Changes Focused on Boosting Game Performance

[CeltEireson]

Point taken, though that's only true at 100% DS.

At 75% DS, old v new is 5.75 v 5.25, at 50% it's 5.5 v 4.5, and at 0% it's 5 v 3...

With the new changes I'd say most folk who specialise in archery (except maybe those with no past lives and no set bonus) will be well over 50% doubleshot by level 20 - 9% past lives, 30% passive bonus from many shot (assuming full BAB), 20% or more from trees, 10% from wallwatch set, 5-7% from item. By level 30 I'd expect most to be over 100%. Though some current builds might struggle to get a good level of doubleshot.

[elvesunited]

There's a lot of vitriol on the performance issue. Let's keep a few things in mind.

1) The devs did not just wake up one morning and decide to make up a lag performance issue just to mess with the player base.
2) This obviously is not the only source of lag
3) When this sort of lag does happen it affects not just the instance but the whole server.
4) As someone who has worked with a lot of legacy software systems I can say that often times you can't fix the issue through hardware and while a massive software overhaul might be able to fix, that can be extremely risky, expensive, and time consuming. ( as in years )


So let's work with what we got. The issue is multiple-procs caused by rapid hits on enemies causing the server queues to back up. They need a reduction of these procs so the options of giving it multiple chances to happen and giving them an increased chance of happening are both not feasible. The issue as described is about procs being applied to enemies. It's quite possible that procs that affect the character when he strikes an enemy aren't nearly as lag producing as those that affect the enemy. The recent news they plan to increase the chance of adrenaline recharge suggests that to be the case. Damage causing procs simply had their damage multiplied. Non damaging procs can probably be improved in a similar fashion by increasing DC or duration or effect based on the number of hits. I get that there are hundreds of these non-damaging procs out there and updating them all would be a herculean task but there are only a handful of commonly used heavy hitters that players really care about. Update those.


For the proposed adrenaline/hunt's end update of lowering the added damage, I get that the amount of damage being pulled down by adrenaline shots and doublestrikes is getting out of hand. ( 850K? ) That said I don't think lowering the damage percent of the adrenaline is the right move. Right now if I'm doing the math right, you get about x10 damage on a double hit. x15 on a triple etc.

( [ weapon damage ] x 1 for single hit ) x 5 for adreline/hunt's end -> x5 damage
( [ weapon damage ] x 2 for double hit ) x 5 for adreline/hunt's end -> x10 damage
( [ weapon damage ] x 3 for triple hit ) x 5 for adreline/hunt's end -> x15 damage

If you lower the percent added damage to +200% that reduces the damage increase to x6 damage on a double hit x9 damage on a triple etc.

( [ weapon damage ] x 1 for single hit ) x 3 for adreline/hunt's end -> x3 damage
( [ weapon damage ] x 2 for double hit ) x 3 for adreline/hunt's end -> x6 damage
( [ weapon damage ] x 3 for triple hit ) x 3 for adreline/hunt's end -> x9 damage

The trouble is that is a nerf for characters with low amounts of doublestrike/doublehit ( like most crossbow users ) Do you really want to make getting large amounts of doublestrike or doubleshot a requirement for getting the highest usage out of adrenaline/hunt's end? Or can we instead add the damage increase? ( Example: +100% for double hit and +400% for adrenaline/hunt's end makes +500% )

[ weapon damage ] x 5 for single hit and adrenaline/hunt's end combined -> x5 damage
[ weapon damage ] x 6 for double hit and adrenaline/hunt's end combined -> x6 damage
[ weapon damage ] x 7 for triple hit and adrenaline/hunt's end combined -> x7 damage

[Pandjed]

The trouble is that is a nerf for characters with low amounts of doublestrike/doublehit ( like most crossbow users ) Do you really want to make getting large amounts of doublestrike or doubleshot a requirement for getting the highest usage out of adrenaline/hunt's end? Or can we instead add the damage increase? ( Example: +100% for double hit and +400% for adrenaline/hunt's end makes +500% )

[ weapon damage ] x 5 for single hit and adrenaline/hunt's end combined -> x5 damage
[ weapon damage ] x 6 for double hit and adrenaline/hunt's end combined -> x6 damage
[ weapon damage ] x 7 for triple hit and adrenaline/hunt's end combined -> x7 damage

If technical possible, they could definitely ignore the multi-hit cap while under the effect of adrenaline or hunt's end, add 400% multi-hit and therefore the scaling would happen on their own, though it would also multiply proc damage, which may not work as intended and would need some more tweaking of the code to include another stacking multi-hit variable only to the base damage. Don't know if it makes stuff worse...

[Malusny]

The 400% to 200% reduction is, yet again, an overshoot compensation that disadvantages players...

As things stand, an adrenaline hit gives 400% damage, plus a DS chance at a non-adrenaline double hit.

With this new proposal you get 200% damage X (1+DS). That's it.

So, just to make up the 200% difference, you need to have hit 100% DS.

Firstly not everyone has 100% DS. I can only achieve that in reaper with the DS boost. So it's a nerf in any non-reaper quest, such as a raid.

Secondly, even if they do have 100% DS, they are still not being compensated for the loss of the extra DS hit they get with current. Effectively, compared to current, the effect of current DS is being removed entirely for an adrenaline hit.

All DS will effectively be used to compensate for the 200% damage nerf, even if you do have 100% DS.

Gain a benefit from these changes? Ooh! Can't have that. We'll nerf you back - AND THEN SOME!

But if you're disadvantaged by these changes? Too bad, they're necessary, deal with it.

Again, an over-hasty, ill-considered, over-zealous nerf that debits players. Why do we never seem to see changes that favour us - why does every solution seem to have to hurt?

I'm sick of this.

I completely agree with you on this. The adrenaline faster recharge rate is not a real problem. Having 100% doublestrike is actually a big problem, can be achieved only in certain classes/enhancements/items all lined up towards it and the net loss of DPS along side of losing 50% procs is a disaster in my opinion.

2-handed builds and some ranged builds are the only fury of the wild users. When it comes to ranged builds it is a different story but for melee's 2-handed builds that will be a disaster since they already got a nerf this year from ability to dmg reduced from 3.0 to 2.5.

Not sure why nerfing 2-handers so much - less procs, less ability to damage (3.0 to 2.5 which is about 16% damage loss overall) and now adrenaline getting nerfed for those who cannot reach 100% doublestrike.


Devs , how are you going to explain that to the players who don't have all the past life and cannot reach even 70 or 80% doublestrike?

Cheers.

ps. I hope you guys reconsider this.

I would keep adrenaline at 300% to have some kind of a balance.

[Malusny]

I completely agree with you on this. The adrenaline faster recharge rate is not a real problem. Having 100% doublestrike is actually a big problem, can be achieved only in certain classes/enhancements/items all lined up towards it and the net loss of DPS along side of losing 50% procs is a disaster in my opinion.

2-handed builds and some ranged builds are the only fury of the wild users. When it comes to ranged builds it is a different story but for melee's 2-handed builds that will be a disaster since they already got a nerf this year from ability to dmg reduced from 3.0 to 2.5.

Not sure why nerfing 2-handers so much - less procs, less ability to damage (3.0 to 2.5 which is about 16% damage loss overall) and now adrenaline getting nerfed for those who cannot reach 100% doublestrike.


Devs , how are you going to explain that to the players who don't have all the past life and cannot reach even 70 or 80% doublestrike?

Cheers.

ps. I hope you guys reconsider this.

I would keep adrenaline at 300% to have some kind of a balance.

Maybe bringing back the 3.0 value on THF would be enough of a compensation?

That would probably allow the less experienced players not to feel that nerf so much if they cannot hit 100% doublestrike?

Cheers.

[SierraSilverfox]

Actually you are being compensated for the extra DS hit, currently with +400% you do 5 + 1 = 6 times damage, with +200% you will do 3 x 2 = 6, and will average higher because of the higher crit chance on the Adrenaline hit.

as it is currently in live, you -do- benefit from adrenaline on all doubleshots fired as you benefit from skills being shot. this WILL nerf a shiradi build from live to the new update by 200% on hunt's end, and you receive nothing in compensation, just like all non damaging procs are going to be obsolete. as it is in live currently, if i hit 300% doubleshot, pop hunt's end and sniper shot, i can drop 4 very high damage crits. that damage is being cut down drastically if hunt's end/adrenaline is cut from 400% to 200%, and honestly all i can say is, it seems that SSG just wants casters to be god of all.

[Malusny]

as it is currently in live, you -do- benefit from adrenaline on all doubleshots fired as you benefit from skills being shot. this WILL nerf a shiradi build from live to the new update by 200% on hunt's end, and you receive nothing in compensation, just like all non damaging procs are going to be obsolete. as it is in live currently, if i hit 300% doubleshot, pop hunt's end and sniper shot, i can drop 4 very high damage crits. that damage is being cut down drastically if hunt's end/adrenaline is cut from 400% to 200%, and honestly all i can say is, it seems that SSG just wants casters to be god of all.

Precisely,

2-handers are getting nerfed all across for some time already.

Performance issues is one thing - good job Devs on working/fixing it... but at the same time you are clearly nerfing melee builds again, 2-handers the most.

[Valerianus]

as it is currently in live, you -do- benefit from adrenaline on all doubleshots fired as you benefit from skills being shot. this WILL nerf a shiradi build from live to the new update by 200% on hunt's end, and you receive nothing in compensation, just like all non damaging procs are going to be obsolete. as it is in live currently, if i hit 300% doubleshot, pop hunt's end and sniper shot, i can drop 4 very high damage crits. that damage is being cut down drastically if hunt's end/adrenaline is cut from 400% to 200%, and honestly all i can say is, it seems that SSG just wants casters to be god of all.

i bet whatever you want* that multivial and meteor swarm will be nerfed again


(*terms and conditions apply. what you want is not to be intended as what you want)

[Epicsoul]

Couldn't that be an example of two different types of lag? I mean yes, certain mob types cause lag more than others but more mobs always equals more lag for me and more mobs of the mob type that is extra laggy will only make it worse. So for example if 4 bugbears doesn't cause me lag but 20 does this doesn't negate that 4 orthons by themselves might cause the game issues because their mob type just has extra stuff causing lag so 20 orthons would be devastating.

Oh I thought of an example:

When I'm fighting a few oozes everything is okay but if my son comes up and cleaves and splits them all suddenly the game lags for a second or two as untold number of oozes spawn. We ran Wizard of Wines recently and this happened at least twice. More mobs = more lag for me.

That's fair. Maybe I should've said there is a correlation, I'm just not sure how much of one. I agree though, more mobs trying to spawn/path/process does increase lag.

[SierraSilverfox]

i bet whatever you want* that multivial and meteor swarm will be nerfed again


(*terms and conditions apply. what you want is not to be intended as what you want)

all that will do is make other casters take over, lol. i'm not even sure where they get their data from, my highest crit on lam with my own self contained set of debuffs was like 100k, just counting fetters. that'd assuming that i have my full debuffs stacked up, meanwhile on live on R8 i popped a chain of 3 17k's and have seen over 150k per individual shot outside of reaper, which assuming 3 hits is 450k, and that was also pretty self contained with my own debuffs and not factoring in full raid debuffs. passive damage is nice, but, sometimes you need that burst damage and while i'm perfectly fine with the changes to manyshot since i do get all of that and more back as passive increase plus a special attack, the damage it does to non-damage procs from Shiradi and other effects along with this supposed nerf to hunt's end just kinda sticks archers behind where they are on live now, and really doesn't correlate to a "buff" in any sense of the word. and yeah, the biggest lag i've seen is from mass aoe in large groups of mobs, as seen when warlocks first caused it using Evards in ES builds that could literally lock up an instance until they die, and also the lag in places like Shroud, neither of these which are being addressed by the doubleshot/doublestrike change, and that's really only going after one potential source of a widescale problem, and it's little individual patches on stuff that causes problems to not really get fixed. it's like putting a bandaid over an infected patch of skin, and not really trying to treat the injury that may need actual stitches. ultimately i'm both happy and torn about this update, and that's a pretty love/hate relationship. but hey, if they don't give anything back to Shiradi's procs to account for the reduced ability to reliably use all of the actual fun stuff in the tree, which yeah, damage is fine but that isn't the fun stuff, it'll just end up being better for chain missile builds on wizards because proccing 20+ missiles per cast in aoe is way better CC than 1 single arrow per hit, especially when you can just cycle SLA and non SLA, heh. Oh, i guess they'll go nerf those spells to the vestiges of hell next since i mentioned it, but hey, maybe they'll at least boost the damage so they're competent with other at-level spells, which they are not currently. it's whatever, though.

[songswrath]

if the amount of procs (die rolls) are a major cause of lag. Why not get rid of the whole confirm to critical hits. That has to save 1/3 of all melee die rolled ?? and change the default setting to hide all combat logs

[DjangoKeys]

So...any word on that third Lam preview?

Are they going to wait until next week to open, or are they going to do it this week with little-to-no advanced warning? I mean, timely announcements aren’t really a thing here, so I figure it’s 50/50 either way at this point.

[Bjond]

Moving calculation invariants to their point of origin should provide the best improvement and can be done without any side effects if you're willing to put enough effort into pre-calculating mult-hit chances correctly.

However, there might be a cheaper way that does enough. Your server might have a HW RNG source. I started seeing those in embedded processors about 15~20yrs ago and they're often far cheaper to use as a random source than any pure SW solution: often a single read instead of 3xread+mult+add+mod+store (assuming LGA, not including instruction reads); in principal, at least 7x improvement for what is likely the hottest code on the server.

Your only issue might be access. Depending on the level of control you have over the server -- ie. using your own OS (which I'd guess is unlikely) versus something else. You might have have to convince the OS to give you direct access; eg. map a device into your address space or alter privileges to permit use of instructions or registers your OS normally protects.

This could be trivial or painful depending on the OS. It wouldn't be worth using if you had to take a system trap or make a driver call or if the HW RNG takes excessive effort to draw from, but my bet is you'd likely see a noticeable speedup immediately without making any other changes if you could reduce all server RNG calls to a single read.

[elvesunited]

So...any word on that third Lam preview?

Are they going to wait until next week to open, or are they going to do it this week with little-to-no advanced warning? I mean, timely announcements aren’t really a thing here, so I figure it’s 50/50 either way at this point.

The DDOcast said there would be a third. The proposed changes for adenaline/hunt's end pretty much demand one.

[mikarddo]

Well it looks like we get another go at it. From the DDOcast, there is going to be a third update lamania release.

The big change will be to adrenaline and hunt's end which are proposed to go down from 400% damage to 200%
with the recharge chance being changed to 100% on a vorpal ( at least for melee )

We'll have to see if they go after any of the other non-damage procs.

I am guessing that will be a slightly bad deal for melee. Getting 100% regen once per hit rather than 2 times 1/3 means more Adrenalines - but getting +200% rather than +400% means only 3/5 the damage on each. Total looks like it will be 1*3 = 3 (new) rather than 2/3*5 = 3.3 (current). A slight lowering of damage but with just advantage in being able to use Adrenaline more often for CC.

For throwers using Adrenaline the story is worse as they had 4-5 hits already. So getting 100% for one hit is worse than 33% for 4-5 hits. So, thats fewer Adrenalines that each do only 3/5 the damage. A major nerf with no upside.

[TPICKRELL]

I am guessing that will be a slightly bad deal for melee. Getting 100% regen once per hit rather than 2 times 1/3 means more Adrenalines - but getting +200% rather than +400% means only 3/5 the damage on each. Total looks like it will be 1*3 = 3 (new) rather than 2/3*5 = 3.3 (current). A slight lowering of damage but with just advantage in being able to use Adrenaline more often for CC.

For throwers using Adrenaline the story is worse as they had 4-5 hits already. So getting 100% for one hit is worse than 33% for 4-5 hits. So, thats fewer Adrenalines that each do only 3/5 the damage. A major nerf with no upside.

The problem is that many of us use Adrenaline as a strategic tool to quickly take out a single dangerous crown or reaper. More frequent but lower numbers significantly nerfs this tool. It means that you have significantly more time with the dangerous mobs aggro before it is dead. While the overall damage is nerfed a bit, the usefulness is nerfed a lot more.

[Firebreed]

Excuse me, but I seem to be misunderstanding something. Forget Adrenaline for a second.

The DDOcast news is Hunt's End is going down from 400% to 200%? Hunt's end has no recharge rate for them to buff, are they straight up meganerfing the ability?

What the?

[mikarddo]

The problem is that many of us use Adrenaline as a strategic tool to quickly take out a single dangerous crown or reaper. More frequent but lower numbers significantly nerfs this tool. It means that you have significantly more time with the dangerous mobs aggro before it is dead. While the overall damage is nerfed a bit, the usefulness is nerfed a lot more.

A fair point. You do get some value from being able to cc more often using Adrenalines (due to the almost sure crits) but your point is still value ofcourse. Also, if you can still achieve the same effect with just +200% because you were overkilling before you will simply get to do it more often - I doubt thats the case where it really matters but it could be in some cases.

That does not change that the nerf to throwers (both Adrenaline and Hunts End) would be much worse, do you agree?

[TPICKRELL]

A fair point. You do get some value from being able to cc more often using Adrenalines (due to the almost sure crits) but your point is still value ofcourse.

That does not change that the nerf to throwers (both Adrenaline and Hunts End) would be much worse, do you agree?

I do agree that throwers get it worse, but they are not in melee range so its somewhat more manageable. The main point is that it is a significant nerf to everyone using Adrenaline! Despite the Devs claim of DPS neutrality.

[elvesunited]

I am guessing that will be a slightly bad deal for melee. Getting 100% regen once per hit rather than 2 times 1/3 means more Adrenalines - but getting +200% rather than +400% means only 3/5 the damage on each. Total looks like it will be 1*3 = 3 (new) rather than 2/3*5 = 3.3 (current). A slight lowering of damage but with just advantage in being able to use Adrenaline more often for CC.

For throwers using Adrenaline the story is worse as they had 4-5 hits already. So getting 100% for one hit is worse than 33% for 4-5 hits. So, thats fewer Adrenalines that each do only 3/5 the damage. A major nerf with no upside.

The trouble is devs have to do something. 850K on an adrenaline fueled doublestrike? We'd be exceptionally naive if we thought that was going to live. The trouble is that this nerfs those builds with low levels of doublestrike or doubleshot. This is why my suggestion is that doublestrike/doubleshot multipler be added to the adrenaline/hunt's end multiplier rather than multiplied by it.

damage x ( 1 for single hit + 4 for adrenaline/Hunt's end ) = damage x 5
damage x ( 2 for double hit + 4 for adrenaline/hunt's end ) = damage x 6
damage x ( 3 for triple hit + 4 for adrenaline/hunt's end ) = damage x 7

instead of

( damage x ( 1 for single hit ) ) x (3 for adrenaline/hunt's end ) = damage x 3
( damage x ( 2 for double hit ) ) x (3 for adrenaline/hunt's end ) = damage x 6
( damage x ( 3 for triple hit ) ) x (3 for adrenaline/hunt's end ) = damage x 9

A character shouldn't be beholden to have large amounts of doublestrike or doubleshot to get full benefit out of adrenaline strike/hunt's end. That would badly favor some builds over other builds. And this also helps limit the astronomical numbers we saw in the last lamannia update better than lowering the damage increase to 200%.

Note: Regardless. increasing the recharge rate of adrenaline is a good thing and probably not lag producing as this is proc that affects the character not the enemy though devs would need to confirm if that is true.