Hold the Line and Parry

[Andenthal]

Regarding Hold the Line and Parry.

I see a lot of players saying “Reforge to get more Parry than Dodge because it is better due to Hold the Line procs.”
At first glance that makes sense, as there are no talents that take of advantage of Dodge over Parry, while HtL takes advantage of Parry over Dodge.
I admit I was dubious at first if increasing your chance to Parry really made a noticeable difference in HtL uptime.
So, I wanted to figure out how much HtL uptime was gained by adding more Parry. Turns out, I think I was both right, and wrong.

The formula I used is the same used for calculating Flurry uptime for Fury Warriors (very similar mechanic) – found here http://elitistjerks.com/f81/t22705-dps_compendium/
As is said in the EJ post, and I will restate here; that is not an exact formula for uptime. But it’s pretty darn close.

We’ll use that formula for calculating HtL uptime.

1–(1–P)^A

Where P = Parry chance and A = the number of attacks you will receive while the HtL buff is active.
We assume a 2.0 attack speed for bosses in most all other formulas, so we assume the same here. You’ll get hit 5 times during a HtL proc, so A = 5.

Just sub in P for varying amount of Parry, and we get this chart for uptime:
(Note: numbers are rounded to the nearest hundredth of a percent)

Parry %	     HtL uptime%:	Uptime (sec)	Extra time 	Parry Rating	Rating/P%	Parry/HtL%	
10	     40.95	        122.85	        10.06	        230	        49	        14.63	        
11	     44.16	        132.48	        9.63	        279	        50	        15.58	        
12	     47.23	        141.69	        9.21	        329	        52	        16.94	        
13	     50.16	        150.48	        8.79	        381	        53	        18.09	        
14	     52.96	        158.88	        8.4	        434	        54	        19.29	        
15	     55.63	        166.89	        8.01	        488	        56	        20.97        
16	     58.18	        174.54	        7.65	        544	        57	        22.35	        
17	     60.61	        181.83	        7.29	        601	        59        	24.28	        
18	     62.93	        188.79	        6.96	        660	        61	        26.29	        
19	     65.13	        195.39	        6.6	        721	        62	        28.18	        
20	     67.23	        201.69	        6.3	        783	        63	        30	        
21	     69.23	        207.69	        6	        846	        66	        33	        
22	     71.13	        213.39        	5.7	        912	        67	        35.26	      
23	     72.93	        218.79	        5.4	        979	        69	        38.33	       
24	     74.64	        223.92	        5.13	        1048	        72	        42.11	        
25	     76.27	        228.81	        4.89	        1120        	73        	44.79        	
26	     77.81	        233.43	        4.62	        1193	        76        	49.35	        
27	     79.27	        237.81	        4.38	        1269	        78        	53.42        	
28	     80.65	        241.95	        4.14	        1347        	80        	57.97        	
29	     81.96	        245.88	        3.93	        1427        	83        	63.36	        
30	     83.19	        249.57	        3.69	        1510        	NA	        NA

Explanation of columns:
Parry% - Parry chance (note1)
HtL uptime – The average uptime of Hold the Line, using the above formula
Uptime (sec) – The uptime of Hold the Line in a 5 min fight.
Extra time – How many additional seconds of Hold the Line uptime are gained compared to the previous Parry chance
Parry Rating – How much Parry rating is needed to get this much Parry chance
Rating/P% - How much Parry rating is needed to get an additional 1% Parry chance
Parry/HtL% - How much Parry rating is needed to increase Hold the Line’s uptime by 1%

Note1: This would be your "absolute" or actual Parry chance, which isn't necessarily what your character sheet displays. All other numbers reflect character sheet numbers.


One thing that I noticed though was how much uptime is actually gained when you put it into secs of combat. While going from 10% Parry to 11% Parry gives over 3% more HtL uptime over the course of the fight. But, that only translates to about 9.6 secs – or about 1 full HtL proc. Offensively, this translates to about 6.5 GCDs, and up to 3 Heroic Strikes (if you’re HSing every 3 secs), for 9 total additional attacks with +10% crit by increasing your Parry chance by 1%.

Defensively that comes out to just shy of 5 boss swings with an additional +10% Crit Block rate. (From here you figure out your chance to get a crit block within those ~5 attacks, which varies depending on your Mastery skill.)

As you can see from the chart, HtL uptimes gains from additional Parry are subject to diminishing returns – which is to be expected given we are using the same formula for Flurry, which has diminishing returns on additional crit. Going up to 25% Parry (which I’m guessing most tanks have in WotLK now), only about ½ of a full HtL proc is gained. That comes out to 3 GCDs, plus 1 HS, and about 2 boss swings with the additional Crit chances (offensive and defensive). Not only is it difficult to increase your Parry chance by 1% at this level, but when you do, the gains are likely not worth it (4 attacks and 2 boss swings)

[Andenthal]

I tried to only include subjective statements in the first post. This will be my personal feelings regarding HtL and Parry

For the last 6 or so weeks of WotLK I imagine most tanks are in the 20%+ range of Parry, likely closer to 25% (unless you've reforged it to Mastery). If so, I would not think that seeking additional Parry for HtL up time it worth it. If you have significantly lower Parry chance (15% or so) the gains are pretty good, as getting 60 or so Parry Rating gives you about 8 secs of HtL uptime.

I'm not sure what levels of Avoidance tanks are at on the Cata Beta, but if they are in the lower numbers (sub 15%), it may very well be beneficial to trade some Dodge for Parry for increased HtL uptime. I have a feeling past 15% and you'd need to figure out if additional HtL uptime is worth the avoidance loss from changing Dodge into a lower amount of Parry.

I was also starting to do the math regarding "regearing" for HtL uptime versus gearing for straight threat (Crit mostly). I'm not done yet, but it seems that until at least about 20% Parry chance, you get more DPS/threat "regearing" for Parry versus regearing for Crit (due to HtL's +10% crit). It's slightly harder comparing regearing for Parry (and HtL) versus gearing for Mastery, as the HtL bonus is not quite the same thing as Mastery. I have a feeling the defense side of it will likely be a "aim for balance" strategy.

[Airowird]

The Flurry uptime and the HtL uptime are totally different
The Flurry one just calculates the chance there is for you to have procced Flurry in the last 3 swings and gives you that chance to have the haste buff on a swing.
HtL doesn't work like that, the occurence that activates the buff is not the one consuming the buff (or even using the buff for that matter!)

Here's what you should have done:
Calculate the odds of HtL being refreshed on each consequent swing after the initial/activating one.
Multiply that with the time between the activating and the refreshing hit.
Add those up and add the chance to not refresh * full duration.
This is your statistical average HtL duration per parry.
Simply add in Parry chance & boss timer and you have your effective uptime (statisticly)

Also, I'm personally not expecting Parry chances above 20% in Cata (at 85), the 15-20% bracket will be the major focus I think/hope

The thing is, the statement you quoted is correct when you clarify it decently:
* Dodge & Parry are on equal DR formulas, so whichever you have less of is the better one to get for pure damage reduction.
* Parry gives extra, while Dodge does not.
* If you have to chose between Dodge being the higher one or Parry, Parry wins.
* It does NOT mean however you should dump all your Dodge for Parry, I'm guessing getting your Parry between 100% & 110% of your Dodge rating is the best choice, depending on the mastery you already have.

[Gasman]

Has anyone done any theorycrafting on how blade ward stacks affect HtL uptime? One would think that HtL would make blade ward a much better tanking enchant than it once was.

[Andenthal]

The Flurry uptime and the HtL uptime are totally different
The Flurry one just calculates the chance there is for you to have procced Flurry in the last 3 swings and gives you that chance to have the haste buff on a swing.
HtL doesn't work like that, the occurence that activates the buff is not the one consuming the buff (or even using the buff for that matter!

I didn't see a huge difference in dealing a crit every X strikes versus receiving a Parry every X strikes - statistically speaking.
I will take your advice into consideration, thank you.

Although, even if the numbers relating to each percentage of Parry are inaccurate (likely by very low single digit percentages) the numbers relating to uptime and seconds of uptime in combat are correct. I should have specified as such in my second post, but that was what I was more interested in finding out. How much longer will HtL be up if I swap my stats around? (or prioritize this piece of gear over that one, etc)

My Warrior has 22.13% Dodge and 21.96% Parry currently (969 and 989 Ratings respectively) - which is about as balanced as you can get without any reforging. If I were to make my Parry 110% of my Dodge I would trade about 80 Avoidance rating. That would give me an average of about 5.5 more seconds of HtL up time in a 5 minute fight. For me personally, it's hardly worth the gear change for something that's not likely to make a noticeable impact on the outcome of the fight.

[Andenthal]

Has anyone done any theorycrafting on how blade ward stacks affect HtL uptime? One would think that HtL would make blade ward a much better tanking enchant than it once was.

There's 2 components to the Bladeward enchant. The first is a stacking buff that has a chance to occur on a damaging attack.
The second is a damage component that occurs on a Parry (after you gain the buff).

You can track the effectiveness of Blade Ward though its damage done. If it is damaging often (not the amount of damage, but how many times it damages), then you are getting good use out of it.

If you get the buff, then the buff fades before a Parry occurs, you're not getting much out of it. It has to do damage (you have to Parry) to affect HtL uptime, simply gaining the buff does nothing.
I think you'll find the latter more true than the former.