concerned about either happen entirely on the client side (non-combat pets, visual stack indications), happen infrequently (most AoE abilities don't have things like healing based on their damage done), and/or don't really require significant additional calculation (the server already calculates how much damage you're doing, so a heal based on that is just one multiplication and an addition to your health pool).

    Other players non-combat pets' position is also displayed on your side, so that is information going through the server.

    Most skills "don't have healing based on damage done", that "happens infrequently" ? I'm not focused on AoE.
    - Templar's Sweeps
    - Nightblade's Killer blade and Grim focus
    - Dragonknight's Burning embers
    And that is only calculated heal-back skills, you can add in skills with tricky descriptions like "on hit, if target is under 50% health, you deal [x]% additional damage and/or get ressource back".

    Yes, that is "just" an addition and a multiplication on top of the rest (well, actually that's a bit more, but ok, let's consider its 2), the calculation itself is not that hard, but consider getting rid of these 2 operations on something like NB's spammable, that's 2 less operations per NB player per second.

    At the end of the day, server-wide, that can be a lot less calculations.
  • the1andonlyskwex
    the1andonlyskwex
    ✭✭✭✭✭
    MerguezMan wrote: »
    Other players non-combat pets' position is also displayed on your side, so that is information going through the server.

    Most skills "don't have healing based on damage done", that "happens infrequently" ? I'm not focused on AoE.
    - Templar's Sweeps
    - Nightblade's Killer blade and Grim focus
    - Dragonknight's Burning embers
    And that is only calculated heal-back skills, you can add in skills with tricky descriptions like "on hit, if target is under 50% health, you deal [x]% additional damage and/or get ressource back".

    Yes, that is "just" an addition and a multiplication on top of the rest (well, actually that's a bit more, but ok, let's consider its 2), the calculation itself is not that hard, but consider getting rid of these 2 operations on something like NB's spammable, that's 2 less operations per NB player per second.

    At the end of the day, server-wide, that can be a lot less calculations.

    I don't think non combat pet locations have to have any server-side interaction at all. They're probably just drawn relative to their owner. Because they're not involved in combat, there isn't really any need for them to be in the same place from everybody's point of view. Additionally, as someone else pointed out, they're already disabled in PvP where the performance problems are.

    Also, I think you're still vastly overestimating the computational cost of all of the damage-dependent skills. Sure, it's an extra calculation every time one of those skills gets cast, but it's still only one calculation per cast. Compare that to the AoE ranging and targeting calculations, which need to happen for every caster-target pair in the zone/instance. If there are 50 potential targets for an ability, the server has to calculate 50 distances, throw away the targets that are out of range, then perform some sort of scoring and assignment to choose among the remaining targets to determine who the ability actually hits. Those calculations alone absolutely dwarf any calculations that occur after targets have been selected.
  • Gorreck
    Gorreck
    ✭✭✭
    I honestly don't know how they can fix it, the strange thing to me is that performance is getting worse and worse.

    I'll use Warhammer Online for an example, on release mass-PvP performance was bad, but to their credit they devs worked and worked on it and in the end it was the probably the best performing MMORPG I've ever seen in mass-PvP (it was, of course, too late to actually save WAR Online by that point).
  • SodanTok
    SodanTok
    ✭✭✭✭✭
    ✭✭
    Where is it stated that their best idea to improve performance is to add cooldowns? Where?
  • Gorreck
    Gorreck
    ✭✭✭
    SodanTok wrote: »
    Where is it stated that their best idea to improve performance is to add cooldowns? Where?


    The are testing on the PTS now and its going live to "test" next patch (notes are on the PTS forum, I think).

    If it ends up sticking many builds and some classes are likely to be decimated.
  • MerguezMan
    MerguezMan
    ✭✭✭✭
    SodanTok wrote: »
    Where is it stated that their best idea to improve performance is to add cooldowns? Where?

    Here:
    https://forums.elderscrollsonline.com/en/discussion/539136/update-on-cyrodiil-performance-upcoming-aoe-tests/p1
    The first round of tests we are planning will focus on Area of Effect (AOE) abilities in Cyrodiil and will make it more difficult for AOE abilities to be the only abilities used, adjusting cooldown, cost and regen values of all AOEs (damage and healing).

    Specific details on the tests we will be running in Cyrodiil are as follows:
    • Test 1 – Shared global AOE cooldown - 3 second timer. This test adds a global 3 second shared cooldown to any AOE ability. This means that when you cast an AOE, you will not be able to cast another for 3 seconds. For example, as a Templar, if I cast Ritual of Retribution, I wouldn’t be able to cast Puncturing Sweep for 3 seconds.
    • Test 2 – Individual AOE cooldown - 3 second timer. This test adds an individual AOE cooldown to each AOE ability. This means that when you cast an AOE, you will not be able to cast that same AOE ability for 3 seconds. For example – as a Templar, if I cast Puncturing Sweep, I wouldn’t be able to cast Puncturing Sweep again for 3 seconds.
    • Test 3 – No cooldown, global ramping AOE cost. This test adds a global ramping AOE cost for each AOE cast. Similar to how streak or roll dodge works, where when you cast an AOE you receive a debuff for 5 seconds, each stack of the debuff increases the cost of any AOE cast.
    • Test 4 – Individual AOE cooldown – 3 second timer, global ramping AOE cost. This test adds an individual AOE cooldown to each ability as in test 2, but also combines that with a global ramping cost from test 3.
    During the times that any of these tests are active, we will be awarding double Alliance Points for anyone active in Cyrodiil.

    After we complete the above tests, we may try other combinations of cooldown, cost, and regen values on AOE abilities. However, we need to run these tests first and then assess the data. We will then let everyone know what we found and how we will move forward. We will be very upfront, but please be aware that if these tests confirm our hypothesis, then chain-casting AOE abilities will no longer form the core of the ESO PvP experience in the way it has for the last few years. We would then go through each class and ensure that there are viable builds for each and make adjustments as necessary.

    You may notice the wording: "the FIRST round of tests [...] will FOCUS on AoE abilities in Cyrodiil", which can imply:
    - there will be more tests of that kind
    - not only AoE abilities can be affected
    - not only Cyrodiil can be affected

    And from the various topics around, I 'm quite sure I'm not the only one having lag out of Cyrodiil.
    Edited by MerguezMan on August 6, 2020 1:36PM
  • the1andonlyskwex
    the1andonlyskwex
    ✭✭✭✭✭
    Gorreck wrote: »
    SodanTok wrote: »
    Where is it stated that their best idea to improve performance is to add cooldowns? Where?


    The are testing on the PTS now and its going live to "test" next patch (notes are on the PTS forum, I think).

    If it ends up sticking many builds and some classes are likely to be decimated.

    The fact that they're testing their hypothesis about performance using cooldowns doesn't really indicate anything about how they plan to solve the problem if their hypothesis is confirmed.

    If you actually read the rest of the announcement, it seems pretty clear that changes to AoE ability costs and/or resource regeneration (particularly for well-geared, well-coordinated, high-CP groups) are likely things that they plan to look at for their final solution.

    I think the cooldowns are just a quick way for the performance team to experiment with reduced AoE without making the balance team do a ton of (potentially unnecessary) work first.
  • MerguezMan
    MerguezMan
    ✭✭✭✭
    Gorreck wrote: »
    SodanTok wrote: »
    Where is it stated that their best idea to improve performance is to add cooldowns? Where?


    The are testing on the PTS now and its going live to "test" next patch (notes are on the PTS forum, I think).

    If it ends up sticking many builds and some classes are likely to be decimated.

    The fact that they're testing their hypothesis about performance using cooldowns doesn't really indicate anything about how they plan to solve the problem if their hypothesis is confirmed.

    If you actually read the rest of the announcement, it seems pretty clear that changes to AoE ability costs and/or resource regeneration (particularly for well-geared, well-coordinated, high-CP groups) are likely things that they plan to look at for their final solution.

    I think the cooldowns are just a quick way for the performance team to experiment with reduced AoE without making the balance team do a ton of (potentially unnecessary) work first.

    Sounds a bit naive to think they would test 4 weeks of cooldowns without considering implementing them.
    Sounds very naive to think once cooldowns are implemented they wouldn't be more widely used to fix performance elsewhere.

    [https://www.youtube.com/watch?v=YPN0qhSyWy8]
  • SodanTok
    SodanTok
    ✭✭✭✭✭
    ✭✭
    MerguezMan wrote: »
    SodanTok wrote: »
    Where is it stated that their best idea to improve performance is to add cooldowns? Where?

    Here:
    https://forums.elderscrollsonline.com/en/discussion/539136/update-on-cyrodiil-performance-upcoming-aoe-tests/p1
    The first round of tests we are planning will focus on Area of Effect (AOE) abilities in Cyrodiil and will make it more difficult for AOE abilities to be the only abilities used, adjusting cooldown, cost and regen values of all AOEs (damage and healing).

    Specific details on the tests we will be running in Cyrodiil are as follows:
    • Test 1 – Shared global AOE cooldown - 3 second timer. This test adds a global 3 second shared cooldown to any AOE ability. This means that when you cast an AOE, you will not be able to cast another for 3 seconds. For example, as a Templar, if I cast Ritual of Retribution, I wouldn’t be able to cast Puncturing Sweep for 3 seconds.
    • Test 2 – Individual AOE cooldown - 3 second timer. This test adds an individual AOE cooldown to each AOE ability. This means that when you cast an AOE, you will not be able to cast that same AOE ability for 3 seconds. For example – as a Templar, if I cast Puncturing Sweep, I wouldn’t be able to cast Puncturing Sweep again for 3 seconds.
    • Test 3 – No cooldown, global ramping AOE cost. This test adds a global ramping AOE cost for each AOE cast. Similar to how streak or roll dodge works, where when you cast an AOE you receive a debuff for 5 seconds, each stack of the debuff increases the cost of any AOE cast.
    • Test 4 – Individual AOE cooldown – 3 second timer, global ramping AOE cost. This test adds an individual AOE cooldown to each ability as in test 2, but also combines that with a global ramping cost from test 3.
    During the times that any of these tests are active, we will be awarding double Alliance Points for anyone active in Cyrodiil.

    After we complete the above tests, we may try other combinations of cooldown, cost, and regen values on AOE abilities. However, we need to run these tests first and then assess the data. We will then let everyone know what we found and how we will move forward. We will be very upfront, but please be aware that if these tests confirm our hypothesis, then chain-casting AOE abilities will no longer form the core of the ESO PvP experience in the way it has for the last few years. We would then go through each class and ensure that there are viable builds for each and make adjustments as necessary.

    You may notice the wording: "the FIRST round of tests [...] will FOCUS on AoE abilities in Cyrodiil", which can imply:
    - there will be more tests of that kind
    - not only AoE abilities can be affected
    - not only Cyrodiil can be affected

    And from the various topics around, I 'm quite sure I'm not the only one having lag out of Cyrodiil.

    I ask again, WHERE does it says imposing AoE cooldowns is their idea to improve performance. I didnt ask what measures they are implementing to test performance impact
  • MerguezMan
    MerguezMan
    ✭✭✭✭
    SodanTok wrote: »
    MerguezMan wrote: »
    SodanTok wrote: »
    Where is it stated that their best idea to improve performance is to add cooldowns? Where?

    Here:
    https://forums.elderscrollsonline.com/en/discussion/539136/update-on-cyrodiil-performance-upcoming-aoe-tests/p1
    The first round of tests we are planning will focus on Area of Effect (AOE) abilities in Cyrodiil and will make it more difficult for AOE abilities to be the only abilities used, adjusting cooldown, cost and regen values of all AOEs (damage and healing).

    Specific details on the tests we will be running in Cyrodiil are as follows:
    • Test 1 – Shared global AOE cooldown - 3 second timer. This test adds a global 3 second shared cooldown to any AOE ability. This means that when you cast an AOE, you will not be able to cast another for 3 seconds. For example, as a Templar, if I cast Ritual of Retribution, I wouldn’t be able to cast Puncturing Sweep for 3 seconds.
    • Test 2 – Individual AOE cooldown - 3 second timer. This test adds an individual AOE cooldown to each AOE ability. This means that when you cast an AOE, you will not be able to cast that same AOE ability for 3 seconds. For example – as a Templar, if I cast Puncturing Sweep, I wouldn’t be able to cast Puncturing Sweep again for 3 seconds.
    • Test 3 – No cooldown, global ramping AOE cost. This test adds a global ramping AOE cost for each AOE cast. Similar to how streak or roll dodge works, where when you cast an AOE you receive a debuff for 5 seconds, each stack of the debuff increases the cost of any AOE cast.
    • Test 4 – Individual AOE cooldown – 3 second timer, global ramping AOE cost. This test adds an individual AOE cooldown to each ability as in test 2, but also combines that with a global ramping cost from test 3.
    During the times that any of these tests are active, we will be awarding double Alliance Points for anyone active in Cyrodiil.

    After we complete the above tests, we may try other combinations of cooldown, cost, and regen values on AOE abilities. However, we need to run these tests first and then assess the data. We will then let everyone know what we found and how we will move forward. We will be very upfront, but please be aware that if these tests confirm our hypothesis, then chain-casting AOE abilities will no longer form the core of the ESO PvP experience in the way it has for the last few years. We would then go through each class and ensure that there are viable builds for each and make adjustments as necessary.

    You may notice the wording: "the FIRST round of tests [...] will FOCUS on AoE abilities in Cyrodiil", which can imply:
    - there will be more tests of that kind
    - not only AoE abilities can be affected
    - not only Cyrodiil can be affected

    And from the various topics around, I 'm quite sure I'm not the only one having lag out of Cyrodiil.

    I ask again, WHERE does it says imposing AoE cooldowns is their idea to improve performance. I didnt ask what measures they are implementing to test performance impact

    Actually it says ZoS optimized performance on ability casting as much as they could, and that anyway players shouldn't be able to cast so many AoEs as now, as it was not their initial design, and as it results in lag when too many AoE effects happen in limited place.

    Their intention is clearly to nerf AoEs in a near future and limit their casting.

    Hard-coded cooldown is clearly an efficient way to do so, and it is exactly what they are testing. Even test 3 gives you a cumulative resource debuff for 5s that applies to all AoEs, so at some point you would have to wait to be able to cast anything.

    Read the post again, it's quite clear:
    However, one foundational issue remains. At some point, we crossed a threshold where most players in PvP were able to cast endless AOE abilities, without ever running out of resources. [...]

    This is not what we intended, but [...] we allowed it. However, as this behavior grew more prevalent, we reached a point where casting so many continuous AOE abilities in such a small area started to overwhelm the server process for that area, leading to situations where the "lag meter" spikes and the server becomes unresponsive for a period of time.

    [...] we have reached the point where we cannot fix this issue around the edges: we need to address the core problem, which we will be starting with Update 27.

    After we complete the above tests, we may try other combinations of cooldown, cost, and regen values on AOE abilities. [...] be aware that if these tests confirm our hypothesis, then chain-casting AOE abilities will no longer form the core of the ESO PvP experience in the way it has for the last few years. We would then go through each class and ensure that there are viable builds for each and make adjustments as necessary.

    So, from what M. Lambert says, AoEs are responsible for the lag in Cyrodiil, so AoE will soon have significant cost increase and/or hard-coded cooldown (What M. Lambert doesn't say, is how laggy the game can also be outside of Cyrodiil, and how his team would fix that).

    And if this does still not answer your question... who's idea could it be if not ZoS' ?
    Edited by MerguezMan on August 6, 2020 6:29PM
  • Gorreck
    Gorreck
    ✭✭✭
    Gorreck wrote: »
    SodanTok wrote: »
    Where is it stated that their best idea to improve performance is to add cooldowns? Where?


    The are testing on the PTS now and its going live to "test" next patch (notes are on the PTS forum, I think).

    If it ends up sticking many builds and some classes are likely to be decimated.

    The fact that they're testing their hypothesis about performance using cooldowns doesn't really indicate anything about how they plan to solve the problem if their hypothesis is confirmed.

    If you actually read the rest of the announcement, it seems pretty clear that changes to AoE ability costs and/or resource regeneration (particularly for well-geared, well-coordinated, high-CP groups) are likely things that they plan to look at for their final solution.

    I think the cooldowns are just a quick way for the performance team to experiment with reduced AoE without making the balance team do a ton of (potentially unnecessary) work first.



    If it works, they'll impliment it as is, mark my words.

    Nothing in my experience of this game suggests otherwise.
  • ealdwin
    ealdwin
    ✭✭✭✭✭
    Gorreck wrote: »
    SodanTok wrote: »
    Where is it stated that their best idea to improve performance is to add cooldowns? Where?


    The are testing on the PTS now and its going live to "test" next patch (notes are on the PTS forum, I think).

    If it ends up sticking many builds and some classes are likely to be decimated.

    The fact that they're testing their hypothesis about performance using cooldowns doesn't really indicate anything about how they plan to solve the problem if their hypothesis is confirmed.

    If you actually read the rest of the announcement, it seems pretty clear that changes to AoE ability costs and/or resource regeneration (particularly for well-geared, well-coordinated, high-CP groups) are likely things that they plan to look at for their final solution.

    I think the cooldowns are just a quick way for the performance team to experiment with reduced AoE without making the balance team do a ton of (potentially unnecessary) work first.

    Agreed.

    I might be wrong, but it seems like that the cooldowns and ramping cost increases are not necessarily the end solution, but rather a method of forcing limitations on AoEs in Cyrodiil for the purpose of testing. ZoS can't exactly just tell people to "please, if you may, stop using AoEs for the next week in Cyrodiil"; no one would listen. So, instead they opted for implementing some measures to force different usages of AoEs.

    The tests are less about potential solutions, but more so about testing a few things. (1) If AoEs are indeed the solution. If there is no performance difference between Week 1 (High restriction on AoEs) and Week 3 (Less restriction on AoEs), then ZoS knows that AoEs are not the problem. If they are a problem, and less AoEs helps performance, then the tests tell them (2) what level of decreasing AoEs ZoS needs to look at. If Week 3 is enough to fix performance issues, then ZoS starts looking at minor adaptations to skills to achieve that level of performance. Or, if Week 3 is not enough, and Week 1 was the best performance, then ZoS is probably going to look at reworking skills to decrease the number of abilities that are AoE.

    The last thing that ZoS wants to do is to do a complete overhaul of skills with the intent of fixing performance, and then have it not work, and have gone through all that work for nothing. I'm not saying these tests are not going to be painful; they will be. But, they do seem to be an attempt by ZoS to test a hypothesis so that they have data to work off of before making any significant changes.
  • SodanTok
    SodanTok
    ✭✭✭✭✭
    ✭✭
    MerguezMan wrote: »
    SodanTok wrote: »
    MerguezMan wrote: »
    SodanTok wrote: »
    Where is it stated that their best idea to improve performance is to add cooldowns? Where?

    Here:
    https://forums.elderscrollsonline.com/en/discussion/539136/update-on-cyrodiil-performance-upcoming-aoe-tests/p1
    The first round of tests we are planning will focus on Area of Effect (AOE) abilities in Cyrodiil and will make it more difficult for AOE abilities to be the only abilities used, adjusting cooldown, cost and regen values of all AOEs (damage and healing).

    Specific details on the tests we will be running in Cyrodiil are as follows:
    • Test 1 – Shared global AOE cooldown - 3 second timer. This test adds a global 3 second shared cooldown to any AOE ability. This means that when you cast an AOE, you will not be able to cast another for 3 seconds. For example, as a Templar, if I cast Ritual of Retribution, I wouldn’t be able to cast Puncturing Sweep for 3 seconds.
    • Test 2 – Individual AOE cooldown - 3 second timer. This test adds an individual AOE cooldown to each AOE ability. This means that when you cast an AOE, you will not be able to cast that same AOE ability for 3 seconds. For example – as a Templar, if I cast Puncturing Sweep, I wouldn’t be able to cast Puncturing Sweep again for 3 seconds.
    • Test 3 – No cooldown, global ramping AOE cost. This test adds a global ramping AOE cost for each AOE cast. Similar to how streak or roll dodge works, where when you cast an AOE you receive a debuff for 5 seconds, each stack of the debuff increases the cost of any AOE cast.
    • Test 4 – Individual AOE cooldown – 3 second timer, global ramping AOE cost. This test adds an individual AOE cooldown to each ability as in test 2, but also combines that with a global ramping cost from test 3.
    During the times that any of these tests are active, we will be awarding double Alliance Points for anyone active in Cyrodiil.

    After we complete the above tests, we may try other combinations of cooldown, cost, and regen values on AOE abilities. However, we need to run these tests first and then assess the data. We will then let everyone know what we found and how we will move forward. We will be very upfront, but please be aware that if these tests confirm our hypothesis, then chain-casting AOE abilities will no longer form the core of the ESO PvP experience in the way it has for the last few years. We would then go through each class and ensure that there are viable builds for each and make adjustments as necessary.

    You may notice the wording: "the FIRST round of tests [...] will FOCUS on AoE abilities in Cyrodiil", which can imply:
    - there will be more tests of that kind
    - not only AoE abilities can be affected
    - not only Cyrodiil can be affected

    And from the various topics around, I 'm quite sure I'm not the only one having lag out of Cyrodiil.

    I ask again, WHERE does it says imposing AoE cooldowns is their idea to improve performance. I didnt ask what measures they are implementing to test performance impact

    Actually it says ZoS optimized performance on ability casting as much as they could, and that anyway players shouldn't be able to cast so many AoEs as now, as it was not their initial design, and as it results in lag when too many AoE effects happen in limited place.

    Their intention is clearly to nerf AoEs in a near future and limit their casting.

    Hard-coded cooldown is clearly an efficient way to do so, and it is exactly what they are testing. Even test 3 gives you a cumulative resource debuff for 5s that applies to all AoEs, so at some point you would have to wait to be able to cast anything.

    Read the post again, it's quite clear:
    However, one foundational issue remains. At some point, we crossed a threshold where most players in PvP were able to cast endless AOE abilities, without ever running out of resources. [...]

    This is not what we intended, but [...] we allowed it. However, as this behavior grew more prevalent, we reached a point where casting so many continuous AOE abilities in such a small area started to overwhelm the server process for that area, leading to situations where the "lag meter" spikes and the server becomes unresponsive for a period of time.

    [...] we have reached the point where we cannot fix this issue around the edges: we need to address the core problem, which we will be starting with Update 27.

    After we complete the above tests, we may try other combinations of cooldown, cost, and regen values on AOE abilities. [...] be aware that if these tests confirm our hypothesis, then chain-casting AOE abilities will no longer form the core of the ESO PvP experience in the way it has for the last few years. We would then go through each class and ensure that there are viable builds for each and make adjustments as necessary.

    So, from what M. Lambert says, AoEs are responsible for the lag in Cyrodiil, so AoE will soon have significant cost increase and/or hard-coded cooldown (What M. Lambert doesn't say, is how laggy the game can also be outside of Cyrodiil, and how his team would fix that).

    And if this does still not answer your question... who's idea could it be if not ZoS' ?

    Chain casting limitations were set upon more than few skills in this game to this day. Yet none with hard coded cooldowns. Meanwhile their test states only that if the hypothesis they are TESTING with hard coded cooldowns proves to be true they will make AoEs chain casting not part of ESO combat anymore. To assume the testing implementations used in test1,2 and 4 will be used to enact this new vision of combat and not implementation of test #3 aka something used in past to affect chain casting on skills like dodge roll or different options that fit ESO combat like general increase to cost or redesign of skills that benefit of chain casting is very very warped and wrong view.

    They can easily rebalance all AoE skills to offer long but weak benefits. Not claiming this would solve performance problem or even help it but its up there with all the other possibilities, one of which and frankly most unlikely, is adding cooldowns to AoE skills.
    Edited by SodanTok on August 6, 2020 11:39PM
  • MerguezMan
    MerguezMan
    ✭✭✭✭
    ealdwin wrote: »
    I might be wrong, but it seems like that the cooldowns and ramping cost increases are not necessarily the end solution, but rather a method of forcing limitations on AoEs in Cyrodiil for the purpose of testing. ZoS can't exactly just tell people to "please, if you may, stop using AoEs for the next week in Cyrodiil"; no one would listen. So, instead they opted for implementing some measures to force different usages of AoEs.

    The tests are less about potential solutions, but more so about testing a few things. (1) If AoEs are indeed the solution. If there is no performance difference between Week 1 (High restriction on AoEs) and Week 3 (Less restriction on AoEs), then ZoS knows that AoEs are not the problem. If they are a problem, and less AoEs helps performance, then the tests tell them (2) what level of decreasing AoEs ZoS needs to look at. If Week 3 is enough to fix performance issues, then ZoS starts looking at minor adaptations to skills to achieve that level of performance. Or, if Week 3 is not enough, and Week 1 was the best performance, then ZoS is probably going to look at reworking skills to decrease the number of abilities that are AoE.

    The last thing that ZoS wants to do is to do a complete overhaul of skills with the intent of fixing performance, and then have it not work, and have gone through all that work for nothing. I'm not saying these tests are not going to be painful; they will be. But, they do seem to be an attempt by ZoS to test a hypothesis so that they have data to work off of before making any significant changes.
    SodanTok wrote: »
    Chain casting limitations were set upon more than few skills in this game to this day. Yet none with hard coded cooldowns. Meanwhile their test states only that if the hypothesis they are TESTING with hard coded cooldowns proves to be true they will make AoEs chain casting not part of ESO combat anymore. To assume the testing implementations used in test1,2 and 4 will be used to enact this new vision of combat and not implementation of test #3 aka something used in past to affect chain casting on skills like dodge roll or different options that fit ESO combat like general increase to cost or redesign of skills that benefit of chain casting is very very warped and wrong view.

    They can easily rebalance all AoE skills to offer long but weak benefits. Not claiming this would solve performance problem or even help it but its up there with all the other possibilities, one of which and frankly most unlikely, is adding cooldowns to AoE skills.

    Guys, face the thing:

    1) Would highly reducing the use of AoE relatively improve performance ?
    YES. Can hardly tell by how much it would improve performance, but it will.
    In previous updates, we've seen changes made to various things, including the bash system, and even if it's probably not bashes that crash the game, it went live. Chances that any relative improvement will be taken are high.

    2) Why test cooldowns on at least 3 weeks on live servers, if that is only for metrics ?
    Would you take the nurse and a box of condoms to the bedroom just to take metrics ? (that sounds totally innocent)
    ZoS can easily know how many AoEs are used now, and what part of server stress it causes for relative calculations. This information goes through the server, and is easily recorded.
    Zos can hardly tell for each possible class, build, player preferences, what woud happen exactly by limiting AoE use. Players may find other ways to trigger equivalent effects, or may turn to single-target spammables that would have significant impact on performance when intensively used (as you don't use that AoE attack, you may use 3 times more that ST attack), or even sets AoE effects.

    3) Can the end result be just a weaker version of current effects ?
    NO. If a skill guarantees you AoE damage/buff/debuff/heal/heal-back, and is (one of) your main spammable(s), setting its effect 10% less than now will not prevent you from using it, and would not reduce server calculations (10% or 20% in a formula does not change the number of operations).
    ie. Being stamina, what would you use to heal instead of Vigor ?
    Some classes will be hit way harder by the change than others. The list of skills that will be impacted by the first batch of tests is already in PTS patch notes.

    4) Can the end result just be a cost increase ? Or something else ?
    MAYBE. But then, that would be a big cost increase, to prevent players going through that limitation using group mechanics, CP, or sets. The end result of such change is that you would avoid to use such high cost skills to preserve resources.
    ie. Would you spam "Siege weapon shield" if it lasts 3 seconds for the same cost ?
    Limiting the number of targets of your AoE may be a solution, but there are 2 negative effects: it may not reduce so much calculations, as the system would still have to check possible targets positions, status, conditions, AND it randomizes the effect, which can feel weird for the player.
    ie. You face 3 enemies, use a capped 2-target effect, the enemy closer to you which is supposedly bigger threat may not be impacted.
    There are not thousands of options to limit skill casting and prevent players from finding a legit workaround. Implementing cooldowns is a reliable one (even if we'd prefer something else).

    Anyway, tests start in 2 weeks, we should have dev feedback in about 2 months. And then we'll have patch notes...
  • ArcVelarian
    ArcVelarian
    ✭✭✭✭✭
    Two words: Better Servers.
    Murphy's Law of PvP : If it can be abused and or exploited, it will be abused and or exploited.
  • Crow_IX
    Crow_IX
    ✭✭✭
    MerguezMan wrote: »
    Dear Zenimax,

    The cooldowns implementation will for sure improve server performance. Having less players would also improve server performance, but I guess it also means you get less money, as players would not spend on crown crates if they don't play the game anymore...

    Dear players,

    The server performance is bad. So far, the best idea ZoS has is to add cooldowns. May we suggest something else ?

    this isn't completely true as experienced in live servers in the early hours or even latest hours, while populations are 1 bar per faction, server stability is still very poor. meaning having less population wouldn't exactly improve anything. nor do i believe aoe's will fix or improve anything as we have had massive aoe metas before with way better performance than we have now. i truly believe it is a coding issue rather than a server overload issue, we have experienced a massive decline in server performance after 1: the file compression update and 2: the addition of cast times. these are just my thoughts though.
    RIP skill based PvP days. . .
  • Mortiis13
    Mortiis13
    ✭✭✭✭
    Only problem I have is that zos isn't known for slight adjustments and takingvPlayer feedback serious , even from those who have an excellent understanding about math and how the game works practically.
    Edited by Mortiis13 on August 8, 2020 10:29AM
  • p00tx
    p00tx
    ✭✭✭✭✭
    ✭✭
    Pets are already disabled in Cyrodil, and the performance in there is still hot garbage. Try going into the snowy north in Cyrodil, then try going to the south, where there is no weather change. There is a massive performance difference, even when you're completely alone. Keeps change shape from gaining or losing adjacent resources, the game slows to a crawl and half of the server disconnects. The floors drop out on final boss in vKA on Xbox and half of the team disconnects (if there is a nightblade on the team. Super weird). Any environmental changes trigger severe frame loss and widespread disconnects. There is obviously something super broken with their environmental coding.

    I guess with that in mind though, Zos should continue going after straw men like CP and AoEs.
    PC/Xbox NA Mindmender|Swashbuckler Supreme|Planes Breaker|Dawnbringer|Godslayer|Immortal Redeemer|Gryphon Heart|Tick-tock Tormentor|Dro-m'Athra Destroyer|Stormproof|Grand Overlord|Grand Mastercrafter|Master Grappler|Tamriel Hero
Sign In or Register to comment.