Horridon’s Last Gasp

Horridon’s Last Gasp is one of the new healing trinkets which drop in the newest Mists of Pandaria (MOP) raid Throne of Thunder.  The trinket has the following stats in the order of LFR/Normal/Normal Thunderforged/Heroic/Heroic Thunderforged*:

  • Static intellect: 1218/1467/1552/1657/1728*
  • Mp2 proc lasting 10 seconds: 1375/1656/1751/1870/1964*
  • 1 RPPM (real procs per minute) (here’s an explanation of what this means)
  • No ICD (internal cooldown)
  • Procs off direct or periodic healing spell

*All heroic thunderforged items have simulated stats.  Values are not officially available on Wowhead.

A lot of conjecture has gone into calculating the Mp5 equivalent by guessing at the average time between procs based on the rppm.  With simulated calculations, the Mp5 of the normal version has been stated to be somewhere in the 770 region making it essentially the same as an upgraded Relic of Chi-Ji.  Since it’s been a week from release, I wanted to look at some real time logs to evaluate exactly how this trinket ranks in comparison to old trinkets.

First off, lets revisit quickly the old trinket values.  All old trinkets had a 15% chance to proc with the exception of the Relic of Chi-Ji which had a 20% chance.  Relic and Qin-xi’s Polarizing Seal were designed to proc frequently (~38-47% uptime) but have lower on-proc stats (spirit/int) while Spirits of the Sun was designed to proc less frequently (~17-24% uptime) but have a higher spirit proc.  [These values were determined by looking at my own World of Logs).]  Based on this, their Mp5 gain was more or less straight forward give the formula for calculating Mp5:

Combat Mp5 = Total Mana *0.02 + (1.1287 *Spirit* Holy Insight%)

According to these values, the Mp5 equivalent for the three trinkets ends up being around the following:

iLvL Name Static Int Static Spirit Proc Int Proc Spirit Static Mp5 Equiv Proc Mp5 Equiv*
489 Normal Qin-xi’s Polarizing Seal 1079 3236 608.93365
493 Normal Qin-xi’s Polarizing Seal(1) 1120 3359 632.072
497 Normal Qin-xi’s Polarizing Seal(2) 1162 3487 655.7747
476 Relic of Chi-ji 1157 3027 0 711.7864375
480 Relic of Chi-ji(1) 1201 3143 0 739.0633542
484 Relic of Chi-ji(2) 1246 3262 0 767.0457083
496 Normal Spirits of the Sun 1152 6908 0 721.949963
500 Normal Spirits of the Sun(1) 1195 7170 0 749.3313889
504 Normal Spirits of the Sun(2) 1241 7443 0 777.8624167
502 Heroic Qin-xi’s Polarizing Seal 1218 3653 687.3783 0
506 Heroic Qin-xi’s Polarizing Seal(1) 1265 3791 713.90275 0
510 Heroic Qin-xi’s Polarizing Seal(2) 1313 3935 740.99155 0
509 Heroic Spirits of the Sun 1300 7796 0 807.2793761
513 Heroic Spirits of the Sun(1) 1349 8093 0 838.0338624
517 Heroic Spirits of the Sun(2) 1401 8402 0 870.0309541
483 LFR Spirits of the Sun 1021 6121 0 633.8323578
487 LFR Spirits of the Sun(1) 1059 6353 0 657.8560642
491 LFR Spirits of the Sun(2) 1099 6593 0 682.7081743
476 LFR Qin-xi’s Polarizing Seal 956 2866 539.5186 0
480 LFR Qin-xi’s Polarizing Seal(1) 993 2976 560.39955 0
484 LFR Qin-xi’s Polarizing Seal(2) 1030 3089 581.2805 0

* Based off 42% uptime on frequent proc trinkets (RoCJ, QPS), 18.5% uptime on not frequent proc trinkets (SoS).

The new trinkets have the more complex proc designs of RPPM. As you could see from the formula above, the number of procs per minute is now affected by haste scaling and a PPM factor.  The new trinkets have the following PPM factors:

  • Stolen Relic of the Zuldazar – 3
  • Lightning-Imbued Chalice – 3
  • Inscribed Bag of Hydra-Spawn – 1.13
  • Horridon’s Last Grasp – 1

There is also a trinket from the Shado-Pan Assault faction called the Soothing Talisman of the Shado-Pan Assault which gives an on-use amount of mana back (29805) every 3 minutes.  The Mp5 calculation returns on average a 827 Mp5 value.  This is misleading because the trinket has a 3 minute cooldown and therefore benefits the fights which closer to multiples of 3 minutes long assuming you use it the first time very quickly.  This is a discussion for another time.  Lets focus on the last trinket above, Horridon’s Last Grasp.

The RPPM values for this trinket suggest that it procs once every minute at least.  I looked up 3 World of Logs reports of people who used this trinket from looking at the wowprogress site at the top geared holy paladins.  (Of course this could apply to other healing classes to a large degree.) Here is a summary in the table below.  Note that everything I say must be taken with a grain of salt as the sample size here is incredibly small to be making any claims.

Fight Length (s) Total Mana Return Equiv Mp5 iLVL Haste # Proc Direct Heals Ticks Total Heal Fight
713 179730 1260.378682 Normal 3509 21 5362 1452 6814 Iron Qon
659 101950 773.5204856 Normal 2281 12 2266 462 2728 Iron Qon
621 177192 1426.666667 Normal 2426 18 4423 842 5265 Iron Qon
621 89424 720 Normal 3044 10 1234 2283 3517 Iron Qon
651 56513 434.047619 Normal 2186 7 4666 504 5170 Iron Qon
712 182160 1279.213483 Normal 2621 19 3885 260 4145 Iron Qon
648 117732 908.4259259 Normal T 2305 14 3345 1036 4381 Iron Qon
649 59568 458.9214176 Normal T 2063 7 4418 1218 5636 Iron Qon
588 96048 816.7346939 Normal 1844 11 4663 1336 5999 Twin Consorts
626 40696 325.0479233 Normal 1416 5 3368 1329 4697 Twin Consorts
541 78840 728.650647 Normal T 2160 9 3313 1025 4338 Twin Consorts

Some general things I found:

  • Procs off eternal flame ticks, beacon transfers
  • Can stack to 3, possible more
  • When stacked, it will refresh the timer in other words, it will again tick 5 times at 3x the mana stated on the trinket for the 3rd stack, I saw a bug where the second stack actually ticked 6 times, not sure if this is a bug
  • Time between ticks ranged from 1 -57 (3 random reports)

Relationship between # Procs

According to the values above, simply calculating the number of procs per minute from the constant RPPM value would most likely produce misleading results.  Recall the formula for calculating procs per minute. The proc chance is dependent on two things, one is haste and the other is the last time you performed a direct healing or periodic healing spell. Haste be simply calculated from the rating above and the time since last attack can be calculated by dividing the ‘Total Heal’ column by the fight length.  Lets call this value Normalized Ticks. Using these in the forumla for RPPM above does not quite give the expected proc values.  Performing a multiple regression using these variables reveals the coefficient for Haste to be ~10 and for the ticks to be -1.6.  The directionality of these variables makes sense as more haste will increase procs, and more Normalized ticks will decrease the number of procs.  The constants suggest that the weight for Haste is much greater than the weight for Normalized Ticks.

It is hard to say anything definite from these cases as there are so many confounding factors.  I can safely say that haste will be fairly important depending on the fight.  I can less confidently state the following. For paladins, above 2200 haste or so, you can pretty safely bet that the normal trinket will be better than the upgraded Relic.  You can make up for a low haste by prioritizing casting more spells/hots but this will not have as big of an effect as increasing haste.  One thing that all of the individuals with a mana return above 100K is higher haste.  What I would really like to do is get my hands on this trinket and try it out at different levels of haste to see exactly what the change is.

Update 1: I recently saw a log of an individual with 1670 haste and the trinket returned over 150,000 mana to him because at one point he had 4 stacks of it, then another 3 in a 5 minute fight. Another pull was around the same time and it return just over 50K to him.  This makes me want to throw all my calculations out the window from the randomness of this trinket.  I have a feeling that the EF style of play is much favored when using this trinket but I’ll have to look into it more to be sure.


What do you think?

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s