1. #1

    protection: capping dodge expertise vs parry expertise

    Hey, for the time being I'm capping hit & expertise for dodge but not expertise for parry, and go for haste afterwards. I'm wondering though, how big would the advantages/disadvantages, of going for full cap on parry expertise, be? My average itemlevel is 500.

    Thanks for any advise.

  2. #2
    Advantages: More reliable HoPo generation
    Disadvantages: Lower DPS, HPS, APS, HoPo generation

    Personally I value expertise hard cap, but I do not value if over haste. So I go Hit>Exp7.5>Haste>Exp15>Mastery>Crit. Often change reforging from mastery to crit on some fights where mastery is not so strong.
    Volun-told - A supposedly optional event, award, assignment, or activity in which a person (or persons) are required to attend either by persons-in-charge nominating them or their peers expecting them to be there. The individual often has no say in the matter, and non-attendance in frowned upon.

    I am so tired of seeing terrible people, being admired, for being terrible people.

  3. #3
    in other words, don't go for parry cap. you will lose more than you gain.

  4. #4
    Quote Originally Posted by dar0 View Post
    in other words, don't go for parry cap. you will lose more than you gain.
    That really depends. In some situations, mainly being in heroic modes undergeared, sometimes you can not afford to miss 1 or 2 hopo generators, it can result in death.
    Though, that can be more an exception than a rule since it is only CS/HotR that can be dodged parried

    Its a reliability vs throughput question.
    Personally, I do not think that the reliablity loss is so big, especially if you are around 10-11 expertise. Maybe 7.5 is not so good, but even if your priortise haste you should be able to reach around 10-12 expertise anyway, since you can't reforge everything to haste.
    And the throughput gain of haste is quite noticeable.

    Also, if you are using SW/HP as T5 talent, especially SW, there is even lower reason to cap expertise.

    It can also be argued in those situations where you "cant afford to miss", having haste>expertise cap will also often save your life since you will cover more attacks with SotR (you should already have HoPo banked for every hard hitting spell so that is irrelevant in haste vs expertise argument)

    Though really, in all other situations where you life hangs in the balance of hitting a SotR or not, there is absolutely no reason to cap expertise.

    I am currently sitting on 12~ ish expertise normally, and around 15 when I equip expertise trinket. But I have prioritised haste on every gem and reforge on my gear.
    Last edited by Firefly33; 2013-03-26 at 12:48 PM.
    Volun-told - A supposedly optional event, award, assignment, or activity in which a person (or persons) are required to attend either by persons-in-charge nominating them or their peers expecting them to be there. The individual often has no say in the matter, and non-attendance in frowned upon.

    I am so tired of seeing terrible people, being admired, for being terrible people.

  5. #5
    Just to offer a different perspective, I like to exp hardcap mostly for the simple fact that I hate missing abilities, seeing an ability get parried and not getting holy power when I expect to annoys me to no end. That said I played around with the haste>exp hardcap build awhile ago and just didn't see the point of it because I was gaining <1% haste while losing hardcap.

    On an up to date note I played around with this prio just now and gained 2.5% haste while sitting at 12% exp. not bad, gonna test this prio system for this weeks ToT farm clear and see how well it goes.

    Firefly, out of these 2 trinket setups which would you run.
    Brutal Talisman of the Shado-Pan Assault + Primordius' Talisman of Rage @ 12% exp, 36.76% haste .or.
    Brutal Talisman of the Shado-Pan Assault + Lei Shen's Final Orders @12% exp, 38.02% haste.

  6. #6
    if you are under the hard cap you might run into the situations were you get an unlucky streak of parry. I have felt this when i sometimes had my crusader strike/hammer being parryed 3 times in a row and that is with 13% exp. it is really sad when that happens.

    you might benefit from having more haste instead of exp but it leaves it all down to luck which is often bad because it can screw you over

  7. #7
    Depends on fight really, also not really sure about the uptime on Primordius. But yeah, depends on fights a bit. Lei Shen have more haste so it is naturally better defensively. Though the difference is not huge with reforging, amd the talisman is most likely (depending on proc uptime) better than Lei Shen for dps.

    I feel like it is safe to say that the difference is not huge. Either choice is fine.

    I got Gaze of the Twins this week, but its uptime is far to low so far. When I get decked into more 435 gear I will be able to get more crit making that trinket better, but this early in the tier, it just procs way to little due to it only proccing on crits. Still get about 40% uptime with only 6% crit on single target fights. So I am going to experiment more with Gaze and crit% before giving a complete verdict. For now it looks a bit weak, but still interesting.

    @hit vs exp. Well, it can be so that you only gain 1% haste, but then you only lose 1% expertise aswell.
    Remember that expertise only affects melee hits. That is SotR, HotR, CS and melee hits. SotR still procs even if you get parried, it is just the damage part that you lose.

    Getting parried sucks, yes, but it is one of those situations, it feels worse, but in reality it isn't. You can't see that 1% haste work in the same way you can see 1 less expertise. You are not going to feel "Hey, I got off 2 more SotRs this fight!", but you will feel "Hey, I got parried twice!". Even though the effect of one is bigger than the other.

    The damage that is affected by expertise is roughly 33% of our damage, give or take a couple of % depending of playstyle, gear, boss, and talents. That means that the other 66%~ is not affected by it at all. While haste affects 90%~ of our damage output, aswell as giving benefit to sacred shield.
    Last edited by Firefly33; 2013-03-26 at 02:06 PM.
    Volun-told - A supposedly optional event, award, assignment, or activity in which a person (or persons) are required to attend either by persons-in-charge nominating them or their peers expecting them to be there. The individual often has no say in the matter, and non-attendance in frowned upon.

    I am so tired of seeing terrible people, being admired, for being terrible people.

  8. #8
    Yea, I'll stick to Primordius' this week while giving haste>exp hardcap prio a go and double check the proc rates, last week I was getting anywhere from 45% - 70% uptime on fights @ 34% haste. Doing a full clear this week should give a better representation of proc rates for each fight though I hope.

    Although running Primordius' without exp cap might be worse for proc rates since it procs from melee attacks which will have a chance to get parried now, don't know for sure, guess I'll found out by wed/sun night.

  9. #9
    With the gear I have now, I reach expertise cap if I use Brutal Talisman, even with the exp7.5>haste>exp15 prio.

    Though that is with GID, considering replacing GID with Gaze for the next week.
    Volun-told - A supposedly optional event, award, assignment, or activity in which a person (or persons) are required to attend either by persons-in-charge nominating them or their peers expecting them to be there. The individual often has no say in the matter, and non-attendance in frowned upon.

    I am so tired of seeing terrible people, being admired, for being terrible people.

  10. #10
    Without taking RPPM procs into account, sacrificing the expertise hard cap to increase haste is a theoretical decrease in Holy Power generation and dps over the course of a several-minute-long fight. This drop becomes more and more severe the more haste you have on top of what you're gaining by giving up expertise. "Reliability" need not enter into the discussion.

    For each 1% haste you gain in preference to hard-capping expertise, you're losing 1.25% expertise. Even if these two values were identical, the drop in expertise would still always have a greater negative effect than a concurrent equal gain in haste.

    Code:
    Exp Rtg	Hst Rtg  Hit %	 Atk Spd  Throughput
    2550	2550	 92.50%	 106.00%  98.05%
    2600	2500	 92.65%	 105.88%  98.10%
    2650	2450	 92.79%	 105.76%  98.14%
    2700	2400	 92.94%	 105.65%  98.19%
    2750	2350	 93.09%	 105.53%  98.24%
    2800	2300	 93.24%	 105.41%  98.28%
    2850	2250	 93.38%	 105.29%  98.33%
    2900	2200	 93.53%	 105.18%  98.37%

    Given that, then considering that losing the hard cap also exposes you to bad-luck RNG streaks, I can't find any reason to gamble for possible (but uncontrollable) short-duration gains at the cost of overall performance in both HoPo generation and dps/threat.

    [Edit: Firefly, you make a very good point (as always) that many of our attacks are spells and do not benefit from expertise beyond what is necessary to achieve 15% spell hit. I'm still torn on the overall effect on HoPo generation with our three main holy power generators being subject to parries -- and our fourth generator generally being a proc from either of those three except on AoE pulls. For dps/threat, haste almost certainly prevails.]
    Last edited by Kaeth; 2013-03-26 at 06:06 PM.

  11. #11
    Fluffy Kitten Rivin's Avatar
    Join Date
    Oct 2009
    Location
    Washington, USA
    Posts
    13,823
    Quote Originally Posted by Kaeth View Post
    I'm still torn on the overall effect on HoPo generation with our three main holy power generators being subject to parries --
    Only two are affected by expertise at all: Crusader Strike and Hammer of the Righteous. Judgment cannot be dodged or parried and Grand Crusader will give you Holy Power even if Avenger's Shield misses or is dodged.

  12. #12
    Quote Originally Posted by Kaeth View Post
    Given that, then considering that losing the hard cap also exposes you to bad-luck RNG streaks, I can't find any reason to gamble for possible (but uncontrollable) short-duration gains at the cost of overall performance in both HoPo generation and dps/threat.
    Actually. That is plain wrong. Maths incoming.
    What you are forgetting is that only 1 of our 3 HoPo generators is subject to parries and only 33%~ ish of our damage output.

    h = haste factor
    m = miss chance
    nAe(r) = value of expertise per rating during normal rotation
    nAh(r) = value of haste per rating during normal rotation
    C = Crusader Strike HoPo regen = 1/4.5
    J = Judgement effective HoPo regen = 1/6.5

    nAe=(1.2*C*(1+h)/340)*r
    nAh=((1.2*C*(1-m)+J)/425)*r

    Now this was the old one. The new one even further devalues expertise because it is only 12% chance to proc GC.

    nAe=(1.12*C*(1+h)/340)*r
    nAh=((1.12*C*(1-m)+J)/425)*r
    should be the new formula then.

    Lets as in your example presume 7.5 expertise and 6% haste. This means

    nAe=(1.12*C*(1+h)/340)*r
    nAe=(1.12*(1/4.5)*1.06)/340= (7.759477e-4)r

    nAh=((1.12*C*(1-m)+J)/425)*r
    nAh=((1.12*(1/4.5)*(1-0.075)+1/6.5)/425)= (9.036903e-4)r

    So what you can see is that at 6% haste and 7.5 expertise, haste provides 16.46% more HoPo. That is a non negliable amount. It also provides far more dps. Do not have the exact numbers here, but a lot higher than 16.46%.

    Now you can plot in your own haste and expertise values to see where you stand.

    If you are using SW, it further devalues expertise.

    h = haste factor
    m = miss chance
    nAe(r) = value of expertise per rating during normal rotation
    nAh(r) = value of haste per rating during normal rotation
    C = Crusader Strike HoPo regen = 1/4.5
    J = Judgement effective HoPo regen = 1/6.5
    awAe(r) = value of expertise per rating during Avenging Wrath
    awAh(r) = value of haste per rating during Avenging Wrath
    awC = Crusader Strike HoPo regen during SW = 1/6
    awJ = Judgment HoPo regen during Sw = 1/3

    nAe=(1.12*C*(1+h)/340)*r
    nAh=((1.12*C*(1-m)+J)/425)*r
    awAe=(1.12*awC*(1+h)/340)r
    awAh=((1.12*awC*(1-m)+awJ)/425)r

    fAe = value of expertise per rating during an entire fight
    fAh = value of expertise per rating during an entire fight

    AW^ = Avenging Wrath uptime
    fAe = (nAe*(1-AW^))+(awAe*AW^)
    fAh = (nAh*(1-AW^))+(awAh*AW^)

    This further devalues expertise a lot!
    Last edited by Firefly33; 2013-03-26 at 06:24 PM.
    Volun-told - A supposedly optional event, award, assignment, or activity in which a person (or persons) are required to attend either by persons-in-charge nominating them or their peers expecting them to be there. The individual often has no say in the matter, and non-attendance in frowned upon.

    I am so tired of seeing terrible people, being admired, for being terrible people.

  13. #13
    Fluffy Kitten Rivin's Avatar
    Join Date
    Oct 2009
    Location
    Washington, USA
    Posts
    13,823
    Well, I can't read that mumbo-jumbo, but even if dropping expertise for haste provides more Holy Power generation in theory, the big reason to cap hit and expertise is reliability. On some fights you might not get parried at all and all that haste is useful, but on some fights you might get every Crusader Strike parried (unlikely, but possible!) and then the value of haste on that fight plummets. With expertise capped, it's always steady.

  14. #14
    Quote Originally Posted by Rivin View Post
    Well, I can't read that mumbo-jumbo, but even if dropping expertise for haste provides more Holy Power generation in theory, the big reason to cap hit and expertise is reliability. On some fights you might not get parried at all and all that haste is useful, but on some fights you might get every Crusader Strike parried (unlikely, but possible!) and then the value of haste on that fight plummets. With expertise capped, it's always steady.
    To which the point is, the difference in reliablity is almost non-existant, while the increase in HoPo generation is quite large. Especially if you reach 11-12 expertise while maxxing haste, which is completely reasonable. And getting a double parry in a row is not the end of the world either.

    I'd rather base my gearing on the rule than the exception. Getting parried is not a big issue. Even if every one of your attacks get parried, you still have time on eveyr single boss fight to bank up 5(6) HoPo for every single "big boss hit", making you able to chain double SotR for when the real damage is. This is not the haste vs mastery argument. This is a much simpler argument. Haste over expertise will allow you to cover a lot more boss auto attacks, with the drawback that a bad, and to further express BAD luck, you could recieve 1-2 more boss melee hits before getting of the next SotR. But you still cover every big hits with SotR regardless. And then remember that the increased haste works in the exact opposite, at points you will recieve 1 less boss melee hit before getting up the next SotR.

    Expertise is obviously more valuable in 25 man since the boss melee hit is bigger. But in 10 man, the reason to go for expertise hard cap is just not as appealing as the haste. Especially if you are using SW.

    At 0 haste and 7.5 expertise, it is almost 20% increased value in haste over expertise for HoPo regen and even bigger dps increase.

    To tell, especially a 10 man, paladin tank to go for expertise hardcap over haste, is kinda like telling a hunter to not get crit.

    "Crit increases your dps the most, but if you get bad luck with crits you may get lower dps". Just no, you do not do that. If one stat benefit you 9 out of 10 times, then you probably go for that stat.
    Haste gives far more HoPo and DPS. The reliability of Expertise does not outweigh that.
    Especially considering that the increased HoPo generation in haste gives a reliablity of its own, in you taking on average less auto attacks between each SotR and you are covering more auto attacks each fight with it.
    Last edited by Firefly33; 2013-03-26 at 07:01 PM.
    Volun-told - A supposedly optional event, award, assignment, or activity in which a person (or persons) are required to attend either by persons-in-charge nominating them or their peers expecting them to be there. The individual often has no say in the matter, and non-attendance in frowned upon.

    I am so tired of seeing terrible people, being admired, for being terrible people.

  15. #15
    @Firefly

    Wow, I'm definitely going to have to blatantly steal some of your formulas to find the expertise rating = haste rating breakpoint (for normalized HoPo gen) for varying preexisting levels of haste. Though, based upon your existing posts, I'm going to guess it's far closer to the 7.5% expertise point than the 15% one...

  16. #16
    Fluffy Kitten Fhi's Avatar
    Join Date
    Oct 2010
    Location
    Carnished Toast
    Posts
    3,219
    I'm actually trying out softcap instead of hardcap and putting that 2550-ish (less cos I is human) rating into haste. In my case it probably doesn't matter as much since I don't do progression raids, but it was largely because of the "feeling" I have with softcap (which Firefly put into numbers).

    Paladin and D3 Moderator | MMO-C Guidelines | Signature and avatar art by Dancing Hare

  17. #17
    Quote Originally Posted by Kaeth View Post
    @Firefly

    Wow, I'm definitely going to have to blatantly steal some of your formulas to find the expertise rating = haste rating breakpoint (for normalized HoPo gen) for varying preexisting levels of haste. Though, based upon your existing posts, I'm going to guess it's far closer to the 7.5% expertise point than the 15% one...
    I believe pre patch, it was about 18.44% haste - 7.5 expertise was the breakpoint. But the change with 20% to 12% chance to proc GC proc slightly changed this, so would assume it is closer to 18.5-18.75% haste now.

    Edit: Just checked, the new value is 23.45% haste. At 23.45% haste expertise is better for throughput at 7.5 expertise rating. So basically, always. Since when you get above 23.45% haste, you will have more than 7.5 expertise rating and before you reach the next breakpoint you will have 15 expertise anyway.

    That said, it becomes a matter of when is that throughput worth it vs reliablity? When you have 16% haste and 7.5 expertise, the throughput difference is almost not there. So you have to calculate your own numbers and find your own comfortzone.

    And as I said earlier, prioritising haste > expertise hard cap doesnt mean you get 7.5 expertise, often you get between 10-14 expertise anyway. Heck, I am even at 15 now if I put my expertise trinket on. Going full haste does not exclude the possibility of more than 7.5 expertise on gear.

    ---------- Post added 2013-03-26 at 08:44 PM ----------

    For refference. If you are using SW. SWs uptime should be somewhere between 12-20% depending on how fast you use it and the fight duration. 16.67% is the average uptime.

    nAe = (1.12*C*(1+h)/340)*r
    nAh = ((1.12*C*(1-m)+J)/425)*r

    nAe = (1.12*(1/4.5))/340) = 0.00073203
    nAh = ((1.12*(1/4.5)*(1-0.075)+(1/6.5))/425) = 0.00090369

    awAe = (1.12*awC*(1+h)/340)r
    awAh = ((1.12*awC*(1-m)+awJ)/425)r

    awAe = (1.12*(1/6))/340 = 0.00054902
    awAh = (1.12*(1/6)*(1-0.075)+(1/3))/340 = 0.00148824

    (Notice how much the value of exp drops and haste increases during sanctified wrath. As expected though. Haste is almost 300% more efective during SW.)

    fAe = (nAe*(1-AW^))+(awAe*AW^)
    fAh = (nAh*(1-AW^))+(awAh*AW^)

    Now we can plot this into a graph.

    fAe = (0.00073203*(1-x))+(0.00054902*x)
    fAh = (0.00090369*(1-x))+(0.00148824*x)

    Where x is the uptime of SW.



    The red graph is the value of haste.
    The blue graph is the value of expertise.

    The left side of the graph is 0% uptime on SW, the right side is 100% uptime. This is ofc at 0% haste and 7.5 expertise.
    As said, during a normal fight the uptime of SW should be between the 10-20% mark.
    Last edited by Firefly33; 2013-03-26 at 07:49 PM.
    Volun-told - A supposedly optional event, award, assignment, or activity in which a person (or persons) are required to attend either by persons-in-charge nominating them or their peers expecting them to be there. The individual often has no say in the matter, and non-attendance in frowned upon.

    I am so tired of seeing terrible people, being admired, for being terrible people.

  18. #18
    Trying this tonight! Thanks for the maths!

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •