Vindication Uptime

[Kihra]

Hmmz, in WOL i have about 85% uptime. I put another ret on it and it became 99%. I tank with a dk in 25 man, so a ret is prolly the best option for it, we have plenty of paladins with DG/AM anyway. Yeah sure he can be dead or on another target, but it's the best option I think.

Yeah 85% seems to be about what you get with really low hit rating doing proper 969. If you pay attention and weave in Avenger's Shield when needed, you can push it up to low 90s uptime. With melee hit cap and and throwing AS out when needed, you can achieve a 96+% uptime. Although Vindication uses spell hit cap, at melee hit cap with AS as a backup, you're pretty much ok.

Still, if I'm ever tanking with a warrior or druid, I'll just have them shout from now on, since there's no point in taking the chance that Vindication will fall off even for a second.

[theckhd]

Hm... now I have to play around with gear and see what level of hit rating I can reasonably reach. This makes the tier gloves look a lot more attractive than before.

[Awyndel]

Well ofc it makes hit a better stat, but I don't see a reason to start stacking it more then usual.

A ret paladin can easily drop impr might and pick up vindication ( asuming you have 2 ) , and if you're tanking with a warrior or druid its even easier. Worst case you drop an aura mastery and let the holy paladins cover those.

Myself I'm picking up hit instead of avoidance anyway. I already use the tier hands instead of the shoulders. And I'm thinking of the 277 off set head once I pick up the 277 tier chest ( not soon ) . 264 shoulders look good as well.

[Iselian]

Awyndel, I know most retribution paladins running with some variant on this spec, usually going with Divine Purpose and Unyielding Faith for the last four. Do the ones you know run with something closer to this? They could always drop PoJ for Vind, and go Tuskarr's to have some run speed, or Imp BoM as you said.

[Bobness]

Alternatively if you have 2 Pala Tanks have one spec into seal of command, it may be anecdotal but i've yet to see a mob i've attacked not have vindication whilst SOC has been present in my rotation.

This obviously will be a problem if threat is sketchy.

[Shathus]

Does this up the value of the crafted boots as well then?

I was looking through some logs to see what kind of uptime I was getting (running with 4-6% hit depending on fight) and was very confused to see only 25-35% uptime. Then I realized the ret pally in our group had it as well and hers overrode mine a lot it would seem. Between the 2 of us it seems be about a 98% uptime, so I guess nothing much to worry about there.

[Racolus]

Meh, asking someone to apply the AP debuff for you is just like asking someone to apply the 'slow' debuff for you, it is viable, and I think it has been argued many times whether we should take JotJ or not.

The point in here is you can apply this sort of debuff yourself, instead of shouting/asking/begging/hoping someone to do it for you.

I'm kinda rely on myself instead of asking someone. Afterall, it is a debuff that makes the boss hit less and I think it is the tank's job.

Besides, there is always some OT role you are all alone with only someone to keep you barely alive. In this case you can't expect the 'applier' to come over and drop the debuff then go back to do his job again.

Jotj and vindy are all kinda 'passively' applied, Jotj kicks in with your Judgement, vindy kicks in with auto attack and hammer. Why bother to ask someone when you can easily do it by simply maintaining your rotation?

[Kihra]

Does this up the value of the crafted boots as well then?

I was looking through some logs to see what kind of uptime I was getting (running with 4-6% hit depending on fight) and was very confused to see only 25-35% uptime. Then I realized the ret pally in our group had it as well and hers overrode mine a lot it would seem. Between the 2 of us it seems be about a 98% uptime, so I guess nothing much to worry about there.

A ret paladin will be melee hit capped, which all by itself gives you 95% uptime or so. Basically I think it boils down to this:

(1) If your co-tank is a warrior or druid, then have them shout. There's just no reason not to have the guaranteed 100% uptime.
(2) If you have a ret, you're fine, but you should probably have a Power Aura so you can Avenger's Shield any unlikely gaps that occur.
(3) If you're tanking a boss mob by yourself, you should strongly consider having a decent amount of Hit Rating to maintain Vindication.

The Lich King fight is a great example of (3). You may end up being the only one on LK, since people have to switch to Valkyrs and Vile Spirits and Raging Spirits. Watching Vindication fall off right as I got a Soul Reaper was exciting.

[Kihra]

I'm waiting for an exciting graph from Theck that shows the expected Vindication uptime for a 969 rotation with specific hit and expertise values.

[Vorianloken]

Very interesting read so far.. I guess that warriors/druids can stop SOME of the qq-ing about us having a "better" demo shout :0

And I have dual hit rings and toc10 heroic boots and ony10 helm, and other stuff here and there so I am never hurting for hit at the moment.

And yes theckd needs to come up with a pretty graph!

[Awyndel]

Awyndel, I know most retribution paladins running with some variant on this spec, usually going with Divine Purpose and Unyielding Faith for the last four. Do the ones you know run with something closer to this? They could always drop PoJ for Vind, and go Tuskarr's to have some run speed, or Imp BoM as you said.

Our rets run with the standard 3.3 aura mastery spec. It's mostly what u linked but with fear reduction for blood queen.

Giving up PoJ is not an option, costs dps and health. we only need one impr might in the raid so thats where i'm gonna take it from.

And about the other posts. Nobody is suggesting not doing it ourselves. We are merely discussing what uptime that gives us and how we can improve it.

I don't think we are going to gear for more hit when it costs health. Maybe when it costs avoidance though. Let's see some pretty numbers . On the other hand if having a ret/warrior/druid doing it double makes the uptime more reliable that would make making a gearplan for hit for this ocasion quite pointless in most scenario's.

[tlitp]

In order to maximize the uptime of Vindication, the stat priority is the following :

Code: Select all

base swing timer >> hit (to softcap) > expertise (to softcap) >> hit (to hardcap) > expertise (to hardcap) > haste

                  delta_uptime
base              92.76 (0.00)
+0.1 swing timer   0.59
+20 hit rating     0.31 (s)
+20 exp rating     0.27 (s)
+20 hit rating     0.15 (h)
+20 exp rating     0.12 (h)
+20 haste rating   0.09

Pretty graphs are kinda superfluous, as the 6/9 combo grants by itself a minimum uptime (no buffs, 0 hit/exp/haste, 1.5 base swing timer) of ~80%.

The code itself :

Code: Select all

q_vind=8.*weapon_base_speed./60; %vindication (8 ppm, not haste-normalized)

ptr1=(1+Reck).*(pph_sov.*SoV_flag+pph_soc.*SoC_flag+pph_sor.*SoR_flag) ...
    ./player_swing_speed; %AA
ptr2=1./t_hotr;           %HotR, ShoR
ptr3=1./t_jud;            %J

s1=(100-boss_avoid)./100;  %AA
s2=(100-boss_avoid)./100;  %HotR, ShoR
s3=(100-boss_miss)./100;   %J
s_vind=(100-boss_spell_resist)./100; %vindication

qq_vind=1-(1-s1+s1.*(1-q_vind)+s1.*q_vind.*(1-s_vind)).^(10.*ptr1) ... %AA
    .*(1-s2+s2.*(1-q_vind)+s2.*q_vind.*(1-s_vind)).^(20.*ptr2) ...     %Hotr+ShoR
    .*(1-s3+s3.*(1-q_vind)+s3.*q_vind.*(1-s_vind)).^(10.*ptr3);        %J

[theckhd]

Well, a pretty graph does show one thing:



So melee hit-capping brings you up from just under 90% uptime to around 93.5% uptime. That's a pretty small difference for a noticeable amount of hit rating.

Note that this assumes the 1% hit bonus from a Draenei (hence the knee in the graph occurs at 230 hit rating rather than 263).

Specifically for tlitp: I've added this code to the matlab sims, I'll upload the new versions a little later this morning. I also fixed one error I found:

Code: Select all

%Vindication
q_vind=8.*weapon_base_speed./60; %vindication (8 ppm, not haste-normalized)

ptr1=(1+Reck).*pph.*(SoV_flag+pph_soc.*SoC_flag+pph_sor.*SoR_flag) ...
    ./player_swing_speed; %AA
ptr2=1./t_hotr;           %HotR, ShoR
ptr3=1./t_jud;            %J

s1=(100-boss_avoid)./100;  %AA
s2=(100-boss_avoid)./100;  %HotR, ShoR
s3=(100-boss_miss)./100;   %J
s_vind=(100-boss_spell_resist)./100; %vindication

vind_up=1-(1-s1+s1.*(1-q_vind)+s1.*q_vind.*(1-s_vind)).^(10.*ptr1) ... %AA
    .*(1-s2+s2.*(1-q_vind)+s2.*q_vind.*(1-s_vind)).^(20.*ptr2) ...     %Hotr+ShoR
    .*(1-s3+s3.*(1-q_vind)+s3.*q_vind.*(1-s_vind)).^(10.*ptr3);        %J

Note that pph_soc and pph_sor as defined in the parryhaste module are scale factors. In other words, to get the proper value of pph_soc, you need pph*pph_soc. I pulled pph outside the parentheses in ptr1 to fix it.

I also fixed a similar error with bph in the dynamic redoubt code, and added a bunch of zeros(size(x)) and ones(size(x)) multipliers in various spots to make many of the min() or max() functions work properly when Gear_Hit is an array.

[tlitp]

I also fixed one error I found: (...) Note that pph_soc and pph_sor as defined in the parryhaste module are scale factors. In other words, to get the proper value of pph_soc, you need pph*pph_soc.

They are scale factors in your version of the code, they aren't in mine. So much for our best efforts to match them.
But, yeah, the graph should come in handy for people that aren't especially fond of maths. I've forgotten to put a big TLDR in there - if you're concerned about Vindication's uptime, look at the weapon swing timer first, not at your hit rating.

[Xenix]

One major annoyance I've discovered is that you can only refresh/overwrite your own Vindication, not someone else's. This means that if mine drops off (say, on LK), and the ret paladin put his up then has to run out of melee, I don't get a chance to start reapplying it until his has already dropped off. That means depending on where I am in the rotation and where my ability GCD's are relative to his could result in a fairly large downtime compared to a theoretical one.