# A Tier In My Eye

Sorry for the slow updates, I’ve been deeply engrossed in a secret side project that I’ll hopefully get around to unveiling/announcing next week. However, for this week’s installment of “what Theck could throw together in his spare time” I bring you an updated analysis of the tier 15 set bonuses.

We’ve looked at this issue before, in a more limited fashion.  This time, we have a model that not only includes Sacred Shield and Seal of Insight, but does a better job of modeling Word of Glory as well.  I’ve updated it to use the same “backward-facing” healing algorithm for WoG as it does for SoI, such that every WoG hits for full strength, but can only eliminate damage taken in the previous melee attack (i.e. in the last 1.5 seconds).  Thus, it only overheals when there’s no damage remaining in that previous attack, and is completely independent of what other healers are doing.

Since we want to look at the 4-piece bonus as well, I’m also having the simulation use Divine Protection on cooldown in all of these simulations, and since we’re dealing with melee I’m assuming it’s glyphed to provide 20% mitigation against everything.

Tier 15 2-Piece Bonus: Shield of Glory

For those that aren’t aware, the 2-piece bonus grants a buff called Shield of Glory that increases your block chance by 40% whenever you cast Word of Glory.  The new WeakAuras I posted last week include a tracker for this, in case you want one.  The speculation has been that it may potentially be worth inserting WoG into the rotation every now and then to gain a passive 40% block chance.  However, our previous simulations suggested that just wasn’t the case.

Let’s see if that predictions look like now that we have a more complete model.  I’m using the Control/Haste set for this one – for stats see the later section on the 4-piece bonus.  First, the 10-man normal boss that hits for 150k after mitigation:

Boss Attack = 150k, SoI model=nooverheal data set setbonus-10000-18

| Set: |      S |    SH1 |    SH2 |     TS |     ST |
| mean |  0.214 |  0.211 |  0.212 |  0.187 |  0.181 |
|  std |  0.106 |  0.102 |  0.105 |  0.107 |  0.102 |
|   S% |  0.522 |  0.522 |  0.523 |  0.324 |  0.353 |
|   HP |   755k |   755k |   755k |   755k |   755k |
|  nHP |  5.034 |  5.034 |  5.034 |  5.034 |  5.034 |
| ---- |  --- 2 | Attack | Moving | Avg.-- | ------ |
|  10% | 37.819 | 38.798 | 38.069 | 30.111 | 28.232 |
|  20% |  6.596 |  5.340 |  6.715 |  5.674 |  4.641 |
|  30% |  0.591 |  0.125 |  0.568 |  0.471 |  0.355 |
| ---- |  --- 3 | Attack | Moving | Avg.-- | ------ |
|  10% | 60.371 | 62.391 | 60.451 | 48.822 | 47.568 |
|  20% | 18.540 | 15.786 | 18.130 | 15.413 | 13.248 |
|  30% |  1.852 |  0.815 |  1.800 |  1.877 |  1.337 |
|  40% |  0.006 |  0.001 |  0.004 |  0.015 |  0.008 |
| ---- |  --- 4 | Attack | Moving | Avg.-- | ------ |
|  20% | 36.461 | 34.535 | 36.098 | 28.422 | 26.059 |
|  30% | 10.131 |  8.544 |  9.942 |  7.234 |  5.641 |
|  40% |  0.584 |  0.456 |  0.424 |  1.021 |  0.574 |
|  50% |  0.005 |  0.009 |  0.005 |  0.066 |  0.024 |
| ---- |  --- 5 | Attack | Moving | Avg.-- | ------ |
|  20% | 52.949 | 51.886 | 52.468 | 42.073 | 39.935 |
|  30% | 21.685 | 19.972 | 21.240 | 14.724 | 12.361 |
|  40% |  3.624 |  2.914 |  3.320 |  3.042 |  1.964 |
|  50% |  0.404 |  0.263 |  0.379 |  0.364 |  0.153 |
|  60% |  0.002 |  0.005 |  0.003 |  0.022 |  0.005 |
| ---- |  --- 6 | Attack | Moving | Avg.-- | ------ |
|  30% | 35.491 | 34.051 | 34.905 | 24.589 | 21.829 |
|  40% | 10.156 |  8.489 |  9.571 |  6.452 |  4.784 |
|  50% |  1.700 |  1.153 |  1.521 |  1.017 |  0.524 |
|  60% |  0.043 |  0.028 |  0.036 |  0.085 |  0.023 |
|  70% |  0.001 |  0.000 |  0.000 |  0.001 |  0.000 |
| ---- |  --- 7 | Attack | Moving | Avg.-- | ------ |
|  30% | 49.033 | 47.968 | 48.496 | 36.357 | 33.367 |
|  40% | 20.545 | 18.725 | 19.756 | 12.933 | 10.640 |
|  50% |  5.634 |  4.616 |  5.325 |  2.939 |  1.991 |
|  60% |  0.650 |  0.511 |  0.620 |  0.380 |  0.184 |
|  70% |  0.052 |  0.032 |  0.048 |  0.022 |  0.009 |
|  80% |  0.001 |  0.000 |  0.001 |  0.000 |  0.000 |

If you recall the old findings, it was highly dependent on exactly how much overheal we assumed for WoG.  At 100% overheal it was disastrously bad, but at 0% overheal it was an improvement. In these simulations, “TS” puts up about 77% overheal while “ST” puts up about 64%.  And remember, this is just the overheal we’re creating ourselves with this queue, so it’s about as best-case as it can get.  It only gets worse if you start insinuating extra overhealing done by healers.

Nonetheless, the set bonus versions fare relatively well here.  These results show that it’s not bad, matching or beating out the simple “Spam SotR” queue (“S”) in most cases.  However, the smarter shifting queue (“SH1″) tends to beat it in all the important categories.  So again, there doesn’t seem to be any strong incentive to use the set bonus as a rotational buff.

If you can target the heal effectively such that the overheal is very low, then the set bonus becomes stronger.  But that’s basically how we use WoG without the set bonus: as an emergency heal when we’re low on health.  So my recommendation remains unchanged: don’t worry about keeping the set bonus up all the time, just keep using SotR and WoG like you normally would.  Of course, special cases (like SotR+1HP-WoG for Dread Thrash) are exempt from that advice.

Since the differences are just amplified in the 250k and 350k swing data, I’ll just present that data without commentary:

Boss Attack = 250k, SoI model=nooverheal data set setbonus-10000-19

| Set: |      S |    SH1 |    SH2 |     TS |     ST |
| mean |  0.237 |  0.234 |  0.236 |  0.206 |  0.200 |
|  std |  0.110 |  0.101 |  0.108 |  0.111 |  0.105 |
|   S% |  0.522 |  0.523 |  0.522 |  0.324 |  0.353 |
|   HP |   755k |   755k |   755k |   755k |   755k |
|  nHP |  3.021 |  3.021 |  3.021 |  3.021 |  3.021 |
| ---- |  --- 2 | Attack | Moving | Avg.-- | ------ |
|  20% | 35.459 | 37.448 | 35.993 | 30.374 | 28.103 |
|  30% | 13.947 | 12.025 | 13.790 | 11.171 |  9.825 |
|  40% |  3.303 |  1.217 |  3.219 |  3.003 |  2.527 |
|  50% |  0.716 |  0.080 |  0.553 |  0.519 |  0.418 |
|  60% |  0.021 |  0.001 |  0.002 |  0.014 |  0.011 |
| ---- |  --- 3 | Attack | Moving | Avg.-- | ------ |
|  20% | 59.506 | 63.282 | 59.920 | 49.323 | 47.801 |
|  30% | 34.636 | 33.267 | 34.531 | 26.334 | 24.242 |
|  40% | 12.076 |  6.691 | 11.722 | 10.644 |  8.773 |
|  50% |  2.535 |  0.956 |  2.223 |  2.915 |  2.086 |
|  60% |  0.237 |  0.097 |  0.178 |  0.386 |  0.248 |
|  70% |  0.002 |  0.001 |  0.002 |  0.014 |  0.007 |
| ---- |  --- 4 | Attack | Moving | Avg.-- | ------ |
|  30% | 54.078 | 54.395 | 54.182 | 42.576 | 40.788 |
|  40% | 30.414 | 27.240 | 30.235 | 22.782 | 20.543 |
|  50% | 14.037 | 11.123 | 13.588 | 10.229 |  8.339 |
|  60% |  4.436 |  3.083 |  4.216 |  3.853 |  2.689 |
|  70% |  0.764 |  0.421 |  0.637 |  1.265 |  0.694 |
|  80% |  0.039 |  0.041 |  0.034 |  0.297 |  0.130 |
|  90% |  0.000 |  0.002 |  0.001 |  0.024 |  0.010 |
| 100% |  0.000 |  0.000 |  0.000 |  0.002 |  0.001 |
| ---- |  --- 5 | Attack | Moving | Avg.-- | ------ |
|  40% | 48.076 | 47.293 | 47.961 | 36.490 | 34.364 |
|  50% | 29.001 | 26.410 | 28.575 | 20.068 | 17.550 |
|  60% | 13.957 | 11.397 | 13.578 |  9.420 |  7.382 |
|  70% |  4.965 |  3.271 |  4.706 |  3.837 |  2.448 |
|  80% |  1.175 |  0.521 |  1.128 |  1.204 |  0.616 |
|  90% |  0.153 |  0.060 |  0.151 |  0.241 |  0.094 |
| 100% |  0.007 |  0.005 |  0.009 |  0.053 |  0.013 |
| 110% |  0.000 |  0.000 |  0.001 |  0.007 |  0.002 |
| 120% |  0.000 |  0.000 |  0.000 |  0.001 |  0.000 |
| ---- |  --- 6 | Attack | Moving | Avg.-- | ------ |
|  40% | 63.435 | 64.507 | 63.513 | 50.915 | 48.899 |
|  50% | 44.899 | 44.256 | 44.658 | 32.381 | 29.659 |
|  60% | 27.162 | 24.543 | 26.683 | 17.553 | 15.031 |
|  70% | 13.311 |  9.897 | 12.802 |  8.124 |  5.989 |
|  80% |  4.733 |  2.563 |  4.468 |  2.973 |  1.849 |
|  90% |  1.136 |  0.482 |  1.057 |  0.806 |  0.391 |
| 100% |  0.179 |  0.057 |  0.143 |  0.193 |  0.064 |
| 110% |  0.014 |  0.003 |  0.011 |  0.033 |  0.008 |
| 120% |  0.000 |  0.000 |  0.001 |  0.005 |  0.001 |
| ---- |  --- 7 | Attack | Moving | Avg.-- | ------ |
|  50% | 58.991 | 59.193 | 58.898 | 45.495 | 42.817 |
|  60% | 41.645 | 40.168 | 41.218 | 28.683 | 25.778 |
|  70% | 25.355 | 22.592 | 24.937 | 15.685 | 13.075 |
|  80% | 12.788 |  9.976 | 12.368 |  7.225 |  5.457 |
|  90% |  5.342 |  3.594 |  5.086 |  2.698 |  1.768 |
| 100% |  1.761 |  0.978 |  1.658 |  0.852 |  0.449 |
| 110% |  0.429 |  0.194 |  0.412 |  0.210 |  0.094 |
| 120% |  0.066 |  0.020 |  0.060 |  0.039 |  0.011 |
| 130% |  0.006 |  0.001 |  0.006 |  0.006 |  0.002 |
| 140% |  0.001 |  0.000 |  0.001 |  0.001 |  0.000 |

.

Finisher = S, Boss Attack = 350k, SoI model=nooverheal data set setbonus-10000-20

| Set: |      S |    SH1 |    SH2 |     TS |     ST |
| mean |  0.247 |  0.245 |  0.246 |  0.217 |  0.210 |
|  std |  0.111 |  0.103 |  0.110 |  0.113 |  0.107 |
|   S% |  0.522 |  0.522 |  0.522 |  0.324 |  0.354 |
|   HP |   755k |   755k |   755k |   755k |   755k |
|  nHP |  2.158 |  2.158 |  2.158 |  2.158 |  2.158 |
| ---- |  --- 2 | Attack | Moving | Avg.-- | ------ |
|  20% | 50.804 | 53.070 | 50.570 | 45.730 | 43.729 |
|  30% | 30.239 | 28.970 | 29.752 | 27.484 | 25.189 |
|  40% | 15.067 | 13.081 | 15.006 | 11.815 | 10.663 |
|  50% |  6.810 |  4.612 |  6.868 |  5.772 |  4.813 |
|  60% |  2.719 |  0.988 |  2.585 |  2.744 |  2.208 |
|  70% |  0.838 |  0.133 |  0.627 |  0.921 |  0.743 |
|  80% |  0.171 |  0.007 |  0.068 |  0.136 |  0.112 |
|  90% |  0.012 |  0.000 |  0.002 |  0.009 |  0.009 |
| ---- |  --- 3 | Attack | Moving | Avg.-- | ------ |
|  30% | 55.882 | 57.944 | 55.703 | 47.208 | 45.410 |
|  40% | 37.334 | 36.158 | 37.386 | 29.144 | 27.209 |
|  50% | 21.831 | 17.103 | 21.445 | 16.996 | 14.874 |
|  60% | 10.587 |  5.848 | 10.326 |  9.534 |  7.601 |
|  70% |  3.906 |  1.635 |  3.545 |  4.186 |  3.105 |
|  80% |  1.102 |  0.381 |  0.837 |  1.234 |  0.848 |
|  90% |  0.169 |  0.074 |  0.132 |  0.365 |  0.212 |
| 100% |  0.002 |  0.001 |  0.002 |  0.016 |  0.007 |
| ---- |  --- 4 | Attack | Moving | Avg.-- | ------ |
|  40% | 57.905 | 58.424 | 57.950 | 46.925 | 45.257 |
|  50% | 42.374 | 40.439 | 42.199 | 32.005 | 29.846 |
|  60% | 28.348 | 24.918 | 28.297 | 21.010 | 18.397 |
|  70% | 15.984 | 12.856 | 15.755 | 11.851 |  9.771 |
|  80% |  7.819 |  5.515 |  7.435 |  6.051 |  4.498 |
|  90% |  2.902 |  1.783 |  2.660 |  2.937 |  1.884 |
| 100% |  0.650 |  0.410 |  0.572 |  1.210 |  0.653 |
| 110% |  0.086 |  0.060 |  0.063 |  0.369 |  0.181 |
| 120% |  0.007 |  0.012 |  0.007 |  0.125 |  0.061 |
| 130% |  0.000 |  0.000 |  0.000 |  0.017 |  0.009 |
| 140% |  0.000 |  0.000 |  0.000 |  0.001 |  0.001 |
| ---- |  --- 5 | Attack | Moving | Avg.-- | ------ |
|  50% | 59.997 | 60.331 | 59.760 | 47.679 | 45.805 |
|  60% | 46.133 | 44.822 | 45.961 | 34.436 | 31.709 |
|  70% | 32.199 | 29.660 | 32.037 | 22.568 | 19.890 |
|  80% | 19.945 | 16.986 | 19.531 | 13.468 | 11.003 |
|  90% | 10.663 |  8.286 | 10.288 |  7.559 |  5.523 |
| 100% |  4.800 |  3.365 |  4.708 |  3.810 |  2.426 |
| 110% |  1.774 |  0.993 |  1.747 |  1.628 |  0.860 |
| 120% |  0.643 |  0.288 |  0.614 |  0.671 |  0.314 |
| 130% |  0.117 |  0.054 |  0.109 |  0.219 |  0.085 |
| 140% |  0.021 |  0.009 |  0.017 |  0.058 |  0.021 |
| 150% |  0.001 |  0.001 |  0.000 |  0.014 |  0.005 |
| 160% |  0.000 |  0.000 |  0.000 |  0.004 |  0.001 |
| ---- |  --- 6 | Attack | Moving | Avg.-- | ------ |
|  60% | 61.948 | 62.917 | 61.881 | 49.198 | 46.570 |
|  70% | 48.615 | 48.429 | 48.479 | 35.854 | 32.936 |
|  80% | 35.005 | 33.118 | 34.640 | 24.029 | 21.086 |
|  90% | 23.003 | 20.044 | 22.601 | 14.894 | 12.082 |
| 100% | 13.353 | 10.307 | 13.060 |  8.336 |  6.174 |
| 110% |  6.957 |  4.520 |  6.769 |  4.171 |  2.680 |
| 120% |  3.171 |  1.692 |  2.995 |  1.909 |  1.060 |
| 130% |  1.057 |  0.502 |  0.991 |  0.756 |  0.348 |
| 140% |  0.261 |  0.095 |  0.201 |  0.238 |  0.099 |
| 150% |  0.059 |  0.015 |  0.039 |  0.072 |  0.024 |
| 160% |  0.011 |  0.002 |  0.004 |  0.017 |  0.003 |
| 170% |  0.001 |  0.000 |  0.000 |  0.003 |  0.000 |
| ---- |  --- 7 | Attack | Moving | Avg.-- | ------ |
|  70% | 62.789 | 63.383 | 62.669 | 49.582 | 46.628 |
|  80% | 50.099 | 49.631 | 49.764 | 36.918 | 33.775 |
|  90% | 37.508 | 35.868 | 37.132 | 25.779 | 22.467 |
| 100% | 25.986 | 23.685 | 25.677 | 16.567 | 13.711 |
| 110% | 16.787 | 14.151 | 16.541 |  9.782 |  7.530 |
| 120% |  9.943 |  7.741 |  9.758 |  5.341 |  3.769 |
| 130% |  5.115 |  3.583 |  4.957 |  2.611 |  1.649 |
| 140% |  2.208 |  1.390 |  2.104 |  1.108 |  0.640 |
| 150% |  0.835 |  0.462 |  0.813 |  0.415 |  0.218 |
| 160% |  0.280 |  0.124 |  0.269 |  0.137 |  0.061 |
| 170% |  0.086 |  0.030 |  0.080 |  0.042 |  0.014 |
| 180% |  0.026 |  0.005 |  0.021 |  0.010 |  0.003 |
| 190% |  0.005 |  0.001 |  0.003 |  0.002 |  0.000 |
| 200% |  0.002 |  0.000 |  0.001 |  0.001 |  0.000 |

Tier 15 4-Piece Bonus: Divine Protection

There’s been mixed reactions to this bonus.  Some players love it, some are hesitant to give up well-itemized (and potentially thunderforged) off-set pieces for two more set pieces given how poorly most of our set is itemized for control/haste.

To test that, I’ve come up with the following simulation.  We’ll compare the data for Control/Haste and Control/Stamina gear sets that do not have the set bonus against a gear set that does use the set bonus, but sacrifices some stats.  The stat differences are based on this comment by Zothor, where he worked out the ideal swap from thunderforged off-set to 4-piece tier.  I have, however, taken the liberty to do some reforging of hit->expertise to maintain caps.  Here are the final stats for the different sets:

|    Set: |  C/Ha |  C/St |  C/Sg | C/Set |
|     Str | 15000 | 15000 | 15000 | 15720 |
|     Sta | 28000 | 34000 | 31000 | 27240 |
|   Parry |  1500 |  1500 |  1500 |  2423 |
|   Dodge |  1500 |  1500 |  1500 |  1614 |
| Mastery |  1500 |  1500 |  1500 |  2137 |
|     Hit |  2550 |  2550 |  2550 |  2576 |
|     Exp |  5100 |  5100 |  5100 |  5100 |
|   Haste | 12000 |  8000 |  8000 | 10788 |

Note that the direct comparison is C/Ha to C/Set; I’ve just included the stamina sets in case the set bonus is more powerful than expected and to have another reference point.

Now, let’s see how they compare.  First, the 150k data:

Finisher = SH1, Boss Attack = 150k, SoI model=nooverheal data set setbonus-10000-23

| Set: |   C/Ha |   C/St |   C/Sg |  C/Set |
| mean |  0.210 |  0.233 |  0.233 |  0.205 |
|  std |  0.101 |  0.106 |  0.106 |  0.102 |
|   S% |  0.522 |  0.482 |  0.482 |  0.511 |
|   HP |   755k |   876k |   816k |   740k |
|  nHP |  5.034 |  5.843 |  5.439 |  4.932 |
| ---- |  --- 2 | Attack | Moving | Avg.-- |
|  10% | 38.550 | 36.768 | 43.808 | 38.757 |
|  20% |  5.233 |  2.554 |  4.759 |  6.061 |
|  30% |  0.116 |  0.015 |  0.025 |  0.179 |
| ---- |  --- 3 | Attack | Moving | Avg.-- |
|  20% | 15.460 | 11.241 | 16.514 | 16.666 |
|  30% |  0.809 |  0.383 |  0.555 |  1.200 |
|  40% |  0.001 |  0.000 |  0.005 |  0.017 |
| ---- |  --- 4 | Attack | Moving | Avg.-- |
|  20% | 34.034 | 31.020 | 36.894 | 34.493 |
|  30% |  8.381 |  4.645 |  7.337 |  8.809 |
|  40% |  0.450 |  0.168 |  0.481 |  0.563 |
|  50% |  0.009 |  0.001 |  0.006 |  0.014 |
| ---- |  --- 5 | Attack | Moving | Avg.-- |
|  20% | 51.302 | 49.340 | 55.036 | 51.044 |
|  30% | 19.624 | 14.006 | 19.151 | 20.041 |
|  40% |  2.826 |  1.518 |  3.128 |  3.318 |
|  50% |  0.243 |  0.035 |  0.185 |  0.283 |
|  60% |  0.004 |  0.000 |  0.003 |  0.004 |
| ---- |  --- 6 | Attack | Moving | Avg.-- |
|  30% | 33.546 | 27.377 | 34.258 | 33.419 |
|  40% |  8.214 |  5.209 |  8.906 |  8.952 |
|  50% |  1.091 |  0.262 |  0.913 |  1.234 |
|  60% |  0.027 |  0.002 |  0.017 |  0.032 |
| ---- |  --- 7 | Attack | Moving | Avg.-- |
|  30% | 47.345 | 42.301 | 49.405 | 46.919 |
|  40% | 18.329 | 13.850 | 19.839 | 18.883 |
|  50% |  4.451 |  2.189 |  4.609 |  4.885 |
|  60% |  0.510 |  0.133 |  0.464 |  0.604 |
|  70% |  0.030 |  0.001 |  0.017 |  0.032 |

It’s hard to tell from this data since there’s so little of it, but it doesn’t look like the 4-piece is bringing any significant advantage to the table.  It provides lower SotR uptime and worse spike presence pretty much across the board.  For the 10-man raider, the 4-piece seems fairly unattractive.  It’s not terrible, per se, but it definitely doesn’t seem to compete with the extra haste and stamina you get from off-set pieces.

However, unlike the 2-piece bonus, the 4-piece is a lot more sensitive to boss attack size.  Let’s see what happens when we crank the dial up to 250k boss swings:

Finisher = SH1, Boss Attack = 250k, SoI model=nooverheal data set setbonus-10000-22

| Set: |   C/Ha |   C/St |   C/Sg |  C/Set |
| mean |  0.234 |  0.256 |  0.257 |  0.221 |
|  std |  0.102 |  0.105 |  0.105 |  0.103 |
|   S% |  0.523 |  0.482 |  0.482 |  0.536 |
|   HP |   755k |   876k |   816k |   740k |
|  nHP |  3.021 |  3.506 |  3.263 |  2.959 |
| ---- |  --- 2 | Attack | Moving | Avg.-- |
|  20% | 37.499 | 28.886 | 32.327 | 35.130 |
|  30% | 12.061 |  6.843 | 12.024 | 11.676 |
|  40% |  1.221 |  0.502 |  1.483 |  2.655 |
|  50% |  0.081 |  0.022 |  0.033 |  0.146 |
|  60% |  0.001 |  0.000 |  0.001 |  0.005 |
| ---- |  --- 3 | Attack | Moving | Avg.-- |
|  30% | 33.369 | 22.621 | 32.503 | 30.202 |
|  40% |  6.660 |  4.112 |  8.332 | 10.238 |
|  50% |  0.957 |  0.401 |  0.900 |  1.598 |
|  60% |  0.100 |  0.030 |  0.048 |  0.166 |
|  70% |  0.001 |  0.000 |  0.002 |  0.012 |
| ---- |  --- 4 | Attack | Moving | Avg.-- |
|  30% | 54.379 | 46.699 | 55.190 | 50.182 |
|  40% | 27.149 | 21.752 | 29.260 | 27.772 |
|  50% | 11.131 |  5.830 |  9.316 | 11.315 |
|  60% |  3.168 |  1.069 |  2.204 |  3.162 |
|  70% |  0.455 |  0.090 |  0.331 |  0.510 |
|  80% |  0.034 |  0.005 |  0.034 |  0.052 |
|  90% |  0.001 |  0.000 |  0.000 |  0.005 |
| ---- |  --- 5 | Attack | Moving | Avg.-- |
|  40% | 47.137 | 41.709 | 49.760 | 45.135 |
|  50% | 26.457 | 18.178 | 24.735 | 25.319 |
|  60% | 11.518 |  6.259 |  9.722 | 10.977 |
|  70% |  3.346 |  1.053 |  2.809 |  3.369 |
|  80% |  0.536 |  0.159 |  0.520 |  0.685 |
|  90% |  0.060 |  0.007 |  0.043 |  0.095 |
| 100% |  0.003 |  0.000 |  0.003 |  0.010 |
| ---- |  --- 6 | Attack | Moving | Avg.-- |
|  50% | 44.100 | 35.111 | 43.671 | 41.135 |
|  60% | 24.655 | 16.643 | 22.668 | 22.762 |
|  70% | 10.041 |  4.423 |  9.212 |  9.649 |
|  80% |  2.647 |  0.988 |  2.574 |  2.927 |
|  90% |  0.502 |  0.081 |  0.375 |  0.627 |
| 100% |  0.056 |  0.007 |  0.041 |  0.100 |
| 110% |  0.001 |  0.000 |  0.001 |  0.009 |
| ---- |  --- 7 | Attack | Moving | Avg.-- |
|  60% | 40.021 | 32.283 | 39.562 | 36.834 |
|  70% | 22.651 | 14.265 | 22.229 | 20.977 |
|  80% | 10.181 |  5.484 | 10.029 |  9.778 |
|  90% |  3.695 |  1.196 |  3.030 |  3.689 |
| 100% |  1.038 |  0.257 |  0.768 |  1.082 |
| 110% |  0.204 |  0.024 |  0.167 |  0.245 |
| 120% |  0.016 |  0.003 |  0.023 |  0.039 |
| 130% |  0.001 |  0.000 |  0.001 |  0.003 |

In this data, we’re seeing an increase in SotR uptime from the set bonus, which is a good start.  It’s still not really giving  a significant improvement in any of the spike categories though, and trailing in most of them.  That said, it’s close enough that you’re probably fine with either choice – the two aren’t so significantly different that either is a poor choice.  It ends up being a trade-off that doesn’t have much effect in the end.  The set bonus increases SotR uptime, which is good, but costs stamina, which is bad, and the two mostly offset one another.

Let’ see if it improves with 350k swings.

Finisher = SH1, Boss Attack = 350k, SoI model=nooverheal data set setbonus-10000-21

| Set: |   C/Ha |   C/St |   C/Sg |  C/Set |
| mean |  0.244 |  0.267 |  0.268 |  0.223 |
|  std |  0.103 |  0.106 |  0.106 |  0.108 |
|   S% |  0.523 |  0.482 |  0.482 |  0.576 |
|   HP |   755k |   876k |   816k |   740k |
|  nHP |  2.158 |  2.504 |  2.331 |  2.114 |
| ---- |  --- 2 | Attack | Moving | Avg.-- |
|  20% | 53.023 | 54.094 | 55.312 | 48.477 |
|  30% | 28.922 | 27.172 | 32.239 | 26.134 |
|  40% | 13.030 | 11.389 | 15.346 | 12.631 |
|  50% |  4.575 |  2.022 |  4.237 |  4.449 |
|  60% |  0.965 |  0.306 |  0.920 |  0.983 |
|  70% |  0.130 |  0.022 |  0.032 |  0.137 |
|  80% |  0.006 |  0.000 |  0.002 |  0.012 |
| ---- |  --- 3 | Attack | Moving | Avg.-- |
|  40% | 36.134 | 32.386 | 39.629 | 33.580 |
|  50% | 17.051 | 10.788 | 17.012 | 15.700 |
|  60% |  5.776 |  3.107 |  5.565 |  5.750 |
|  70% |  1.608 |  0.667 |  1.416 |  1.586 |
|  80% |  0.370 |  0.048 |  0.257 |  0.396 |
|  90% |  0.070 |  0.002 |  0.032 |  0.069 |
| 100% |  0.001 |  0.000 |  0.003 |  0.010 |
| ---- |  --- 4 | Attack | Moving | Avg.-- |
|  50% | 40.315 | 35.246 | 41.916 | 35.838 |
|  60% | 24.832 | 18.204 | 24.447 | 21.789 |
|  70% | 12.821 |  7.603 | 12.312 | 11.274 |
|  80% |  5.436 |  2.222 |  4.236 |  4.977 |
|  90% |  1.731 |  0.507 |  1.307 |  1.713 |
| 100% |  0.397 |  0.097 |  0.324 |  0.455 |
| 110% |  0.062 |  0.014 |  0.054 |  0.092 |
| 120% |  0.014 |  0.000 |  0.003 |  0.019 |
| 130% |  0.001 |  0.000 |  0.000 |  0.003 |
| ---- |  --- 5 | Attack | Moving | Avg.-- |
|  60% | 44.695 | 37.974 | 45.599 | 38.878 |
|  70% | 29.605 | 22.269 | 29.850 | 25.662 |
|  80% | 16.860 | 10.325 | 16.311 | 15.070 |
|  90% |  8.185 |  4.229 |  7.607 |  7.605 |
| 100% |  3.365 |  1.154 |  2.908 |  3.188 |
| 110% |  0.982 |  0.279 |  0.828 |  1.121 |
| 120% |  0.261 |  0.037 |  0.180 |  0.317 |
| 130% |  0.044 |  0.004 |  0.036 |  0.075 |
| 140% |  0.006 |  0.000 |  0.002 |  0.011 |
| ---- |  --- 6 | Attack | Moving | Avg.-- |
|  70% | 48.266 | 40.831 | 49.405 | 41.463 |
|  80% | 33.005 | 24.500 | 33.392 | 28.608 |
|  90% | 19.919 | 12.898 | 19.867 | 17.715 |
| 100% | 10.243 |  5.314 |  9.942 |  9.496 |
| 110% |  4.460 |  1.690 |  4.110 |  4.265 |
| 120% |  1.621 |  0.369 |  1.392 |  1.605 |
| 130% |  0.482 |  0.072 |  0.338 |  0.543 |
| 140% |  0.082 |  0.005 |  0.056 |  0.139 |
| 150% |  0.014 |  0.000 |  0.006 |  0.018 |
| 160% |  0.001 |  0.000 |  0.000 |  0.003 |
| ---- |  --- 7 | Attack | Moving | Avg.-- |
|  80% | 49.407 | 41.995 | 51.019 | 42.359 |
|  90% | 35.667 | 27.965 | 37.047 | 31.028 |
| 100% | 23.619 | 16.265 | 24.129 | 20.791 |
| 110% | 14.079 |  8.029 | 13.905 | 12.506 |
| 120% |  7.655 |  3.461 |  7.154 |  6.817 |
| 130% |  3.511 |  1.270 |  3.141 |  3.372 |
| 140% |  1.318 |  0.405 |  1.151 |  1.427 |
| 150% |  0.423 |  0.088 |  0.393 |  0.500 |
| 160% |  0.116 |  0.012 |  0.104 |  0.154 |
| 170% |  0.028 |  0.001 |  0.020 |  0.046 |
| 180% |  0.006 |  0.000 |  0.001 |  0.007 |
| 190% |  0.000 |  0.000 |  0.000 |  0.001 |

Better… but… still a little disappointing.  We now have significantly higher SotR uptime with the set bonus active thanks to the higher incoming damage rate.  However that high uptime isn’t translating to a huge decrease in spike presence.  At best, it’s a complete wash.  It just doesn’t seem strong enough to have a big effect.

That may seem puzzling – after all, 5% more SotR uptime seems like it should be worth something, right?  But the reasons for the lackluster performance are pretty easy to explain.  I think there are two primary reasons that the set bonus falls short.  First, we lose about 2% of our total health by using the set pieces instead of thunderforged off-set items.  That in and of itself is a fairly noticeable loss given how strongly stamina performs in the simulation.

But the more subtle reason is in the timing of the set bonus.  You receive that extra holy power only while you already have a 20% mitigation cooldown active – in other words, during a time when you’re already less susceptible to spikes.  The big spikes aren’t happening when you have Divine Protection up because you have Divine Protection up, so having higher SotR uptime during Divine Protection just serves to lower total damage taken.

In a sense, instead of applying that holy power to the top spikes, or the tail of the distribution, you’re just taking a chunk out of the middle and shifting it further down.  You’re turning “not deadly” sequences into “still not deadly” sequences, which explains why it has very little effect on the dangerous spikes at the top of the distribution.

The haste set, on the other hand, gives you more holy power to use mitigating the large spikes at all times, which translates into better spike mitigation over the entire distribution, including (and especially at) the top where the deadly spikes are.

Conclusions

So  the updated simulation seems to reinforce our earlier impressions of the set bonus.  In short, the 2-piece is a nice bonus for when we have to WoG, but not worth using as a maintenance buff.  The 4-piece is similarly not bad, but also not effective enough to warrant giving up well-itemized thunderforged off-set pieces.  The latter is especially worth noting if you’re in a situation where you have access to much higher-ilvl off-set pieces (i.e. heroic and/or thunderforged off-set vs. normal-mode tier).

This entry was posted in Tanking, Theck's Pounding Headaches, Theorycrafting and tagged , , , , , , , , , , , , , , . Bookmark the permalink.

### 13 Responses to A Tier In My Eye

1. Zothor says:

Ugh. I hate when the numbers support the gut-check “that…. doesn’t seem particularly good” feeling. How could Blizz not have seen this coming when they introduced offset gear with extra stamina and armor, and often better itemization?

I seriously hope Thunderforged gear goes away next patch if they’re not going to give us a way to make Tier gear comparable. Le sigh.

• Theck says:

I have to agree. It feels odd that tier struggles to compete with off-set loot. Though thunderforged tokens would be odd, so maybe it would be easier to have an item that allowed you to trade a tier item for a TF version (like the Firelands system for upgrading pieces).

2. Wrathblood says:

Excellent points Theck, but I have a couple comments.

First, on the 2-piece (which I continue carrying a, somewhat diminished, torch for) am I crazy or is the TS rotation actually stacking up very competitively with SH1 and SH2?

Second, great point on the weakness of 4-piece (by definition, it always occurs when you have a cooldown up). Plus, while some boss damage spikes occur over a period of time, plenty more are simply hard single shots (Talon Rake, etc). You put DP up to help survive it and then you get the HoPo. And its nice to have gotten the HoPo for more SHoR uptime, but its uptime in effect after the big hit has already passed. Of course you’ve presumably pooled your HoPo ahead of time to ensure you’ll have it up for the big hit, and the follow-on HoPo helps you recover from the big hit. But the HoPo is still less valuable than it would have been 5 seconds earlier.

• Theck says:

Can you be more specific? I mostly look at the 3- and 4-attack categories, and none of those seem to favor TS or ST over SH1. They come close in some cases, but never seem to provide an advantage. As you get into longer strings they can pull ahead a little bit, but it’s questionable how meaningful that is (Meloree has argued in the past that the 3- and 4-attack categories are all that matters; 2 in a row is throughput damage, 5+ without healer reaction is massive failure on the part of healers, tanks, or both).

3. Thiron says:

Would anything change for 2-set in case of add fight, like Tortos or Horridon, with a lot of small hits?
Also, 2-set seems worth using in some more…specific cases. For example, during Holy Avenger as extra “cooldown”. Or – if you’re capped on Holy Power while you are *not* tanking, but are going to taunt in 5 seconds. Yes, tankswaps are what makes it useful IMO.
Still, I’m fine with giving up tier pieces for everyone else. It’s more beneficial for the raid if others take it. It also gives me a EPGP (DKP-style system) points “boost” that I need to get the haste plate ahead of DPS ^_^

• Theck says:

I think the answer to that is a little more complicated. Certainly, more attacks incoming makes an RNG-based effect more likely to give a consistent benefit. But keep in mind that SotR scales very well with multiple mobs too. If SotR “wins” against a single target, it probably wins against many targets as well.

The main difference/advantage I can see with the 2-piece is its length. As you said, it’s something that you can cast several seconds before a taunt to get 10 seconds of “free” block. I say “free” because as long as you cast it a few seconds before you taunt, that Holy Power wouldn’t have given you any benefit via SotR, assuming you were capping HP.

Similarly, during Holy Avenger, you could keep SotR and the 2-piece active. As long as you don’t need the full 30+ seconds of SotR coverage (i.e. maybe you’ll be taunted off of before the SotR coverage from HA will end), then that’s a decent survivability gain.

4. blizzhoof says:

I think the 4-pc can be incredibly useful on 25-man H Durumu, H Dark Animus, and H Lei Shen (phase 1) to help do a lot more damage and to keep the buff up, but that’s about it.

Specifically I’d like to use it on DA while eating one Anima Ring debuff each time he casts it or eating Decapitate and continuing to tank Lei Shen. Basically, I think it can be incredibly useful when you are trying to maximize damage while maintaining good survivability.

Would you mind running it with the kind of numbers I’m seeing from those types of bosses?

For data purposes
Highest I found with one stack of Anima Ring:
[19:56:59.142] Blizzhoof afflicted by Anima Ring from Dark Animus
[19:57:09.276] Dark Animus hits Blizzhoof 842132 (A: 130811) <- Obviously without ShoR up.

Random hits without the debufff:
[20:01:52.803] Dark Animus's Anima Ring fades from Blizzhoof
[20:02:19.287] Dark Animus hits Blizzhoof 388780
[20:02:26.791] Dark Animus hits Blizzhoof 285836
[20:02:41.657] Dark Animus hits Blizzhoof 64718 (A: 177911, B: 103984)
[20:02:45.754] Dark Animus hits Blizzhoof 117194 (A: 50226)
[20:02:47.641] Blizzhoof afflicted by Anima Ring from Dark Animus

Durumu melees for about 400k as far as I can tell and Lei Shen is about 350 to 400k as well.

• Theck says:

The normal Durumu/Lei Shen melees of 350k-400k won’t look appreciably different from the 350k data set. The 900k+ hits from gimmicky effects may look different, I can certainly try those just to see if anything changes. But I’m not sure this simulation is a good representation of that either. You’re only going to be able to cover 10 seconds with Divine Protection, so the simulation is inherently averaging over 10 seconds with it up and 50 seconds with it down. If you use Divine Protection to specifically offset a period of high intake, then the value of the 4-piece will be a decent bit higher, both because it will give you more HP per second and because it will concentrate SotR uptime into a period of high danger/spikiness.

• blizzhoof says:

Yea, I guess my post wasn’t super clear. Basically I’m wondering if in the gimmicky situations the 4-pc will give me higher survivability than using thunderforged off-pieces. I’m pretty sure that it will (as well as give a big dps benefit), but I was just curious if my intuitions were accurate.

Also, you said “averaging over 10 seconds with it up and 50 seconds with it down.” If you’re running the 4-pc, then you’re going to be running Unbreakable Spirit. Have you run the simulation with Unbreakable Spirit?

• Theck says:

I don’t think I did, no. But keep in mind that until 5.4, it only reduces the cooldown to about 45-48 seconds. Still a fairly significant buff, but not double.

5. Thels says:

Thanks a lot for the update, Theck!

I can finally put some tier pieces in the bank, without feeling worried that it might’ve been the wrong choice after all.