[H2P] AS x SL - When to pick each talent?
[H2P] AS x SL - When to pick each talent?
by Anshlun » 29 Jan 2017, 14:38
Introduction
Due to popular request, last week I dug deeper into the differences between AS and SL and how they behave in different situations. The hotfixes pre-NH affected how the talents competed against each other a bit, showing a couple different interactions between them and I'll highlight the most interesting ones in this post.
Terminology
Before going to the interesting data, it is important that you understand some terms that will be used that you might not be familiar with.
Stat Budget - Refers to the amount of secondary stats you have. For instance, if you have 10k haste, 5k critical strike, 4k mastery, and 0 versatility, that means you have a 19k Stat Budget (10+5+4+0). This is important when we talk about stat distribution.
Stat Distribution - Is how you distribute your Stat Budget among different stats.
Balancing
After the raid hotfixes that were applied on January 17th, in an overall scenario, AS and SL are actually really close together, with both on their best form performing really alike. Both talents now have their strengths highlighted with different stat distributions and fight styles.
In order to measure how well the talents perform in different stat distributions and how to relate to them easier, we created the Crit-Mastery Distribution ratio or CMD ratio for short. The CMD ratio is easily calculated, you divide your critical strike rating by the sum of critical strike rating and mastery rating:

That ratio is used to calculate the balance between your crit and mastery, which is important for you when deciding if you should use AS or SL, as seen below.
Also, for those not willing to calculate it, I've added a command to PriestBot:
PS.: For realms with spaces in the name, like Area 52, replace spaces by dashes, like area-52.
Damage Comparisons
Disclaimer: The methods used here to evaluate damage were the usual methods used by us, which can be found in our previous work, like the Raid Stat Weights work.
All the tests were run based on a 26k Stat budget profile, averaging 905 ilvl with 4-piece bonus. Changing ilvl and trinkets brings only minor changes to the graphs and can be ignored.
To understand the differences between AS and SL, we will be looking into three different major situations: Single Target, Two Targets and Two Separated Targets. You might ask, "Ansh, why the targets being separated or not matters? Not like Mind Sear is relevant, LOL" Why yes, that's true, Mind Sear doesn't, damage wise, influence that much the results, BUT, having the targets together, means Mind Flay will generate more Insanity, thanks to Mind Sear, which leads to longer Voidforms, which leads to more damage to your DoTs via Mass Hysteria, meaning more value to Mastery and SL. Also having targets together means you will spend less globals refreshing your DoTs because Void Bolt will extend the duration from both.
Single Target
For single target, the situation of both talents is really alike, as seen below:

´
As we can see, changing from 10k to 12k haste barely impacts the final DPS result (both graphs are in the same scale), but changing the balance between crit and mastery can impact the DPS up to around 8%.
The plateaus found on the graphs mean that dropping mastery in favor of crit, for that talent, in that interval, doesn't impact its DPS, meaning they have the same value. It is important to notice that the highest DPS value from AS isn't when you put every point into crit. That happens because of the quick Diminishing returns that you have with crit as a stat in general and how nice Shadow scales with Mastery now even without SL.
Reading from the results shown, if your CMD is above 0.45, for single target, you'll get better results running AS, otherwise, run SL.
Two Targets Stacked


For two targets, the game changes a bit. Now we start seeing SL starting to win even with high CMD, considering high haste, while, with a lower haste level, the split between them stays around the 0.4 marker. Extrapolating the results from two targets to more targets, you can expect SL to have the same trend, but more accentuated, being better faster the higher your haste is, since a bigger slice of your damage will be done by your DoTs.
Two Targets Separated


As I mentioned at start of this section, when the targets are separated, there's a different curve for both, with AS winning in any scenario as long as CMD is above 0.25. Extrapolating this for more targets, the CMD threshold should increase a bit, because more damage is being dealt by the dots that scale better with Mastery than Crit, but still AS should be slightly better in most CMD values.
Conclusion
Overall, Mastery is better stat both gameplay wise and damage wise for Shadow than Crit. So you should be aiming for it, while respecting the values mentioned in the Stat Weight Discussion post. Use this research as a way to know what is best for your right now rather than a guide to what you should aim for.
In Nighthold and Dungeons alike, most of multitarget scenarios are stacked on top of each other, so the results for Two Targets Stacked represent better the average scenario for those scenarios and is a good guideline to follow.
Due to popular request, last week I dug deeper into the differences between AS and SL and how they behave in different situations. The hotfixes pre-NH affected how the talents competed against each other a bit, showing a couple different interactions between them and I'll highlight the most interesting ones in this post.
Terminology
Before going to the interesting data, it is important that you understand some terms that will be used that you might not be familiar with.
Stat Budget - Refers to the amount of secondary stats you have. For instance, if you have 10k haste, 5k critical strike, 4k mastery, and 0 versatility, that means you have a 19k Stat Budget (10+5+4+0). This is important when we talk about stat distribution.
Stat Distribution - Is how you distribute your Stat Budget among different stats.
Balancing
After the raid hotfixes that were applied on January 17th, in an overall scenario, AS and SL are actually really close together, with both on their best form performing really alike. Both talents now have their strengths highlighted with different stat distributions and fight styles.
In order to measure how well the talents perform in different stat distributions and how to relate to them easier, we created the Crit-Mastery Distribution ratio or CMD ratio for short. The CMD ratio is easily calculated, you divide your critical strike rating by the sum of critical strike rating and mastery rating:

That ratio is used to calculate the balance between your crit and mastery, which is important for you when deciding if you should use AS or SL, as seen below.
Also, for those not willing to calculate it, I've added a command to PriestBot:
- Code: Select all
!cmd character realm region
- Code: Select all
!cmd anshlun whisperwind us
- Code: Select all
!cmd anshlun whisperwind us 1
PS.: For realms with spaces in the name, like Area 52, replace spaces by dashes, like area-52.
Damage Comparisons
Disclaimer: The methods used here to evaluate damage were the usual methods used by us, which can be found in our previous work, like the Raid Stat Weights work.
All the tests were run based on a 26k Stat budget profile, averaging 905 ilvl with 4-piece bonus. Changing ilvl and trinkets brings only minor changes to the graphs and can be ignored.
To understand the differences between AS and SL, we will be looking into three different major situations: Single Target, Two Targets and Two Separated Targets. You might ask, "Ansh, why the targets being separated or not matters? Not like Mind Sear is relevant, LOL" Why yes, that's true, Mind Sear doesn't, damage wise, influence that much the results, BUT, having the targets together, means Mind Flay will generate more Insanity, thanks to Mind Sear, which leads to longer Voidforms, which leads to more damage to your DoTs via Mass Hysteria, meaning more value to Mastery and SL. Also having targets together means you will spend less globals refreshing your DoTs because Void Bolt will extend the duration from both.
Single Target
For single target, the situation of both talents is really alike, as seen below:

´

As we can see, changing from 10k to 12k haste barely impacts the final DPS result (both graphs are in the same scale), but changing the balance between crit and mastery can impact the DPS up to around 8%.
The plateaus found on the graphs mean that dropping mastery in favor of crit, for that talent, in that interval, doesn't impact its DPS, meaning they have the same value. It is important to notice that the highest DPS value from AS isn't when you put every point into crit. That happens because of the quick Diminishing returns that you have with crit as a stat in general and how nice Shadow scales with Mastery now even without SL.
Reading from the results shown, if your CMD is above 0.45, for single target, you'll get better results running AS, otherwise, run SL.
Two Targets Stacked


For two targets, the game changes a bit. Now we start seeing SL starting to win even with high CMD, considering high haste, while, with a lower haste level, the split between them stays around the 0.4 marker. Extrapolating the results from two targets to more targets, you can expect SL to have the same trend, but more accentuated, being better faster the higher your haste is, since a bigger slice of your damage will be done by your DoTs.
Two Targets Separated


As I mentioned at start of this section, when the targets are separated, there's a different curve for both, with AS winning in any scenario as long as CMD is above 0.25. Extrapolating this for more targets, the CMD threshold should increase a bit, because more damage is being dealt by the dots that scale better with Mastery than Crit, but still AS should be slightly better in most CMD values.
Conclusion
Overall, Mastery is better stat both gameplay wise and damage wise for Shadow than Crit. So you should be aiming for it, while respecting the values mentioned in the Stat Weight Discussion post. Use this research as a way to know what is best for your right now rather than a guide to what you should aim for.
In Nighthold and Dungeons alike, most of multitarget scenarios are stacked on top of each other, so the results for Two Targets Stacked represent better the average scenario for those scenarios and is a good guideline to follow.
Re: [H2P] AS x SL - When to pick each talent?
by Jockel » 30 Jan 2017, 09:00
What i've found very interesting about the graphs for me was:
- Going up from 10k haste to 12k haste is actually a (probably very very small) DPS decrease for almost every scenario
- Using AS can be worth it in quite a lot of scenarios
- Gear to go for, should probably still be Mastery, since I hate the travel time of AS
- Ain't separated fights quite a lot of the fights in NH? Chronomatic, Aluriel is pretty balanced I guess in terms of target stacking, Etraeus is seperated, Guldan quite balance with probably stacked targets early in fight, Elisande mixed, High botanist mostly stacked, can't tell about Tichondrius
- Without 4pc the value of crit should be even higher right now
-
- Going up from 10k haste to 12k haste is actually a (probably very very small) DPS decrease for almost every scenario
- Using AS can be worth it in quite a lot of scenarios
- Gear to go for, should probably still be Mastery, since I hate the travel time of AS
- Ain't separated fights quite a lot of the fights in NH? Chronomatic, Aluriel is pretty balanced I guess in terms of target stacking, Etraeus is seperated, Guldan quite balance with probably stacked targets early in fight, Elisande mixed, High botanist mostly stacked, can't tell about Tichondrius
- Without 4pc the value of crit should be even higher right now
-
Re: [H2P] AS x SL - When to pick each talent?
by Noexit » 30 Jan 2017, 11:39
Its hard to directly compare because this is assuming a higher ilvl and 4 piece bonus, something that is surprisingly hard to get it seems (or I just have really bad luck). Even with bonus rolling on tier bosses I am still at 1 piece lol.
But this at least makes me stress even less about hitting that 12k haste number, something I have struggled to do with my gear options. I am sitting somewhere between a .4 and .5 CMD depending on which Legendary I am wearing currently. So almost right at the convergence point on the graphs outside of the 2 separate target ones.
But this at least makes me stress even less about hitting that 12k haste number, something I have struggled to do with my gear options. I am sitting somewhere between a .4 and .5 CMD depending on which Legendary I am wearing currently. So almost right at the convergence point on the graphs outside of the 2 separate target ones.
Re: [H2P] AS x SL - When to pick each talent?
by Treid » 30 Jan 2017, 21:02
Does this assume Shoulders and Bracers on?
Re: [H2P] AS x SL - When to pick each talent?
by Anshlun » 31 Jan 2017, 03:10
Treid wrote:Does this assume Shoulders and Bracers on?
For all TC material in this forums, if it doesn't explicitly say it assumes any legendaries, it doesn't have any legendaries.
Noexit wrote:Its hard to directly compare because this is assuming a higher ilvl and 4 piece bonus, something that is surprisingly hard to get it seems (or I just have really bad luck). Even with bonus rolling on tier bosses I am still at 1 piece lol.
But this at least makes me stress even less about hitting that 12k haste number, something I have struggled to do with my gear options. I am sitting somewhere between a .4 and .5 CMD depending on which Legendary I am wearing currently. So almost right at the convergence point on the graphs outside of the 2 separate target ones.
Not really hard to compare because changing ilvls just means you have a small budget of stats to work on, but the interaction between them is the same. Having 4pc or not is just a matter of time. No reason to gear towards not having it, knowing you will get it in a week or two at this point.
Re: [H2P] AS x SL - When to pick each talent?
by Gomes » 31 Jan 2017, 11:16
Anshlun wrote:Not really hard to compare because changing ilvls just means you have a small budget of stats to work on, but the interaction between them is the same. Having 4pc or not is just a matter of time. No reason to gear towards not having it, knowing you will get it in a week or two at this point.
It would maybe suggest though, that without 4 piece, talenting AS could possibly be ahead of SL on a large number of fights in NH.
Re: [H2P] AS x SL - When to pick each talent?
by michel1056 » 31 Jan 2017, 14:58
Sorry to ask anshlun but wat is PriestBot?
Re: [H2P] AS x SL - When to pick each talent?
by Mak0 » 01 Feb 2017, 11:53
It's the bot for the discord channel.
Basically you can pm the bot the command explained here with the right input and it will display the ratio for your character so you can read the charts more easily.
Basically you can pm the bot the command explained here with the right input and it will display the ratio for your character so you can read the charts more easily.
Re: [H2P] AS x SL - When to pick each talent?
by michel1056 » 01 Feb 2017, 14:48
Thanks for the information :)
I will join that channel right away XD
I will join that channel right away XD
Re: [H2P] AS x SL - When to pick each talent?
by LaChoz » 07 Feb 2017, 16:50
once again a great tool for helping our fellow SP's ! many thanks
Users browsing this forum: No registered users and 0 guests
- Board index
- All times are UTC - 8 hours [ DST ]