Over the last 3 years of playing LE beta I’ve had an irritation that’s become a pearl. Instead of leaving it well alone I’ve decided to finally share it here and exorcise myself from the burden of frustration and absolve myself of any future guilt. I really like Tempest Strike. Or rather I really want to like Tempest Strike. Because of that I’d regret not providing any feedback toward the skill on the forums before any Shaman or Tempest Strike changes make it into the game.
The Primalist is my favorite Class, the Shaman is my favorite Mastery, and Tempest Strike is the skill that first jumped out to me as an archetype that I’d love to play within that Mastery. Unfortunately there are a number of factors that limit both the skill’s effectiveness and it’s general feel that I believe need to be discussed.
Identity and Theme
TS’s theming is probably the strongest aspect of the skill. Having the Primalist swing their weapon down so hard that spells start flying out is fantastic. It covers the primary damage types for the Shaman and really has a great player fantasy that it provides. Hits a nice hybrid sweetspot between melee and caster that I personally find very appealing and I’ve heard many, mostly newer players, take great interest in the skill as well.
In terms of identity there are a number of things we can see that the skill is trying to do. Based on the skill tree the game is telling players that it’s a damage skill. Many ways to alter how the skill deals damage and just a couple utility nodes thrown in. It’s clear the skill is meant to have an element of randomness. The twist or hook the skill gives to the player is that it has an alternative stat scaling built into it with how it treats attack speed. I’d also argue that it is meant to be a Shaman skill, as it’s not usable while transformed for a Druid and has little overlap with melee or companion scaling for a Beastmaster.
These core aspects of the skill are compelling, but the specifics of how everything works together lead to a number of problems.
Performance and Functionality
When a new player decides to play with TS they get promised something interesting and it holds up for quite a while. Through the leveling process the player uses the tools their given to deal spell damage while in melee range. Perhaps they decide to use Maelstrom and have
Tornado follow them. Now they’re free to use TS while these skills are active and get a bit more spell damage while being in melee range without spending any extra mana. Because of the pacing and over low difficulty level experienced throughout the Campaign and Normal Monolith timelines this build performs without a hitch. Overall a player using TS up until around level 70 shouldn’t have too many issues with the skill and might even think it’s fairly good. In a scenario like this Tempest Strike is functioning exactly as it should be by facilitating the player’s ability to deal spell damage while getting up close and personal with enemies.
There are several issues that this scenario build will run into and all of them have to do with how it manages to scale into endgame. In brief the issues that end up lacking are damage output, consistency, play speed, and skill tree cohesion.
Problems and Shortcomings
When looking at how TS falls short of its contemporaries it’s important to note that there are many factors that all end up contributing toward something that’s more than their parts. As previously mentioned, the core elements of the skill are compelling but the way they’re implemented ends up having negative synergy.
The greatest of these problems lies in the alternative stat scaling for Attack Speed. The concept of taking a stat and changing how it provides power to a skill is fantastic. On paper, if the math on the power exchange comes out favorable or neutral there shouldn’t be any problem with what is altered. The issue with this particular exchange is that it’s not just about the math. As a player progresses through an ARPG/dungeon crawler one of the ways the game communicates to them that their power is growing is by increasing the overall speed at which the player is able to navigate through game. Movespeed and action speed are incredibly powerful tools to make the play experience smooth and engaging. By removing the player’s ability to increase their action speed on a skill the game prevents the feel of the skill from ever leveling up beyond what it was when the player first used the skill. In addition to the overall feel of the skill suffering, it also prevents the player from engaging with combat mechanics effectively. You’re ability to reposition yourself to avoid telegraphs and any melee attack become limited while the game continues to scale in terms of power and speed.
The second most glaring issue is the consistency of the skill. Simply put, there is not a way to make the skill do something every time you use it. There are a number of ways to build the skill and each of them have benefits, but none of them work all the time. When used as a supplemental damage source this is a nuisance. If the player attempts to use TS as a main damage source this becomes a massive problem. This is also the first case of negative synergy as you’re unable to make up for a lack of consistency by improving the rate at which you roll the dice. Best case scenarios are a 9% chance for complete failure by scaling all proc chance, 14% by scaling two, or 29% chance of failure by maximizing the proc chance of a single proc. While complete consistency might not be necessary depending on how a skill is designed, there is some expectation of a payoff given the risk taken on.
The third issue feeding into the negative synergy of low attack speed and consistency is that the overall damage output is subpar. The big issues here are that the damage effectiveness for the procs are low (100% for lightning, 100% for physical, 70% for cold) and the skill tree doesn’t contain many sources of multiplicative damage. The melee hit and subskills are no exception to this. By scaling attack speed you can achieve a large amount of damage penetration that helps but does not overcome these problems. Again, if the rate at which you could proc the spells could be increased then the relatively low damage could be offset. For those that are skeptical that the damage is low in comparison I’ve been playing a TS build and recording footage that will be included at the bottom of this post as an example.
When it comes to using a melee skill to supplement the spell damage being dealt, or even to generate mana, Swipe ends up being the superior choice by giving more mana based on attack speed and a 35% global damage multiplier for your other spells.
The last two problems relate to the identity and theming of the skill rather than the numeric performance. The skill tree fights itself. There are three main sections of the skill tree, but they’re all effectively the same. Enhance proc A, enhance proc B, enhance proc C, and allow for synergies between them. These elements are fine, but perhaps take up too much space. A fourth branch also allows you to enhance all procs simultaneously, this too is fine. The issues come up when you look at the downsides on nodes (reduced proc chance), amount of points required to invest in a proc, overall low power of node effect, and the emphasis on the skill failing in order to succeed. Two nodes in particular only take effect when the skill fails to proc a spell. One provides all proc chance to the next attack and the other provides mana. In no world does a player want a consolation prize for a skill fizzling out rather than a way to prevent the skill from fizzling. Honorable mentions include minor melee damage if a skill procs, all proc chance on low health, and a chance to add 6 melee damage to one random companion for 4 seconds if a proc occurs. Many of the ideas in the tree are good, but it limits itself too much by reducing the skill’s power in some areas in order to gain power in others.
The final problem is that this Shaman skill is strong when played as a Beastmaster. This is mostly tied to the fact that Shaman has not been updated for a long time and its passive tree doesn’t provide terribly strong incentives. In the Primalist’s current state the BM provides a considerable amount of scaling toward TS that the Shaman just isn’t capable of. A plethora of attack speed (Aspect of the Viper, Aspect of the Shark), crit chance and multiplier (Aspect of the Lynx), and Frenzy totem (more attack speed and crit multiplier). But what about Avalanche, what about a bit of spell damage? The raw stats in Shaman provide an overwhelmingly weaker amount of damage to TS and Swipe will support your Avalanche or other spells better than TS will.
In terms of gameplay there is no problem with BM using TS effectively. However, thematically it feels like a failure when the class it resonates with is unable to scale it effectively.
Overall the confluence of these shortcomings and the hindered game-feel of the skill prevent it from being a reasonable option to use for characters in endgame. So what could be done?
Desired Improvements
In general there are many things that could be done to improve the skill and with so many moving parts it’s hard to know which is the correct one. The following is a listing of my personal opinions on what I believe the appropriate changes would be to improve the performance of the skill and, more importantly, how it feels to play.
Address low attack speed problems. There are many ways to address this. Which method is appropriate depends on what other changes happen to the skill’s RNG and damage values. Possible changes include:
- Complete removal of AS penalty. With no other changes to damage or RNG the skill would still not be strong, but it would be playable and feel fun.
- Add an AS effectiveness stat (as per added damage effectiveness). This could limit the speed, but not completely kill it.
- Add a short cooldown (1s), but allow the attack animation to scale with AS. No effective change to baseline damage while improving its feel. CDR allows some scaling of proc rate.
- Add skill tree nodes that improve the baseline attack rate.
Address the consistency of the skills RNG. Again, many solutions to this could be proposed, but I firmly believe that using a skill should always produce an effect.
- Alter baseline proc chance to 100% with a weighted pool of which spell will proc (33/33/33). Skill tree nodes could affect the weighting of this pool or allow for a chance at additional procs beyond 100% chance.
- Change alternative stat scaling to Attributes. 1 Strength = 1% earth spike chance. Attunement for lightning bolt, dexterity for northern winds.
- Change alternative stat scaling for Attack Speed to increased proc chance rather than penetration.
- Add skill tree nodes to allow secondary procs from other effects e.g. cast a proc spell on crit, etc.
- Increase proc chance based on time since last use. Buff and consume playstyle similar to Lagon’s Slumber on Maelstrom.
Improve the effectiveness of the Skill Tree.
- Mana gained on proc rather than failure. Perhaps even giving a base mana regen buff for each unique spell procced in the last 4 seconds. The skill should not provide an incentive to remove everything that makes the skill cool.
- Improve values of proc chance increases and multiplicative damage.
- Reduce infighting penalties of synergies. Synergy nodes could improve the skill without reducing the odd man out proc. The benefit of disabling a proc should be much higher.
- Increase variety of the different areas of the tree to make the procs different playstyles rather than flavors.
- Improve the damage effectiveness of all aspects of the skill, particularly for Perfect Storm.
- Add ways of effectively scaling the melee portion of the skill, or perhaps add the ability for the spells to scale off melee damage.
- Add ways to increase the AoE of the skill. At the very least all of the spell procs need to be given the area tag to scale with the new area of effect affixes.
- Add and improve utility aspects. Stronger and more buffs for totems/companions. Striking a totem to get an effect rather than chance to spawn totems, etc.
Final Thoughts
If TS is meant to be a damage skill, the player needs to have some ability to use it as a damage source in isolation. At best, the current form of Tempest Strike can be used for a small amount of supplemental damage alongside other spells (Maelstrom, Tornado, Avalanche). However, Swipe is better suited both for playstyle and for overall effectiveness.If TS is meant to be a utility skill with a side of damage, it falls short on providing utility in general and the skill tree does not reflect this intent, particularly when compared to Swipe (mana gain, % multiplicative damage buff, % hp regen, culling strike, companion abilities, etc).
I don’t believe skills need to have a downside in order to be thematically fitting, enjoyable, and still have balanced gameplay. If a skill is going to be given a downside, the payoff needs to be worth it, but the downside should still not hinder how fun the skill is to use or the players ability to navigate the game effectively.
Example TS Build Footage
This build has 250% attack speed (penetration), 98% spell crit chance, 500% crit multiplier, 169 flat spell damage (low compared to a staff, but worth the stats gained in crit, AS, and multiplier). Gear is not minmaxed at all and damage could be improved with idols and other sources of additional % increased damage, but this level of offensive stats on most any other build would see incredible returns on investment.I tried playing for consistency with all proc chance, tested pure earth spikes since it can get more multiplicative damage, then settled on earth/lightning with charged stone feeling the overall best.
The build is also intentionally not dealing damage with anything other than Tempest Strike as an example of what it can do in isolation.
Video of low corruption Shade kills comparing setups: https://www.youtube.com/watch?v=wjMYa-CG1BE
I’d add that when duel wielding and using Tempest Strike, that swinging with each weapon hand one after the other is going to look and feel much better as far as core gameplay feel. Currently you just hit with both hands at the same time and it looks and feels janky.
).