Killing Machine issues

Death Knight
Prev 1 2 3 4 6 Next
Average boss fight is 5-10 minutes.
Hit a dummy for that amount of time, log your dps.

Reset your meter, do the same thing again, this time use a different rotation style. Try waiting 1 GCD for KM procs, try sitting on recharging runes, try not waiting at all... Do whatever you want.

Spoiler alert:
Your DPS will be the same everytime. Frequently, the longer you wait, your dps will go slightly down.


a !@#$load of time on dummies is how I came to the conclusion that the arguments for waiting were, in fact, correct and effective, thanks

Plus, the rotation is harder to manage the more you wait or the harder you focus on waiting. There's less focus involved if you just burn runes/rp and yet you get the same (or higher) dps.


I don't care which is harder, I care which is more effective.

Your 'gut feeling' is wrong on this one. I really wish putting a lot of effort into handling KM procs would yield higher DPS. I really think putting the extra effort into your rotation should reward you.

However blizzard says no. Doing more work doesn't benefit you in this case.


It's not a gut feeling. It's having the evidence presented to me, then proceeding to experiment with it. Admittedly I've gotten like 5 or 6 ilvl worth of gear since I last tested it since I've been dual wielding anyway, but the vast majority of people looking for this info are well below my gear levels anyway.

just a thought: tell your simcraft to delay oblit by 1 global less than your rune regen rate only when dd fF uU and rp <32

Are we changing our story now?
Implying you're going to be dry on your next GCD; that is, you can Oblit once, and you will have a gap following that, is not the same as "Always wait on Dd Ff Uu".


Already responded showing you how I've accounted for what you're saying all along, but I should also clarify: that was instructions on how to attempt to get simcraft to actually model some aspect of this discussion. We can't get simcraft to wait intelligently. Hence all the math shown being utterly inaccurate and ineffective representations.
Hey look everyone, Lailala is topping wol parses!

oh wait.

Show me the parse of anyone who's ever ranked (hint: i have in the past, on multiple fights) who's doing this. I'm not. Then again, I've stopped caring about logging dps on this char for a long time.
I'm a tank, lol.

I do get to dps sometimes, though. But our ranks are all over the place cuz of like 5 people logging.

last progression fight I dps'd:
http://i.imgur.com/NuZ1d.png

with a running of pheromones during execute range

http://www.youtube.com/watch?v=lKO-Cf2czlo < vid to prove when I ran
Hey look everyone,I'm typing nonsense!
oh wait.
Show me more nonsense

Sure.
Lailala's average dmg from abilities his/her last raid. (whole raid)

Plague Strike; M.H. + O.H. = 22k
Obliterate; M.H. + O.H. = 50k
Howling Blast; = 49k
Frost Strike; M.H. + O.H. = 61k

The real issue is obvious, 2h vs. DW
HB + PS (71k) > Obliterate (50k) = Why cast obliterate?
Hey look everyone,I'm typing nonsense!
oh wait.
Show me more nonsense

Sure.
Lailala's average dmg from abilities his/her last raid. (whole raid)

Plague Strike; M.H. + O.H. = 22k
Obliterate; M.H. + O.H. = 50k
Howling Blast; = 49k
Frost Strike; M.H. + O.H. = 61k

The real issue is obvious, 2h vs. DW
HB + PS (71k) > Obliterate (50k) = Why cast obliterate?


Wow, thanks for entirely failing reading comprehension. You gave me a good laugh though. /pat

And to clarify:

If some amazing new-fangled theory out there proved to be true, then lots of ppl would know about it. There's no super-secret method out that that only a handful of ppl are doing.

"But how do you know he is saving KM for oblitzzzz????//??/one!"

Easy, show me a parse that has a substantial amount of more oblit crits than fs crits.
Hey look everyone,I'm typing nonsense!
oh wait.
Show me more nonsense

Sure.
Lailala's average dmg from abilities his/her last raid. (whole raid)

Plague Strike; M.H. + O.H. = 22k
Obliterate; M.H. + O.H. = 50k
Howling Blast; = 49k
Frost Strike; M.H. + O.H. = 61k

The real issue is obvious, 2h vs. DW
HB + PS (71k) > Obliterate (50k) = Why cast obliterate?


rime
also gcd capping as dw, anyway


Sure.
Lailala's average dmg from abilities his/her last raid. (whole raid)

Plague Strike; M.H. + O.H. = 22k
Obliterate; M.H. + O.H. = 50k
Howling Blast; = 49k
Frost Strike; M.H. + O.H. = 61k

The real issue is obvious, 2h vs. DW
HB + PS (71k) > Obliterate (50k) = Why cast obliterate?


Wow, thanks for entirely failing reading comprehension. You gave me a good laugh though. /pat

And to clarify:

If some amazing new-fangled theory out there proved to be true, then lots of ppl would know about it. There's no super-secret method out that that only a handful of ppl are doing.

"But how do you know he is saving KM for oblitzzzz????//??/one!"

Easy, show me a parse that has a substantial amount of more oblit crits than fs crits.


feng is the best one I can think of to get reliability, lots of dual wield, but

first 2h is 32% oblit crit 27% fs crit, so not really substantial
second 2h is 37% oblit crit 27% fs crit http://worldoflogs.com/reports/rt-ix9dzjji4sbc3xbv/details/12/?s=4168&e=4478
3rd 2h is 33% oblit crit, 26% frost http://worldoflogs.com/reports/rt-qq0ckrcpr5zqu53q/details/9/?s=4668&e=5043
4th is 33% oblit crrit, 22% frost crit http://worldoflogs.com/reports/rt-glht5gpshma38o1v/details/16/?s=1035&e=1396
5th is back to 5% difference

so, yeah, 3 of the top 5 are about doubling the difference in crit rate of the other 2

Wow, thanks for entirely failing reading comprehension. You gave me a good laugh though. /pat

/mission accomplished
THough, to be SAFE.
we should discuss the 45% rate at which we actually get rime.

Wow, thanks for entirely failing reading comprehension. You gave me a good laugh though. /pat

/mission accomplished
THough, to be SAFE.
we should discuss the 45% rate at which we actually get rime.


you really need a discussion to do this?

look at the numbers you posted

Plague Strike; M.H. + O.H. = 22k
Obliterate; M.H. + O.H. = 50k
Howling Blast; = 49k
Frost Strike; M.H. + O.H. = 61k
HB + PS (71k) > Obliterate (50k)

add 45% of a howling blast (.45*49k=22050)

72050 vs 71k

feel free to get specific, you'll find rime wins, even if by a tiny amount, and that's not considering the benefits of less globals used when you don't get rime
12/17/2012 12:19 PMPosted by Babelon
/mission accomplished


lol <3 then

12/17/2012 12:18 PMPosted by Lailala
so, yeah, 3 of the top 5 are about doubling the difference in crit rate of the other 2


in lfr hero'ing it up atm, so can't look - are they all roughly the same? or are the 2 who are getting great crit rates head and shoulders above the other 3?

if all the same, or close, then it proves it's not worth messing with.

otherwise, if those 2 blow the last 3 outta the water, i will look at it more

forgive the complete lack of punctuation and capitilzation
1. 5%
2. 10%
3. 6%
4. 10%
5. 5%

they're alternating

and that doesn't prove it's not worth messing with, it proves that both are capable of top parses, and it does what was asked specifically:

Easy, show me a parse that has a substantial amount of more oblit crits than fs crits.


The point was that there are people doing it, and successfully. If you honestly believe 5% more oblit crits isn't worth bothering because other people have gotten lucky enough to parse similarly without it, I'm not sure what to tell you.
12/17/2012 12:53 PMPosted by Lailala
The point was that there are people doing it, and successfully. If you honestly believe 5% more oblit crits isn't worth bothering because other people have gotten lucky enough to parse similarly without it, I'm not sure what to tell you.


12/17/2012 12:53 PMPosted by Lailala
1. 5%


My entire point is that, unless it's pulling waaaaay ahead, I'm not gonna bother (just like I don't bother with plague leech for 2h normally).

Is it a gain?

Perhaps.

Is it a substantial guaranteed gain?

No way.
It's great for you that you don't care what's actually better or want to take the effort to optimize yourself.

However, it's not ok to make those statements and then end with a definitive "factual" statement that's not true.

If you don't care, and cannot and will not contribute, don't post in discussions about it, it's that simple.
12/17/2012 03:45 AMPosted by Lailala
Didn't think so. The reason waiting has become currently effective is that the lack of haste supports it. We're going to have downtime, we need to optimize when and why it occurs. Of course, jumping 30-40 ilvls over most people, and even 15 or so over people like me that have had nice luck with drops, is going to change the results here. That's kind of the point.


This makes virtually no difference. Haste doesn't change your downtime that significantly.

At 5000 haste rating, with no haste procs, you will have 17% downtime. At 6000, ~16%. At 7000, ~14.4%.

12/17/2012 03:45 AMPosted by Lailala
Secondly, your sims are nearly... no, screw that. Your sims are flat out worthless. You can't just tell it to delay oblit for a second or 2 with no ability to tell it why or when to do so and assume that's some kind of valid representation of how waiting for procs actually works out.


Why thank you for assuming that my sim is the spreadsheet I linked.

My sim is about 3.5k lines of code in MATLAB that was used to provide defensible claims for the bulk of the claims made about Blood, Frost, and Unholy DK's for the large majority of the Beta testing period because SimulationCraft didn't have DK support. To call it worthless is nothing short of laughable and shows that you have absolutely no idea what it is nor what I did.

I had it delay Obliterate if a swing was coming up within a certain timeframe (variable), if it didn't have KM already, and if it wouldn't waste runes/RP. I also tried numerous other priority scenarios, far more in-depth than what you've even tried to do on SimulationCraft.

12/17/2012 03:45 AMPosted by Lailala
Lastly, you're just making assumptions all over the place that make absolutely no sense whatsoever.


And what are you bringing to the table? A quote from an EJ thread and that's it?

12/17/2012 03:45 AMPosted by Lailala
Your concern needs to be with regeneration, not expenditure. As long as something is regenerating, you're not costing yourself by waiting.


No, it doesn't.

Time is a relatively fixed resource.

The boss has a finite amount of HP and he will die at a certain time. Your own DPS contribution relative to the raid is not that large (and even less in 25's), thus, small changes in your DPS will not greatly affect the time at which the boss will die. If your raid DPS is 1.5 mil / sec, you're doing 80k DPS of that, and the boss has 500 million HP, the boss will die in roughly 333s. If you increase your DPS by 4k, the boss will now die in 332s. If you drop your DPS by 10k, the boss will die in ~336s. Time is relatively fixed. Ability throughput absolutely matters.

Thus, while you may not be wasting runes, you are wasting ability throughput because you are trying to cram that same number of abilities into a shorter window. By stalling, you will push abilities out of that timeframe. Thus, you will lose DPS. Now, you may offset that loss with an increase in OB crits, but the question is which effect is stronger?

12/17/2012 03:45 AMPosted by Lailala
You're literally flat out ignoring the reason this works, which is that we're not sitting on constant intervals of 1-2 seconds like your math assumes.


Again, I have seen no evidence that this works, and neither have you been able to provide me with any.

12/17/2012 12:18 PMPosted by Lailala
feng is the best one I can think of to get reliability, lots of dual wield, but


This is not the crit rate. You are showing the %DPS by ability breakdown.

Crit rates would be the following for the links you posted:

2nd: 42% OB, 38% FS
3rd: 44% OB, 38% FS
4th: 43% OB, 35% FS

So, no, the difference in crit rate between the two abilities isn't even that large.
It's great for you that you don't care what's actually better or want to take the effort to optimize yourself.

However, it's not ok to make those statements and then end with a definitive "factual" statement that's not true.

If you don't care, and cannot and will not contribute, don't post in discussions about it, it's that simple.


It's not optimal. You proved it yourself.

I'm done w/you though. Read the post above mine.
My entire point is that, unless it's pulling waaaaay ahead, I'm not gonna bother (just like I don't bother with plague leech for 2h normally).

Is it a gain?

Perhaps.

Is it a substantial guaranteed gain?

No way.


Honestly I tend to follow this way of thinking myself in general. Example: Could I do 1-3% better tanking heroics if I put a lot more effort in and took some more active talents? Sure. But if I can do my job just fine without stressing myself out, I'll do that instead. As long as I am doing well I'm content. But then again I'm a casual, I don't mind being less than perfect if it keeps my stress levels down and I'm doing well enough.

Strive for mediocrity! Huzzah!

And I reference heroics only because it's what I'm currently doing on this character.

Point being, it's a game. As long as you're doing a good job and you're having fun, just keep doing what you're doing and don't mind what other people think. *Unless they're your guild. Then you may have to mind a little bit :P
Didn't think so. The reason waiting has become currently effective is that the lack of haste supports it. We're going to have downtime, we need to optimize when and why it occurs. Of course, jumping 30-40 ilvls over most people, and even 15 or so over people like me that have had nice luck with drops, is going to change the results here. That's kind of the point.


This makes virtually no difference. Haste doesn't change your downtime that significantly.

At 5000 haste rating, with no haste procs, you will have 17% downtime. At 6000, ~16%. At 7000, ~14.4%.


15% less downtime is not significant? Is this a joke?

12/17/2012 03:45 AMPosted by Lailala
Secondly, your sims are nearly... no, screw that. Your sims are flat out worthless. You can't just tell it to delay oblit for a second or 2 with no ability to tell it why or when to do so and assume that's some kind of valid representation of how waiting for procs actually works out.


Why thank you for assuming that my sim is the spreadsheet I linked.

My sim is about 3.5k lines of code in MATLAB that was used to provide defensible claims for the bulk of the claims made about Blood, Frost, and Unholy DK's for the large majority of the Beta testing period because SimulationCraft didn't have DK support. To call it worthless is nothing short of laughable and shows that you have absolutely no idea what it is nor what I did.

I had it delay Obliterate if a swing was coming up within a certain timeframe (variable), if it didn't have KM already, and if it wouldn't waste runes/RP. I also tried numerous other priority scenarios, far more in-depth than what you've even tried to do on SimulationCraft.


My "assumption" was based on this statement:

The reason you can't find them is because the Beta forums are gone, where I ran multiple different ability priorities through my sim to replicate this behavior to no apparent gain, or depending on the stall time, to a loss.


You state that you're using a stall time variable, i.e. not being able to get it to check on the fly.

Now, if you've got all these super precise and perfectly executed sims, feel free to post them. I don't believe for a second that you did hundreds of hours of work on this and didn't save any results whatsoever, only posted them on the beta forum and deleted anything esle.

12/17/2012 03:45 AMPosted by Lailala
Lastly, you're just making assumptions all over the place that make absolutely no sense whatsoever.


And what are you bringing to the table? A quote from an EJ thread and that's it?


Uh, logic and, you know, actual support for my claims rather than "I checked this one time and I did it perfectly I promise therefore I'm right and you're wrong, screw evidence."

12/17/2012 03:45 AMPosted by Lailala
Your concern needs to be with regeneration, not expenditure. As long as something is regenerating, you're not costing yourself by waiting.


No, it doesn't.

Time is a relatively fixed resource.

The boss has a finite amount of HP and he will die at a certain time. Your own DPS contribution relative to the raid is not that large (and even less in 25's), thus, small changes in your DPS will not greatly affect the time at which the boss will die. If your raid DPS is 1.5 mil / sec, you're doing 80k DPS of that, and the boss has 500 million HP, the boss will die in roughly 333s. If you increase your DPS by 4k, the boss will now die in 332s. If you drop your DPS by 10k, the boss will die in ~336s. Time is relatively fixed. Ability throughput absolutely matters.

Thus, while you may not be wasting runes, you are wasting ability throughput because you are trying to cram that same number of abilities into a shorter window. By stalling, you will push abilities out of that timeframe. Thus, you will lose DPS. Now, you may offset that loss with an increase in OB crits, but the question is which effect is stronger?


Firstly, no, you're assuming a person will make mistakes. You still create the same amount of RP over the same time period, therefore the same amount of frost strikes and procs. You need to prove that using an ability at say, second 5 instead of second 8 definitively creates more runes-- it doesn't, unless you overcap in the process.

Sure, when you have good rng on procs, you'll get more resources in that 1 window that one time, on average it comes out the same.

12/17/2012 03:45 AMPosted by Lailala
You're literally flat out ignoring the reason this works, which is that we're not sitting on constant intervals of 1-2 seconds like your math assumes.


Again, I have seen no evidence that this works, and neither have you been able to provide me with any.

12/17/2012 12:18 PMPosted by Lailala
feng is the best one I can think of to get reliability, lots of dual wield, but


This is not the crit rate. You are showing the %DPS by ability breakdown.

Crit rates would be the following for the links you posted:

2nd: 42% OB, 38% FS
3rd: 44% OB, 38% FS
4th: 43% OB, 35% FS

So, no, the difference in crit rate between the two abilities isn't even that large.


You're right I screwed up, sad part is I knew that wasn't the crit % but for some reason was thinking I had set it to show crit% lol.

However, there is quite a range even just there, from 4% to 8%. There's also ones like this:

http://worldoflogs.com/reports/oyc728eqh7b11pvy/details/8/?s=5628&e=5993

with a solid 8% difference in crit rate as well

However we really need to be looking at the amount of km procs they got, as well. Gimme a minute.
Here's my last word on the matter:

Are you an alt of Sarosha?

You argue numbers similarly.
#1: 43 procs, 23 oblit crits
#2: 61 procs, 30 oblit crits
#3: 47 procs, 19 crits
#4: 65 procs, 34 crits

Long story short, the top parses go down based on who wasted the most km procs.

The exception to this comes at the 4 spot, where he wasted a lower % of procs, but was also in the longest of the fights, which lowered him back down as the overall impact of army and heroism diminished.
popcorn.gif

Join the Conversation

Return to Forum