Prot gearing confusion.

[Deleted]

But this doesn't explains at all why on SimCraft you get lower HPG with exp hard-cap, than with lower expertise, but higher haste.

Why are you using SimCraft for situations that can easily be calculated by hand? Simulations ain't perfect.

Exp Hard Cap > Haste:
https://www.dropbox.com/s/ydz9fvfrvv...exp_h_cap.html

CS: 185.0 times, all of which hit.
J: 121.5 times.

Why did SimCraft cast CS more than 1.5 times as often as J? If CS casts are correct, we should be casting J 123.33 times.

Haste > Exp Hard Cap:
https://www.dropbox.com/s/3din2pnuy7...aste_prio.html

Haste went from 1.4559 to 1.5 exactly, which means we should be casting all spells 1.5/1.4559 times as often, or 1.0303 times as often.

CS: 1.0303 * 185 = 190.61, which seems to match.
J: 1.0303 * 121.5 = 125.18. But we get 127.1 casts.

Note that 127.1 * 1.5 comes down to 190.65, so here we do have 2 Js cast for every 3 CSs.

For some reason, the amount of Js that were cast in the first simulation ended up lower than they should be.

- - - Updated - - -

Minutiae is what internet forums do best...

The problem is that it's a miniscule advantage in DPS at quite a costly advantage in survival. It's like asking a Hunter: Would you like to do 1% increased dps at the chance of dropping dead in the middle of a bossfight?

[Deleted]

Why are you using SimCraft for situations that can easily be calculated by hand? Simulations ain't perfect.

Simulations are supposed to be better as they take more factors into the results. But then again, this is something that Theck should know, as he is the maintainer of SimCrafts paladin module. Whenever Simcraft is wrong or not.

Maybe the difference is that the sims used Theck's "shifting queue", which might wait for CS if it's cd is less than 0.5s, although this shouldn't normally happen, as it doesn't uses the non-haste-affected GCD for SS and lv90 talents.

[Nairobi]

So, basically what we can take out of this (and what I thought was already known) was that:

1) Always have at least 7.5% expertise.
2) Aim to get 15% expertise whenever possible, though not at the cost of haste.
3) Expertise gains value relative to haste when haste > 31%, but is slightly less valuable than haste below that.
4) The performance deltas in #3 are EXTREMELY small.
5) People love to argue on the internet.

Most of this has been a moot point, given the fact that in current gear levels (even if you only do normal modes) you're likely already tripping over expertise. This is further trivialized by the fact that we're looking at DPS and HPG variances at the 2nd and 3rd decimal places between the 2 camps of people (exp > haste vs haste > exp).

[Promdates]

The problem is that it's a miniscule advantage in DPS at quite a costly advantage in survival. It's like asking a Hunter: Would you like to do 1% increased dps at the chance of dropping dead in the middle of a bossfight?

I'd do this. But what's the chance of dropping dead?

[Theck]

PS: Those plots are assuming that we use Judgment on CD, right?

Nope, they assume a standard CS-J-X-CS-X-J-CS-X-X rotation, thus CS used every 4.5 seconds and J used an average of once every 6.75 seconds.

[trystero]

The problem is that it's a miniscule advantage in DPS at quite a costly advantage in survival. It's like asking a Hunter: Would you like to do 1% increased dps at the chance of dropping dead in the middle of a bossfight?

I don't know how costly it is - there's a point at which, yeah, you would take that marginal DPS increase. Are we talking about randomly dying every 10 pulls? 100? 15,615?

FWIW, I agree with you about expertise, as does everyone else, even those arguing against you . Wanko hard caps expertise. Firefly does too. Bizarre thread here, where one side doesn't practice what it preaches and seems to be arguing for the sake of it. If he didn't strike me - and everyone else - as "that guy" who'd argue with you for ordering fries at Wendy's, pointing out that in his personal tests McDonald's offers 1.3% more volume per dollar, I'd say he's trolling all of you with inflammatory passages like this,

That going for expertise hard cap generates more Holy Power is just a straight out lie. I really dont see anything mindboggling about this however. The situation that I just metioned was my own way of min-maxing. That is something that I do not expect the general paladin population to do. The only thing minboggling here is your need to fabricate lies to support your claim. Which agains show how weak your standpoint is. Fact remains that haste>expertise with realistic haste and expertise levels yields the highest throughput. The only time that expertise>haste in those scenarios (for 10 mans atleast) is if you are terribly bad at managing your holy power. In which case, yes, I would suggset expertise cap to solve your own problems with managing your resource system.

[Deleted]

I don't know how costly it is - there's a point at which, yeah, you would take that marginal DPS increase. Are we talking about randomly dying every 10 pulls? 100? 15,615?

FWIW, I agree with you about expertise, as does everyone else, even those arguing against you . Wanko hard caps expertise. Firefly does too. Bizarre thread here, where one side doesn't practice what it preaches and seems to be arguing for the sake of it. If he didn't strike me - and everyone else - as "that guy" who'd argue with you for ordering fries at Wendy's, pointing out that in his personal tests McDonald's offers 1.3% more volume per dollar, I'd say he's trolling all of you with inflammatory passages like this,

We hard cap it, as we can't get more haste from dropping it lower. After 530-ish ilvl it is rather hard not to hard cap exp.
I could reforge it into crit, but that would be a tiny bit extreme even for me. (Nothing impossible, but not there yet. :P )

[Theck]

Maybe the difference is that the sims used Theck's "shifting queue", which might wait for CS if it's cd is less than 0.5s, although this shouldn't normally happen, as it doesn't uses the non-haste-affected GCD for SS and lv90 talents.

The "shifting queue" refers to SotR usage, which should be off-GCD and independent of rotational constraints.

My best guess is that the reason for the difference in SimC is the non-haste-affected GCD for SS and L90 talents. CS is almost a 3-second cooldown at 45.59% haste (3.09s), so something like CS-SS-X might be happening. The combined duration of the 1.03-second CS GCD and the full 1.5-second SS GCD gives you 2.53 seconds, leaving 0.56 seconds remaining on CS's cooldown. However, once you add reaction time and/or latency into that, it would almost certainly be below 0.5 seconds, triggering the delay condition on J.

If it's happening for both SS and LH/ES, that could explain having one or two "missing" J casts, on average. That should disappear if simming with ptr=1, though I don't remember whether SimC v530-6 has those changes or not. v530-7 definitely should. Alternatively, you could try changing the delay condition on Judgment from 0.5 seconds to something smaller, i.e. ~0.35 seconds.

[Deleted]

The "shifting queue" refers to SotR usage, which should be off-GCD and independent of rotational constraints.

My best guess is that the reason for the difference in SimC is the non-haste-affected GCD for SS and L90 talents. CS is almost a 3-second cooldown at 45.59% haste (3.09s), so something like CS-SS-X might be happening. The combined duration of the 1.03-second CS GCD and the full 1.5-second SS GCD gives you 2.53 seconds, leaving 0.56 seconds remaining on CS's cooldown. However, once you add reaction time and/or latency into that, it would almost certainly be below 0.5 seconds, triggering the delay condition on J.

If it's happening for both SS and LH/ES, that could explain having one or two "missing" J casts, on average. That should disappear if simming with ptr=1, though I don't remember whether SimC v530-6 has those changes or not. v530-7 definitely should. Alternatively, you could try changing the delay condition on Judgment from 0.5 seconds to something smaller, i.e. ~0.35 seconds.

Oh, great!
I someway remembered that they where affected with haste too, and you didn't touched it, as they will change next patch too anyway. But I guess I remembered it wrong. ^^

I will check it.

[trystero]

We hard cap it, as we can't get more haste from dropping it lower. After 530-ish ilvl it is rather hard not to hard cap exp.
I could reforge it into crit, but that would be a tiny bit extreme even for me. (Nothing impossible, but not there yet. :P )

In that case, this thread has been going back and forth about something that's not even feasible?* That one of the "sides" in this debate cannot exist?

???????????????????

This rivals the type of stuff I experience at my job, and I work in academia, where I'm very accustomed to pedantic arguments (you can't put this flyer up, because the arrangement of color and line does not conform to the Hobbes' Standard, studies show that should expend at least 15% but no more than 25% of your classroom time to designated 'group work', this Camtasia presentation that no one's even going to watch must remove the school logo from in its PowerPoint slides because the logo is a sporting figure and this is an intellectual, not athletic, enterprise - yes, that one was actually real).

*Unless T16 gear is different, and you'd be able to maximize haste and have low expertise?

[Deleted]

In that case, this thread has been going back and forth about something that's not even feasible?* That one of the "sides" in this debate cannot exist?

???????????????????

This rivals the type of stuff I experience at my job, and I work in academia, where I'm very accustomed to pedantic arguments (you can't put this flyer up, because the arrangement of color and line does not conform to the Hobbes' Standard, studies show that should expend at least 15% but no more than 25% of your classroom time to designated 'group work', this Camtasia presentation that no one's even going to watch must remove the school logo from in its PowerPoint slides because the logo is a sporting figure and this is an intellectual, not athletic, enterprise - yes, that one was actually real).

*Unless T16 gear is different, and you'd be able to maximize haste and have low expertise?

T16 has even more expertise (especially if you want to use the tier set), and most will reforge out from expertise, so.... Sorry?

Umm... yeah, this is kind of silly. ^^ But that was given when we argued about 0.005%'s!

[Promdates]

The having not enough expertise to reach cap is very possible while leveling and gearing up an early (re: 470-500) paladin. Right now in T15 it's something more difficult to get rid of excess haste, I think I'm personally around 15.5+ (my ret set sits at 11+ just because I don't/can't reforge out of it due to my prot set doubling up for ret). In T16 though, I'm sure there is an item build where you could have very minimal expertise.

edit: I just put this together. http://ptr.wowhead.com/list=204833/full-haste It has 46.55% haste (19782), 21.09 mastery (12653) 7.69% hit and 15.01% expertise. I picked the highest or second highest haste item in spots I could, and then filled in with normal non-haste available items in other spots. I did end up using 4 mastery gems (two of which were JC gems), 1 Expertise/Haste and 1 Expertise/Hit to get those numbers. Could replace the mastery gems with crit gems if you really wanted to. With Thok's you end up at 13,791 mastery and 21562 haste. I originally wanted minimal expertise, but had somewhere around 18% hit, so had to put it somewhere.

edit 2: I made one set with the 4pT16+Ret Tier pants+Tank legendary cloak. It's less mastery than the other set, but it's around the same haste/hit/exp levels.

[Deleted]

2) Aim to get 15% expertise whenever possible, though not at the cost of haste.

I have no idea what makes you say that, as pretty much all the statistical information in this thread points out that capping expertise at the cost of haste is beneficial.

Either way, the entire discussion was about what to recommend to new tankadins, who are still busy learning the class and gearing up. For them, capping expertise is certainly advise-able, even if it comes at the cost of haste.

(It would also apply to us better geared and more experienced tanks if such an option existed, but that's not what the discussion itself was about.)



My hunter example was merely a fix to Firefly99's hunter example, which basically asked "Do you want to do more DPS, if it's only a tiny DPS increase with no downsides?" which is not representable to the situation at hand. Not Hard Capping Expertise will decrease your survivability, and thus increases the chance of you dropping dead in the middle of a bossfight. How increased this chance is, is of course hard to specify, as it depends on how hard the boss hits, and how good your healers are, not to mention how close you get yourself to executing the perfect rotation.

[Nairobi]

I have no idea what makes you say that, as pretty much all the statistical information in this thread points out that capping expertise at the cost of haste is beneficial.

Taken out of context, I can understand your confusion. My statement implies that you should favor haste while working towards 15% expertise. Let me give you an example:

You get shoulders with 1100 haste and 900 mastery, and a red socket (Str bonus). You're currently ~ 700 rating shy of expertise cap. Would you:
A) Reforge the 1100 haste into 440 expertise and stick a 320 exp gem in the socket
B) Reforge the 900 mastery into 360 expertise and stick a 160/160 exp/haste gem in socket

A will give you 660 haste, 900 mastery and 760 expertise. B will give you 1260 haste, 540 mastery and 520 haste. Basically picking up 600 haste for 240 expertise.

Personally, I'd choose B and then play with other pieces to get the remaining 160 expertise (like swapping in another orange gem). This would put you exactly at the exp cap, and with a leg up on haste over the setup listed in A. Obviously, this ficticious example is contrived for argument's sake only, but since that's what we're doing here, I thought I'd at least clarify my point.

Either way, the entire discussion was about what to recommend to new tankadins, who are still busy learning the class and gearing up. For them, capping expertise is certainly advise-able, even if it comes at the cost of haste.

(It would also apply to us better geared and more experienced tanks if such an option existed, but that's not what the discussion itself was about.)

I don't think anyone is arguing that. As stated by...well, pretty much everyone who's bothered to respond to this exercise in futility, we already are capping expertise. And looking forward to t16, we're going to be swimming in it, making it one of our prime stats to reforge OUT of. The biggest issue with your argument is that there are so few examples of any situation where you actually have to reforge out of haste (rather than mastery, avoidance, or crit) to get to expertise cap. Most of what we have available/desireable is Haste/Mastery (and we value mastery less than haste, we can all agree), Haste/Crit (crit is the obvious reforge choice) or Haste/Accuracy (in which case we move the hit out past cap, or we already have expertise on the item). Unless you're really unlucky with a lot of D/P pieces, it's not likely or possible that you're going to have to reforge Haste to meet exp cap.

I will concede that gemming is the one exception to this rule, in that we can usually pick up a nontrivial amount of expertise from swapping in orange for yellow, but once you pass ~535ish ilvl, that's likely to be overkill.

[Adelphos]

Ok I started this thread to help me with prot stats so I am well informed for raiding (currently 87) and while all of you have helped me alot and sometimes brain freeze on the math. I do have a problem that keeps popping up.

While doing dungeons and Im tanking my gear is 95% haste mastery. I have never had a problem with aggro other than dps zerging right when I pull and I don't feel as though I'm squishy I dont get huge spike damage.

But every 3 dungeon or so I get vote kicked cause I have dps gear on and not tank Im a noob and go roll ret. I talked to some of these people who supposedly play prot pallies saying avoidance stats are the best.

So are they wrong or am I gearing wrong too early?

[rhandric]

Ok I started this thread to help me with prot stats so I am well informed for raiding (currently 87) and while all of you have helped me alot and sometimes brain freeze on the math. I do have a problem that keeps popping up.

While doing dungeons and Im tanking my gear is 95% haste mastery. I have never had a problem with aggro other than dps zerging right when I pull and I don't feel as though I'm squishy I dont get huge spike damage.

But every 3 dungeon or so I get vote kicked cause I have dps gear on and not tank Im a noob and go roll ret. I talked to some of these people who supposedly play prot pallies saying avoidance stats are the best.

So are they wrong or am I gearing wrong too early?

Most of the population (that is, most non-raiders, and even some raiders) haven't adapted to prot paladins using haste gear, and thus erroneously see a tank taking haste gear as being wrong. They're mistaken, but there's not much you can do about it.

[Deleted]

So are they wrong or am I gearing wrong too early?

They are wrong. Terribly wrong.

[Deleted]

Ok I started this thread to help me with prot stats so I am well informed for raiding (currently 87) and while all of you have helped me alot and sometimes brain freeze on the math. I do have a problem that keeps popping up.

While doing dungeons and Im tanking my gear is 95% haste mastery. I have never had a problem with aggro other than dps zerging right when I pull and I don't feel as though I'm squishy I dont get huge spike damage.

But every 3 dungeon or so I get vote kicked cause I have dps gear on and not tank Im a noob and go roll ret. I talked to some of these people who supposedly play prot pallies saying avoidance stats are the best.

So are they wrong or am I gearing wrong too early?

They are completely wrong. DPS gear is pretty much tank gear for paladins now (bar crit rating)

I have never experienced complaints on this matter to be honest. The only thing I gotten frowned upon is the use of intellect shield, but that is usually explained in 10 seconds and after I started tanking nobody ever complains on it. I do not understand why you get votekicked if you do not experience deaths and keep aggro.

[rhandric]

They are completely wrong. DPS gear is pretty much tank gear for paladins now (bar crit rating)

I have never experienced complaints on this matter to be honest. The only thing I gotten frowned upon is the use of intellect shield, but that is usually explained in 10 seconds and after I started tanking nobody ever complains on it. I do not understand why you get votekicked if you do not experience deaths and keep aggro.

My guess? Other str plate wearer in group...

[Adelphos]

They are completely wrong. DPS gear is pretty much tank gear for paladins now (bar crit rating)

I have never experienced complaints on this matter to be honest. The only thing I gotten frowned upon is the use of intellect shield, but that is usually explained in 10 seconds and after I started tanking nobody ever complains on it. I do not understand why you get votekicked if you do not experience deaths and keep aggro.

Thats what I said. Look at the forums they say haste/mastery. Are they all wrong and your right? To which I get no response but a vote kick. Not everytime but more than I feel is right.