Vengeance Demon Hunter 7.3

[Valks 2,375]

On 20.11.2016 at 0:58 AM, Guest Oddtobias said:

Are the talent recommendations viable for mythic+ in 7.1?

They should be, yes. They are evaluated on their strengths in PvE and, if for some reason they have a different use in a raid or dungeon, it is normally listed. 

[Valks 2,375]

On 22.11.2016 at 7:09 PM, Narabug said:

Please compare the two talent choices.  Back up your findings.  Use numbers, compare logs.  On the guide, and everywhere else I've seen you note it, you've made it very clear that you simply like Last Resort better.  You are starting with 'Last Resort is better' and going from there.  Stop.  Start with 'Which choice provides more benefit to the group?'

Just stepping in here to leave a message as a mediator, I guess you could say? Anyway:

Generally, when a reviewer is chosen for their prowess in game, we will leave a note essentially describing their achievements. If you look at any guide by Furty, it very clearly states at the top that he raids in Serenity and describes them as one of the top guilds in the world. 

For Munkky, we chose him as a reviewer for the guide because of his reputation and prowess as a theorycrafter. He is excellent with the information that he provides from a theorycrafting PoV and, I hope you can agree, much of the guide is truly excellent. We never wanted to claim that he was a top player, but wanted instead to point to his abilities as a theorycrafter.

It might be that he has made a small oversight when it comes to the practical use of a talent such as Soul Barrier on certain fights, but I think his evaluation of the talent in the guide isn't unfair. He hasn't listed it as useless and he has said very clearly that it is definitely viable in certain situations.

I do think Last Resort does require your healers to almost play to it in a way. Your healers, during periods of intense healing, could simply allow you to "die" and let it proc rather than needing to prioritise you and allowing a DPS to die. I'm not a DH player nor do I claim to be one, I'm just trying to ensure all sides of this are seen.

Anyway, we'll look into the talent comparison!

[Narabug 0]

4 hours ago, Blainie said:

I do think Last Resort does require your healers to almost play to it in a way. Your healers, during periods of intense healing, could simply allow you to "die" and let it proc rather than needing to prioritise you and allowing a DPS to die. I'm not a DH player nor do I claim to be one, I'm just trying to ensure all sides of this are seen.

Anyway, we'll look into the talent comparison!

Actually a note on this, as I was discussing with a friend last night - If the most common use of Last Resort is to intentionally pop it on CD and pretend that it is a 15s invulnerability on a 3 minute cooldown, then I think that a reasonable argument could be made for taking Nether Bond instead, as it will accomplish more-or-less the same on a 2 minute cooldown instead of 3 minute.  It would also be significantly more reliable.

As I see them, the final tier of talents can be reasonably outlined as follows, taking all simulated numbers out of the equation:

• Last Resort - Drastically Improves TMI within a small time window of under 5 second; Longest CD.
• Nether Bond - Significantly Improves TMI with moderate time window of 15 seconds; Moderate CD.
• Soul Barrier - Improves overall TMI throughout the fight; Low CD.

So, if you think you are in an encounter where you have the potential to be killed faster than you or any of your healers can respond, Last Resort is a solid choice.  If, however, you simply want an ability to give yourself a 15s window of significantly-increased survivability, then you should be taking Nether Bond.  Finally, if you believe that there are no 'one-shot' or extremely heavy tank damage windows in the encounter, then Soul Barrier offers the best sustained mitigation/healing.

The problem with the assessment overall is that Tanks simply are not the most likely to die in a given encounter in this expansion.  By taking additional pressure off of yourself, Soul Barrier allows the healers to focus on the people who are most in-need of healing - the DPS.

[Valks 2,375]

3 hours ago, Narabug said:

The problem with the assessment overall is that Tanks simply are not the most likely to die in a given encounter in this expansion.  By taking additional pressure off of yourself, Soul Barrier allows the healers to focus on the people who are most in-need of healing - the DPS.

Going to pass all of this feedback on. Thanks for contributing!

[Guest Etsunari]

Just smiling when re-reading your responses to my posts in september.

Happy to see some eyes opening, at least.

And, yes, Soul Barrier is better than Last Resort. All arguments are already here.

About Versa, Crit and Haste.
How many versa you need to have a 15% healing returning from your damages ? 15x400 ? 6000 ? Assuming your doing active damages all the time. Why 15% ? Because it's what granted you your fire damages as self-healing and you don't need to be active, you just need to burn, be in fire. And again, two ways for that : crit and haste. With what : legendarie and enchants, food (*), gems... I manage to lower CDs to 1.8 sec, meaning I'm almost doing fire all time, even when switching bosses, adds or as 2d tank, my Spikes are up all time and I can shield with 6 lesser souls each time. Can you compete that with Versa ?...

Please don't just ignore the possibility of another way to improve DH Vengeance. Try it.

None of the simulation tools are mesuring that. You just can't simulate it. Because we don't have Patchwerk-like bosses in Legion. They bump or we need to switch on 2 bosses or they move, jump... No full dps all the fight, then versa isn't as good and pretty as we show it on simulations. Another indication that simulation are wrong is that nobody find the same numbers when running a simulaton... strange no ?

My live experience (+alpha +beta) is that the prio should be something like : Haste (capped at +/- 20% to touch the 2sec CD) then Crit > Mastery = Versa. Other stats are bonuses (like Leech, Armor)

 

(*) Why, after years, we have now fire food ? just because it's fun ? Come on.

[Valks 2,375]

On 11/29/2016 at 6:55 PM, Guest Etsunari said:

Just smiling when re-reading your responses to my posts in september.

My responses? 

On 11/29/2016 at 6:55 PM, Guest Etsunari said:

And, yes, Soul Barrier is better than Last Resort. All arguments are already here.

I have passed on everything that has been said, but it's not my place to make any changes to the guides.

On 11/29/2016 at 6:55 PM, Guest Etsunari said:

My live experience (+alpha +beta) is that the prio should be something like : Haste (capped at +/- 20% to touch the 2sec CD) then Crit > Mastery = Versa. Other stats are bonuses (like Leech, Armor)

I'll pass this on to the writer/reviewer.

[Valks 2,375]

On 11/29/2016 at 6:55 PM, Guest Etsunari said:

Another indication that simulation are wrong is that nobody find the same numbers when running a simulaton... strange no ?

Can't edit for some reason, but also, this isn't really that strange. Simulations will find different results when using different inputs. If you use a custom APL, you'll get a different result. Different gear? Different result. Different sim style? Different result.

[Guest Riadle]

Someone has a stats weight or pawn link for dps stat priority? Tnx

[Guest Just_A_Guy]

On 11/24/2016 at 3:35 AM, Blainie said:

chose him as a reviewer for the guide because of his reputation and prowess as a theorycrafter. He is excellent with the information that he provides from a theorycrafting PoV and, I hope you can agree, much of the guide is truly excellent. We never wanted to claim that he was a top player, but wanted instead to point to his abilities as a theorycrafter.

Quote from Munkky on reddit:

" I also want it to be known, that I do ALL of my own math. I DO NOT use sims. I DO NOT trust AMRs sims, not that I have a grudge against them. I don't trust sims for tanking because it is so hard to quantify everything and talent choices, comp, cds, dps of the group, the fight, etc. all factor into it. "True" weights would be derived from multiple encounter types and scenarios. I do not like to do feelcraft, and instead do all of my own math and use testing from both my own chars as well as those of my friends that I trust. "

You chose wrong.

You can't say someone who thinks their napkin math is more reliable than a computer that can run tens of thousands of iterations on the math is a person with theorycrafting prowess.  I mean, that is just straight up ignorance.  I'm glad he has so much faith in himself that he thinks his mathematical skill is superior to a machine designed to do more math in ten seconds than a human can do in ten lifetimes, but he's just straight up wrong.  I mean seriously.

Let's compare:

Simcraft: open source platform with fully transparent and customizable scripting that has a multi-year, multi-expansion proven history of being a reliable means of testing and validating gear/talent/rotational hypotheses.

Munkky's napkin math: Yeah...

If this dude's methodology was used for "theorycrafting" a dps class that class would be bottom tier and considered unplayable.  The only reason he is getting away with any of this is because some people are bought into his fallacy that tanks can not be simmed.  Of course, since a sim is effectively just rapidly iterated math, which Munkky himself admits to using, only with orders of magnitude less precision.

Final word: Vengeance Demon Hunter is currently considered to be a bad tank by the community at large based on the advice in this guide that is being followed by the vast majority.  Yet the ones who reject the false information that this hack of a theorycrafter, who is now more interested in proving he is right than finding out what is right, consistently surprise groups with amazing survivability and performance.

If you all value the reputation of this site at all I would highly recommend you find yourself a new Vengeance DH dude, 'cause this one isn't cutting it.

[Munkky 4]

Here is the reply to the above posters thread on the same Reddit thread I posted my statement in.

Quote

Well, the hostility is really unnecessary.

But, in general, simulations are good about comparing a difference in 2 (or a very small amount) of variables under tight controls with very precisely defined parameters. They are what you can use to verify or check an assumption, but they are not good at generating a solution unless the problem is trivial.

We can do enough simulation runs to solve a problem like the optimal strategy for tic-tac-toe or even checkers, but playing a wow class has too many moving parts. Unless you are willing to pay a LOT of money for a huge computing cluster, you are not going to derive the best talent setup and rotation and stat priority against a given fight given any combination of legendaries, etc, because there are too many variables involved.

As such a simulation will not usually spit out a general axiom you can use, because you have to control many of the variables. You won't be able to say Talent A is better than Talent B from a sim, because you ran the sim for X rotation, on Y boss pattern, with Z amount of movement and so on. Unless you do all the permutations of X/Y/Z/etc, all your claims after a given sim have a lot of caveats.

In order to actually mathematically prove that talent A is better than Talent B, you have to do "napkin math" and write your proof by hand. That's how math works. That is how creating theories and proving them happens in the world of math.

 

This is someone who I have never spoken with and have no idea who they are. Sims for tanking haven't hardly been used in years. If I remember correctly, the only 2 tank specs that even kind of have sims good enough to run are Druid and Paladin. I am also friends with the person who worked on both of those sims, and have worked with him on Vengeance related TC. I wrote up my statement to explain to the public my situation and what I do. I am putting myself out there. I did not put anything hostile in there and the fact that "my haters" immediately try to flame me and talk about things they don't know anything about (in this case, tank sims) just reinforces my guides. 

I also find the comment about " Yet the ones who reject the false information that this hack of a theorycrafter, who is now more interested in proving he is right than finding out what is right, consistently surprise groups with amazing survivability and performance. " to be very funny. The "top DH" seem to be using the methodology that I suggest, and I also actually work with several of them. The people that "reject the false information" are more often than not less experienced players who out gear the content they are doing, then try to use that as an argument. I also have had numerous players come to me, thanking me for helping them learn the class. They are often 7/7M at this point, working on ToV progress. 

So final thoughts. Sims aren't that reliable. I work with the top DH players and have spent the past 10 months emotionally and mentally attached to this spec and to this community. I have put in more sleepless nights and time and energy into this than my "haters" will lead you to believe. I am also one of the most open minded people who do Theorycraft, which can be plainly seen in the sheer amount of viable talents I have as well as the fact that I actually put out all 3 stat priorities. 

Edited December 21, 2016 by Munkky

[Guest Just_A_Guy]

Telling someone they are wrong is not hostility.

Every single thing you just described as being a benefit of doing math by hand the computer can do with more accuracy due to iteration volume that you can never match with a pencil.

All you said was there are variables so you can't sim. This is simply false. Every bit of math that you can do by hand you can also program the computer to do. Every variable you can change with your pencil you can also change on a computer. That you do not know how to write and modify scripts to simulate different variables does not mean it can not be done.

Quite literally the ONLY thing different between what a computer can do and what you can do by hand when it comes to math is that the computer can execute orders of magnitude more volume and thus account for the inherent variance due to RNG present in WoW.

I'm sorry if you find what I am about to say "hostile" but you are just flat out wrong in your premise that you can do math by hand that a computer cannot do.

To your final point, that is great that you "work with top DH" but does not change the fact that DH is considered by the community to be a bad tank and the ones I've played with who don't follow your guide surprise groups by not being squishy and frustrating to heal.  

That is great that you think you're open minded and are very emotional about the DH class but at the end of the day all that matters here is results.  Telling people you emotionally attached to the community means nothing and appears to be a blatant attempt to garner support and make people agree with you because you seem like a good guy and calling you out makes you sad.  You probably are a good guy and your friends probably think you're pretty swell, but for what you're doing here all that matters is results.

if your goal is to address the cold, hard data and theorycraft DH to its maximum potential, why does it matter how emotional you are about the class and how much people should appreciate you and connect with you because of it?  Either your numbers are right or they're not.  And based on the obvious weaknesses in your approach, I suspect they are not.

It appears to me at this point based on your conclusions, and by your own description of yourself, that your ego is too invested in your guide for you to be objective towards the facts and seek truth rather than approval.  I stand by my original statement.

[Narabug 0]

16 hours ago, Munkky said:

the fact that "my haters" immediately try to flame me and talk about things they don't know anything about (in this case, tank sims) just reinforces my guides. 

I also find the comment about " Yet the ones who reject the false information that this hack of a theorycrafter, who is now more interested in proving he is right than finding out what is right, consistently surprise groups with amazing survivability and performance."  to be very funny. The "top DH" seem to be using the methodology that I suggest, and I also actually work with several of them. The people that "reject the false information" are more often than not less experienced players who out gear the content they are doing, then try to use that as an argument.

I have put in more sleepless nights and time and energy into this than my "haters" will lead you to believe. I am also one of the most open minded people who do Theorycraft, which can be plainly seen in the sheer amount of viable talents I have as well as the fact that I actually put out all 3 stat priorities. 

I don't know why you call people who disagree with you "my haters," but perhaps that speaks fairly directly to the point of view that you have towards people who present information that does not agree with your personally-derived conclusions that are based on your personally-crafted and undocumented ideas.

Notice how you are responding to the point made in the second paragraph.  You're given an example of a person who knows a DH that, by his account, does not follow your guides or talents, and consistently surprises groups with amazing survivability and performance.  Your response, rather than asking any specifics that may help to improve your guide, is to immediately discredit the information by making something up.  In this case, they must surely be less experienced, and just out-gear the content that they are doing, in an effort to make a personal argument against you, presumably.

Your sleepless nights really don't matter, though I do sympathize with it, as years ago I used to spend a lot of time working on a different class.  What does matter is that the information that you are presenting is valid and can be backed up with facts.  I honestly wouldn't care if Sco said that some talent was the best choice if he couldn't back it up with facts, examples, math, logs, and simulations.  (Though that is just as example, because Sco doesn't get too into the class theorycrafting).

From what I have seen thus far, your mode of operation truly seems to be: Decide which talents I like the best, then prove that they are the best by attacking individuals that use something else or simply stating that the 'best' players use them.  It's circular logic.  Once you've decided on your desired 'best' talents, then you're never going to move past it because you are consistently working to reinforce your own first guess rather than discover the truth.  That's not how you arrive at a solid results.  That's not a scientific approach.  It is a religious approach and it uses the same arguments that someone defending a religion would use.

Looking back through this thread, there are numerous posts where individuals have specifically given you their reasons for seeing things the other way, and every single time, your response is to attack the person or discredit their experience.  You have not once presented any numbers on how you arrived at specific stat values.  Your talent choices completely leave out any numbers.  If there truly is some underlying math that you are doing, why not place it in the guides so that people can actually see how you're arriving at your conclusions?

[Munkky 4]

I am done with these arguments. They are getting us nowhere and are simply taking shots at each other. I have the same open door policy I have had since I started doing Theorycrafting for Demon Hunter. If you see something wrong with the guide, give me proof that it is wrong and I will change it. Saying that you used sims is not valid proof. If you wish to argue about math, then that is fine, do the math yourself then argue with me. Do not simply argue just because you do not like me.

If you wish to speak with more experienced players than myself about tank theorycrafting as a whole, please refer to this link.

https://discord.gg/7ncTW9q

They will be able to aid you in all things tank related. Also please ask in there about sims and especially AMR there. Maybe some others can shine a light on this hot subject for you.

Have a nice day.

Edited December 22, 2016 by Munkky

[Narabug 0]

3 hours ago, Munkky said:

I have the same open door policy I have had since I started doing Theorycrafting for Demon Hunter. If you see something wrong with the guide, give me proof that it is wrong and I will change it. Saying that you used sims is not valid proof. If you wish to argue about math, then that is fine, do the math yourself then argue with me. Do not simply argue just because you do not like me.

Please see this entire thread.  As I stated, multiple people have presented valid points to suggest that the data in this guide is inaccurate.  While Soul Barrier is going to be the moot discussion come 7.1.5, your response to it in this thread is a great indicator as to why nothing else can be trusted.  It seems to me that your 'open door' policy consists of 'I am open to people who agree with what I randomly decide is the best.'

People have presented arguments, math, and logic, and you consistently ignore it or attack the person who is delivering the information.  As I said a month ago, and in the post above, you should not be selecting what you like best, then finding ways to dismiss other things.  You should be considering both on equal grounds.  If your stance on one single talent is so clearly flawed, then why should anyone trust the stat weights, or any other talents?

You seem to really be taking these discussions personally, and it results in escalation.  In my case specifically, it started with me posting arguments for Soul Barrier with math, logs, and logical reasoning.  You then attacked my gameplay and progression as a way of shutting down the argument without actually addressing it.  Someone have a differing point of view to yours is not a personal attack.  Attacking someone personally as a result of their opposing position is a personal attack.

Edited December 22, 2016 by Narabug

[Munkky 4]

1 hour ago, Narabug said:

Please see this entire thread.  As I stated, multiple people have presented valid points to suggest that the data in this guide is inaccurate.  While Soul Barrier is going to be the moot discussion come 7.1.5, your response to it in this thread is a great indicator as to why nothing else can be trusted.  It seems to me that your 'open door' policy consists of 'I am open to people who agree with what I randomly decide is the best.'

People have presented arguments, math, and logic, and you consistently ignore it or attack the person who is delivering the information.  As I said a month ago, and in the post above, you should not be selecting what you like best, then finding ways to dismiss other things.  You should be considering both on equal grounds.  If your stance on one single talent is so clearly flawed, then why should anyone trust the stat weights, or any other talents?

You seem to really be taking these discussions personally, and it results in escalation.  In my case specifically, it started with me posting arguments for Soul Barrier with math, logs, and logical reasoning.  You then attacked my gameplay and progression as a way of shutting down the argument without actually addressing it.  Someone have a differing point of view to yours is not a personal attack.  Attacking someone personally as a result of their opposing position is a personal attack.

I just looked at every post that I have made in this forum. None of them were attacking anybody. You mistook my comment towards your LR vs SB argument. You showed logs to prove your point. I then took those logs, and used them to prove my point. None of that was an attack on you, none of that was saying you are bad at the game or anything. I used the exact same resource as you and found other things that you didn't include in your analysis. I gave you the link to the site, explained a few other things as to counter your argument, none of which were volatile. You are the one who started this drama. I wish to end it and want no part in any of it. If you have reasonable issues with my guides, then say so and give me proof. That's all I've ever asked.

[Rudy199 8]

As of right now for 7.1.5 the level 108 & 110 talent choices are pretty clear-Spirit bomb and Soul barrier

[Valks 2,375]

Capping any further discussion here. I'm not trying to stop people giving feedback (not my intention at all and please do continue with feedback), but the discussion has shifted away from the initial point pretty drastically. 

If you wish to continue it, please do it elsewhere, even in another thread on the forum. Any further comments on the discussion above will just be hidden. 

Thanks for understanding. If you have feedback or questions on the guide itself, please feel free to comment and I can still help you, of course! :)

[Valks 2,375]

On 12/23/2016 at 0:35 PM, Rudy199 said:

As of right now for 7.1.5 the level 108 & 110 talent choices are pretty clear-Spirit bomb and Soul barrier

I'm sure the guide will get an update before 7.1.5 drops to account for any changes, don't worry! :D

[Valks 2,375]

The guides are currently being updated for 7.1.5! I just wanted to leave a few notes for all of our users that might be getting ready to leave a comment:

• Some pages do not require updates. The "Last Updated" stat does not mean the page is out of date, it might just not have needed any changes. If you feel that it does, leave a comment telling us what needs changing and why.
• The process will not happen immediately - some guides will be updated faster than others simply due to the number of resources available. Be patient, they'll be ready when you need them!
• There might be some continuity errors when making small adjustments to large guide pages. If you do find one of these, just let us know in the comments and we'll get it fixed ASAP.
• "Why have you not taken into account X buff to X ability?" - remember, just because something got buffed, it doesn't mean it is now automatically better than the other options!

As always, we want to thank you all for being patient while we get things updated and I'm always available to help you all if you need it :)

[Kirby 0]

On 1/9/2017 at 3:21 PM, Blainie said:

The guides are currently being updated for 7.1.5! I just wanted to leave a few notes for all of our users that might be getting ready to leave a comment:

• Some pages do not require updates. The "Last Updated" stat does not mean the page is out of date, it might just not have needed any changes. If you feel that it does, leave a comment telling us what needs changing and why.
• The process will not happen immediately - some guides will be updated faster than others simply due to the number of resources available. Be patient, they'll be ready when you need them!
• There might be some continuity errors when making small adjustments to large guide pages. If you do find one of these, just let us know in the comments and we'll get it fixed ASAP.
• "Why have you not taken into account X buff to X ability?" - remember, just because something got buffed, it doesn't mean it is now automatically better than the other options!

As always, we want to thank you all for being patient while we get things updated and I'm always available to help you all if you need it :)

Is there any way we can get the Pawn Stat Weights for 7.1.5?

[Valks 2,375]

54 minutes ago, Kirby said:

Is there any way we can get the Pawn Stat Weights for 7.1.5?

The biggest issue with providing weights, especially for tanks, is that they can vary incredibly depending on item level and current stat levels. This is even more true for tanks vs. DPS given that there is no differentiation between the type of target being hit by DPS. Tanks take a variety of damage (magic, physical, burst, DoTs, etc.) which means you'd need to take into account even more variables for reliable weights.

Creating them for general use is a real headache!

[Guest Mevkire]

Wondering why anyone would ever need or want the t19 4 pc set bonus, currently my demon spikes has a 12.8 second cd each charge has a 6 second duration 2 charges 12 seconds down time .8 seconds, now using the talent feed the demon and having consumed at least 1 soul fragment what would reduce the cd to 11.8 seconds leaving me with -.2 down time so now the ability is always active with minimal requirements and with most of the set having horrible stat priorities on it it self why would we want to force ourselves into a 4 set with pretty much no affect?

[Orthios 271]

9 hours ago, Guest Mevkire said:

Wondering why anyone would ever need or want the t19 4 pc set bonus, currently my demon spikes has a 12.8 second cd each charge has a 6 second duration 2 charges 12 seconds down time .8 seconds, now using the talent feed the demon and having consumed at least 1 soul fragment what would reduce the cd to 11.8 seconds leaving me with -.2 down time so now the ability is always active with minimal requirements and with most of the set having horrible stat priorities on it it self why would we want to force ourselves into a 4 set with pretty much no affect?

For abilities with multiple charges, only 1 charge can be recharging at a time.  So, you'll only have the near 100% uptime through 3 uses:

Using 13s/charge here:

t=0: Use first charge

t=6: First charge wears off, has 7s left on cd, use second charge.

t=12: Second charge wears off, first charge has 1s left

t=13: First charge finishes recharging, second charge starts recharging, use first charge.

t=19: First charge wears off, Second charge still has 7s left.

t=26: Next charge available, 13s until next available charge.

 

So on and so forth.

[Valks 2,375]

12 hours ago, Guest Mevkire said:

why would we want to force ourselves into a 4 set with pretty much no affect?

Just in addition to what Orthios has said, it's also worth mentioning that the actual increase from the 4P is fairly huge. With up to a 3s reduction with each SC, you could be achieving upwards of a 25% reduction depending on AM pain usage etc.

[Guest Shambala]

Hey mates, just looted http://www.wowhead.com/item=137430&bonus=1826 (885)
Useless for VDH now, am I right? :(