Get the Desktop App for Battle.net Now
- All your games in 1 place
- Log in once
- Automatic game updates
Guide for Affix & Slot Efficiency in RoS [last updated in 24/05/14]
This guide's main goal is providing information for players that want to optimize their gear through the best possible affixes for each slot.
Please keep in mind that different builds have different priorities and they'll obviously reflect on your itemization. For example, a build that revolves around weapon procs (or procs in general) will want certain amount of attack speed, even though the affix on its own is pretty inefficient when it comes to increasing the player's damage. On the other hand, builds that rely on abilities limited by cooldowns might find the affix really attractive, even though it stacks multiplicatively.
Before anything, I suppose I should briefly speak about the Paragon Tree:
- 12% Movement Speed = 120 Attribute = 24 Max Resource
- 5% Critical Chance = 50% Critical Damage = 10% CDR = 10% attack speed*
- 250 AR = 25% Life = 25% Armor
- 10% RCR = 50% Area Damage
We can draw some quick conclusions from this comparisson:
1. 24% movement speed is worth only 240 STR (or VIT) and in most cases will be the affix to be rerolled on whatever slot it shows up.
2. Asides from weapons and diamond helm, 10% CDR in Paragon beats every armor slot, this is relevant due to the fact that CDR scales multiplicatively. Also, 10% attack speed can't be achieved anywhere else but in Paragon.
3. AR-Life%-Armor% mitigation efficiency will depend on the base values of the barbarian. Usually, AR is better because it grants a base increase to the affix, while all others are multiplicative.
4. The Paragon Tree is the most efficient place for getting Area Damage: you can only achieve 20% AD (24% on weapons) per affix slot and you'll always be sacrificing something for it. Best case scenario, you're trading 8% RCR for 20% AD - half of what you get from Paragon.
@Nubtro has a fine compillation with them here and that's what I'll use as base:
In order of properly weighting one or more affixes, it is required to know their base values.
For example: you're in doubt what would increase your DPS the most: 50% CDam or 700 STR.
More often than not, people would answer CDam and that's not always correct. Yes, barbarians have some serious scaling with Critical Hits but you need to look at your base values: for a character with 8k STR, 40% CC, 400% CDam, 700 STR increases the average damage by 8.64%, while 50% CDam by 7.69%.
With that in mind, and so that I could make some generalizations, I created a character with easy enough base values, which are as follows:
Crit Chance: 40%
Crit Damage: 400%
Attack Speed: 15%
Skill Bonus on Gear: 15%
Elemental Skill Bonus: 20%
Bonus vs. Elite: 15%
Total Health: 400.395
Total Armor: 15.000
Average All Resists: 600
When creating tables, I'll always put a baseline for this Dummy in bold.
We have three options here, the first two limitless on Paragon:
1. Vitality [mitigation]: usually seen as a crappy option because the mitigation provided by increasing your health pool works against the player, for it will require an increase in healing in the same amount otherwise it'll start feeling weaker and weaker. The exception to this are the barbarian skills & passives that scale with larger health pools: Revenge, Rend and Inspiring Presence. Finally, Tough as Nails also provides an additional mitigation scaling with Vitality through increasing Armor by half the value.
2. Strength [offense + small mitigation]: asides from the obvious benefit of increasing damage, Strength also increases Armor by the same value, which can be further enhanced by the Armor% skills and passives. It's better than nothing.
3. All Resistance [mitigation]: that's your choice, unless already running with some really insane mitigation. The logic is simple: on sockets you'll be trading 2.8 points of STR (or VIT) per point of AR, while on any other slot, the trade falls between 4.1 and 7.5.
Even though STR doesn't have diminishing returns, it requires the player to progressively stack greater amounts in order of gaining the same damage increase. With that in mind, it is worth considering that the player might reach a point where the increase in damage from 500-750 STR is so low that another affix becomes more attractive.
This table shows base values of STR and the relative increase in DPS from +500 STR:
9.900 STR = 05.0% increase in DPS
8.000 STR = 06.2% increase in DPS
4.900 STR = 10.0% increase in DPS
2.400 STR = 20.0% increase in DPS
And this, using an increase of +750 in STR:
9.900 STR = 07.5% increase in DPS
8.000 STR = 09.3% increase in DPS
7.400 STR = 10.0% increase in DPS
3.650 STR = 20.0% increase in DPS
2.400 STR = 30.0% increase in DPS
As said before, the higher the base value, the smaller the increase in damage from sucessive increments of the same value. However, even at 9.9k STR, 500 and 750 STR still grant a decent increase in DPS...we shall see how that increase stands against the other affixes later.
More attack speed, more ticks on WW/RLTW, more procs - everyone knows the importance of attack speed for barbarians, issue is how much you would end up sacrificing for hitting something like 2 attacks per second (1.4 weapon + (4 * 7% armor) + DW).
It's worth mentioning that attack speed increases DPS, not brute damage output, thus being a bad trade-off for skills with a set amount of ticks per second (Rend/EQ/Avalanche...).
This table shows base values of %APS and the relative increase in DPS from +7% APS:
57.0% APS = 4.5% increase in DPS [DW+Paragon+Item+WotB]
32.0% APS = 5.3% increase in DPS [DW+Paragon+Item]
25.0% APS = 5.6% increase in DPS [DW+Paragon]
15.0% APS = 6.1% increase in DPS [DW] <- starts losing to +500 STR at 8k STR
07.0% APS = 6.5% increase in DPS [1 item with APS]
00.0% APS = 7.0% increase in DPS
@zylog had already pointed out how ineffective attack speed is and it can be seen that for our Dummy Character, at 8k STR and 15% APS, increasing 500 STR or 7% APS grants nearly the same increase in DPS. However, if Dummy had anything higher than 15% APS or used WotB, things would be ugly.
Can't think of any barbarian that wouldn't go %CC on Gloves & Amulet. As for Bracers, the only damage-related contender is +Elemental Damage and it shouldn't be hard getting both. Helm's choice should be even easier - unless the player would want a %APS legendary with Socket.
Since 10% CC is no-brainer on its gear slots, this table shows certain base values of %CC and the relative increase in DPS from +6% CC with 300% CDam:
75.0% CC = 05.5% increase in DPS [tons of buffs]
60.0% CC = 06.4% incresse in DPS <- almost loses to +500 STR at 8k STR
55.0% CC = 06.8% increase in DPS
50.0% CC = 07.2% increase in DPS
45.0% CC = 07.7% increase in DPS
40.0% CC = 08.2% increase in DPS
35.0% CC = 08.8% increase in DPS
Now with 400% CDam:
75.0% CC = 06.0% increase in DPS [tons of buffs]
60.0% CC = 07.1% incresse in DPS
55.0% CC = 07.5% increase in DPS
50.0% CC = 08.0% increase in DPS
45.0% CC = 08.6% increase in DPS
40.0% CC = 09.2% increase in DPS
35.0% CC = 10.0% increase in DPS
Finally, at 500% CDam:
75.0% CC = 06.3% increase in DPS [tons of buffs]
60.0% CC = 07.5% incresse in DPS
55.0% CC = 08.0% increase in DPS
50.0% CC = 08.6% increase in DPS
45.0% CC = 09.2% increase in DPS
40.0% CC = 10.0% increase in DPS
35.0% CC = 10.9% increase in DPS
Conclusion: the higher the base %CC, the less valuable +6% CC becomes; at the same time, the higher the base %CDam, the more valuable +6% CC becomes.
How does +6% CC fall against +500 STR? With the exception of cases where the Crit Damage was too low (~300%), 6% CC should always beat 500 STR. At absurd Crit Chance levels (~75% CC and higher), it should start losing to 500 STR, even at 500% CDam.
Another Amulet no-brainer. Since weapons only roll up to 35% CDam, Gloves also seem to be a no-brainer: leaving their final slot for either %APS, %RCR, %CDR. Rings are a slightly harder choice because of the Avg Damage affix and %Life (if someone were to stack health).
This table shows values of %CDam and the increase in DPS from +50% CDam with 40% CC:
600.0% CDam = 05.9% increase in DPS
550.0% CDam = 06.2% increase in DPS <- as much as +500 STR at 8k STR
500.0% CDam = 06.7% increase in DPS
450.0% CDam = 07.1% increase in DPS
400.0% CDam = 07.7% increase in DPS
350.0% CDam = 08.3% increase in DPS
300.0% CDam = 09.1% increase in DPS <- loses to +750 STR at 8k STR
With 50% CC:
600.0% CDam = 06.2% increase in DPS <- as much as +500 STR at 8k STR
550.0% CDam = 06.7% increase in DPS
500.0% CDam = 07.1% increase in DPS
450.0% CDam = 07.7% increase in DPS
400.0% CDam = 08.3% increase in DPS
350.0% CDam = 09.1% increase in DPS <- loses to +750 STR at 8k STR
300.0% CDam = 10.0% increase in DPS
Critical Hit Damage is a must-have on all slots and that's about it. Strength stacking might be somewhat competitive to it but neither Gloves nor Rings have any other strong contenders to these two affixes. Considering how Crit Chance and Crit Damage scale with each other, the first affix also continues being a top priority.
Amulets already have their priority of offensive affixes but Rings still lack one (the first three being STR, CDam, CC). However, to properly compare Average Damage, it's also required to know the average base & black damage values of certain weapons [legendaries, of course]:
- Skullsplitter: 316 - 585 (avg 450.5)
- Disemboweler: 353 - 526 (avg 439.5)
- Dire Axe: 249 - 461 (avg 355)
- Limb Cleaver: 249 - 461 (avg 355)
- Tsunami Blade: 168 - 392 (avg 280)
- Punyal: 107 - 321 (avg 214)
* Black Damage: 1199 - 1490 (avg 1344.5)
Even though a max-rolled damage is rare, we'll work with it for the best case scenario:
Tsunami Blade: 7.57%
Limb Cleaver: 7.23%
Dire Axe: 7.23%
While at it, I'll also throw the relative gains from socketing a +280 Damage Ruby:
Tsunami Blade: 17.67%
Limb Cleaver: 16.87%
Dire Axe: 16.87%
Basically...what can be concluded is that the smaller (and faster) weapons benefit the most from the Average Damage affix. Also, remember that this is the best-case-scenario and that the further away from best-rolled damage, the more attractive it becomes (a Tsunami Blade legendary with 1050 - 1350 damage receives an increase of 15.17%).
As for 2Hers...I won't even bother comparing all of them because it became rather obvious that +Damage is only for fast weapons with smaller damage numbers.
Finally, Ruby isn't competitive with Emerald on socket (unless for builds with low CC). On the other hand, +Damage on Rings is great, being competitive with the top affixes of the slot: CDam, CC and STR.
A no-brainer on Shoulders & Diamond in Helm for %CDR builds, mostly because the other item slots have much more attractive affixes. Also, since %CDR has diminishing returns, it might be a good idea trading a Weapon affix (% Damage, % APS, Vitality...) for a higher %CDR value.
In terms of Paragon points, this affix has the opposite treatment of %CDR: its Paragon counterpart (Area Damage) is actually 5 times more efficient, making it way less attractive on its respective slots (On Paragon: 1% RCR = 5% AD; On Items: 1% RCR = 2.5% AD). Personally, I don't see much utility for the affix on barbarians: it messes with the fury dumping and with the healing from Life per Fury Spent.
Read %RCR section: Paragon points are the best source of this affix.
Life % rolls on Helm, Chest, Pants, Belt, Shoulders and Rings. Despite the affix increasing the health pool by the same percentage, characters with higher Vitality will obviously receive higher increments of health from it.
Since when you're considering Life %, you're actually aiming for increasing your health pool, the most sensate thing to do would be comparing it against Vitality itself.
Table with base values of VIT and the relative increase in mitigation from +500 VIT:
5.000 VIT = 10.0% increase in mitigation
4.500 VIT = 11.1% increase in mitigation
4.000 VIT = 12.5% increase in mitigation
3.500 VIT = 14.3% increase in mitigation
3.000 VIT = 16.6% increase in mitigation
Table with base values of Life % and the relative increase in mitigation from +15% Life:
40% Life = 10.7% increase in mitigation
25% Life = 12.0% increase in mitigation
20% Life = 12.5% increase in mitigation <- as much as +500 at 4k VIT
15% Life = 13.0% increase in mitigation
10% Life = 13.6% increase in mitigation <- as much as +750 at 5.5k VIT
Helms can roll up to 750 VIT so they'll hardly ever lose to Life %. Rings are much better sticking to offensive affixes, while Shoulders and Chests can boost the damage from cooldown-gated skills. Belts have nothing to compete against Life% (unless you're considering LpFS) but Pants got the sockets slot to fill, which will require mitigation comparissons against AR and Vitality.
Plus, while at it, don't forget that mitigation though higher health pool makes your non-scaling healing methods relatively less effective.
Not much to say, table with +100 AR at different base values:
1.000 AR = 07.4% increase in mitigation
900 AR = 08.0% increase in mitigation
800 AR = 08.7% increase in mitigation
700 AR = 09.5% increase in mitigation
600 AR = 10.5% increase in mitigation <- starts losing to +500 VIT at 4.5k VIT
500 AR = 11.8% increase in mitigation
All Resistances stacking will always apparently lose to its mitigation counterparts, but the mitigation it provides (in the same fashion as Armor) actually reduces the damage taken without increasing your health pool, which is a win. Still, trading 5 Vitality for 1 point of AR is pretty difficult - I'd much rather stick to socketing Diamonds and working with Paragon.
Meh. I felt like a noob when a couple of weeks ago discovered that armor reduced all damage taken (and not only melee or whatever I thought that happened). Sucks I'll have to create table for both 595 and 397 values given that the affix might be competitive on certain slots.
Table with +397 Armor at different values:
20.000 Armor = 1.69% increase in mitigation
17.000 Armor = 1.94% increase in mitigation
15.000 Armor = 2.15% increase in mitigation
12.000 Armor = 2.56% increase in mitigation
10.000 Armor = 2.94% increase in mitigation
Table with +595 Armor at different values:
20.000 Armor = 2.53% increase in mitigation
17.000 Armor = 2.90% increase in mitigation
15.000 Armor = 3.22% increase in mitigation
12.000 Armor = 3.84% increase in mitigation
10.000 Armor = 4.41% increase in mitigation
Well...I guess I was wrong. Even at extremely low base values and with extremely high Armor % bonuses, +595 Armor doesn't grant much mitigation beyond 5.5%. You're better sticking to what's already given from wearing...armor...plus strength.
My newest favorite offensive affix. What's not to love about a new damage multiplier? Bracers are a no-brainer (STR+CC+Elem) and it's a heavy candidate for final piece on Amulet (STR+CC+CDam+Elem): I really don't think %APS, Avg Damage or %CDR can compete against that one. The only requirement is using skills of its respective element.
I don't even feel like creating a table for that one for it follows the same logic as everyother scaling. Difference is that it has the huge 20% damage base.
Table with values of Elemental Damage % and the increase in damage from +20%:
115% Elemental Damage = 9.3% increase in damage = +750 STR at 8k STR
100% Elemental Damage = 10.0% increase in damage = +750 STR at 7.4 STR
80% Elemental Damage = 11.1% increase in damage
60% Elemental Damage = 12.5% increase in damage
40% Elemental Damage = 14.3% increase in damage
20% Elemental Damage = 16.7% increase in damage
10% Elemental Damage = 18.2% increase in damage
From what I've read, this bonus is added to other ordinary buffs (such as BR, Insanity, Berserker Rage, etc). Multiplying the damage of Primary Skills is meh, the other skills are great, specially if you take into consideration their slots. Boots are no brainer and for the final Helm affix, it competes against mitigation bonuses (VIT & Life%) and attack speed (from few legendaries), I would go for it. Cooldown Skill Damage also competes against mitigation for final affix: it's either Life% or VIT.
Table with base values of Damage % (+skills) and the increase in damage from +15%:
105% Damage = 7.3% increase in damage (Item + Insanity + Berserker + Marauder's)
65% Damage = 9.1% increase in damage (Item + Insanity)
30% Damage = 11.5% increase in damage (Item + Marauder's)
15% Damage = 13.0% increase in damage (Item or BR:Marauder's)
10% Damage = 13.6% increase in damage (Battle Rage)
Items with Movement Speed are vital, aren't they? NO, they.are.garbage. 50 Paragon into Movement Speed and reroll all that crap from your gear. Why? Because 25% Mov. Speed = 250 Attribute, which is worse than an ilvl60 legendary stat roll.
Bonus vs Elites (primary affix), vs Demons/Beasts... (secondary affix) all scale in the similar fashion as %Elemental Damage and %Skill Damage in terms of damage with the difference of having diminishing returns.
Another important thing to note is that these bonuses are exclusive in the damage calculations, meaning that they aren't added to other bonuses but, instead, multiply themselves. Example: 30% vs Elites & 20% vs Demons would result in a 56% damage increase against Diablo.
• Damage vs Elites legendaries [* might be higher]:
• Damage vs Demons legendaries:
http://us.battle.net/d3/en/artisan/blacksmith/recipe/demons-aileron [15% on 4pc]
• Damage vs Beasts & Humans:
See Elemental % and Skill % Damages tables.
To be honest I don't even know if it this method is any reliable but to me it seems to illustrate the differences between affixes & item slots well enough in such a manner that I could create a ranking with the top affixes for each slot (sort of a TL;DR version):
○ Shoulders: STR + VIT + (CDR / Cooldown Skill / Life% / AR)
○ Helm: STR + CC + Socket with CDR + (Spender Skill / VIT / Life%)
○ Chest: STR + Socket with 3 AR + Cooldown Skill + (VIT / Life%)
○ Pants: STR + Socket with 2 AR + VIT + (Life%)
○ Belt: STR + VIT + (Life %) + (AR / LpFS)
○ Boots: STR + VIT + Spender Skill + AR
○ Bracers: STR + Elemental% + CC + (VIT / AR)
○ Gloves: STR + CC + CDam + (VIT / CDR)
○ Ring: STR + CDam + CC + (AVG / CDR)
○ Amulet: Elemental% + CDam + CC + (STR / AVG / CDR)
○ Weapon: Damage Roll + STR + Socket with CDam [130%] + (% Damage / CDR)
Last affixes are what I consider easily replaceable by legendaries with more offensive affixes (such as CDam and Elemental%). CDR% is also replaceable on builds that want to rely on attack speed. I wouldn't replace Skill-Specific Affixes though, 15-30% increase on their damage is pretty big.
Edited by Eduw#1663 on 5/23/2014 11:43 PM PDT
Thanks for putting this together. Do you suppose that with strength being uncapped by paragon points, the ultimate endgame would favour some non-strength roll items in some slots in favour of % scaling stats? If this is the case, I wonder what the point of transition in terms of paragon levels this occurs at.
Any idea what amount of average damage rolls are available on rings at level 70?
When I was playing beta, they went higher than 60-120. It still seemed though dex >> average damage on jewellery though.
Good job putting this together (>^^)>
It will be a good reference to keep~
Edited by VocaloidNyan#1582 on 3/19/2014 2:22 PM PDT
Pretty much nailed everything. Don't leave bonus to enemy types out though. Bonus vs elites/demons ect stacks on top of everything else as well and should scale very nicely.
Nice write up, covers pretty much dead on how I planned on setting my character up :D
Good post. You might want to add 15% Rend damage on Chest/Shoulders.
When playing barb, has anyone gotten something else than Physical, Lightning, Cold, Fire elemental damage on the legendaries that roll a random element?
Stone of Jordan
Eye of Etlich
I would imagine we shouldn't, zapro, due to Smart-Drop and the fact that we don't have any attacks that are arcane, poison or holy. I feel like the answer to that question was obvious and you had something else on your mind to ask...
Edited by sfxer#1565 on 3/19/2014 4:20 PM PDT
Absolutely. We'll eventually reach a point where the % damage increase comming from an X ammount of Main Attribute won't be as attractive as an additional 10-15% damage to a single skill, 15-20% to elements and so on.
Say you have a 15% Scaling Damage Piece wants to compare against STR:
Say you've got 5900 STR (60 damage multiplier). In this situation, to have the same increase in damage as from the initial affix (15%), you'd need 900 STR. Math:
x * 60 * 1,15 = 69x
x * (60 + y) = 69x -> y = 9
Seeing as we'll most likely be running with 7K+ STR, odds are that these Scaling Damage Affixes will progressively replace some of its affix slots. To be honest, I hadn't really put much thought about it but I assume that we could probably create a curve with the 'breakpoints' where STR starts losing in terms of additional damage when compared to the traditional Scaling Damage numbers, 10-15-20-30%: basically whenever one of these affixes are worth more than 750 STR, it becomes a win for them. Quick table for reference:
10% - After 7.40k STR (75.0x) [increase difference]
15% - After 4.90k STR (50.0x) [+5% = -2.50k STR]
20% - After 3.85k STR (37.5x) [+5% = -1.35k STR]
25% - After 2.90k STR (30.0x) [+5% = -0.75k STR]
30% - After 2.40k STR (25.0x) [+5% = -0.50k STR]
As you can see, the higher the % bonus, the sooner your character is likely to hit the breakpoint. To be honest, taking these numbers and the fact that we'll be running with 7k+ STR into account, I might have to re-evaluate the attractiveness of the stat by itself (ie, one could easily trade 750 STR for any Elemental% whenever over 4.9k STR : the same could be done with less straight-forward affixes, like %APS/%CC/%CDam but that would require building tables of their own).
[Probably gonna update the categories with breakpoints alike in the future]
Looking over how damage calculation is done, it seems like the stats attractiveness (based on minimum damage) depends on the weapon being used (base minimum physical damage, sum of minimum physical damage on weapon, % damage increase, sum of minimum physical damage on gear, "Adds % to Elemental Damage", and finally miminum elemental damage on the weapon). So yea...would require some in-depth information of the typical numbers gotten for ilvl70 weapons.
Thanks. I didn't really want to get that deep but I might create a section for them.
I know Damage to Demons (same for other damage to niche) is a Secondary Affix but, asides from one multiplying the other, the first and Bonus to Elites both suffer diminishing returns and wouldn't be too difficult evaluating - specially now that I've realized how to 'weight' STR vs Scaling Bonuses [thanks to @zylog].
Thanks for the information, fixed. I sort of got unlucky with my crafting and couldn't get it to roll...same for Furious Charge as Cooldown Affix.
Great Post. I really wish I had the understanding of the game you obviously have. Someone should suggest a sticky for this post and hopefully Blizzard take some suggestions on board.
It appears they have nerfed CD, CC and AS for Strength, increase in X damage and area damage.
This post would seem to argue to improve the caps for CD ect...
I like these skills better they are more concrete.
Maybe some things to add-
Perhaps chance to block, rolls on:
Rings: 13% on 2.0 justice lantern
Helm: 8% (HOC)
Shields: 8% (SS, haven't looked at rares tbh)
Belts: 8% (Kotuurs brace, only one I've found)
Elemental damage can also roll on rings -soj (15-20%) Hellfire?
Crowd control reduction? Not sure if you want to include this, can roll on:
Helms: 10-20% (sages orbit, blind faith)
Ammies: 6-10% (checked with my 1.08 crafted ammy, do not have any from 2.0)
Weapons: 10-20% (blackguards)
Rings: justice lantern (35-50%)
There's elite damage that rolls on legs also (and some rare weapons):
Rings: SOJ (25-30% @lvl60), unity, (12-15% @60)
Area damage on weapons:
Hack (14-20%) Pretty sure i had a hallows wep with 18 also
Edited by RagingKoala#1984 on 3/19/2014 7:14 PM PDT
One thing that bothers me is items like mempo which have the unique attack speed roll as primary, but at the same time take up one of the 4 primary stats.
the 7 ias takes up a primary stat. Why not get this on a different slot instead? 750 vitality as a primary stat seems to be better than 7 ias in the helm slot, and you can get 7 ias on say a ring. Probably even those legendary bracers too or belt.
Edited by VocaloidNyan#1582 on 3/19/2014 8:34 PM PDT
Thanks for the feedback, I'll try replying in parts. =P
I've never planned getting too deep into mitigation and I honestly have no clue how to evaluate %Block Chance. Either way, if I (or we) were to add it, we would probably need in-depth information related to ilvl70 shields (block chance & block amount) and some monster damage parameters (ilvl70 T1-T6). It's probably the hardest stat to weight.
Crowd Control Reduction is a secondary affix with diminishing returns. I'd rather stick to primary affixes for now, but I can already tell you that it's one of the most attractive of the ones available (specially Justice Lantern's due to the diminishing returns).
As for Elemental Damage on rings, from what I've seen it's exclusive to SoJs. You'd have to weight it against other Ring affixes, it's probably much better but I'm working on updating the comparisson tables based on @zylog's question.
For +Area Damage on weapons, I've seen around and you're probably right. Issue with the slot is that it is pretty effective on Paragon points already and a damage increase on a static 20% chance isn't that reliable. Either way...going to update with up to 18% - it's probably higher on ilvl70.
Yep, that's the issue with guaranteed roll legendaries. I can count at least 3 legendary helms with attack speed and, with the exception of Andy's which you'll always reroll the fire damage taken, it's a quite attractive reroll piece. The only exception to this rule is for characters with already really really reeeeally high Main Stat. I'm working on %APS vs Attribute benchmark but I can already tell you that it 7% APS isn't better than STR until above 12.05k STR.
Threats of violence. We take these seriously and will alert the proper authorities.
Posts containing personal information about other players. This includes physical addresses, e-mail addresses, phone numbers, and inappropriate photos and/or videos.
Harassing or discriminatory language. This will not be tolerated.