I’ve been reading through the thoughtful feedback on here, and I really appreciate all the responses from Blizzard — it’s very helpful to know what the devs are thinking, and what they’re looking at.
And now, I would like to share my own feedback, along with a few notes that I’ve read on the Hunter forum. There are many excellent feedback threads on that forum; I will be referring to a few through my post, although I will be paraphrasing them.
Before I start, I want to lay out my premises, my intentions, and my approach. I am taking the devs at their word; when Dragonflight (DF) was first announced, a series of video interviews were shared with us. In one such interview, the devs emphasized that they are focusing on the class identity over the spec identity. This is the main premise I am operating from, and it is my intention to analyze the trees (including the class tree) in order to determine whether the class identity — and to a lesser extent, spec identity — is clear and easily seen, or not. Lastly, I am the sort of person who “looks at the big picture” — some people might would feel that I am missing the tree for the forest, but this is how I see things.
I will talk about a few things in my analysis of the trees: The nodes, the pathing, the themes, and lastly, I will compare and contrast the Hunter trees to other class trees. I will also cite other feedback threads, as I mentioned earlier.
Compare and Contrast: Hunter Talent Trees and the other Classes
I will start with the compare and contrast; I have complied numbers on all the talent trees currently available in Alpha and put them into collapsible sections; I will post this in the following post. A note before I begin, though — I counted the nodes on one half the week of August 2-8, and the other half the week of August 9-15. I acknowledge that by August 16, some of the numbers may be a little different, but the overall picture remains largely the same.
What I did was I counted the number of nodes in each tree — the general class tree, and the spec trees for each class. I counted how many nodes are in each tree, and I counted the multiple-point nodes (2-point and 3-point). The reason I did this was due to seeing quite a number of Hunters — on both the Alpha forum and the Hunter forum — expressing concern over “bloat” (the number of multiple-point nodes). I wanted to see if there was a legitimate cause behind their concern… and the numbers I found was quite interesting. There definitely are a few outliers. I will now share the numbers before addressing my own thoughts.
In this post, I will be sharing the end result of the numbers that I collected; if you want a more thorough break-down, please see my next post, which will be solely dedicated to sharing the numbers that I collected from counting each of the available tree (sans Monk and Demon Hunter, which did not yet have their Alpha trees at the time of this posting). Specifically, what I am sharing right now is the percentage of nodes that are multiple-point.
Death Knight Trees
Class tree — 9%
Frost tree — 29%
Blood tree — 31%
Unholy tree — 29%
Class tree — 20%
Balance tree — 35%
Feral tree — 30%
Restoration tree — 33%
Guardian tree — 24%
Class tree — 20%
Preservation tree — 23%
Devastation tree — 28%
Class tree — 37%
Beast Mastery — 51%
Survival — 44%
Marksmanship — 35%
Class tree — 38%
Frost tree — 29%
Arcane tree — 15%
Fire tree — 27%
Class tree — 20%
Retribution tree — 16%
Holy tree — 15%
Protection tree — 28%
Class tree — 29%
Shadow tree — 24%
Discipline tree — 23%
Holy tree — 19%
Class tree — 24%
Subtlety tree — 14%
Assassination tree — 29%
Outlaw tree — 19%
Class tree —22%
Elemental tree — 25%
Enhancement tree — 21%
Restoration tree — 23%
Class tree — 33%
Affliction tree — 38%
Demonology tree — 30%
Destruction tree — 41%
Class tree — 9%
Protection tree — 11%
Fury tree — 13%
Arms tree — 10%
Several things immediately came to me when I first looked at the numbers I wrote down — there appears to be three different brackets of bloat (or, to use another term, “cost”). For the sake of clarity, I am labeling these three brackets as “Cheap,” “Middle,” and “Expensive.”
The trees which consists of 19% or less multiple-point nodes are “cheap.” The trees which consists of between 20% to 29% multiple-point nodes are “middle” (I have seen people refer to this as the ‘sweet spot’). Lastly, the trees which consists of 30% or more multiple-point nodes are “expensive.”
There are 11 specs which meet the “cheap” criteria:
- Warrior Class + DK Class trees — 9%
- Warrior Arms tree — 10%
- Warrior Prot tree — 11%
- Warrior Fury tree — 13%
- Rogue Sub tree — 14%
- Mage Arcane + Pally Holy trees — 15%
- Pally Ret tree — 16%
- Rogue Outlaw + Priest Holy trees — 19%
There are 20 specs which meet the “middle” (sweet spot) criteria:
- Druid Class/Evoker Class/Pally Class trees - 20%
- Shammy Enh tree — 21%
- Shammy Class tree — 22%
- Evoker Pres/Priest Disc/Shammy Resto trees — 23%
- Druid Guardian + Priest Shadow trees — 24%
- Shammy Ele tree — 25%
- Mage Fire tree — 27%
- Pally Prot + Evoker Dev trees — 28%
- Rogue Sin/Mage Frost/Priest Class/DK Frost/DK Unh trees — 29%
Lastly, there are 13 specs which meet the “expensive” criteria:
- Hunter BM tree — 51%
- Hunter SV tree — 44%
- Lock Destro tree — 41%
- Lock Aff + Mage Class trees — 38%
- Hunter Class tree — 37%
- Hunter MM + Druid Balance trees — 35%
- Druid Resto + Lock Class trees — 33%
- DK Blood tree — 31%
- Druid Feral + Lock Demo trees — 30%
I will now discuss my thoughts about these numbers.
The first thing which leapt out to me is the sheer difference between all four Warrior trees (which ranges from 9% to 13% for multiple-point nodes) and all four Hunter trees (which ranges from 35% to 51% for multiple-point nodes). The reason this is significant, because with a lower percentage of multiple-point nodes, the wider a player can cast their net — or in other words, the player can spec into more nodes in the tree; naturally, with a higher percentage of multiple-point nodes, the player is stuck with speccing into less nodes.
Another thing that struck me as unusual is that there are several clear outliers when we look at the big picture: Hunter, Warlock, and Warrior. Hunter and Warlock are the only classes in which all four of their talent trees are “expensive,” whereas Warrior is the only class in which all four of their trees are “cheap.” Druid, Rogue, and Paladin also stand out; with Druid, 3 out of 5 trees are “expensive,” whereas with Rogue and Paladin, 2 out of 4 trees are “cheap.”
(Death Knight and Mage are the only classes with trees in all three brackets. The rest are mostly in the “middle” brackets with one or two trees in either the “cheap” or “expensive” brackets, aside from the outliers that I mentioned.)
Lastly, I noticed that there are several trees in which half or more of their first tier is taken up by multiple-point nodes:
- Sin Rogue tree — the first tier has 9 nodes, and 5 nodes are multiple-point: 56%
- MM Hunter tree — the first tier has 11 nodes, and 6 nodes are multiple-point: 55%
- BM Hunter tree — the first tier has 12 nodes, and 6 nodes are multiple-point: 50%
- Feral Druid tree — the first tier has 9 nodes, and 4 nodes are multiple-point: 45%. The reason Feral get a mention is because 2 of the 4 nodes are three-point nodes.
As I have mentioned before, the higher number of multiple-point nodes in a tree means a player cannot talent into as many nodes. The first tier is important, since it gives a player the first impressions of a tree. When a player specs into a Fury Warrior, they are able to choose 8 out the 10 nodes in the first tier before unlocking the second tier. In comparison, a BM Hunter can only choose 5 to 6 nodes out of the 12 nodes in the first tier before unlocking the second tier. This can lead to a more frustrating experience for the player, especially if they have already seen the “cheap” trees.
Or, to look at this in another perspective: In order to reach the second tier, a Fury Warrior must choose which 2 nodes they are not taking, whereas a BM Hunter must choose which 6 to 7 nodes they are not taking. The penalty is much larger in trees with a higher prevalence of multiple-point nodes. The reason I used the specific word of “penalty” is because Blizzard has talked about the new talent trees as important for players “making meaningful choices.” However, it can be argued that the choices are not as meaningful for some class trees — as has been mentioned, with “cheap” trees, the players can snatch up many more nodes, whereas with the “expensive” trees, the players must choose which nodes would hurt the least to lose. The experience is quite different between the three brackets.
I understand that the trees cannot be perfectly equalized. However, I think that Blizzard could (and should) aspire for a better balance; with the three outliers (Hunter, Warlock, Warrior), they should aim to get at least 3 trees — ideally all four trees — into the “middle” bracket (between 20% to 29% for multiple-point nodes) — especially Hunter, with SV being at 44% and BM at 51%!
Thoughts on the Class Tree
While looking over the Hunter trees, I noticed several oddities; some of the posters have mentioned them, but I still wanted to put them all together into one post.
Firstly, when looking at the Hunter class tree, I noticed that there are several nodes that affects an ability that a player could take without taking the said ability.
For example, there is Tranquilizing Shot; a player could spec into Improved Tranq Shot without taking Tranq Shot — this is because a player could take unlock Improved Tranq Shot by pathing through Rejuvenating Wind.
Likewise, players could spec into Binding Shackles (which improves onto Binding Shot and Intimidation) without taking Binding Shot or Intimidation — this is possible by a player choosing Scatter Shot over Binding Shot, or Hi-Explosive Trap over Intimidation.
The location of some of the nodes seemed to be rather inconsistent; the trap nodes are a little scattered — Hi-Explosive Trap shares a choice node with Intimidation, despite sharing two very different roles (one is a knockback that does damage, while another is a stun). The node for reducing the cooldown of traps (Improved Traps) requires that a player spec into either Camouflage (stealth) or a choice node for two different versions of Survival of the Fittest (damage reduction). This node — Improved Traps — is on the other side of the tree when compared to Hi-Explosive Trap.
The next talents that directly impact traps are at the bottom — Nesingwary’s Trapping Apparatus and Steel Trap. Now, a player could technically path into those two nodes through Hi-Explosive Trap or Improved Traps — and they could just as easily path into those two third-tier trap nodes without the two trap nodes in the second tier. The reason I found this odd is because when I looked at the other class trees, I noticed that Blizzard has a tendency to lump nodes of the same category together. So I found it unusual that the trap nodes are so scattered throughout the Hunter class tree.
(The one trap-related node in which I thought the placement is decent is the Tar Trap; Tar Trap is required to reach the Hi-Explosive Trap node, which seemed appropriate to me. However, a player could reach the other trap nodes without taking Tar Trap.)
Another choice which seemed odd to me is the placement of Tranq Shot in the “pet side” (or pet path) of the tree. In order to take Tranq Shot, a player has two options: Spec into Tar Trap or Improved Pet Mend. The Tar Trap requirement makes sense, since Tar Trap and Tranq Shot are both utility-based. However, Improved Pet Mend is related to pet management and survival. Next, the only way a player could reach the node Beast Master (which increases the pet’s damage) by taking either Improved Tranq Shot or Binding Shackles —and yet once again, I find it odd that for a player to reach more pet-related nodes, they can only reach it through taking utility-related nodes.
When I looked at the other side of the class tree, I noticed that the rightmost path seemed to go back and forth between shot-based damage and survival nodes. This path starts with Kill Shot and Improved Kill Shot, but then it switches to Natural Mending (shortens the cooldown on Exhilaration), Camouflage, and Survival of the Fittest? The next node a player could take on this path is Born to be Wild (shortens the cooldown of speed and damage reduction talents)… then finally a player could start speccing into damage-boosting nodes once again with Master Marksman, which is required for Explosive Shot/Barrage and Serpent Sting.
I find it an odd choice to place survival nodes in the same path as damage nodes, especially when I compare it to the other class trees — for example, on the Druid class tree, there are very clear, separate paths for healing spells, damage spells, beast attacks (cat and bear), utility nodes, and PvP-based nodes.
The middle of the Hunter class tree is also odd to me — for example, in order to reach the Keen Eyesight node (which increases crit by 2%/4%), a player must take either Binding Shackles (utility) or Agile Movement (increased running speed). In order to take these nodes, a player must spec into either trap nodes or utility nodes (those specific utilities are commonly used in PvP). Why would a player be required to spec into PvP abilities in order to get to a node that benefits players in both PvP and PvE setting?
Now, I always believe in giving credit where credit is due: For the most part, I find the themes in the Hunter class tree to be pretty solid! There’s the pet nodes, there’s the utility nodes (including the trap nodes), and there’s the “special shot” nodes. The tree is pretty strong when it comes to establishing the Hunter identity — it’s just the placements that doesn’t really make sense, and the pathing can be rather odd.
Of course, there’s the whole thing where a player could take Improved Tranq Shot without taking Tranq Shot, or take Binding Shackles without taking Binding Shot or Intimidation. That needs to get fixed.
Now, let’s go to the spec trees! (Edit: Originally, I was going to talk about all three specs, but then I decided to only do MM in this post, since it’s already a… long… post. I will discuss BM and SV another day.)
Thoughts on the MM Tree
I noticed that in the MM tree, a player could take Razor Fragments, which affects Trick Shots, without taking Trick Shots. Yes, Razor Fragments also works with Deathblow (which is a requirement for unlocking Razor Fragments), but it is still a little odd to be able to reach Razor Fragments before reaching Trick Shots.
Likewise, a player could take five nodes affecting Rapid Fire without taking Rapid Fire itself — in part because these nodes are located on the middle or the right side of the tree whereas Rapid Fire is on the left side:
- Focused Aim (increases Rapid Fire damage by 5%/10%)
- Hunter’s Knowledge (increases crit chance of Rapid Fire by 5%10%)
- Deathblow (Rapid Fire have a 10% chance to grant Kill Shot)
- Double Tap (which essentially doubles Rapid Fire during the duration)
- Trueshot Aura (reduces cooldown of Rapid Fire during the duration)
I also identified seven nodes which affects Multi-Shot — and a player could take six without ever taking Multi-Shot. They are:
Precise Shots (Aimed Shot boosting the damage of Multi-Shot by 35%/70%)
True Aim (increases damage of Multi-Shot by 25%)
Lethal Shots (Multi-Shot have a 30% chance of reducing Rapid Fire cooldown)
Calling the Shots (Multi-Shot reduces the cooldown of Trueshot Aura by 1.5/3 seconds)
Bombardment (Every 6/3 Arcane Shot/Chimera Shot grants Multi-Shot the ability to trigger Trick Shots regardless of number of targets)
Salvo (every 45 seconds, Multi-Shot will also cast Explosive Shot)
Now, I will grant that most of those Rapid Fire-affecting and Multi-Shot-affecting nodes also affect Aimed Shot or Arcane Shot or Chimera Shot — for the most part. But two of these nodes cannot function without Multi-Shot: Bombardment and Salvo. Those two nodes are utterly pointless if a player doesn’t have access to Multi-Shot. Players can reach these nodes by pathing through Deathblow, Razor Fragments, and Volley — thereby bypassing Multi-Shot. (Which also results in a player missing out on Trick Shots, since Multi-Shot unlocks Trick Shots.)
A player could easily miss Multi-Shot, since Multi-Shot is erroneously placed in a circle-shaped node, as opposed to a square-shaped node. A player who isn’t paying attention could easily miss out on Multi-Shot, and path all the way down to Salvo… without ever taking Multi-Shot.
I think the issues with Rapid Fire and Multi-Shot could be easily solved by Blizzard moving those two up to the first tier — right below Aimed Shot. When one looks at the other classes’ spec trees, one would notice that many of those trees have abilities right up in the beginning, so those abilities are the first thing that a player grabs before getting nodes which improves on the said ability. By putting Rapid Fire and Multi-Shot right under Aimed Shot, Blizzard could ensure that a player grabs Rapid Fire and Multi-Shot before they unlock all the nodes which improves those two abilities.
If Rapid Fire and Multi-Shot are left in their current location, then the pathing doesn’t make sense — the nodes which affect Rapid Fire are scatted all over the tree, and likewise with Multi-Shot. This is how a player could take five nodes affecting Rapid Fire without taking Rapid Fire, or six nodes affecting Multi-Shot without taking Multi-Shot.
However, it’s not all doom and gloom — despite the odd placement of some nodes, and pathing issues, I find the themes to be pretty clear and solid in the MM tree. It’s obviously the tree for “fancy shooting” and “sniping.” The nodes are pretty distinct and does fit the MM flavor quite nicely.
Closing Thoughts and Concerns
As you can see, the Hunter trees do indeed suffer from “bloat” — they are expensive with their high number of multiple-point nodes, and they have some unusual node placements which enables a player to spec into numerous nodes affecting an ability… without ever taking the said ability. Also, some paths can be a little confusing; for example, one path starts out as a damage path, then abruptly switches over to a survival/utility path, before reverting back to damage. Or there’s the pet path which suddenly changes into a trap/utility path, and then changes back into a pet path near the end. There are several examples of this throughout all the trees.
There are a few more things I want to discuss, but this post is already super long as it is! So I’ll stop for now, and continue tomorrow or on Wednesday.
Up next will be my thoughts about the BM and SV trees — but in the meanwhile, there are some really excellent feedback threads on these specs over in the Hunter forum. I’ll share them so you can take a look until I post about BM and SV.
BM Feedback thread: [Feedback - Build 44999] Hunter Class Talents + Beast Mastery
SV Feedback thread: [DF] Hunter Class and Survival Talent Tree Feedback
SV Feedback thread: Alpha Survival Hunter Questions for the Developers