3.2 LoO BiS?

[Aedolfmawini]

I highly recommend all tankadins on the PTR take the time to put in a ticket on the ilevel 226 libram so that it doesn't make it to live in the current broken form. LoO has had its effectiveness doubled, the same should be done to other block value librams.

[Cidx]

As I thought would happen.

LoO has received a 10 second duration while LotSS received a 20second duration.

With a current 696 rotation LoO is back superior to LotSS in 3.2 providing more BV per rotation. Since the increase of duration both will stay up a full rotation till the skills (Judgement for LoO or HS for LotSS) are used again.

0.0 - Judgement
1.5 - ShoR
3 - HS
4.5 - HoTR
6 - Cons
7.5 - ShoR
9 - Judgement

[Kvaern]

As I thought would happen.
LoO has received a 10 second duration while LotSS received a 20second duration.

It's a constant source of bitter amusement to me just how clueless the guy doing librams must be.

Hopefully someone with just a little knowledge about paladin tanking will come along and fix it like he did the first terrible incarnation of LotSS.

[Candiru]

Well, at 20s duration its actually much easier to swap it in for libram-swapping macro now right?

Although, that's not what they intended I am sure...

[Annothori]

I was under the impression that libram swapping was completely useless because of the GCD restrictions.. I made a few macros to swap from the Obstruction to Sacred Shield and back to get max Block value, but was always incurring extra GCDs I didn't want.

[Meyrinn]

As much as LoO is better then LotSS, with the change to ShoR, I think the BiS librams will be the 3.2 librams. 200 Dodge rating is better than LoO for damage reduction/avoidance, and I think 200 Str from the Ret Libram will be better threat than LoO.

[mclem]

A tiny but relevant update: there's been a blue post on the forums which at least acknowledges that they're aware of the problem:

http://blue.mmo-champion.com/13/1869800 ... hield.html

[Sanctify]

Looks like they nerfed LoO... the extra block value only lasts for 5 secs, not 10 secs as stated on the tooltup.

[steadypal]

Looks like they nerfed LoO... the extra block value only lasts for 5 secs, not 10 secs as stated on the tooltup.

its still 10 seconds for me

[Nadir]

I'm three badges (two dailies) away from testing the new 3.2 libram. If it has a 45 second ICD, I'll probably sell it back and save up for the new ilvl 245 armor/sta ring.

[steadypal]

I'm three badges (two dailies) away from testing the new 3.2 libram. If it has a 45 second ICD, I'll probably sell it back and save up for the new ilvl 245 armor/sta ring.

whoa whoa, i only have 16 badges, how are u having 22 already?

[Cidx]

I have 16(18 soon) as well. However I'm not digging the new dodge libram I mean about 3% dodge before DR. Not to mention Blizzard is aware of the libram issue so we might be seeing a buff to LotSS.

[Nadir]

I'm three badges (two dailies) away from testing the new 3.2 libram. If it has a 45 second ICD, I'll probably sell it back and save up for the new ilvl 245 armor/sta ring.

whoa whoa, i only have 16 badges, how are u having 22 already?

I got an extra six due to a server transfer. One daily away now.

[Nadir]

Slightly off topic, but from my initial testing, it appears that the Retribution libram effectively grants 100% uptime of 200 STR while Holy Vengeance/Blood Corruption is ticking. Therefore I'm probably going to stick with the DPS libram and buy the tanking one second.

Some initial observations about Libram of Defiance (tanking one):
1. Does not have a 45 second ICD.
2. When using Hammer of the Righteous every 6 seconds, I would never see Evasion proc on the first Hammer after the previous Evasion proc.
3. The earliest I was ever able to refresh Evasion is with 8 seconds left on the 18 second timer. The suggests a 9 or 10 second ICD.
4. Wowhead lists the proc chance at 80%.

If we use a 9696 rotation:

0.0 Holy Shield
1.5 Hammer of the Righteous (80% chance)
3.0 Judgement
4.5 Shield of Righteousness
6.0 Consecration
7.5 Hammer of the Righteous (0% chance, on theoretical ICD)
9.0 Holy Shield
10.5 Shield of Righteousness
12.0 Judgement
13.5 Hammer of the Righteous (80% chance)
15.0 Consecration
16.5 Shield of Righteousness

... repeat

I'm not quite sure how to calculate the average up time. You'd have to model:
18 second Evasion duration
6 second interval between proc chance if the previous proc failed
10 second interval between proc chance if the previous proc succeeded
Chance to have Hammer of the Righteous miss, be dodged , or be parried (this one might be irrelevant based on the Libram's wording)


Given that the ICD is 10 seconds and Evasion lasts 18 seconds, then I don't believe that we can substantially increase the average up time by modifying 9696 so that the 2nd Hammer is outside of the ICD window. In an ideal 9696, we're using Hammer every 6 seconds, with an Evasion proc chance every 6 or 12 seconds depending on if the previous proc failed. I suppose we might be able to slightly increase the average up time by using Hammer every 10 seconds per successful proc and 6 seconds per unsuccessful proc. That might get a bit complicated in an actual raid setting, however...

[theckhd]

I'm sure there's an analytical way to calculate LoD uptime, but it will involve conditional probabilities, which is more thinking than I'm able to put into it this early in the morning. Plus, that's not really my specialty. My initial estimate would be based on the fact that the probability of the buff falling off once it triggers is 1-(0.2)^2 = 0.96, since you always have two chances to refresh it after it triggers. This is just a rough guess though, since it's just the probability of falling off, not the actual uptime.

Luckily, my specialty is brute-force numerical simulation, which I can do this early in the morning. I wrote a quick script to estimate LoD uptime: calc_lodefiance.m

Code: Select all

%number of iterations to run
N=10000000;

t=0:6:6*(N-1);
icd=zeros(N+1,1);
buff_up=zeros(N+2,1);

for m=1:N;
    %if the proc isn't on ICD
    if icd(m)==0
        %roll random number
        a=rand;
       
        %libram procs (80%)
        if a<0.8
            %put trinket on ICD for 10 seconds (just one iteration)
            icd(m+1)=1;
            %register buff as active for 18 seconds (three iterations)
            buff_up(m)=1;
            buff_up(m+1)=1;
            buff_up(m+2)=1;

        %libram doesn't proc   
        else
            %nothing to do here

        end
       
    %proc is on ICD   
    else
        %nothing to do here
           
    end       
end

uptime=sum(buff_up(1:N))./N

For 1000 iterations, the value hops around in the 96-99% range, but once you increase the number of iterations to 10^7, it stabilizes between 97.77% and 97.78%.