Stat Weights for MoP Level 90

[derevka]

Lightwell ticks don't proc Echo of Light and they don't benefit from your Haste.

EOL no, Haste yes. You can gain ticks.... and you could do that on the Cata client, not just MOP client. Which of course increases the importance of haste, at least to the reachable breakpoint.

[Revulva]

I will add a check for haste thresholds on Lightwell. It doesn't really change the weight for Haste, it just gives a high value to one particular value of haste: 16.67%.

Pretty soon we'll have our level 90 optimizer working, and I can start playing around with gear to see what levels of haste are realistically obtainable, so we can see if recommending the 16.67% haste threshold as something people should go for (maybe at the expense of some intellect even) makes sense.

[LordMonkfish]

I don't feel that the spreadsheet is currently taking Spirit Shell properly into account.

At the moment, you simply have the option of selecting an 'uptime' for Spirit Shell, not reflecting the typically higher throughput spell selections when this cooldown is up.

The spreadsheet would more properly account for Spirit Shell's worth by including it in the spell usage sheet, where you could input the healing done by the spell (and overheal) from the log for an encounter.

The reason I think this would be a significant change is that Spirit Shell is making up a gigantic amount of my healing (and I'm sure most discipline priests).

For any fight where there are periods of raid-wide damage that Spirit Shell can be used to absorb, it'll probably be somewhere in the region of 40% of our total healing done, hence why mastery is looking such an attractive option currently.

[Revulva]

So, you're suggesting that when you use Spirit Shell, you are only using Prayer of Healing... or, only using Flash Heal and Greater Heal?

It only works with 4 spells... and the only low-throughput spell would be Heal.

If you look on the "Theoretical Stat Weights" tab, you can see some stat weights I calculated for each individual spell. I did this by only using that spell in a fight. So, if you look at Prayer of Healing - you can see that Mastery doesn't pull ahead even if you use Prayer of Healing exclusively while Spirit Shell is up.

I am not arguing that you do tons of absorption with Spirit Shell, I'm just pointing out that the math indicates that you would increase the throughput more by getting Crit. I will double-check the Spirit Shell calcs and also add an assumption that you never use Heal.

[Revulva]

Actually... looking at how I implemented it, there is a bug. I'm massively over-estimating Spirit Shell. By adding in a spell breakdown that splits up the Spirit Shell uptime per spell, it will reduce the overall effectiveness.

I will post when I fix it and update the numbers.

Trust me, I don't have anything against Mastery - I just like doing theorycraft and analysis - and the math doesn't point to Mastery being the best throughput stat.

I feel like a lot of stat priority suggestions for healers have been based too much on anecdotal evidence and looking at healing meters for a few raids... which is a massively inadequate sample size. A more complete analytical model can help us maximize performance.

[Revulva]

Ok, I fixed the calculations for Spirit Shell in the spreadsheet. I created an input area on the "Special Calcs" sheet that lets you pick a breakdown of spells used during SS as well as the percentage of max SS uptime you achieve in the fight. I am optimistically assuming that you would only cast Heal, FH, GH, or PoH during SS, and that you would cast back-to-back for the whole duration.

I was able to calculate some thresholds of spell usage where Mastery becomes more valuable than Crit for Heal/FH/GH/PoH. What I mean is, lets say you cast the spell 100 times. For Mastery to be stronger than Crit, you will need to cast it at least X times while SS is active.

Heal: 100%+ (Crit > Mastery always)
Flash Heal: 60%
Greater Heal: 100%+ (Crit > Mastery always)
Prayer of Healing: 25%*

*This assumes you hit 5 targets every time. The number goes up if you hit less targets.

So, as I have mentioned before, The relative value of stats can change dramatically for Discipline based on how you use your spells.

I tried creating a theoretical breakdown of spells, just to see how it would turn out. A "problem" I am having with the model is that the Spirit Shell calculation is extremely optimistic. Lets say you used Flash Heal 10 times and Prayer of Healing 15 times in a 5 minute fight... right now my calculation would assume that you only used those spells with Spirit Shell up, since that would have been possible. Thinking to myself... I know that I would use those spells at other times during the fight, at least once in a while. So, I also implemented a max percentage of the spell usage to allow during Spirit Shell.

This should give the necessary flexibility to calculate when Mastery will become stronger than Crit, and vice versa.

So, some notes on how to decide if, based on your spell usage, Mastery or Crit is stronger:
1.) How often do you use Flash Heal, and, how often is that done without Spirit Shell active? You need to limit your usage of this spell outside of SS to make Mastery more effective than Crit.
2.) How often do you use Prayer of Healing? Mastery is good for this spell. As long as you are using it about 25% of the time with Spirit Shell, Mastery will give you better throughput than Crit.
3.) Mastery is awesome for PW:Shield. The more you use PW:Shield, the better a stat it is.
4.) Mastery is weak for Heal, GH, Renew, PoM, Penance, level 90 Talents. If you aren't using PW:Shield, Flash Heal (with SS), and PoH for 60% or more of your spells (not total healing, this is spell usage), Mastery is going to even out or fall behind Crit and Haste.
5.) Divine Aegis scales better with Crit than with Mastery.

Lots of interesting stuff I'm finding as I refine these calculations. I will be interested to see what is actually possible in different situations, based on the limited Mana. Mastery is very effective for PW:Shield, FH, and PoH - which are all costly spells. Will we have the mana to keep up a spell usage pattern that allows those to make up well more than half of our spell usage? I can't answer that yet, since I don't have a level 90 disc. priest on beta.

[LordMonkfish]

Thanks for the speedy work, I look forward to playing with the spreadsheet when I get the chance.

I used your spreadsheet throughout Firelands and enjoyed the way it takes your spell usage into account first, and then gives you stat weights - which I feel works very well for healers.

At the moment, when I'm using spirit shell (SS), it's mostly as a raid-healing CD and I'll more or less cast Prayer of Healing for 100% of the CD's uptime. I'll simply pop SS and rotate casts of PoH on each raid group, perhaps throwing 1 Greater Heal on the tank for good measure.

I use a atonement and PoH heavy "rotation" as disc, typically (of course it can vary). Keep PW:S on tanks at all times and any player who will take a predictable and significant amount of damage of single-target damage (think the pounce the Shannox dogs did).

Anytime I am not casting PoH I am casting Holy Fire & Penance on CD with smite in between, popping SS prior to raid damage and proceed to coat the raid with it. I'm interested to see whether this style of healing benefits mastery due to the high PoH usage, or whether my similarly high atonement usage will change things up.

[Revulva]

I think I need to add in Holy Fire, I still only have Smite, I believe. I'll add that in sometime today.

I'm glad at least a couple folks are making use of the spreadsheet ;) I'll probably have to rely on using information from other people's parses in MoP... I think I'll be playing a Warlock, at least to start. Don't tell anyone...

[LordMonkfish]

If you add Holy Fire, remember to add an option for a glyphed version (removes the cast time)!

I'll be happy to provide parses if it'll help the work you're doing - this sort of spreadsheet helps support different styles of play.

I remember playing with different Holy playstyles in a previous version of the spreadsheet and was fascinated by how crit and haste overtook each other in different situations.

Shame to lose you as a priest, but warlocks are looking great it has to be said. I really enjoy the changes to shadow and may be going shadow as a main spec for the first time in my raiding career, but the pull of casual theorycrafting (or at least, experimenting on the work others have done) for healing priests brings me back into a healing frame of mind!

[Revulva]

I have implemented Holy Fire and fixed up a few small bugs with Divine Aegis and Spirit Shell calcs (small enough that results weren't affected).

I must say... Discipline Priests, from a stat priority theorycraft perspective, are one of the best designed and most interesting specs in the game. There is no garbage stat. You can play the class with multiple healing patterns and styles, each favoring different stats. This is the kind of design that I would love to see in every class/spec in the game.

Holy's mastery doesn't really have a chance for this type of interesting interaction. If they buffed it, it would be a clear winner. There isn't a way to get the interesting balance/interaction that Disc. has on a mastery that is just a flat healing increase, more or less. Besides being theoretically weak, Holy mastery is "slop healing" that doesn't affect actual spell usage. If you can't tell, I'm not a huge fan of the design of Echo of Light ;) I like the design of the spec, from a play perspective, just not the mastery.