Tuesday, August 14, 2007

Light's Grace Macro

I was playing around with a macro to keep Light's Grace up on me today. In general, Flash of Light is a superior spell to downranked Holy Lights. If you compare a downranked Holy Light to a Flash of Light with equivalent mana consumption (for example, HL4 vs FoL6), you'll see that the Flash of Light has a higher base heal-per-second and--more importantly--a higher coefficient-per-second.[1]

(Coefficient-per-second refers to the amount of +healing from your gear that is applied per second. If a spell has a higher coefficient, it means that there is always a point where it will heal for more than a spell with a lower coefficient.)

However, a downranked Holy Light does have the advantage that it will give you the Light's Grace buff, which can be invaluable when you need to throw a big heal quickly. A two second HL11 can be the difference between a dead tank and a boss kill.

So I decided to try out a macro that casts Flash of Light most of the time, but casts a downranked Holy Light just enough to keep Light's Grace up:

/castsequence reset=10 Holy Light(Rank 5), Flash of Light, Flash of Light, Flash of Light, Flash of Light, Flash of Light

What this does is cast HL5, 5xFoL7, and then repeat. You can play with the number of FoLs and the reset timer to ensure that Light's Grace never drops off. I personally don't constantly spam, so I have the HL5 coming a bit earlier than necessary. You could add a couple of FoLs to squeeze out the maximum effect.

I tried out on Gruul, and it seemed to work reasonably well. It felt a little bit easier to nail the OT with a 2s HL11 right after coming out of a Silence. The majority of our healers are paladins, so the Hateful Strike, Silence, Hateful Strike sequence is a killer near the end of the fight.

[1] Technically, if you bring Blessing of Light into the picture, FoL6 has a lower base heal-per-second than HL4. But FoL6 becomes strictly better at about +900 healing, which is relatively trivial for a raid healer to achieve.


Anonymous said...

I'm a new Holy Pally.....and I wish I could understand that.

Being a retnoob from 10 to 70 pre-Kara was so much easier.....SoR,JoR,CS rinse and repeat.

I will have to try it next Gruul run.


Kashinboner of Aman'thul

Thomas said...

From the experience I've had, FoL is in general a much more efficient way to waste mana than HoL except in a few situations. At around 1700 healing (raid buffed of course), a target with Blessing of Light (along with Libram of Souls Redeemed from the opera event in karazhan) will actually get more bang for the buck with Holy Light rank IV.

Benefit, more heal/mana.
Downside, long cast.

Honestly, if you're solo healing or healing an MT its good to have Light's Grace up, but in a balanced 25 man you'll have resto shammys and druids who can NS a huge heal faster than a 2 sec rank XI Holy Light.

Still, some fights (such as if you're the main healer on the mage tank for High King Maulgar) its nice to keep up the buff as much as possible, and it is useful to a great extent. Mad props for figuring out a macro for it.

GSH said...

thomas, I believe that is incorrect. For high values of +heal, FoL is superior on a per-second basis. The key is to compare spells with similar mana consumption. Don't look at FoL7 versus HL4, look at FoL6 vs HL4.

GSH said...

Heh, and as to a balanced 25-man, our healing force was 5 paladins and 2 druids. :)

Galoheart said...

Interesting post been a Paladin myself just not the everyday dedicated healing kind.

However everyday you learn something new.

Thomas said...

WARNING: major theorycrafting ahead.

Firstly, the spells I'm testing:
FOL VII: 180 mana, 458-513 heal, 1.5 sec
HL IV: 190 mana, 322-368 heal, 2.5 sec

My +healing is currently at 1546 completly unbuffed (which is how I will be doing this test). Armory Profile

I tested this by casting the certain rank of heal 30 times (excluding crits) and recording the average and extrema. The average is the average of all heals in the test, not the average of the two extrema. To mantain an even casting time I clicked off Light's Grace after every cast of Holy Light. Everything is rounded to 1 decimal.

Without Blessing of Light:
Lowest: 949
Highest: 998
Average: 961.2
Average HpS: 384.5
Average HpM: 5.1

Lowest: 1262
Highest: 1319
Average: 1291.4
Average HpS: 860.9
Average HpM: 7.2

So Flash of Light easily wins in all respects; both mana per heal, time per heal, and base healing were all greater.

With Blessing of Light/Libram of Souls Redeemed:
Lowest: 1721
Highest: 1760
Average: 1737.5
Average HpS: 695
Average HpM: 9.1

Lowest: 1584
Highest: 1635
Average: 1601.3
Average HpS: 1067.5
Average HpM: 8.9

So as you can see, FoL will remain the most time efficient heal even with the extra 685 healing to Holy Light from BoL/Libram. However, the average heal per mana was slightly lower than the downranked Holy Light.

This test is nowhere near perfect, it would need a much larger sample space from multiple Paladins before I would even consider using HL over FoL, however, it does show that on a target with BoL you can get nearly the same or better mana efficiency while keeping up Light's Grace by using Rank IV Holy Light.

Interestingly, if you factor in the reduced cast time of HL thanks to Light's Grace the average HpS of HL IV with BoL goes to 868.8, making it close to equal with FoL VII without BoL.

As it remains though, FoL is in general a better waste of mana than HL IV except situations when you need to keep Light's Grace up, in which case simply throwing out a HL every 15 seconds then reverting back to FoL is your best option (as illustrated by the cool little macro above).

Please excuse the extremely long comment.

GSH said...

thomas, try your experiment with Flash of Light Rank 6 instead of 7. I don't really like HpM as a metric, and I'll see if I can put up a post saying why. It's an accurate number, but I find it lacks context.

As for your numbers, it's odd, but I can't get my theorycraft to match your numbers, unless the Healing Light talent is only affecting the base portion of the heal. Which used to be the case, but I thought it had been changed.

Steven said...
This comment has been removed by the author.
Steven said...

This tidbit might help ...

Both Blessing of Light and Libram of Souls Redeemed do not use Coefficients on your heal spells.

They are a flat phat +Heal to HL and FoL reguardless of rank. If you read the tool tip of Libram of Souls Redeemed it doesn't say "up to 105 healing" it states " by 105"
or something to that effect. The catch being it does nothing if the Target doesn't have Blessing of Light on them.

Its why it pains me to put Kings on Tanks and not Light sometimes.

Thomas said...

I got a marginal upgrade during our SSC run tonight (hooray for trash drops) so I'm running about 32 healing higher than I was last time I tested. As its FoL, the +healing coefficient is ~42.86%, so effectively each heal was 13.7 larger (so I simply subtracted 14 from each heal, its not perfect but I don't quite have the time at the moment to redo the first test. The average is still all the heals added together after adjustment). All the original numbers are in ().

Anyway, my results for FoL Rank VI:

FOL Rank VI: 140 mana, 356-396 heal, 1.5 sec

Without Blessing of Light
Lowest: 1135 (1149)
Highest: 1177 (1191)
Average: 1154.5 (1168.5)
Average HpS: 769.7 (779)
Average HpM: 8.2 (8.3)

With Blessing of Light
Lowest: 1465 (1479)
Highest: 1507 (1521)
Average: 1481.9 (1495.4)
Average HpS: 987.9 (996.9)
Average HpM: 10.6 (10.7)

This I found very interesting. With Blessing of Light/Libram Rank VI FoL seems to give the most bang for the buck out of everything I've tested, at least mana wise. Still doesn't seem like much application in a raid situation though, with the amount of damage bosses do I think its perfectly acceptable to sacrifice the extra mana efficiency (which can be easily compensated for in regen gear) in exchange for the extra 70 HpS Rank VII is giving.

As for the Healing Light problem, I was also under the impression that it no longer solely affected the base healing, but of course, bBizzard does a lot of strange things. I'll look for any info I can find on it.

GSH said...

steven, you are correct. The way BoL works is the major thing keeping downranked HL even close to FoL.

I'm finding that if I don't have the 30% multiplier, my expected value is below thomas' range. If I do have the 30% multiplier, it's above thomas' range. So either my math is messed up, or it's only applying to part of the healing.

GSH said...

Bah. Was using the wrong value for Healing Light. For some reason I was thinking it was 30% instead of 12%. It's applying properly to +healing, as far as I can see.

Aelwen said...

I don't think that macro will work as you want. As far as I know, the reset timer don't consider when you cast the first spell in the sequence, it sees when you click the macro, so everytime you click it the timer will reset. For example, if you cast HL then FoL it'll reset 10sec after FoL not HL, unless you cast another FoL within 10secs. If you cast the second FoL the sequence will reset 10sec after the second FoL.

GSH said...

aelwen, you're probably right that the reset timer should be shortened by a few seconds (maybe to 5-7). It really depends on how long you think you'll go without casting a FoL.

If you're a spammer, the reset doesn't really matter. The thing is that you want to avoid casting the Holy Light as much as possible. Too short a reset time will lead to casting a HL5 when you more than enough time to finish out the sequence.

Anonymous said...

Great site, wish you had an RSS feed.

GSH said...

Hmm. There is an RSS feed. The link must have gotten lost at some point. I've put a link up on the sidebar.

Anonymous said...

Overall your assumptions don't bring into account MP5 versus Spell Crit and Illumination. With BoL up not FoL Rank 6 is less HPS than HL Rank 4. With the assumption that you have 25% crit with Holy Light you'll find that you'll have significantly more mana at the end of the fight and at about 1500 + healing you'll be hitting your target for about 2.8k per heal.

Much theorycraft aside, you make a lot of assumptions in your post.