Feedback: Oracle in The War Within

The goal for Hero Talents gameplay is for them to make you ‘what you are, but more.’ The problem is, I’m not a PI machine, and I never wanted to be one. I don’t want to do external PIs, I hate doing external PIs, I don’t want to receive toxic whispers about my use of PI, I don’t want to make suboptimal plays for my character in order to PI other players’ characters, and I don’t want my spec tuned around an external buff.

I would understand dealing with all of this if the priest was a full-on support class like Aug, but since it’s not, I’d rather see PI deleted completely from the game. Holy/Disc is not a buff machine that dabbles in healing. They are healing specs; that’s what they are. More would be healing throughput/options, not a whole tree around gimmicky externals.

This sentiment isn’t unique to me personally, and doubling down on PI for a hero tree seems wildly out of touch to me. Don’t replace it with an RNG 1-minute PI stand-in ability; just get rid of it or make it self-cast only.

My only hope is that this is a rough draft and it’s not too late to reconsider if the PI is staying in the game as it is, because it’s not fun for anyone except the receiving player, and it’s toxic for everyone else in the group/raid.


the talent tree idea sounds really good, but it definitely needs to be simplified.

make the power infusion just a single button you don’t choose the spell, but the spell changes depending on the target you use, if you use it on a tank then it receives an increase in healing, if you use it on an int class a buff to intellect, or buff for agility and etc.

I hope they change the increase in physical damage and magical damage for just a stat buff, some classes with mixed damage will be greatly harmed by this talent.


The concept is intriguing enough – doubt I’d play it and eager to see what Archon holds, but the option of a support aspect is cool – but basing it around a cycling buff brings over the absolute worst aspect of Blessing of Seasons.

And just to be clear my precious angelic feathers are mine as I have no other mobility so the likelihood I’d be giving them to someone else just to get them back for me again is absolutely minimal.

The problem with being a buff bot is that if it’s tuned up, we have no choice but to play it and live vicariously through the power we lend, while watching our other play options languish. Or it’s undertuned and just a waste of talents.

Hopefully the complexity of this can be reduced and Archon offers some compelling enhancements to the priest’s actual power.


Another note; having separate Magic and Phys damage buff on rotation requires the priest to have knowledge of the damage profile of every other spec in the game

Not an issue for some, huge gap for a majority


You should really watch the video they did that accompanies the article.
Really sheds a light to your questioning and might make some more arise.

They said it rotates the buff every second (they aren’t against modifying anything as it stands, they said).

Yup they acknowledge that in the video.
They said another version of premonition on the backburner is increasing main stat instead.

1 Like

This one might need a redo, the idea of seeing in the future and reacting based on your foresight and premonitions is a cool concept, but the implementation and the effects don’t reflect that fantasy or concept.

Obviously, tuning will come (right?) but conceptually quite a few of these nodes just need to be redone.

(please remove PI)


I’ll start with the good, because it’s going to be very short.

The Twisted Fate “Pick a Card” mechanic of how you choose what Premonition to use is neat, and it conveniently shows that Blizzard is aware of League of Legends so I can use it for examples later.

The 3 “base” buffs all being easily attributable in logs (assuming they just work across the board instead of some convoluted whitelist/blacklist) is great, that’s directionally correct.

And now for all the many bad things.


Blizzard can pretend there’s a “game 1” and “game 2” all they want, but in practice there is only the actual game being played by players. Being able to use logs for analysis during progress is incredibly important, and unattributable buffs make this much less reliable and more difficult. It sucks not being able to tell how much of your damage was actually you, it sucks not being able to tell how much damage your external buff actually provided. Holy paladins on live right now have Blessing of Summer, which lets them easily check “hey, was it a good choice to buff the person I buffed?” on in-game damage meters (and by extension logs too). That’s good and how it should be. Immediate feedback is immensely useful as a player, and being able to trust the tools we use is important. On the other hand there’s Power Infusion, which there’s no way of determining the value of outside of simulations. You can’t just check “hey did this guy actually gain a lot of damage from the PI? Should I buff somebody else instead?”

On farm, chasing ranks is fun to a lot of players, and unattributed externals make this significantly less fun because they add an external factor. This sucks when playing a spec that rarely gets externals, because the couple of people who do get them will just stomp everybody else, and it also sucks for specs that often get them, because now if you don’t get them, you will never ever compete. This also includes buffs like Power Infusion, Blessing of the Seasons, Innervate, Wildfire, the Amidrassil healer head enchant, the Everbloom Leaf, Broodkeeper’s Promise etc, not just Premonition.

The 3 base buffs from Premonition should be easily attributable, which is nice, but then you get to the other 2 effects (mana and haste) and it all falls apart. Those cannot be attributed. They’re much too complex because they change decisions on what spells to cast and the number of spells you can fit into other windows.

Externals as a puzzle

The one good argument for externals being interesting is that on progress you can assign them in different spots and on different people to get past specific checks, depending on who has CDs available at the time. For example on Raszageth, I’d give PI to 3 different people over the course of the fight. Well, planning requires predictability right? Enter Clairvoyance, where you’ll randomly titanforge a megabuff that goes on twice as many people (2 of which are randomly selected). Just… what? You can now no longer plan around the external, because sometimes it’ll just end up many times as powerful (2x as many targets and 5x as many buffs per target), meaning you absolutely annihilate the check you were trying to beat and now wasted other resources. Big RNG swings are not fun, regardless of whether it’s on progress or on farm.

The other side to the puzzle coin however is what we’ve seen with Augmentation in raids. It becomes a spreadsheet + reminder Weakaura game. It may be difficult to optimize, but nearly all of it happens outside of actual gameplay. “Well, your first buff should be haste on the mage during CDs, then you give the druid mana, then the warrior gets physical damage, next the mage again, now you wait a bit longer and give the warlock magic damage, mana for the druid again”. This is taking the puzzling too far.

Good players will optimize it and get a lot of value, while weaker players will grossly misuse the spell and be a burden. Neither group has fun, and it can’t ever be balanced for both at once, as we’ve seen with Augmentation. Either it’s balanced around the high end, and weaker players should just never play it, or it’s balanced around some level of weaker player/worse usage and it’s the most broken thing imaginable at the high end where it gets used optimally.


Again, just like Augmentation, the buffs are multiplicative with other effects. This encourages stacking the entire pile of externals (Augmentation, Oracle, regular Power Infusion, Windfury, Blessing of Summer etc) on the smallest number of people possible to leverage multiplication. It specifically encourages stacking them on the most overtuned and burstiest spec available, with bursty specs already being the most powerful archetype in the game.

Healers don’t matter

This tree being as focused on external power as it is means that it’s going to be very difficult to justify pick either of the other trees (depending on spec). That’s because generally healing checks are not very tough to meet, meaning it’s incredibly valuable for healers to be able to give up a bit of their own power in order to power up DPS instead. That’s not fun for the healers who want to actually spend their power budget on their own healing.

Back to League of Legends

So the “Pick a Card” thing is… borrowed from Twisted Fate in League of Legends. That same game can also provide solutions for many of the earlier points, because it turns out that game actually has a wide variety of supports, and those supports have a lot of examples of abilities that:

  • Can be attributed
  • Are not multiplicative
  • Have decisions to be made while playing

Let’s try some examples:

  • Old old Mordekaiser W. This placed a buff on your target and yourself, dealing damage around them and healing based on a portion of that damage. So you want to put it on somebody who’s going to be in range of enemies, and ideally somebody who needs healing. So usually melee or tanks, but sometimes there’ll be ranged stacked near some mobs too (think roots on Tindral)

  • Nami E. You buff an ally, causing their next 3 attacks to deal some extra magic damage (a flat amount, not multiplicative) and slow the enemy. Nami also has a passive that gives movement speed to any ally she hits with a spell. So here you have 3 “modes”, with the primary one being the damage, but sometimes you’ll want the slow on a specific target or help your friend move a bit faster.

  • Milio W. A circle on the ground that (slowly) follows a friendly target. Allies in the circle get healed over time and gain extra range. Milio also has a passive that causes each tick of healing to cause the ally’s next attack to put a minor DoT on the target of the attack. Again you have 3 “modes”: attack range (incredibly powerful in both LoL and WoW), healing and the DoT.

In every case these buffs scale based on the caster’s stats, not the victim’s. This solves the Augmentation issue of “mastery and intellect are the only stats that do anything”. The buffs can also be trivially attributed because they’re their own damage events. This could even be done in a way where you still sabotage in-game meters (for the people who want to “make their friend’s DPS number bigger” as part of their support fantasy), but logs can attribute correctly. You don’t need to sabotage logs in order to have external buffs.


Delete externals, they’re an unbalanceable, unfun mess that obscures information players have relied on for analysis for decades. If they have to exist, make them attributable rather than things like haste, mana or CDR.


I don’t find buffing to be that exciting of a play-style, it’s difficult to gauge whether deficiencies in your performance are yours or your allies’ and it just lacks the gameplay impact compared to hitting the boss with big spells or watching the raid’s health bars fly up during a ramp. I don’t hate the idea like some but just find it very hard to be excited by the idea of playing even the best possible version of this kind of tree.

Looking at it from the PoV of someone raiding on Disc, the biggest issue to me is the mental overhead involved. PI is fine. My weakaura lights up and then I press a button. No problem. I can do that no matter what else is going on. Having to juggle cards and contemplate the ideal target while precariously greeding Smite casts so I can get my pet up for the ramp for the raid damage in 18 seconds sounds pretty miserable and likely to cause more frustration than anything else when you stand in something because you had to juggle 8 different wildly unrelated plates at once. Disc is not the place to tack on more mental overhead. It’s one of the more difficult specs to raid on as is.

I also think that there is going to be a ton of friction with how the buffs are used for optimizing damage vs. healing. A lot of the personal throughput for this tree comes through Twins of the Sun Priestess, but the buffs are going to do basically nothing unless you get them at exactly the right time. The issue is that usually your DPS are not sending CDs at the same moment your ramp lands, and if you don’t have those buffs when your ramp lands you get essentially zero throughput value. With regular PI, you’re going to at minimum get some extra haste to farm pet CDR if nothing else.

This tree feels like it’s trying to add a minigame to a spec that most certainly does not need one. It muddies the direction of the spec and steals your attention for upkeep that simply doesn’t feel meaningful enough to be rewarding. Based on what I see so far, I am dreading the idea of playing Oracle.


LOL this is so funny. I haven’t touch my spirest for ages and decided to do LFR for giggles sine she is only 450ilvl. Totally forgot to PI others and this warlock whisper saying I didn’t PI anyone. i apolgize and mention I haven’t play my priest for awhile so totally fogot to do it. He rage and said I should quit the game.

I said in raid chat xxxx told me to quit the game since I didn’t PI. I apoglize since I havne’t play my priest for awhile and forgot about it. Now since xxxx is being an A hole he will not get PI from me. Another priest in the group also said they won’t PI xxxx anymore even though he was top damage. Kinda of funny and the warlock left after throwing a few mean remarks.

1 Like

Premonition: This is a cool direction to go in, if (and only if) the throughput gains are properly attributed through log hooks. Buffs like a flat % damage increase or a % versatility increase are awesome. However, giving out a % haste buff is NOT great, because haste cannot be re-attributed, and so the priest has no way of measuring the effectiveness of their buff. Discretionary and un-attributable damage buffs make the game less fun in a meaningful, tangible way for players that care about analyzing and improving their gameplay.

(This feedback applies to non-Premonition PI as well. Twins self-buffing the priest with haste is perfectly fine.)

Clairvoyance: For a hero spec themed around prediction, “low chance” randomness seems like a flavor miss. It would be nice if it were actually possible to predict the buffed Premonitions, maybe by e.g. making the buff activate when you cast the Premonition charge granted by Grand Reveal.

A targeting mechanism for the extra Premonitions other than “nearby allies” would be great, otherwise it will suffer from the exact same design flaws that early proximity-based iterations of Augmentation’s Ebon Might had. The gameplay required to ensure that the right people get your buffs is not enjoyable for the person giving them or the people receiving them.

I would like to second the feedback given in the wowhead interview. The current build with the RNG aoe buffs, the complexity of the buffs, the weird interactions with atonement and the complex rotation. The inequity of the buffs for holy vs disc are all very concerning. I would like to see one of the other options for oracle explored, or a large revision to this one.


Here’s my feedback: nuke this from orbit. Do not pass go, do not collect $200. Just send this Oracle tree and everything about it straight into the shadow realm (and let the record show that this does not, in fact, mean “give this to Shadow as well”) and don’t look back as you come up with something that is distinctly not whatever on earth this is.

I don’t think I’ve seen many folks who have expressed anything but frustration and disappointment about the direction this specific tree is going in. It seems incredibly cumbersome to keep track of what buffs you’d want and when; it’s designed in such a way that making the buffs very predetermined as they are now makes it annoying and encourages too much micromanagement to keep track of these (some of these buffs are VERY powerful on paper) and making them RNG instead would just make Premonition a frustrating gameplay loop for completely different reasons.

There’s a lot of things wrong with Power Infusion as an external, but it’s somehow better than this. This takes everything bad about PI and makes it ten times worse.

TL;DR: Terrible idea and I don’t have faith that this will be implemented well. Go back to the drawing board with this one.


The game needs to be slowed down and easier to understand. Not faster and more complicated.


There’s a few things I like in there. I like the concept of 1-minute PI-(ish) in theory. Waste No Time or Miraculous Recovery (instant heal/PoH/Radiance after Premonition or improvements to PW:Life) seem fun to play off of, as does the Twins treatment for Angelic Feather.

There’s more I don’t like or seems off.

  • 150 stacks of Grand Reveal seems rather high, especially for Disc.

  • The whole mechanic of Premonition seems awkward. I hope, at a minimum, the healing buff is the first one to pop up since it’s the most critical time-wise, but also

  • Having the healing buff competing with throughput buffs will also feel bad

  • The buffs don’t even seem better than PI, even given their increased availability, especially the physical buff for non-Warriors

  • Perfect Vision reducing the CD of Premonition to 50 seconds is gonna desync it from most CDs, so that doesn’t really seem to be a competitive option.

Can’t say I’m excited about this one so far

1 Like

So I haven’t dabbled in healing for a long time. So take my feedback with a grain of salt.

However, I think the approach to premonition would be better if it was simplified to a more familiar spell mechanic we have had since vanilla. That spell mechanic being the Weakened Soul debuff you get when you Power Word: Shield someone.

So here is how I would incorporate that effect…

  • Premonition
    Replaces Power Infusion
    30 second cooldown
    • Increases your target’s healing received, damage dealt and reduces damaged received by 15% for 10 sec. Premonition also gives the target Dispirited for 60 seconds, preventing the target benefiting from Premonition.

Then other talents can affect the power and how often you can use premonition.

Things like…

  • Reduces active Dispirited debuff duration on all targets by 1 second each time Prayer of Mending bounces or when you cast Power Word Shield.

The intent/goal would be to have a button you press somewhat often (30 sec) to supercharge someone for 10 seconds. But they have the fatigue type debuff similar to Sated and Weakened Soul to prevent constantly buffing the same target over and over.

However, an issue arises when you have 2 Oracles in the same group/raid which could step on each others toes with the debuff unless the debuff would be Priest specific.

Meh, not the greatest but that’s my .02c for now.

I wonder what would happen if you actively “ignored” spending a talent point in PI?
Would this entire tree just not work? Or somehow still work?

1 Like

Putting aside the insane amount of power you’re suggesting on a 30sec CD.

How would that work with premonition affecting priest when he casts it?

Does he get dispirited too?

I don’t really understand the point of Premonition in its current form. So every 50-60s you can give a single dps player +10% dmg for 10s every 1 minute. That is weaker than Mark of the Wild which is +3% vers to the whole party.

If they want to add a cool support thing they should make it so casting Power Infusion (or Premonition) casts it on your whole party. Nerf it so its 10% haste or damage/healing for 10s or something.

Or give us different variations of Premonition.

  • Defense version works like Zephyr. 20% damage reduction for 5-10s to the party.
  • Offense version is +10% damage increase to the party for 10s.

Remove all the PI mini games. Have the paths focus on different aspects of the spec. Healing/absorbs on the left, mobility in the middle, and right can be utility.


Obviously tweak the numbers as needed. But in regards to your question, I really didn’t factor in all the other talent synergies. If they still work then cool, otherwise I would just like to pose the question… should we have a copy of the ability at all?

The way I see it, if you’re required to pick Power Infusion in the first place and then this outright replaces Power Infusion, then it should just be “better” by all accounts. It should be a superior version which should allow a higher budget of usage and power to make it stand out against Power Infusion.

Personally, I hate that twins exists. I hate having to pick a target and maintain proper uptime of PI with the correct target as Shadow which means I have to spend extra time being in range and if I choose to macro it with my DPS cooldown, I need to use a focus macro on a friendly target for PI which then means I can’t use it for a enemy for my Silence or Mind Control etc.

Meaning I would rather Twins just go away completely.

Then this short cooldown improved version can only be used on 1 target at a time and that’s okay because it’s such a short cooldown imo.

Solo you would be forced to not be able to spam on yourself because of the debuff and so it will be balanced around that as well.

Still bouncing around in my head but generally my approach is to totally rethink the purpose/place of how Power Infusion should be handled both in function and placement with things like twins adding another dimension alongside the main mechanic of the Oracle Hero Path spell and it’s supporting talents.

I think that would be a great approach.

This is similar to my healing priest mechanic that I was suggesting for Discipline.

Mantra of Concentration

  • Place a Mantra of Concentration on a party member. Each time they attack, 15% of the damage they deal heals them and refreshes their atonement. Does not stack with other Mantras.

Mantra of Fortitude

  • Place a Mantra of Fortitude on a party member. Each time they are attacked, 15% of the damage they take heals them and refreshes their atonement. Does not stack with other Mantras.

Honestly looks dope! Lets see how it plays out in game but on paper it looks really fun!