I sort of feel people are dismissing a little gem:

[Aesir's Edge]

Gemmed by a jewelcrafter it sports:

162 Str
129 Stam
70 crit rating
60 ArP
937 top damage, 3.5 speed

Compared to [Worldcarver]:

137 Str
139 Stam
70 haste
87 expertise
949 top damage, 3.4 speed

and [Earthshaper]:

129 Str
129 Stam
66 hit rating
81 crit
1005 top damage, 3.6 speed


it seems a viable alternative. It's slightly lower in dps, but still has a comparable swing damage when you factor in the 60-70 extra ap. ArP is an excellent stat for Blood and decent for Frost, so in case you're bloated on hit or you don't really value haste and expertise that much, it's not to be dismissed so easily.

Stat weights can't in my opinion be trusted in a vacuum. Expecially the dps value, it's clocked at 9.89, and I seriously don't think that 10 dps difference can ever become a 90 dps difference.

Comparing:

[Aesir's Edge]

[Rune Edge] (not the greenie sigh)

[Lotrafen, Spear of the Damned]

my conclusions would be:

Aesir vs Rune Edge:

Rune Edge has 10 ArP, 10 crit rating + 120 agility, since at lvl 80 a dk needs 62.5 agility for 1% crit and 49.52 crit rating for 1% crit, we can napkin that into 2% crit.

Aesir's Edge has, for a a Blood DK, 176.5 Str, which means 353 ap. Rune Edge has 160. That's a 193 ap difference.

So essentially you're looking at 10 dps, 2% crit and 10 ArP vs 193 ap.


However, considering a raid environment, 120 agility become 132 (that's 0.2% more crit chance). 176.5 Str become 194,15 with Kings, and assuming 50% Fallen Crusader uptime, they end up being 220 Str.

So as soon as you factor raid buffs you have:

10 dps, 2.2% crit and 10 ArP vs 280 Ap.

Using that stat conversions you linked we have:

(10*9.89) + (2.2*45.91=101.002*1.56) + (10*2.75) = 98.9+157.56+27,5 = 286.66 compared to Aesir's 280.

Considering that in my opinion the dps value is highly overextimated (Aesir's melee strikes will hit harder, all strikes will hit harder, all ap powered spells will hit harder, what Rune Edge gains is 1 strike every 35 essentially), I'm still persuaded Aesir's Edge is an incredibly good weapon considering it drops from a 10 men version and stacks up to some of the best 25 men loot.


Edit: I'm not advocationg A'sE is a BiS weapon. But if we are limiting ourselves to BiS items, then anything but Vold'rethar should be discarded. Just pointing at a very viable option. For example it's the weapon I'm hunting for as Frost before Vold'rethar. Crit rating is highly devalued by KM for me, while AP is amazing.

I agree. I have a hard time believing that a lower DPS weapon can truly outperform a higher one in the first place. After those modifications, you basically end up with what is on the first post, only swapping the neckpiece to properly achieve the physical hit cap. Also consider that the first post currently uses an "old" version of the Warhelm, it lost a little hit/chunk of stamina to gain some strength (wowhead is now updated properly).

drop locations and difficulty modes would be fantastic for this thread.

I've been closely following itemization and stat weight discussion in this thread and in spec threads. There's really one thing which I cannot understand. Almost everyone is saying that you should only gem for str but yet same time people are willing to take inferior items to certain slots just to cap for example expertise.

If expertise is that valuable then why it's so bad sin to gem for it? Of course this is entirely depandant of available gear and possible gear choices. Simplified you find BiS item for slot, then find out optimal gear to cap hit and then fix rest with sockets. Rather than trying to find gear set which caps both hit and exp close enough not wasting too much stats and then gem for str alone. Just doesn't make sense... This might not affect at all to BiS list in the end but still..

I'm not really sure we can assume similar Stat weights for Frost and Unholy. Unholy for example would ignore ArP for both the RP dump and the rune abilities most of the time, while Frost still uses it for a large part of its dps. In the same light, probably Frost would value expertise differently because around 50% of its damage totally ignores dodges and parries. And KM does devalue crit in a way. At the same time, basically every Frost ability crits for 245% damage, unlike Unholy, so once again Unholy may be more biased toward AP.

I'm not really being useful here, but personally I wouldn't gear exactly in the same way as Frost and Unholy. Actually gearing for Frost is a big mistery because all while UH and Blood are extremely settled in the magical or physical field, Frost is in hybrid territory.

Well as someone pointed out earlier, stat weights are useful as an indicator, but you can't necessarily use them as if you were in a vacuum. And while Lootrank is ok for finding pure 'Best in Slot' items, it doesn't appear to be much use in putting together an optimal set, which is what this thread is more concerned with.

I've now included an 'hit/expertise capped' set in the OP separately from the main 'pure best' set. Statistically, on average the capped set will do lower overall damage given a large enough sample size and vacuum-like test conditions, and so is not truly the optimal gearset. However some people may just prefer the more stable, less streaky damage you can get when fully hit & expertise capped. Not to mention the implications a missed strike can have on your rotation (which is admittedly a much bigger concern for Blood & Frost since Unholy has plenty of free GCDs to rectify errors).

I am of course more than happy to construct a separate table for Frost, but without stat weights it's going to be difficult. Getting those needs someone like Methods to do the maths because it's not really my forté, especially when my personal experience of the spec is quite limited.

This was posted by Methods in the Frost Thread yesterday. Hopefully it may help to determine a seperate BiS List for Frost.

Ah yeah, just copy-pasted it and forgot to change it. Fixed now, thanks.

Ok cool, I'll see what changes that makes to Tier 8 and get cracking on a separate Frost set.

As I mentioned a few pages ago, I've been working on an Excel spreadsheet to make calculating the best sets for all of these variables (Race / Auras, Specs, Levels of Gear) quite a bit easier. I think I'm finally at a point where I'm comfortable releasing something with the caveat that it is still very much in development. With that said, I've developed this spreadsheet from scratch primarily to support this thread:


Description:

• The Death Knight Gear Optimizer is a comprehensive gear selection spreadsheet developed in Microsoft Excel 2003/2007 designed to aid every Death Knight in making informed gear decisions. The program uses a variable weighting system to assign values to various sets of gear and exhaustively searches the selected items to provide a guaranteed optimal gearing solution. The optimizer can be used to find an overall Best-in-Slot wish list, or it can be used simply to figure out how best to utilize that newest upgrade.
  
• The gear optimizer provides intuitive ways to select items for calculation and takes into account almost every factor imaginable. An option for Smart Gemming is also available, which guarantees that the Death Knight Gear Optimizer not only provides you the best gear, but also the most appropriate way to gem it!
  
• Please note that this program is still under development. Please send any questions or bug reports to Zerack at AftermathLB via a Private Forum Message.

Download Link:

• Download Here - DK_Optimize_v1.4.2.zip (~310KB)

Features:

• Exhaustive gear search provides an optimal result everytime.
• Native implementation of racial bonuses from Axe Specialization to Dranei Hit Auras.
• Support for Hit and Expertise derived from talents.
• Calculation options to require specific Tier set bonuses in order to maximize gearing potential.
• Intuitive results screen shows both optimal gear and total statistics.
• Provides results for Two-Handed and Dual Wield gear sets.
• Calculation estimates let you know how long you have to wait.
• Weighted procs for trinkets allow accurate calculation of every slot.
• Gemming heuristics calculate the optimal gemming for every set to guarantee the best from every socket.

Examples:

• The home page, where the most important selections are made. Choose your race, stat weights, tier options, talent options, and more.
  
  

• An example of an item screen, where you can enter and update statistics for every item to analyze. Check boxes allow you to quickly choose which items to process without deleting valuable item information.
  
  

• The easy to understand calculation screen lets you know how long you need to wait for the calculations to finish. You can choose to cancel at any time to tweak your selections from the home screen and start over.
  
  

• One possible set of gear on the results screen. The gear is displayed in a familiar arrangement. Gems are colored according to the color of the gem, while the borders reflect the colors of the sockets that they are in. This example shows a result using naive gemming.
  
  

Version Notes:

• June 16, 2009 - Version 1.4.2
  • Added Trinket: Comet's Trail.
• June 10, 2009 - Version 1.4.1
  • Added all known Algalon 10 and 25 loot.
  • Added community calculated set weights for Blood and Frost specs.
• May 26, 2009 - Version 1.4
  • Added the ability to weight set bonuses and compare then natively.
  • Added several leather items that are BiS competitors.
  • Added two known Algalon 10-man drops.
  • Updated several items to reflect the 10-man hard mode iLevel increase.
• May 13, 2009 - Version 1.3.1
  • Fixed a bug that would occasionally cause expertise to be slightly overvalued.
  • Fixed a bug that caused some drop down boxes to function incorrectly.
  • Updated several items to reflect re-itemization.
  • Updated trinket values to approximate the average effect for each trinket.
  • Added correct socket bonuses for most hard mode drops.
  • Added calculation and display of post-enchant statistics for all sets.
• May 06, 2009 - Version 1.3
  • Fixed a bug that would cause certain gems to not be used in optimizations.
  • Added option to force hit and expertise caps in optimal sets.
  • Added logic to intelligently consider Crusher vs. Precision for glove enchants.
  • Added additional items to most slots by default.
  • Added information about drop location for all items.
  • Added the ability to load and store predefined weight sets.
• May 04, 2009 - Version 1.2
  • Implemented Smart Gemming
  • Fixed a bug that would cause incorrect values to be displayed on the results screen.
  • Fixed a bug that incorrectly double-counted certain gem values.
• April 30, 2009 - Version 1.1
  • Fixed a bug that would cause racial expertise bonuses to incorrectly influence the score of a set.
• April 29, 2009 - Version 1.0
  • Initial Release

Additional Comments:

To take a step back from the official-ese spouted above, this is a pretty cool tool. I'm always bothered when we post BiS sets that someone finds simply because there's very little we can do to 100% PROVE that this set it optimal for given conditions. This spreadsheet does away with that by iterating all the sets from gear you select and choosing the best one. As a note - the statistics on the final page are only stats from gear and enchants. They will not reflect the actual stats you see on your character screen in game. The program does intelligently choose the cloak, glove, and boot enchant, but that is all. The rest of the enchants are there for show / completeness as they are more or less a static gear decision.

The only other thing that I can think of that's a little unsatisfactory right now is the way that Dual Wielding is handled. I haven't implemented separate weights for OH DPS values, nor do I calculate changes in value for offhand racial expertise. All things that can be implemented later, most of which were excluded due to my own lack of knowledge. Finally - the spreadsheet is most definitely not lying when it says that calculating smart gemming will greatly increase computation time. I've managed to trim down the number of sets that it tries to gem very significantly, but doing gem logic even with smart socket selection for a gamut of more than 2-3 gems is very time intensive. It works perfectly, but be ready to wait some time for the results.

Please enjoy, and let me know if you have any problems. I'm more than glad to work to improve this and make it a great resource for the DK community here at EJ.

@Katsapliock - Thanks for the feedback. To hit your points in list form:

1. In regards to time: You can do quite a number of items if you turn smart gemming off. Even though the gemming is done after the item calculations, keeping track of 'candidate' sets to gem is memory intensive, and thus slows the process down. Smart Gemming really only takes huge amounts of time when you start doing things like turning on JC Gems AND permitting 4-5 additional gems. Unfortunately there's not much else I can do to optimize this process unless someone comes up with a really slick gemming heuristic.
2. I can certainly mark them in future releases as 10HARD / 25HARD for clarity. It's a simple text change that you noted you can make yourself if really needed. Having it default would be no problem.
3. I most likely skipped it over since I believe I mostly pulled in only 226 items and up for my testing purposes. I'll probably be going back and doing a more complete default-loot fill in later.
4. I'm not entirely sure that I understand your question on this one. Unless you have some odd glitch that I'm unaware of, there is a box displaying total contribution from gear statistics on the right hand side of the results page. Perhaps if your resolution is smaller you might have to scroll to the right slightly? I'm fairly certain that this is what you are asking for.