Magicka DPS Set Bonus Comparison - MATH (pre-HotR)

dpencil1

@ZOS_RichLambert


I made pretty graphs for you!

dpencil1

The OP has been updated with Comparison G, a new full set of calculations where all stats we have been considering are buffed, similar to the fully buffed set combination comparison. The Cumulative Comparison chart was also updated to include this data, which bumped Necropotence into the bottom spot. Updates to the OP are called out within it.

Morgul667

This is great, thanks

I wonder what this would become with Ilambris undaunted set

dpencil1

Thanks @Morgul667
Assuming you would be using a staff rather than dual weild, you'd need to settle for one 5-piece set to use Ilambris. You won't benefit fully from Netch because you'll want some source of fire damage to proc the fire portion of Ilambris. Spinner may be the best choice if you know you're not going to be over the penetration cap. Mother's Sorrow would not be a good choice as the sole 5-piece, and you're losing the operational benefit of Julianos as a 5-1-1 because your monster pieces can cover that. Scathing Mage may be a good choice, perhaps paired with 4 Infallible Aether and Precise weapons. With that set it all comes down to uptime. You need more than 60% or it is a loss compared to Julianos. If you are a dedicated pet sorc, the jury is still out.

WildWilbur

Wow, dude, you have way too much time... But very insightful!

Morgul667

Thanks

Sru

Just .. awesome! Thank you

catch22atplay

Let's just compare Necro vs Julianas for a moment here. With my sorc and 157% magicka. 20% increase to spell damage +8% for 4 sorc spells slotted that each increase sp by 2%.

Necro
2pc 1096 magicak
3pc 1096 magicka
4pc 1096 magicka
5pc 3150 magicka

1096*3*1.57=5,162
3150*1.31=4,127

Total magicka=5,162+4,127=9,289 magicka

this equates to about 885 spell damage sort of

Julianas
2pc 1042 crit value=4.76% crit chance
3pc 1096 magicka
4pc 1042 crit value=4.76% crit chance
5pc 299 spell damage

4.76*2=9.52% crit chance
1096*1.57=1,721 magicka
299*1.28=383

1,721 magicka is approx 164 sp
383+164= 567sp
crit is a bit more difficult as it depends on all your stats.
4.76% crit= 144 to 174sp i'll average it to 159. The 28% increase is already included
159*2=318sp
567+318=885sp

Well would you look at that. Seriously i didn't fudge the numbers. I have my own spreadsheet that calcs this out and this is what i can up with. No my math is not 100% accurate. At least not the crit% is = to x sp. I did have to increase crit damage to +30% or 80% if you prefer to get the high of 174sp in case you were wondering. Now can Julianas beat Necro sure it can and vise versa. All depends on your stats, the spells you use, and if you use a pet, etc.

Note that i ignored pet damage that only increases with magicka. So a Pet build Necro wins for me in my scenario.

Not sure why people like netch's. It increases shock damage by 400. Ok maybe that's better then 299sp*1.28=383 sp. But that's 17sp. Now if you use Netch and by using it you loose Mettle's +6% magicka you loose out by quite a bit. For me this is -4.4% damage.

Geez so many variables. Glad i have a spreadsheet. Seriously though the math isn't that hard. The OP just gets more off as it gets updated.

Zedrian

@dpencil1 Thank you for these comparison for the next patch. It appears to me the Effective Spell power calculation on UESP does not take into account proc damage, and only considers Spell Power that will affect skills tooltip. When you replace a 5/5/1 piece (5 julianos, 5 necro, 1 ilambris) with a 5/4/2 (5 Necro, 4 moondancer, 2 ilambris) the effective Spell power obviously drops. Do you have a method or suggestion how to consider proc damage when comparing sets?

Dymence

dpencil1 wrote: »

While I'm at it, might as well actually do Scathing too. I don't know the most common uptime so someone can correct me if you have a lot of experience parsing with it. I know that below 60% it would have a lower 5-piece bonus than Julianos and that the 2-4 piece bonuses are identical, so I'm going to go with 80% uptime and see what we get. I will use Comparison D for this one, since we really need the extra Crit chance to reasonably get to 80% uptime.

516 x 0.8 = 412

Against our baseline we would get:
31096 Magicka, 2412 SD, 69.7% Crit, 80% Pen, 50% Crit Dmg

31096 ÷ 10.5 = 2961 + 2412 = 5373
0.697 x 0.5 = 0.3485 + 1 = 1.3485
5373 × 1.3485 = 7245 x 0.8 = 5796

5796 - 5051 (Comparison D baseline value) = 745 over base

Compare that to Julianos with 2% magicka for completing the 5-1-1 at 688 and 100% Lightning damage Netch's Touch at 765. A very strong showing indeed, assuming you could consistently land 80% uptime.

Average scathing uptime would be between 64 and 72%, off the top of my head.

dpencil1

@catch22atplay
You lost me at "157% magicka." I think you are making an unfounded assumption to get to that %. Around the max you should be getting is 33%. Also remember Julianos has one max magicka bonus as well. I'm happy to entertain your ideas but you need to show your work for how you arrive at your starting position more clearly.

@Dymence
Thanks. Do you know what sort of crit chance people usually have to get that 64-72% uptime?

@Zedrian
Great question. The main problem is that you can't say that some Effective Spell Damage equals some specific dps. It's a general guild for what would be more or less powerful, but there are other factors that will come in to play to determine the actual dps someone does. The only thing I know of to do is actually test dps on a dummy/raid with different setups to compare how proc sets compare. What I've seen is that procs have a tendancy to do between 1-2.5k extra dps. The fact that they do not crit, combined with the big buff to crit bonuses makes proc sets someone less desirable than they were previously.

One example how higher Effective Spell Power doesn't necessarily equal more dps: on my Sorc pet build, I would have a higher ESP with Bound Aegis, but I get higher single target dps with Twilight Matriach instead. Lots more variables come in to play to determine actual dps.

Dymence

dpencil1 wrote: »

@Dymence
Thanks. Do you know what sort of crit chance people usually have to get that 64-72% uptime?

Not sure about the actual crit chance people have, but it's safe to say they run a monster set with Moondancer, Scathing and Thief mundus. So you're only really getting the crit from Scathing and Thief at that point. Some might also run Infallible instead of Moondancer and Nightblades get slightly higher crit due to passives.

dpencil1

Ok, that looks like about 70-75% then. So the 80% uptime I used is probably the maximum to reasonably expect if someone specs fully into crit.

Subversus

Can someone help us less clever individuals understand what will be BiS for pve next patch? Is MS/julianos/1undaunted going to be competitive or is netch/MS going to win?

Septimus_Magna

Subversus wrote: »

Can someone help us less clever individuals understand what will be BiS for pve next patch? Is MS/julianos/1undaunted going to be competitive or is netch/MS going to win?

Julianos+MS will be better

dpencil1

@Subversus
There is not going to be one BiS or go home this patch. The difference between all of these set combinations is very small. If you look at the OP, you'll see that I mention above the Combined Set (fully buffed) chart that I remove 10000 points from each to better show their difference. This is what that same chart looks like with the full values applied:



So the difference from the highest value and the lowest value:
Spinner/Mother:11012
BSW/Necro: 10428
is only 584...that's only a 5.6% difference.

And if we exclude the lowest performing set in the list, Necropotence, the gap is even smaller:
Spinner/Mother: 11012
Julianos/BSW: 10733
Only 279!...that's only a 2.6% difference!

So just about all of the combinations listed will be viable. It will come down to what will work best for you given other factors you can use to increase the effectiveness of those sets, and what your overall playstyle is, as well as group composition.

Subversus

@dpencil1 one word, WOW! This is amazing from the developer team, I have never seen such balance in regards to set drops. This is amazing, actually! Thank you for your outstanding work in getting this data.


@ZOS_GinaBruno I never thought I'd say this, but in this regard you actually did a good job.

Texecutioner187

To make sure I understand these charts... the points and numbers in the first one is total spell crit derived from the set, while the latter charts is spell power increase from the combo?

Moloch1514

So....revert Necro to 4000 and address the shield stacking in PvP via Battle Spirit. Done.

@ZOS_GinaBruno @ZOS_RichLambert should take a look at these charts.

dpencil1

jgruberman wrote: »

To make sure I understand these charts... the points and numbers in the first one is total spell crit derived from the set, while the latter charts is spell power increase from the combo?

@jgruberman
Nope. The best way to understand what I did is to actually go backwards in the OP. I didn't want to put the important results way at the bottom, so I put the original calculations and first set of charts down there and put the final results on top. Basically, the way Magicka, Spell Damage, Crit Chance, Crit Damage, and Spell Penetration all work together is calculated in a certain way which leads to a number called the Effective Spell Power. I used a baseline set of parameters for each of these values and then added the various set bonuses to them, making note of the difference in ESP from the baseline. I then went through and buffed each of the variables independently to see if this would change certain sets' relative power, which in some cases it does. These are the charts Comparison A-G. Then I took all those numbers and added them together to get a kind of average. That's the Cumulative Comparison. Then I went back through and recalculated everything, this time in pairs, but I only did a no-buff version and a fully buffed version instead of bumping each individual variable. Those two versions are the LOW stat Combined Comparison and the HIGH stat Combined Comparison. The only set that really sees a dramatic change between those comparisons is Mother's Sorrow, since it basically just amplifies whatever raw power you are bringing to the table.

So Mother's Sorrow will actually be a bad choice for lower level/cp players and a good choice for max cp high dps players.

Aurielle

dpencil1 wrote: »

jgruberman wrote: »

To make sure I understand these charts... the points and numbers in the first one is total spell crit derived from the set, while the latter charts is spell power increase from the combo?

@jgruberman
Nope. The best way to understand what I did is to actually go backwards in the OP. I didn't want to put the important results way at the bottom, so I put the original calculations and first set of charts down there and put the final results on top. Basically, the way Magicka, Spell Damage, Crit Chance, Crit Damage, and Spell Penetration all work together is calculated in a certain way which leads to a number called the Effective Spell Power. I used a baseline set of parameters for each of these values and then added the various set bonuses to them, making note of the difference in ESP from the baseline. I then went through and buffed each of the variables independently to see if this would change certain sets' relative power, which in some cases it does. These are the charts Comparison A-G. Then I took all those numbers and added them together to get a kind of average. That's the Cumulative Comparison. Then I went back through and recalculated everything, this time in pairs, but I only did a no-buff version and a fully buffed version instead of bumping each individual variable. Those two versions are the LOW stat Combined Comparison and the HIGH stat Combined Comparison. The only set that really sees a dramatic change between those comparisons is Mother's Sorrow, since it basically just amplifies whatever raw power you are bringing to the table.

So Mother's Sorrow will actually be a bad choice for lower level/cp players and a good choice for max cp high dps players.

Interesting... Right now, I'm running 5 Julianos / 4 Necropotence / 2 Willpower and it works pretty well (not BIS obviously, but I don't run vet trials, so I don't really care). I'm currently grinding out Mother's Sorrow so I can go 5 Julianos / 5 Mother's Sorrow / 1 Monster (yet to be determined) when the buff to Mother's Sorrow hits. As a CP 260, am I wasting my time? At what point, CP-speaking, is it worthwhile to introduce Mother's Sorrow?

MLGProPlayer

Moloch1514 wrote: »

So....revert Necro to 4000 and address the shield stacking in PvP via Battle Spirit. Done.

@ZOS_GinaBruno @ZOS_RichLambert should take a look at these charts.

It would be nice if ZOS used math to arrive at balance decisions. They need to hire guys like dpencil1 and Asayre.

catch22atplay

dpencil1 wrote: »

@catch22atplay
You lost me at "157% magicka." I think you are making an unfounded assumption to get to that %. Around the max you should be getting is 33%. Also remember Julianos has one max magicka bonus as well. I'm happy to entertain your ideas but you need to show your work for how you arrive at your starting position more clearly.

157% if you misunderstood is a 57% increase. Did you miss this part below;

Taking off 1 light piece with Divines (using thief mundus) and which only has a 5 pc power bonus and has 351 magicka enchant i loose 551 magicka

551/351=1.5698 so magicka was increased by 56.98%.

You also get Attribute increases by % based on how many points put into each color and maxxing at 100 points per color. This is factored in before the ones already mentioned. Although i didn't know the total % increase was as high as i've shown. BTW for 100 points spent in a color it's 100^0.56=13.1826%. Again that's prior to the ones already listed.

An old formula i found is Stat Pool=((Base+AP+Gear)*%CP)+Food+Mundus(Divines))*Skills

For me this should be 113.1826*1.31=148.27%

Not sure if the Skills part is additive or multiplicative. But either way it still falls short.

Ya math a bit off there. So that old formula appears to be incorrect or i'm missing something. Maybe it's now 100^0.65 . This puts us at 157.14% when i'm looking for 156.98% or the formula is now completely different. /shrug

This was Asayre's formula btw. But it does seem to need updating. Did we all forget how we got the extra magicka/stam and health via cp spent? Do not ignore this as this must be added. I used removing 1 light piece of armor with 351 magicka and lost 551 magicka. So we lost 57% or in other words we gain 57% to equipping it. If we modify the formula using 100^0.65 instead of 100^0.56 we get the following;

Stat Pool=((Base+AP+Gear)*%CP)+Food+Mundus(Divines))*Skills
so gear was 351 then we times it by %cp which is 100^.65 and that is 19.95% then we times it by our skills % of 31%
we now have 351*1.1995*1.31=552 magicka, ya shoot me i'm off by 1 magicka

What other math would you require. I remove a piece of gear and see how much magicka i loose. I then find an old formula to figure out why there is a discrepency and learn that cp effects magicka increase. I then change the formula to more closely represent my findings after removing said gear. Sure maybe i misunderstood how his formula worked. But removing the gear and seeing how much i lost tells us exactly what we need to know.

BTW i did include the magicka when i did the math for the Julianas.

Subversus

@dpencil1 quick question, do the setups you provided involve vMA weapons or infused regular + monster pieces?

idk

Am I missing it or are basic aspects of a build taken into account in the analysis? CP choices as well as the obvious choice of a mundus to optimize the build with any of these sets goes a long way into figuring out which sets will fair better than others. With the changes to the game it is no longer Theif and sharpened weapons and what all is chosen will have a strong impact on how CP is allocated.

dpencil1

Some quick responses here:

@Aurielle
The low stat baseline I used was 30k magicka, 2k spell damage, 40% crit, 80% penetration, and 50% crit damage. In that scenario, not including CP, mundus, traits, etc. Mother's Sorrow does very poorly.
With the buffed stat baseline, I was using 40k magicka, 3k spell damage, 60% crit, 90% penetrarion, and 70% crit damage. How you get these kinds of increases are up to you, CP, mundus, trait, other gear paired with the set, etc. The point is that when stats are boosted significantly like this Mother's Sorrow does very well. It basically amplifies whatever damage you are already producing. Less starting damage, less effect. More starting damage, more effect.

@catch22atplay
Thanks for the clarification. Let me stew on this for a while and get back to you.

@Subversus
Nope. These calculations don't touch on those things at all.

@Giles.floydub17_ESO
No, those other parameters are not considered. I am aware that they will have an impact on a complete build, and you can purposefully spec yourself into a situation which makes one set or combination of sets perform better than the others. The point here was to examine the sets themselves and what raw power they bring. How you choose to combine and amplify that power is up to you.

geonsocal

@dpencil1 ...wow, kinda weird there's another dpencil out there, huh?

I thought torug's pact was gonna be the new sexy crafted set alongside julianos for magicka aficionados???

any chance spinners with torug's pact > spinners with julianos?

Savos_Saren

Thanks @dpencil1 for adding the base for Silks of the Sun in your comments! I always appreciate your "maths" (even if you're a despicable Sorc)

Texecutioner187

Giles.floydub17_ESO wrote: »

Am I missing it or are basic aspects of a build taken into account in the analysis? CP choices as well as the obvious choice of a mundus to optimize the build with any of these sets goes a long way into figuring out which sets will fair better than others. With the changes to the game it is no longer Theif and sharpened weapons and what all is chosen will have a strong impact on how CP is allocated.

As long as the stats were derived from having the same CP and Mundus stone, won't the comparisons be the same as far as which stack up where? Not being smartassed, I'm actually asking

idk

jgruberman wrote: »

Giles.floydub17_ESO wrote: »

Am I missing it or are basic aspects of a build taken into account in the analysis? CP choices as well as the obvious choice of a mundus to optimize the build with any of these sets goes a long way into figuring out which sets will fair better than others. With the changes to the game it is no longer Theif and sharpened weapons and what all is chosen will have a strong impact on how CP is allocated.

As long as the stats were derived from having the same CP and Mundus stone, won't the comparisons be the same as far as which stack up where? Not being smartassed, I'm actually asking

But some of those sets would bring about different choices with CP and with Mundus. Some stats we get from set bonus cannot be derived from CP while others can which has a further impact. Not even getting into two sets offer advantages to specific classes, that is a different story.