Exactly. In many cases, this is another situation in which we have to just chalk it up to personal preference (unless of course you're at 50 hit rating and 5 expertise). If you can manage your rotations well and keep track of misses and dodges, and the occasional miss or dodge don't bug you, then you'll more than likely experience a dps increase by prioritizing str over spell hit and expertise cap.

With the complexity of fights these days, however, keeping track of missed attacks becomes harder and harder, which will almost certainly result in totally fudged rotations and significant dps loss. Fights are sooo much less stressful when you don't have to worry or watch for misses and dodges. But, again, once you approach the caps, it becomes a playstyle issue.

One of the main perks of DW is that, in terms of theoretical DPS, sHit is very close to Haste and Crit and not cripplingly behind ArP.

You cannot trade significant amounts of non-Str stats for more Str through gear selection. After you cap Expertise, trade Crit/Haste first and ArP second for sHit through gear selection, then gem one or two Str/Hit where there are yellow DPS socket bonuses if needed to round out the numbers.

You give up very little theoretical DPS that way and the smooth rotations are almost certainly worth more DPS in practice. On the other hand, if your gear selection is limited and you cannot cap sHit without trading down extensively in ilvl or gemming extensively for Hit, then it becomes a personal question of how good you are dealing with misses.

An interesting topic in my mind is where to aim in gearing for Hit. Since we have talents that lower the soft cap for strike hits but have a much higher roof on white damage then 2 hander counterparts.

There are many stat values in this thread and I have seen them all from page one to this point and I myself (not a math person) am still confused. what is the "ideal" range for hit?

I see many long time DK members have a very large range from 180 hit up to 450 range. I myself currently run 320 hit, I miss on average 12% white strikes with no dodges due to exp cap.

Is there any conclusion as to what is the ideal number range to aim for? If I missed the answer in this thread somewhere, please feel free to point me in the right direction. Thank you.

Aside from getting hit capped for specials and spell hit capped, hit beyond those points will lead to a lower DPS than gearing for other stats (STR, ArP). The amount of DPS gained from extra hit is low, since it will only affect your white damage, while STR would affect all your damage, and ArP will affect a higher percentage of your damage.

As stated previously, gearing for specials is a must. Gearing for spell hit leads to a non-variable rotation (ie. personal preference).

Well, technicly the DPS gain from hit vs. str is qualitatively different, in that hit raises the floor (but not max) dps you can do, while strength raises both floor and max.

If you really wanted to have a very high top end damage, you could simply ignore hit and expertise altogether, as you could always "luck out" on a run and not miss anyways. The cost of course would be the opposite end of things, when your DPS suffers due to an unusually bad run of misses.

I mention this because we tend to treat one 7000 dps spec/rotation as being equivalent to another, when in fact they are only equivalent over very large sample sizes. We dont always treat behavior over small sample sizes with the necessary respect.

Mik scrolling combat text is close to it. But none will work with ToT perfectly. ToT works differently than skills like Mut, stormstrike and whirlwind, in that there is sometimes a huge delay between the two strikes. The others I listed always show both strikes at once.

The OH with ToT is usually delayed which leads to delayed combat text and procs for Rime, RI, FC.

EDIT: Sorry i mistook the quote for it regarding spell hit cap, but the poster was refering to special hit cap and expertise cap.
Either way, the essence of the post still stands. I think that in general a lot of people overestimate the occurance when a couple of percent under the hitcap.

You are overestimating the numbers quite severely here.

We're talking about a 2-3% spell miss chance, not a 20%+ chance to miss your spells.
We're talking about missing 1 icy touch in every 5-6 boss fights, missing 1 rime HB every 2 boss fights, and missing Pestilence every 3 bosses.

This is a general train of thought I see very often, also relating to dual wielding and miss. People are thinking that the effect of not having a cap is huge, yet if you check what the numbers really represent its not so bad after all.

For GoD its already known to be a higher dps loss, but nontheless you can take a shot at a perfect fight without miss.
If you ignore hit (after special cap) and go full for strength you will see a DPS increase in every fight, when you go for hit you won't see a dps loss (this is the difference that Sakuratei talked about on the last page) once every 3 bosses.

And this is where the simulator is right, to put it in crude imaginary numbers (just to make the point). If the strength you gain gives you an extra 100 dps on each boss and the lack of hit drops your dps by 200 every 3th boss. Then ignoring the spell hit cap and aiming for full strength is better.

There are some parts i disagree on though, and that why i personally think getting the spell hit cap is of a higher priority than ignoring it. A bit for the same reason as you mentioned in your second paragraph.
The simulator is the ideal raider. It only needs a milisecond to consider its options after a miss. If i were playing and i would miss my IT i might not notice it before i recast my PS, in which case i have to juggle around a bit with runes, which might lead to a much bigger dps loss than a simulator would have.
Also, I prefer reliability over top dps. I'd rather do 5.5k dps every single fight than 6.0kon 2 fights and 5.0k on the other 2. But again thats a preference call, some might like it the other way around.

However, a maybe even more important question in this regard is: Does it matter much ?

I don't know about your situation but i seem to be well provided in hitgear. A lot of the BiS gear also has hit, and if you need to fill a yellow socket your best option (pre-spell hit cap) is a 10 strength / 10 hit gem.
Even if you tend to ignore prioritizing hit over strength, it still does not take away that pre-hitcap it is the 3th best stat we can get !
I probably couldnt even reach as low as 160 hitrating if i tried.

What?

Prioritizing Str over hit and expertise will result in higher dps in fights when the observed miss rate is sufficiently higher then the average miss rate. Wether this is advisable is another discussion entirely, but if you want the absolute highest dps parse possible for a fight, stack str and prayer will over sufficient time yield the highest dps parse. Your average DPS may go down, and down substantially, but there is that chance you really hit one out of the ballpark.

My point wasn't that this is a smart way to go about your business, but that we dont talk about is that any given boss fight is a small sample, and that certain stats and builds have a greater potential variability, and that variability leads to a different range of possible dps outcomes.

What this could mean in terms of analysis, is instead of simulating 1 500 hour long fight, we could simulate 5000 6 minute fights, and graph the resulting DPS curve in 25 or 50dps increments (ie, counting the number of runs that fall within those small bands of dps). The resulting curves would not only tell us the average DPS per run, but also give a sense of what range of values you might see in a generic patchwerk style boss fight.

Specifically, you would expect large fluctuations of Str and Hit to effect the shape of the curve, not just the average value of the entire run.

In addition to seeing changes in the way str and hit affect the curve, if we ran the simulator for shorter intervals a couple thousand times, specs with cooldowns like blood would probably see significant changes. For example, hysteria's uptime would change drastically because at the end of a 6 minute fight it would JUST be off cooldown but the "fight" would end and you wouldn't be able to use it a 3rd time, only twice in 6 minute fight and would have 16.7% uptime. If the fight was extended by 30 seconds then it would go up to 23.1% uptime.

This does happen in a real raid often, but it happens both ways, sometimes you just miss it, sometimes you get your last one in exactly 30 seconds before the boss dies, and sometimes you even hold it back to stack it with heroism or something else at the end. Maybe have the sim run a 4 min, 5 min, 6 min, 7 min, and 8 min fights a couple thousand times.

I have a question.

I have been wielding the [Aledar's Battlestar] togheter with the [Peacekeeper Blade] (As I havent been able to get another decent offhand.) Last night I got my hands on the [Lionhead Slasher]. As I belive its an upgrade over my current weapons, but how long will it hold. As its way to fast to be ideal for an Dual Frost DK.

Would slower weapons of Ulduar 25 man standard, be an upgrade or is the [Lionhead Slasher] one to keep untill ToTC or ICC drops a slower weapon for me ?

You could probably switch to an unholy DW build for the time being since you have a nice offhand like that or even one of the other unholy DW builds that doesn't require a slow MH. It's a great weapon to have but you just have to decide on whether you'd like to switch up your spec to really use the weapon effectively.

Tonight our guild did XT hard for the first time along with hodir hard among other bosses. I used this build which gave me great results. Our other frost DW dk was kinetica in the raid. I'd definitely say that using the GoP build/rotation yielded a higher dps for myself compared to when I was supplying IIT. If you get an enhance shammy supplying 20% haste or another DW DK go ahead and try out this build. Here's the WoL report from tonight along with the night before with our ignis speed kill.

XT hard
Ignis Speed Kill
BTW I realize that Ignis is a little skewed due to KM Rime procs also hitting the adds around our tank. We used 1 tank and he tanked the boss along with the adds. We all stood in the water while the boss was facing away from us while also standing in the water. Hopefully these logs can give you all an idea of how GoD build is performing vs IIT builds.

Here's my armory and yes I'm aware I have too much hit but I'm waiting for some new boots to drop before I decide to reenchant my boots.

Also something to note was I typically would start the fight by getting greatness to proc + FC and then I'd pop UA and disease the boss so that I'd be at my max for dot damage during the fight.

I have been playing as Blood Spec on my DK since WoTLK Launch, but recently I felt like trying out Dual Wield Frost with the addition of ToT in 3.2. I wanted to find out a few things, but to be perfectly honest, I don't have the willpower to trudge through a 61 page thread looking for information, so if someone could answer a couple of questions for me, I'd greatly appreciate it.

1. I assumed that since I'd be using 2x2.6 speed weapons and all my strikes hit with both weapons, that RoTFC would have the same uptime if it was on my offhand weapon, so I should us RI on my main hand for higher damage, is this correct?

2. I really don't like the idea of playing without Subversion, I've been playing with it all this time I'm not sure I could manage my threat without it, are there any viable Frost DW specs that include this talent? I came up with a spec on my own, but I'm not sure how it would perform.

3/52/16

3. Quick question I forgot to add, sorry if it is really obvious, but BP or UP?

Thanks in advance.

[Grimhorn Crusher] is easy access too since it only costs 25 [Champion's Seal].

1. Yes to the weapon speeds.
What enchant you put on what weapon doesn't really matter. THe damage from RI is seriously neglectable.
2. Looks good if you don't have to provide the haste buff
3. BP

EDIT:
The BiS thread still presents agility as best cloak enchant, but I guess we can assume that haste is best now?

Some more data points:

Algalon-25 - 6816 DPS. (Actual DPS to algalon 6447)
Spec: Sub-UH w/ IIT
Glyph: GoD
Rotation: IT PS OB BS BS - dump - OB OB BS PEST - dump (the 2nd rotation in the used another PS/IT with greatness/racial/SoA/UA/wrathstone procced to max rolling disease damage)

Vexaz-25 HM - 6008 DPS.
Spec: Sub-UH w/ IIT
Glyph: GoD
Rotation: IT PS OB BS BS - dump - OB OB BS PEST - dump (the 2nd rotation in the used another PS/IT with greatness/racial/SoA/UA/wrathstone procced to max rolling disease damage)

Algalon-25 - 6551 DPS. (Actual DPS to algalon 6072)
Spec: Sub-Blood
Glyph: HB
Rotation: HB OB BS BS - dump - OB OB OB - dump (HB to keep disease up, if rime procced replace with OB, single disease rotation expect for the opening it/ps)

Buffs:
Standard raid buffs
Flask
Food

Sigil:
SoV

Enchants:
FC (main)
RI (off)

Weapons:
239-pvp axes

The GoD spec could have popped a point from BCB into 5/5 KM. I'm running basically the same gear in both parses. I rolled diseases in the GoD spec, averaging about 1050 / tick. The HB spec was running 1 disease most of the time, HB to keep it up.

Edit: Added actual damage to Algalon to give more realistic single target DPS.

Sorry, but why do you pick up both GoD and IIT ?

I dont see the use of investing 6 talent points in 5% haste. And GoD doesn't refresh Icy Talons so, you can't use GoD if you need to supply the haste.

GoD spec doesn't refresh IIT like foxx said so it would be 6 points just for 5% haste.

I find RI better on the main hand since it stacks up faster for me due to the delay in ToT strikes.

@Vaedron: We all know that those 6 talent points provide 25% total personal haste. That is not at all what Foxx2405 was talking about. His post is referring to Saabik using Glyph of Disease with IIT. Basically, the issue is that a well executed GoD spec -shouldn't- be using Icy Touch apart from initial application. IIT haste buff is achieved through the "application" of diseases, and not the "refreshing" of diseases the GoD provides.

So what we have is:

IT cast, IIT buff is given to the raid

Raid has: 20% haste
Player has: 20% haste + 5%haste

Pest is used to refresh diseases and IIT buff expires

Raid has: 0 haste buff
Player has: 0 + 5% haste

In order to use a GoD spec, IIT just simply isn't viable. If you don't have a shammy to provide the haste buff, then the IIT spec w/GoD is virtually useless for your raid if its only going to be up for 20 secs.

Some fights I get an enhance shaman (or WF period), some I don't. This is a fight where I was able to ignore IIT.

For more clarification:

Talent Calculator - World of Warcraft
(no IIT)
+1 subversion (3% crit OB/BS)
+1 hungering cold (neat utility)
+1 KM (more DPS, though I often do not find the GCD to use my rimes)
+2 Imp Frost pres (neat at best)
+1 DC (very minor DPS)

vs
Talent Calculator - World of Warcraft
(w/ IIT)
+5 IT (utility when no sham available, at a loss of personal DPS)
+1 IIT (5% haste)

Certainly the former offers a slight increase in DPS, and would be ideal to max DPS. However, those 6 points aren't very efficiently spent for DPS increases as there are a lack of places to put them. The later allows me to have more rDPS/gDPS in 10m and garbage heroic daily groups. You could pickup some nice things such as lichborne, more range, free interrupts, etc. which are neat, too.

I will attempt to venture into a higher DPS next week.

Yesterday night, we went to Algalon the Observer 25man and had him down to 3% when he began to cast a fourth Big Bang... and right after he finished his cast, he despawned for no reason

Algalon 25man - 7000 DPS

I was running a 3/52/16 Specc with GoD. At that point I only had die T9 Spaulders, no Gloves, was wearing T8 Chest and Gloves, Thorim HM Helm and Leviathan HM Leggins. I also didn´t have the OH weapon currently shown in the armory but the mace from ToC Heroic with 163 DPS.
These infos are just for a better analysis of the DPS.

Actually my DPS will go up by like 200 or 300 with the new items for the next week. Just wanted to share this with you regarding the fact that there are not that much high DPS parses from DW Death Knights at the moment.

While this is a good parse, be weary of how much damage you actually did to Algalon. I checked the 10 attempts you had and your DPS is significantly lower when you didn't get to use HB on multiple adds (especially phase 2). If you take the damage by actor value, you had 88.9% of your damage on this attempt to Algalon, making the actual DPS to Algalon 6271. I didn't realize this made such an impact until I saw this so I went and looked at my numbers as well, they were inflated a bit so I edited my post.

In most cases I wouldn't discredit AE damage, but in fights where it isn't a goal to kill the adds it probably shouldn't be counted (my parse made the same mistake)

I couldn't find your deduction, in fact I couldn't find any proper calculation of the benefit of weapon speed, just simulation.

A simulation is sort of a black box, nobody can understand exactly how the weight came to be, we just trust it because we hope we fed it the right data. A proper calculation, aka theorycrafting, however is transparent and clean, so here's an attempt, feel free to nitpick.

I'll assume raid buffed, thus about
35% crit,
30% haste (7% personal haste is not unreasonable) and
15% miss chance for white attacks for simplicity. If you want to assume a different hit, just change the miss % in the auto attack calculation, assuming special hit capped should be a given.

I will only compare average weapon damage to average dps.

I will consider neither Boss level (glancing blows etc) nor Armor Penetration, Armor Penetration would only have an effect on Frost Strike (I'm not too familiar with boss armor after sunder and a reasonable amount of armor pen). Glancing blows could have a noteworthy effect, but I'm not sure exactly how the glancing blow mechanics work, so I'll assume that the damage difference can be ignored without influencing the results too much, especially since I always assume Full Subversion and no BCB in my examples. Boss level related reductions to crit would have the opposite effect (since specials scale better with crit) and I know even less about that, overall it should alleviate the absence of glancing blows.

Attack power does not change this calculation at all since Scaling is normalised. Main Hand or Off Hand is also irrelevant since the 50% damage penalty is consistent in all relevant aspects.

I will calculate for both Subversion and BCB, and I assume an Unholy subspec at least including necrosis.
I will first calculate the value of an average point of damage for every ability that uses Weapon Damage, then determine to how much dps that translates and compare to the benefit of dps for auto attacks. Numbers are rounded to 2 decimal points.

Obliterate:

100% weapon damage, 25% diseases, 15% frost fever, 15% (24)% extra crit, 45% crit modifier

(1 * 1.25 * 1.15) * (1 + ((0.5 * 1.45) OR (0.59 * 1.45))) ~= 2.48 OR 2.66 (depending on Subversion)

Blood Strike:

44% weapon damage, 25% diseases, 15% frost fever, 0% (9%) extra crit, 45% crit modifier

(0.44 * 1.25 * 1.15) * (1 + ((0.35 * 1.45) OR (0.44 * 1.45))) ~= 0.95 OR 1.04

Plague Strike:

50% weapon damage, 15% frost fever, 6% extra crit, 30% crit modifier

(0.5 * 1.15) * (1 + (0.41 * 1.3)) ~= 0.88

Frost Strike:

60% weapon damage, 15% frost fever, 13% extra magical damage, 10% razorice, 10% Black Ice, let's say 55% crit with KM (just a qualified guess), 45% crit modifier

(0.6 * 1.15 * 1.1 * 1.1 * 1.13) * (1 + (0.55 * 1.45)) ~= 1.70

In 20 secs, the normal Frost DW rotation (without GoD) uses 1 Plague Strike, 2 Blood Strikes, 4 Obliterates and has space for ~5 extra globals, of which we assume 4 go to Frost Strikes

So every point of average damage yields:
(2.48 * 4 + 0.95 * 2 + 0.88 + 1.70 * 4) / 20 ~= 0.98 dps without subversion and
(2.66 * 4 + 1.04 * 2 + 0.88 + 1.70 * 4) / 20 ~= 1.02 dps with subversion


Auto Attacks:

20% (35%) extra Necrosis/BCB, ~30% haste, ~15% miss chance

((((1 * 1.2) * 1.3) * ((1 / 1.15) + 0.35)) ~= 1.90 OR 2.07 dps (BCB calculation is simply adding 15% of auto attack dmg with haste and miss, but without necrosis or crit, since it doesn't scale with either.)


This means that a weapon needs to have from 1.86 to 2.11 times more average damage than it loses dps compared to another weapon to be better, disregarding stats on the weapon (which are usually in favor of the one with the higher dps).

Applying this to the example you listed [Lionhead Slasher] vs [Reaper of Dark Souls], we get (206 + 384) / 2 = 295 average damage and 196.7 dps vs (281 + 422) / 2 = 351.5 average damage and 130.2 dps.

Translated that means 55.8 average damage vs 66.5 dps, or 56.92 dps increase vs 126.35 dps loss, assuming no BCB and full subversion. Apparently [Lionhead Slasher] is clearly superior no matter the circumstances, the gap only widens if you consider other talent specs and the higher stats on the weapon.

Comparing [Lionhead Slasher] to [Grimhorn Crusher] we get 78 avg dmg vs 53.4 dps, 79.56 dps increase vs 101.46 dps loss, again with subversion and without BCB. [Lionhead Slasher] is still better.

Comparing [Lionhead Slasher] to [Aledar's Battlestar], again full subversion, no BCB: 129.5 dmg vs 33.4 dps. This time it's 132.09 dps increase vs 63.46 dps loss. So the slower weapons quickly catch up, but I'm pretty sure if we'd use the heroic version of [Lionhead Slasher], it would be better again.

Summarizing we can say that if you choose between similar quality items, the slower one always wins, but if you have vast differences in quality (more than 2 tiers), the faster one is probably better and the better weapon is always better equipped in the main hand, regardless of speed.