DiabloFans

Hi everyone,

I just wanted to get some advice on the playstyle of dmo engery twister.

Me and my friends completed our Group Setup (monk, dmo, globe barb, WD) and tried grifts starting at 75, tooks us 35 minutes.. it was horrible. I was always resource starved and especially the rift guardian took forever.

I get it this setup is for pushing grifts not for farming. We need a perfekt rift with super high densitiy to my twisters can wreak havoc.

We were optimizing and learning to play this but we always had cave maps with almost nothing in it. So most of the time we were not able to finish in time.

When we got a good rift, we went through in a breeze. However even if we finished the trash after 5 minutes on 75 , it took me 5 minutes to down the guardian.

After that we swapped the support WD to a LON Crusader.

With this setup we were able to complete 79 and are going higher today.

So my questions now:

How to kill rifht guardians with engery twister DMO Group spec? (only dps in group)

Is this really a push build depending super hard on the grift level (density) and is thus not really viable for speed farming?

If yes, what would be a viable setup for speedfarming?

iam now 550 . not seen one

got one from a friend though

neat

you're not the first one with the idea of paragon brackets, nevertheless you wrote it out very well.

Also i like this idea very much and from a technical side it is not hard to implement.

The only fukin argument blizzard will come with against this - if they even bother commenting on any of these - it increases ui clutter und confuses new players bla bla..

iam totally on your side bagstone, however others have other oppinions.

Your argument is undeniably correct - and i cannot see me not agree with this argument - and its implication, that it's hard for me to understand how others can ignore this problem or argument against it.

This seems to be the case for you also since you wouldn't repeat yourself

The problem is many people cant see more than their own little world. If its not affecting them (by not pushing leader boards) the argument is invalid. That is their point of view and how we perceive it. However what they really mean is: is does not affect them and thus they dont care.

You cannot help someone who does not want to be helped. He cannot see his advantage in changing the situation. So why try to convince a lost cause?

You made your point, its valid.

But sadly its really not affecting the larger part of the playerbase. So , yes, the problem is not realized in the major part of the community so blizzard does not think its a problem.

Just take hearthstone as an example: i !think! that every registred user in this board has at least played hearthstone, and 50% who are regular posters have also been rank 20 or higher in hearthstone.

However with rank 20 you already are better than 50% of all players.

So with this in mind, how representative is the discussion in this board in relation to the playerbase? non existent.

And even here its only a few people who are saying its affecting them.

That is because one reason: There are only a few people competing for the high ranks really doing this extreme paragon farming.

I would say that its not more than 500-1000 people globally who are really trying it.

How i get my number? Each leaderboard has 1000 places, 80% of them are happy with rank 200 >1000

why 80%? well because iam easy 200 and i dont care to do more.

Its not like i have a chance with a job and a girlfriend etc etc.

Would I like a chance to be participating in the top 200 ? yes of course. But this is not because the game is limiting my possibilities.

Time is the limiting factor now.

And paragon gives people with time an advantage.

Thats all thats to say. D

Ultimatley this thread is about changing the game in a way to give people with a real life the opportunity to compete in d3.

Bagstone, its us who are whining that there is a problem with paragon, not them being ignorant of a problem.

We care, but we cannot totaly forget our own advantage. I already said, its a problem but would you still see this as a Problem if you had 24 /h a day for playing d3? you would jump the advantage train at level dem shit outta dose paragons:D

I would..

So if you say: why are all people ignoring this problem, ask yourself. would it be still your problem if your personal circumstances changed? (e.g having no real life anymore)

So its not a flaw of the game, but a result in the inequality of reallife - everybody has different amount of times.

That being sad:

its all not counting in bots. BOTTING HAS TO DISSAPPEAR.

But this problem still exists without botting., but without botting i wouldnt see it as a problem. Because well if gabynator had to be 24h before his pc since theres no botting, he would forget to eat, that would be nice.

i want you to elaborate nachten

i guarantee you that there will be no patch to resolve this before the next season, because its not easy to fix.

haha thats not as bad as previous exploints but still this is kinda sad

GR 10 ? timer expired? how is this even possible?

i dunno much of it is experience... In general i port 30 seconds into a map and look whats going on there. If it looks promising i just start killing.

Any groups smaller then 15~20 mobs is not worth it. especially at 60+

Quote from BlacKcuD»

Quote from Jamoose»

*TL:DR at the bottom*

The game is currently almost unplayable in high greater rifts like 60+ or so and especially in groups of 4 players. I’m not sure if this is the case on all realms but at least on EU it is.

Some builds suffer very badly from this even in solo like the Helltooth WD for an example. Some people claim that if you have 0% area damage then you won't lag in solo but that's just a workaround and not a fix.

In the past, the argument against investing time into fixing this issue was the fact that only a very small percentage of the player base experienced it because it was only evident on high end greater rifts. But today, the lag starts occurring even on 60+ which is not considered that high anymore and i'm pretty sure that more and more people will get to that level and experience the horrible lag.

To the people that are rushing to post things like "no lag here lol noob", please understand that this issue is not always reproducible as it really depends on several factors like the time you play at, what build(s) you are running, the map layout / type you're at and the mob type you are facing. A possible solution to this problem would be to divide all the numbers in game by 100 and display them in the client as strings with added 0's just to keep the consistency.

TL:DR - The game lags horribly on high gr and a possible fix would be to simply divide all the damage and health points numbers in game by 100 so the server calculations, database queries and server-client traffic won’t be as heavy anymore.

Posted also on:

Official forums (US): http://us.battle.net/d3/en/forum/topic/18695103869#1

Official forums (EU): http://eu.battle.net/d3/en/forum/topic/15162513563#1

Reddit: https://www.reddit.com/r/Diablo/comments/3j368b/fix_the_lag/

tl;dr you are wrong/misinformed.
The size of the numbers doesn't matter. The amount of numbers matter, e.g. how certain effects are calculated. Some culprits have already been identified, e.g. Helltooth+Acid Rain. This has been mentioned couple of times in the Reddit AMA.
Source

IF you had read the thread you would have seen that we already talked about this in depth . no need to repeat what already has been said.

Quote from Wyand1337»

Quote from Jamoose»

So more players hitting more mobs with many different types of attacks that have procs like area damage, along with constant damage over time ticking, all of that creates insane lag. It looks like their only chance to fix this shit would be to actually design the game mechanics differently from now on i.e get rid of most DOTs and limit procs to once per x amount of time etc...

I just wish they address this problem as a whole and not just provide small fixes that are usually addressing only certain a build.

Yes, but my guess would be that compared to area damage even DoTs are no big problem.
But their approach to DoTs is another issue. Watching the hp-bars of mobs drop, while a DoT-effect is applied, it seems that there's a heap of unnecessary calculations, as their hp seems to drop almost in a continuous manner instead of discrete packages like for example in WoW. That might be a bit more "realistic", but what's the point of recalculating a mobs hp 20 times per second (or however fast DoTs seem to "tick" in diablo 3) from a gameplay perspective.
Still.. that's nothing compared to the shit that's going on with area damage. I don't know if there's efficient ways to quickly determine whether a mob is inside the area-damage range of another mob that just triggered it... but the first idea that would come to my mind to do this (using something like 2D cartesian coordiantes) would be quite a mess.

The recalculation for dot damage on the HP Bar is done clientside and is not affecting lag. i can guarantee youthis. Its just some kind of interpolation. Nothing fancy.
The client knows how much damage the dot does per second, so he can just devide it by 60 to update the hp bar every frame (@60fps). Even if its a little unpresice it gives a better impression of the dot nature. but be assured the server is not involved in visualizing the dot damage as its done now.

Here is a little info from the reddit AMA:

BY: (blizz_rogerhughston)

It is also worth saying that most of our damage and health calculations on the server are done in floating point numbers. Computationally the expense between large and small floats is the same, and most of the expense on the server is usually the raw quantity of the calculations we are trying to do rather than values stored in the floats themselves.

Thats somehow confirms my assumption that the size of the numbers is not the problem but the frequency of state changes (the raw quantity of calculations)

thus jamoose your suggestion will likely not influence performance greatly. Especially since the server is working with floats.

you are totally right jamoose

However reducing the max density would also change the balance of the game. Imagine archon set not being able to reach > 150 stacks due to density.How would that build even work on higher grifts.

Reducing the density also means reducing the max grift level you can reach.

However . it is a solution.

Quote from Jamoose»

Quote from Jalatiphra»

Quote from Jamoose»

If they patched the game so an RG on GR 70 has 10 billion hp and not 10 trillion hp for an example, and then toned down the players' damage accordingly, wouldn't that reduce server stress? See, i didn't mean that they would add another calculation of divison / multiplication along the way, but rather to reduce all of the numbers in their inception...

no that would not make a difference. as i said, the size of the numbers is not imporant.
If they reduce damage from the players and hp of guardians the situation would still be the same except numbers being smaler. that would not make it better.
you can reproduce the lag on grift 35 or 25. Just fish a high density rift and do not equip a weapon. It will lag like fuck

If the numbers get so big that they are no longer a 32 bit integer, and considering the fact that D3 runs on 32 bit only... Wouldn't that mean that the compiler is emulating 64 bit types and operations onto them, therefore creating a much heavier process on the server in terms of memory usage specifically?

Yes and no. Its true diablo is a 32 bit operation, thus it cannot compute numbers bigger than 32 bit without increased effort.. However
that number is 2147483647 when its signed integer. But make the damage numbers signed is most like not the case since damage is not negative.
So it will have an unsigned integer which maximum damage number is : 4.294.967.295
For comparision the max int for 64 bit unsigned is: 18446744073709551615 . Thats a lot of damage.
Now begins the fun part: You clearly have seen damange number higher then 4.294.967.295 so its safe to assume that a bigger datatype is used.
On 32 bit applications you can have a 64 bit integer too. but its stored in two memory locations, split up in the first and second half. so they are effectively two unsigned ints.
However this comes with a big performance hit. And clearly that is a problem in d3. Because it makes frequent use of this.
Allowing numbers to not get bigger than 4.294.967.295 or convert to 64 bit application, might increase performance. Granted this is exactly what you said earlier (reducing the numbers to gain speed up) However i did not understand you in that way that you referred to this technical limitation because your appraoch was a little bit too vague.
If you change your statement to : change d3 to 64 bit or make sure damage does not get bigger than uint.max then i would have totaly agreed.
While deviding 100000 damage by 100 it would not make a difference to work with either 100000 or 1000. While 14.294.967.295 /10 would make a difference because now the number fits in a 32 bit number again, However you do not solve this problem by division but rather make sure numbers do not get that high.
Also i assume the performacne gain from this would be marginal. The bigger problem is the many state changes and server client synchronisation.
But i have to correct myself. yes you are right , making numbers smaler can increase performance. but this has to be done in another way than you suggested and is only the case for certain big numbers. Which your initial post did not make clear (you probably do not have the required background knowledge).

I am sure that your "make numbers smaler" appraoch would only influence performance by a bit because i can reproduce the lag problems without putting out so big numbers (as i said go to 25 grift and dont equip a weapon - low damage - still lag)

Hmm i cant find the edit button:

of course you are right saying that lag is more promiment in higher grift levels. but you can reproduce the issue on lower grifts also under certain circumstances.

From a technical view this problem is not due to the rift level but to the number of calculations needed. and those correlate with the grift level currently. but you can create the same situation on any grift level if you force these amout of calculations in lower grifts by changing gear.