I did some analysis of the Attributes in LE :
For each attribute, I counted for each mastery the number of skills using it for scaling, the number of nodes increasing it and the number of nodes using it.
I know that table is really basic and doesn’t tell that much info. I did some more precise tables with the decomposition per advanced mastery (I can share them if needed). They are better but still not precise enough as it would require for each unlock(able) skill and node to analyse the value against the cost for a build in a mastery.
Yet, we can already see some things.
For memory, the theorical main function of the attributes is to give a defensive boost:
- Strength = +5% armour
- Dexterity = +5 dodge rating
- Intelligence = +4% ward retention
- Attunement = +5 elemental protection
- Vitality = +15 void, necrotic & poison protection (bigger boost as no skill scales on it)
But truly, nobody really cares about that defensive buff and only cares about the buff on the skills. That defense boost is a little extra that nobody would care about if it wasn’t there.
Excepted for Ward Retention.
I will talk first of the Vitality as I believe it is the biggest issue.
In itself, it is some protection with an extra step. Which makes it harder to understand what it does. And also mean that nobody will increase it using Affixes but will directly increase the associated protections.
Through the nodes, it’s not that easy to increase. Shaman has no node for it. Beastmaster, Druid and Spellblade can access to only one node. Only Void Knight and Necromancer can access to 3 nodes (with 1 on the tree of another advanced mastery).
The two nodes using Vitality are on the Spellblade (second part) and Lich (first part) trees. But as you don’t increase Vitality, the effect is not worth the investment (even with points in Vitality, the effect wouldn’t be worth it). Points are better used elsewhere.
Vitality could easily be removed.
Another easy issue is Dexterity.
I know Rogue is not there yet and should use more that attribute.
Even though, there are still issues.
Biggest is with Primalist: only one skill using that attribute, no node to increase it nor use it. Serpent Strike needs to be reworked to remove the use of the Dexterity.
Other class are a bit better but still not great with the attribute.
Only one node to increase Dexterity for the Sentinel, and on the first part for the Forge Guard tree. For three skills using that attribute.
The Acolyte still miss the Warlock. But there are Harvest and Reaper Form that use Dexterity. And no node associated to the attribute.
Last is the Mage, with three skills using Dexterity for scalling, one node to increase it and one node using it for some extra wards.
I don’t believe Dexterity is worth the investment on any mastery as it is now. All skills scaling on Dexterity also scales on Intelligence or Strength, which are the main attribute of Acolyte/Mage or Sentinel.
Dexterity could be great for Rogue. But it wouldn’t change anything for the other classes if it disappeared.
Then we see that Acolyte and Mage have no point increasing any other attribute than Intelligence.
Acolyte has 2 skills using Dexterity, but both also use Intelligence. None use another attribute.
Mage has 3 skills using Dexterity, but both also use Intelligence. None use another attribute.
And we see that Primalist and Sentinel have two ways: full Strenght or full Attunement.
Primalist has 1 skill using Strenght and Dexterity, which needs to disappear. And 4 skills that use Strength and Attunement. Not enough to justify to go hybrid when focusing on only one attribute he could fill the 5 specialization slots.
Sentinel has 3 skills using Dexterity, all also using Strenght. With only one node to increase Dexterity, it’s more effective to go all Str than going Str/Dex.
Sentinel has also 7 skills using Attunement and Strenght. Finally an occasion to use two attributes? Not so fast!
Void Knight and Paladin has each two skills Str/Att (from sentinel and own tree), and the possibility to get three others from the Forge Guard tree. While the Forge Guard has five skills using Str/Att. So we could imagine Forge Guard is the best option to go Str/Att.
But Void Knight have 2 (“natural”) + 2 (from other trees) nodes to increase Str and 2+2 nodes to increase Att. Paladin have 2+1 nodes to increase Str and 2+0 to increase Att. While Forge Guard have 3+1 nodes to increase Str and 1+1 nodes to increase Att.
So while FG is the better equiped with skills to go the Str/Att way, its nodes are better suited for a Str only build. And it is the VK and the P that have less skills adapted for it that have access at nodes better suited for an hybrid build.
Anyway, the VK, FG and P respectively have 10+3, 13 and 10+3 skills using Str and 5+5, 5+4 and 4+5 skills using Att. So, naturaly, they are more equiped to go on a full Str build.
But, as I said at the beggining, the analysis is not precise enough. Further analysis is required to see the cost of unlocking a skill on another tree, the cost of improving a node on another tree, the benifit on a skill from an attribute and the benefit on an attribute of a node.
The only things I wanted to show were that:
- even though each mastery is concerned by 3-4 attributes out of 5, in most cases only 1 matters, and maybe 2 in rare cases;
- the main effect of an attribute is to improve a skill and we would mostly have the same effect with a unique attribute (and some rebalance for the skills that use two attributes);
- with the exception of the ward retention, the defensive bonus is just a nice extra, not something we can build around (resulting in us not caring about it being present).
I believe the attribute system is too complexe for what is does. It needs to be simplified or the attributes needs to be made more important.
And if they are made more important, we need to have more control on the attributes we want to improve. Not having the game say through the trees “you play that mastery? this is your only attribute”.
There is an attribute to improve elemental resistance, but it is not used to improve elemental resistance when playing a Mage.
It’s long. And it’s not my native language. Feel free to ask me to clarify something as I’m sure I didn’t managed to say everything I wanted to say or how I should have said it. It’s not easy to organize everything correclty.