Revisiting Allagast

TL; DR

Allagast’s Masterpiece is a solid concept, but there is room for significant improvements.

The main issues with the set are:

  • Modest defenses and lack of sustained healing, resulting in punishing gameplay, along with modest damage in return.
  • Awkward conversion on TSS, with loss of DoTs and lack of relevant items supporting an aether/lightning damage profile.
  • Lackluster skill support for the Arcanist mastery, to the point that the set works better without it. Allagast is having an identity crisis. :P

(In other words: the Arcanist half of the set doesn’t work nearly as well as the Inquisitor half, and on the whole, we find much glass and little cannon.)

Underlying my feedback here is a desire to bring Allagast up to speed in endgame content while strengthening its identity as an aether/lightning set based on Storm Box and Trozan’s Sky Shard. Accordingly and in no particular order, my minimum recommended changes are:

  • Increase point bonuses to the TSS skill line
  • Increase flat aether damage on TSS
  • Add flat electrocute to TSS (and change the global conversion from cold to lightning, preferably)
  • Replace Arcane Will bonuses with effective defensive options such as Mirror or Maiven’s
  • Grant 8-12% CDR as set bonus, not as modifier to Deadly Aim (but moving modifiers from DA to Maiven’s would be nicer)
  • Add some form of sustained healing or improve existing ones (a leeching tether as set proc would be lovely)

More comprehensive feedback can be found throughout the post. For convenience, I include here a summary of actionable recommendations and the reasoning behind them. A mod implementing most of these changes is also available for testing, and preliminary results are encouraging.

TSS (tested ✓)
The star of Allagast’s Masterpiece doesn’t shine. Change the conversion to retain DoTs, which are important to a kiting caster. Adding flat RR improves damage and moves a buff to Arcanist.

  • Retain 100% cold to aether in TSS, but make the global conversion from cold to lightning.
  • Add 200-450 flat electrocute.
  • Grant a few points to Frozen Core.
  • Increase flat aether damage bonus.
  • Add 15-20 flat RR for 3-4 seconds. Remove Deadly Aim modifiers.
    Note: this indirectly improves survivability by making Tree of Life more appealing in comparison to Elemental Storm and Ultos. Consider granting flat RR to scepter and damage as set bonus to avoid excessive stacking by dual-wielding (although that didn’t seem to be a problem during testing).

Global damage (tested ✓)
Kiting casters cherish DoTs for both damage and survivability, and DoTs are meant to be stacked. A global conversion to lightning retains some of Frozen Core’s damage, and makes cold devotions and skills more palatable. We should also bolster aether damage, since TSS items outside Allagast only support lightning.

  • Convert 66-75% of cold to lightning; optionally, convert the remaining to aether.
  • Add +100-200% aether damage as set bonus or to a Maiven’s modifier.
  • Add +30-100% longer electrocute duration as set bonus or to a Maiven’s modifier.

Defense (tested, save for health regeneration as an alternative)
We are overreliant on a yet-insufficient Bat for sustained healing and on delicate Aeon/WoR/Mirror control for burst healing and survival. Low armor and physical resistance are a bad match for AoM/FG content. Arcane Will is plainly ineffective. We should improve passive defenses to make Allagast less punishing compared to builds of similar if not far superior firepower, aiming for the following set attributes:

  • Remove Arcane Will bonuses. Grant 2-5 points to Maiven’s Sphere.
  • Add 8-16% physical resistance. Consider removing chaos resistance, which can easily be found in supporting items as well as Steel Resolve.
  • Grant 8-12% CDR.
  • Add a small healing proc with low CD (cf. new Clairvoyant’s Wand, Bonemonger’s set).
    My preference would be to have a set proc that grants one or two additional tethers with lifesteal but lower damage, which would fit the set thematically. The tethers can be casted through summons.
  • Alternatively, add meaningful health regeneration bonuses (but this would be a better fit for Trozan’s Skybreach, which could also use better defenses).

I believe physical resistance should be offered as set bonus, while CDR and regeneration (if using) can be divided between set bonuses and new modifiers for Maiven’s, leaning towards Maiven’s. See the Defense sections for details.

Replace Deadly Aim with Maiven’s Sphere (tested ✓)
Allagast supports Inquisitor very well, but at the expense of Arcanist. To even things out, we should scrap Deadly Aim modifiers and shift them to Maiven’s Sphere, a fundamental Arcanist skill with barely any modifier support. Grant two or more of the following to Maiven’s Sphere:

  • 12% CDR
  • +30-100% electrocute duration
  • +100-200% aether damage
  • High health regeneration (if using).

See the Maiven’s Sphere section for details.

Inquisitor primacy and identity crisis
The Inquisitor side of Allagast is dominant in all regards; the set works better without Arcanist. Consider:

Testing suggests that these combined recommendations would suffice to make Mage Hunter competitive:

  • Improve TSS damage.
  • Add flat RR to TSS to compensate for Arcanist’s lack of RR.
  • Scrap Deadly Aim modifiers, which only serve to make combinations other than Mage Hunter all the more effective.
  • Move CDR and +x% damage bonuses to Maiven’s.

Item support
TSS can only be hard-capped with cold/lightning items. A few simple updates would help bolster the aether component of Allagast’s TSS as well as the Storm Box transmuter, which is otherwise a damage loss.

  • Sprinkle +x% aether damage on TSS items.
  • Sprinkle +TSS bonuses on aether and aether/lightning items.
  • Adjust Allagast bonuses accordingly.

You can skim this post for list items to get a closer look at these suggestions and others. For a more thorough commentary, read on.

Build report

With minor variations, the Allagast build that I played in AoM (campaign, roguelike dungeons, crucible) and FG (crucible only) can be found here: https://www.grimtools.com/calc/lNkKl0qN
As you can see, it deliberately focusses on the main Allagast skill lines and is not specifically optimized for clearing Crucible. There are a fair few item and devotion variations which I didn’t have the chance to test – most notably the new Tree of Life – but I submit that mine is a thematically consistent take that should give a reasonable representation of Allagast’s capabilities. So, feedback given in this OP is primarily but not entirely based on my experience with this specific build.

I’ll start by saying that, after a few teething troubles, my experience with Allagast was remarkably less negative than elsewhere reported on this forum. The build is fun and satisfying to play, damage isn’t all that bad, and the introduction of mobility runes greatly improved survivability, as expected.

Survivability, however, remains the most glaring issue: sustained healing is the eternal plague of non-WD casters, and Allagast is no exception. Bat feels required: without Bat, even fights against plain hero mobs can be awkward and clunky. Allagast is the nth non-WD caster I play that turns out to be an obligate Bat parasite. :P (And an envious parasite at that, as Bat is far more effective in builds with high WD or good pierce/vitality damage or conversion.)

In all challenging content, gameplay requires rather more care than most builds, as it relies heavily on Aeon, Mirror, WoR, and nullification, which is to say user-controlled, time-sensitive active defenses.

This weakness is most noticeable in the Crucible, where damage is passable but survivability is subpar: in my test runs with 4 blessings + Vanguard, clearing rates depended heavily on which nemeses spawned, ranging from deaths on waves 160 and 170 against fast physical hitters or the odd trap/nuke combo, to good runs that hovered around 8:00 minutes of smooth or not-so-smooth kiting. I haven’t kept a test diary, but I seem to recall 5 failures out of two dozens or so runs since FG was released; not a terrible result, but not too good either, considering that many tanks can clear Crucible faster by reliably face-tanking most anything.

To expand on these results a bit, I’ll say that Allagast fares well when dispatching mobs or bosses, but becomes unreliable against physical damage parties with high movement speed or charge skills (i.e. pretty much the entirety of the AoM/FG nemesis lineup :P). Whenever these charming fellows come along you’ll generally find yourself traipsing and dancing until you mistime the Hourglass reset, or the casting queue eats it, or choose the wrong moment to cast an arpeggio of debuffs, at which point you are unceremoniously trampled. There’s precious little room for error.

In this regard, Allagast feels like so many other non-WD casters without lifesteal: much glass and little cannon, too often finding himself at low health with every defense offline and nothing to do but wait for the heavy hitters - and they do hit heavily, with caster armor and little physical resistance - to swat you.

I haven’t played SR yet and won’t be able to try it before a few weeks, but if map and battle design are akin to that Crucible when it comes to bosses and nemeses, there should be little reason to believe Allagast will fare better.

What works, concisely

To start with, here is what Allagast already gets right:

  • Gameplay is delightful. (It’s my favorite caster bar none.)
  • Storm Box just works. Perfect CD reduction, wonderful proc/devotion trigger, good damage and RR modifiers, abundant external item support.
  • 2 extra projectiles for TSS. Sweet grapes of wrath.
  • Good resistance coverage, both for damage and CC (stun, slow, trap).
  • Solid OA/DA.

In the “not wrong, but not right either” camp, we have:

  • Deadly Aim modifiers. While unexciting, the damage and duration modifiers don’t particularly cry out for revision. (The CDR bonus, on the other hand, is positively yowling for 100% uptime.) Even so, all global effects (RR and buffs) granted by Allagast fall on the Inquisitor side, a mastery that already has little trouble playing well with others. I strongly recommend scrapping Deadly Aim modifiers altogether and replacing them with support for TSS and Maiven’s.
  • Good RR, if a teensy bit skewed towards lightning due to the mandatory Censure pick and devotions like Elemental Storm and Ultos.
  • Some casting speed wouldn’t be amiss. I understand that the tether scales with it, and we have a fairly busy casting schedule besides.
  • Energy regeneration isn’t ideal for prolonged combat. Soft-capped Mental Alacrity would help, but it’s hard to find the points for it. Tree of Life would also work, but skipping Ultos has a noticeable impact.

What doesn’t work, verbosely

Now then, for advice actual, let’s pore over Allagast’s woes.

Awkward support for Arcanist skills

Allagast’s Masterpiece on its own grants precious few points to TSS. Quoting an image previously posted on this board:

Well, that looks more like a sketch than a masterpiece. :P The TSS skill line receives 2-0-4 points from Allagast’s, compared with 6-3-3 from Trozan’s set (including +1 to Arcanist) and 7-6-6 from the upcoming Iskandra’s (including +2 to Arcanist).

The obvious recommendations here would be:

  • Increase +TSS bonuses to at least 4-0-4, or 3-2-4 if adopting a cold to lightning conversion. (More on that in the Damage profile sections.)
  • Remove Arcane Will, offering points in Mirror or Maiven’s instead. (More on that in the Defense sections.)

Damage profile: lightning bias from supporting items

To hard-cap the relevant TSS nodes, it’s currently necessary to seek out skill bonuses from extra-Allagast items that support only cold/lightning, leading to a strong bias towards lightning damage. Accounting for higher and more reliable elemental RR, the damage distribution is skewed in favor of lightning by a factor of (ballpark) 3 or 4 to 1 in TSS alone – including Storm Box would significantly increase it.

I would personally favor a more even damage distribution between lightning and aether, depending on Storm Box transmutation and devotion choices. To accomplish this, I’d start by introducing a +x% aether damage bonus:

  • Add +100-200% aether damage as set bonus and/or as a modifier to Maiven’s

Beyond Allagast, consideration should be given to the absence of items that support both TSS and aether damage: TSS support is mostly tied down in cold/lightning sets, which doesn’t leave many compatible items worth mentioning. Introducing new items requires considerable work, and so we have two avenues for immediate improvement:

  • Add +x% aether damage to some TSS items. Notable candidates:
    • Rolderathis’ tome (a perfect fit for Allagast)
    • the Eastern Promise set
    • rings and amulets: Coven’s Sky Seal (also an excellent fit), Starfire, Starfury Emerald
  • Add TSS bonuses to some aether or aether/lightning items. Notable candidates:
    • rings and amulets: Glyph of the Storm Witch, Eternal Haunt, Screams of the Aether, the Aetherstorm set
    • pants: Arcane Currents, Wraithborne, perhaps Aleksander’s Legguards (he is, after all, Mr. Aether Meteor)
    • Wraithwalkers boots

Ideally, we should be able to hard-cap both TSS and Storm Box without acrobatics while retaining an aether damage bonus of around +2000%.

Damage profile: TSS, RR and Inquisitor primacy

To improve the damage profile of TSS specifically, I invite you to consider a straightforward buff that’s been applied to many other skills:

  • Introduce 10-20 flat RR for 3-4 seconds to TSS (c.f. Doom Bolt and Reap Spirit modifiers) on set completion.
  • Remove Deadly Aim modifiers.

Our aether damage would benefit from flat RR more than lightning, both directly and by making Elemental Storm less attractive. Having flat RR without Elemental Storm would also allow us to pursue Tree of Life without unpalatable sacrifices, improving survability.

To compensate for the additional RR, I would strongly recommend removing the hefty but lightning-specific Deadly Aim modifiers. (In fact, I recommend scrapping the Deadly Aim modifiers altogether, as previously mentioned.) Shifting some damage boosters away from Inquisitor and compensating for Arcanist’s lack of native RR makes Allagast stand more clearly as a dual-mastery set rather than yet another stormbox proc-all, thus strengthening its identity. As an example of this problem, consider things like John_Smith’s BoxMan or thejabrixone’s Allagast vindicator, or a simple jury-rigged Allagast purifier, which are currently more effective than by-the-book Allagast, and distressingly so. (I don’t mention these builds because I’m out for blood or crying nerf, but to make the point that the set is not working out as intended. Flexibility is good, but for Allagast to live up to its promise, the mage hunter version should at least meet a baseline of performance that approaches that of other masteries.)

Damage profile: lopsided conversion and loss of DoTs

The foremost offensive deficiency of Allagast is the decision to convert cold to aether without addressing the loss of DoTs.

I must emphasize the importance of DoTs in a fragile kiting build with little to no lifesteal. Storm Box – while hitting over time by itself – is the only source of electrocute in Allagast, and most such electrocute is applied by the tether, a secondary effect that can be hard to target and control in the chaotic conditions of endgame content.

Adding a second source of electrocute for stacking would be highly desirable, and TSS is the ideal choice. It would suffice to introduce a simple electrocute modifier and shift our focus to the global conversion.

As it stands, it should be easy to see that changing 45% of cold into aether, while thematically appropriate, is of dubious benefit; moreover, there are no items or skills to increase its value (the only existing one, Dreeg Veilweaver Tome, is not up to par). What skills would work at half conversion and without DoTs? None of the Mage Hunter’s. What devotions? Amatok’s Blizzard would hardly be worth 4 devotion points, Tsunami is a tough sell compared to Sailor’s Guide or Eel, and Attack Seru is a bit of a reach given the necessity for Aeon’s Hourglass or Tree of Life.

The immediate remedy would be to globally convert cold to lightning, which leads us to a more comprehensive solution:

  • In TSS, retain the 100% cold to aether conversion.
  • Add 200-450 electrocute damage to TSS, depending on global conversion values.
  • Globally, convert 50-100% of cold to lightning; optionally, convert the remaining to aether as a cherry on top. These conversions can be divided between set pieces and set bonuses.
  • Grant a few points to Frozen Core.
  • Adjust conversion values and flat damage modifiers according to the desired aether/lightning/electrocute distribution.

Defense, or lack thereof: Arcane Will? Arcane Will.

Now, this here is the most egregious offender and then some. Of all defensive skills in the Mage Hunter tree, Allagast goes and grants a staggering +7 to Arcane Will. Just so.

Well.

… Well, let’s take a look at the results. Raising Arcane Will from a typical 3-4 one-pointer to 10-11 grants an extra 100 DA, +100% to all damage, and 7 energy/second. Arcane Will can reach 100% uptime, provided you remain below 70% health, which is about 7-8k for most Mage Hunters.

Well, yeesh. I know of no other set or group of items that pushes a circuit breaker – to use the term generously :P – so hard to so little effect. It’s whimsy, especially in a fragile caster set. Let’s just clean this up and forget about it, shall we?

  • Purge Arcane Will. Redistribute points among more qualified candidates: Censure, Mirror, Maiven’s/Conversion, WoR/Vigor. Anything really. :P

Jokes aside, I understand the desire to try something different. However, Allagast is missing the basics, and those need to be covered first. Players who wish to invest in Arcane Will can always do so, and Arcane Will could perhaps find support in an item or a set with better baseline defenses. Allagast just isn’t that set.

Defense: sustained healing

Even kiting casters need more than spaced-out bursts of healing, especially in an endgame where auras, area effects and bullet storms abound. Lifesteal has long dominated this space, but Allagast has no weapon damage and no leeching skill. Moreover, healing skills often restore modest amounts of health. Consider a Mage Hunter with WoR: even though 30% health restored every 10 seconds may sound like enough, the results come in at 3-4k hp. Respectable, but there’s more than a few enemies that deal that much damage in a single hit.

That leaves us with Bat, an ever-present fallback that works wonderfully against mobs. Unfortunately, low weapon damage renders Bat marginal against low numbers of powerful enemies with high leech resistance, where Allagast needs it most. I am very much against buffing Bat, or adding pierce/vitality conversions to improve its yield. Regardless, it is good stick to measure by: any adopted solution should yield an amount of healing that either compounds Bat or, in the interest of moving beyond its dominance, subsume it.

The purpose of the following proposals is not to guarantee survivability on their own, but to provide a baseline amount of healing that can be complemented by devotions and items – currently, we have no option but to take defensive devotion routes on a build that already lacks damage. I submit two options for set completion, starting with my first choice: a small on-attack healing proc with low CD.

What I would personally love to see is a set proc based on the Lightning Tether template: a “corrupted tether” with lower damage and lifesteal. It would be fun to have lifesteal that rewards positional gameplay and movement rather than the usual “be a good leech and stick to your enemy” approach. :P

Unfortunately, I understand that tethers can’t be granted as item skills. In that case, the tethers could be casted by a summon instead. For example (beware, napkin math ahead):

  • 20% chance to summon wisp on attack, maximum of 2, 8 seconds duration. Each wisp casts one corrupted tether:
    • 60 aether damage, 40 lightning damage
    • 8% lifesteal

I’d roughly aim to offer half the healing provided by the Bonemonger’s “Wake of Bones” skill – or more, if we want to go beyond Bat. The extra damage wouldn’t come amiss, either. (These values assume that tethers scale with casting speed, ranging from 3 ticks/second at 100% speed to 6 at 200%.)

Defense: overall fragility, resistances, and CDR

Allagast offers solid resistance coverage, with clear attention given to traditional Arcanist weaknesses such as pierce, poison, and bleeding. My only reprieve here is that the 3-piece chaos resisistance bonus feels unnecessary: not only is it given natively by Steel Resolve, but it can also be found in a number of obvious item pairings (notably Trozan’s hat, the Aetherbolt amulet, the Arcanoweave belt, and the Gildam Arcanum medal).

That 3-piece resistance bonus could be put to better use in the one area where Allagast, like many other casters, is found wanting: armor and physical resistance. For the latter, I feel that the new Iskandra – which gives out 15% – has the right idea: there’s no need to be stingy. All low-armor builds greatly benefit from physical resistance, so just give some.

I might be wrong here, but AoM and FG brought a significant increase in enemies that deal heavy or very heavy physical damage, and the buffs to armor values for casters weren’t enough to compensate for it. Of course, non-channelling casters like Allagast don’t need to stand in melee range, but consider that the very same enemies who can run them down in three hits also boast charging skills or unmatched movement speed (a remarkable change from vanilla); much endgame content is also tied to small arenas where multiple such opponents fight you at once, applying CC or constant damage on top of their nukes. Kiting consistently in such conditions is hard, and you will be hit, often for a third of your health or more.

Ultimately, gameplay feels a bit too unforgiving when compared to sturdier builds, which often end up dealing more damage as well. This leads us to CDR and the necessity of sustained healing.

Currently, a precious 10% CDR is granted only as modifer to Deadly Aim, an on-attack skill with uptime around 60%, provided that you are consistently landing crits. Even though Storm Box is an excellent trigger for it, Deadly Aim works better for offensive bonuses that let us mount our advantage, rather than for defensive ones that are needed to survive: those call for consistency. Consider that when things go south, something as fragile as Allagast must be able to run without losing on defense: it can’t always afford to kite or sneak in a spell here and there, because there’s a high chance a fast enemy will swoop in and wallop him for 5k of damage.

So, when it comes to life-savers, I’d recommend keeping things simple:

  • Grant 10% CDR plainly as a 3 or 4-piece set bonus.
  • Add 8-15% physical resistance.

Alternatively, I’d recommend moving either or both of these bonuses to something with 100% uptime (in line with Star Pact), which brings us to another major proposal.

Maiven’s Sphere and Inquisitor primacy

Inquisitor doesn’t need Allagast for modifiers. Inquisitor already features a grand variety of buffs and debuffs and enjoys great item support for them. With Allagast, we should instead take the opportunity to give some much-needed attention to the Arcanist side of Mage Hunter; without Arcanist modifiers, the set works better for vindicators, apostate, and purifiers.

Maiven’s Sphere of Protection is an Arcanist skill that has seen few to no modifiers despite being a pillar of the mastery: for all we dislike the hefty damage loss, high damage absorption is pretty much a must-have when you can only gain a total of 900hp from the mastery itself, and CC resistance is similarly needed.

So then, I warmly recommend the removal of Deadly Aim and the addition of Maiven’s Sphere from the set completion bonus of Allagast’s Masterpiece. I’d consider keeping physical resistance as a set bonus, and bestowing some powerful boons upon Maiven’s:

  • 10-12% CDR, compensating for the loss of Star Pact.
  • +150% aether damage, mitigating the need for aether TSS support outside Allagast (although some would still be warranted).
  • +150% electrocute damage.
  • +30-75% electrocute duration.

On a different note, I’ll add that an empowered Maiven’s Sphere would also be more thematically appropriate than the rather uninspired boost to Deadly Aim: here stands Allagast, master of the arcane, his breath and touch imbued with runic powers that mortal men shouldn’t wot of; the very air blazes as lightning coils around him and the skies are cloven asunder by his emerald wrath. Or some such. :P

This second post follows the state of Allagast across GD releases. All new releases are tested in-game before comment.

Patch Notes

Here is the lost and found for Allagast, based on the build in the OP (which should be pretty typical, insofar as these changes are concerned):

Version 1.1.2.0

  • Maiven’s Sphere: 1-3% less damage penalty, 2-3% more absorption
    A slight buff in our range for Maiven’s, albeit not substantial enough to change the gameplay experience on its own. Allagast remains too fragile and not very powerful.

  • Maiven’s Sphere: 20% trap resistance
    A very welcome buff that unfortunately doesn’t affect Allagast much: we already had it nigh-capped from Vigor and Hourglass, which are more or less required. Even so, it does introduce some potential wiggle room.

  • Aura of Censure: lost 5% elemental RR
    A well thought-out nerf, which unfortunately impacts the already not-so-powerful Allagast. Mitigated by the changes to Maiven’s.

  • Arcane Will: 55 extra DA
    The buff to Arcane Will, while welcome, does not alter my recommendation: AW remains unsuitable for heavy piece or set bonuses to Allagast. I still feel that no more than 2 points should be granted to it, and the remaining distributed to skills of global utility. I suspect AW is overall more beneficial to battlemages (if you can find 'em :P), high-lifesteal Spellbreakers or PRM/AAR, and the like, who can better afford a hit or two at lower health due to their much faster health replenishment rates.

  • SFX: aether-green TSS
    Yes, it’s strictly cosmetic, but I have to say TSS looks pretty in green. :P

Version 1.1.2.2

  • TSS: 5% higher damage
    A meaningful buff that nudges the impact of TSS. It doesn’t alter the general status of TSS (which I feel remains underpowered), but is a clear step in the right direction.

Other relevant threads

Any revision to Allagast should take into consideration potential changes to Arcanist skills or the Trozan’s Skybreach set. Here are threads which overlap with this one:

Discarded from OP

TSS: double conversion

On TSS alone, shift conversions to retain DoTs, making aether spike rather than lightning.

  • Convert 100% lightning to aether
  • Convert 100% cold to lightning

Discarded because it didn’t work out in testing. We can’t hard-cap TSS and find enough +x% aether damage for the aether component to be worth its weight, at least not without heavy revisions to supporting items.

This was an interesting and detailed read. I’d gave you PhD in Allagast advanced technic, if I could :slight_smile:

I agree on almost everything that has been developed. I think this set is the typical example of how the meta/state of the game has evolved into a very pure defensive approach where damage mitigation is (too much?) important. We also see the critical place of adcth nowadays and casters without %WD, as you stated, suffer from this greatly. And of course, the lack of physical resistance/armor, that I understand from a lore point of view, that forces many casters to go for Valor Pants + Final March boots each time.

The new Iskandra (there are no polished build yet) feels the same as when I tried Allagast for the first time: some nice damage output, some potential but the inherent squishiness and lack of passive defence hurts. It only hinders build diversity for those sets because defence becomes the top priority via devotion paths especially (the “bat” problem you mentioned).

The weird conversion and absence of DoT synergy (this has been addressed partially in 1.1.1.0 (?), the patch before FG) furthermore increases the difficulty for casual players to grasp the power of the set imo. This is a double damage set and we know how good double damage is in GD …

You make a good point regarding skill points and chaos resistance bonus.

I really love the gameplay of Allagast: Storm Box + TSS, compared to Iskandra where you don’t know what to do between two TSS. It’s quite interesting to notice that the only Allagast build published yet is an Apostate relying on pet procs …

Thanks for the detailed and interesting thread.

Belzzzz, thanks for the kind words. I agree on Allagast gameplay: Storm Box + TSS is a match made in caster heaven, especially with Rolderathis’ tome in hand. I enjoy it greatly, and that’s why I started this thread.

Apropos, I updated my feedback for version 1.1.2.2, which saw a buff to TSS damage, and also took the liberty of developing a small mod to test potential changes to Allagast.

Allagast test mod

My recommended changes can now be tested through the Allagast’s Crayons mod. The mod introduces a revised version of Allagast’s Masterpiece, called Allagast’s Doodled-Over Masterpiece. In the current release, the difference between base Allagast and the doodled-over Allagast are:

  • 3-piece bonus:
    • Removed 28% chaos resistance
    • Added 10% physical resistance
    • Added +80% aether damage
  • 4-piece bonus:
    • Removed 15-80 lightning damage to TSS
    • Added 16 RR for 3 seconds to TSS
    • Removed Deadly Aim modifiers
    • Added Maiven’s modifiers: 10% CDR, +150% aether damage, +150% electrocute with 50% longer duration
  • TSS modifier from the scepter:
    • Increased aether damage to 160, from 120
    • Added 300 electrocute over 3 seconds
  • Skill bonuses:
    • Removed +7 Arcane Will, removed +2 to Deadly Aim
    • Added +4 to Maiven’s Sphere: 2 on chest, 2 on shoulders
    • Added +2 to Frozen Core: 2 on shoulder
    • Increased Trozan’s Sky Shard to +3, from 2: 3 for set completion
  • Global conversion:
    • Removed 45% cold to aether
    • Added 65% cold to lightning (45% from the scepter, 20% from the shoulders)
  • (Optional, separate mod) Added a summon with a leeching tether, 20% on attack, maximum of 2, 6 seconds duration:
    • 60 aether damage
    • 40 lightning damage
    • 60 electrocute damage over 2 seconds
    • 8% ADCtH

These changes are substantial, and testers are very welcome to try them out.

Installation and usage

Main Campaign:Allagast’s Crayons v2.4.zip (24.4 KB) or Allagast’s Crayons v2.4 (with healing source) (24.4 KB)
Crucible: Crayons Crucible v2.4 (280.8 KB) or Crayons Crucible v2.4 (with healing source) (281.4 KB)

In a modded game, all blacksmiths can craft Allagast’s Doodled-Over Masterpiece, the modified version of Allagast’s Masterpiece, for 1 scrap and 1 iron a piece.

To test the set in campaign, extract AllagastCrayons.zip to the mods folder in the Grim Dawn installation directory; the mod will then be available as a custom game. To use an existing character in custom games, copy its folder from save/main to save/user and mark it as “In Mod” with GD stash.

To test the set in Crucible, go to the Grim Dawn installation directory and backup the survivalmode2 folder. Extract CrayonsCrucible.zip in the folder, replacing database/survivalmode2.arz and resources/text_en.arzwith the modded versions. Existing characters can play the modded Crucible normally; note, however, that they can’t bring the modded items back into the main, unmodded campaign. I recommend cloning a character and using it in Crucible only. (Addendum: this archive also includes a revised Iskandra set I’ve been working on, to spare me the effort of maintaining two separate Crucible mods. Feel free to ignore it.)

Changelog

v2.4: Moved optional healing source to a separate version. Minor parameter tweaks.
v2.3: on TSS, decreased RR duration from 4 to 3 seconds.
v2.2: on TSS, decreased flat electrocute bonus from 420 to 300.
v2.0: included a source of sustained healing: leeching summons with tethers.

Preliminary test results

Maiven’s and physical resistance

Predictably, offering points in Maiven’s and introducing some physical resistance makes Allagast less prone to squashing, although it remains far from sturdy – and that’s fine. The lack of Arcane Will bonuses is unnoticeable in practice. Some form of sustained healing would still be welcome, otherwise players will be forced into defensive devotions routes from the get-go.

Sustained healing from summoned wisps (optional extra)

The experimental wisps spawn on attack, cast a single leeching tether while trying to maintain their distance, and expire after about 7 seconds. Their yield is reasonable, albeit situational.

Starting at around 8% lifesteal from 100 aether/lightning damage, the wisps can effectively replace Bat in campaign content and in Crucible to a lesser extent, with some tradeoffs: the double tether is more effective against powerful individual targets and works much better at a distance, while Bat remains unmatched at close range and for large bursts against mobs. (My impression is that Bat still heals rather more overall, although not necessarily when you need it to.)

Combined with flat RR from TSS, persistent healing does a decent job at opening up alternative devotion routes, both offensive and defensive: it’s easier to consider things like Imp + Amatok + Spear + Ultos for procs galore, or Crown + Ultos for maximum RR, or Giant’s Blood + Hourglass + Tree of Life + extra physical resistance for resilience, and so on. There isn’t as much flexibility as I’d hoped due to Spear and Hourglass (and possibly Bat itself) remaining too good to ignore, but I’m still pretty happy with the result.

TSS

Including DoTs, the revised Allagast has 15-20% higher TSS damage, bringing the spell into nuke territory. Although the modifier values may look high, remember that we are working with an underperforming set without external item support, and that even with an additional source of healing we remain pretty fragile.The electrocute modifier for TSS can be adjusted according to the global conversion from cold to lightning; at 65%, 300 additional electrocute works well.

Although TSS now feels very good to use, I am not all that satisfied by the damage profile: the addition of effective electrocute, while warranted, makes the set even more biased towards lightning. I suspect that adding +x% aether bonuses in supporting items remains warranted, even if we decreased electrocute and further increased flat aether.

Tested content

  • Crucible: 7:00 with 3/4b+1vb, down from 7:50 with unrevised Allagast. Increased TSS damage and flat RR make a difference. The build is sturdier and the wisps work well enough in the brawl-in-a-bowl that is Crucible; you can still be bursted down, but gameplay isn’t nearly as punishing.
  • Lokarr: 35-40 seconds, down from a wobbly 50-55. The improved defenses feel very noticeable against Lokarr, although you still can’t facetank him (not that we’d want that).
  • Mad Queen: 13 seconds, down from 15-17 or so. (Not a very meaningful variance, since damage spread alone can make up for it.)

The times suggest that the revised Allagast Mage Hunter is now competitive with Apostate, Purifier and Vindicator.

I regrettably can’t play SR yet. If anyone feels like taking the revised Allagast for a spin, I’d welcome their contribution.

You could perhaps copy the conversion on the epic sets Daega and Festerblaze. They have two pieces (head, shoulder) with 20-30% conversion and another 50% on the set bonus for all parts. It would neatly convert the cold damage while not making individual pieces overly strong.

Yes, that’s part of the original recommendations (“convert 50-100% of cold to lightning […] these conversions can be divided between set pieces and set bonuses”). Mind that moving conversions to set bonuses would merely push the issue around, as Allagast completion bonuses are already on the stacked side; it’s hard to avoid, unfortunately, if we want the Arcanist side to work well.

Maybe we could move the +80% aether damage to 2 pieces and increase conversion values at 3. Having a 100% global conversion might also allow us to skip the electrocute modifier for TSS, although that would make the build even more point-hungry. Regardless, it’s mostly a matter of minor adjustment.

The test mod now includes an experimental source of healing, starting with version 2: up to two on-attack summons that cast leeching tethers. They work well enough in campaign and Crucible, but only provide baseline healing; anything other than the most reckless setups will still want another source of sustenance (or at least Aeon’s Hourglass).

On the subject, the modded Allagast has now been tested in Crucible as well as campaign, and it does very well; in fact it might be wise to tone down some of the changes a bit. The Preliminary test results from my third post have been updated accordingly.

A few notes on the leeching summons as tested.

Tethering wisps, or: the Aethercoil

For testing, the Doodled-Over Allagast mod offers wisps that each cast a single leeching tether:
Allagast’s Aethercoil
20% chance on attack, 2 summon limit, 8s duration, 2.5s cooldown

Pet skill: Corrupted Tether, 6s maximum duration

  • 60 aether damage
  • 40 lightning damage
  • 60 electrocute damage over 2 seconds
  • 8% ADCtH

(The test wisps use the Elemental Seeker mesh for visibility reasons, but ideally they’d be aether-themed.)

Why summons?

First off, why resort to summons for healing? They would not be my preferred solution, but they offer some advantages… foremost of which, actually being implementable. :P

The reason I would like to use leeching tethers is that tethers are the most distinctive feature of Storm Box, which makes them a good fit for Allagast thematically. Other templates would certainly work just as well, much like a plain “health restored” proc, but they wouldn’t be as interesting.

I also feel that any granted skill should be automatically casted, as Allagast builds already have nearly a skillbar’s worth of active skills (see the test build), and adding more would be burdensome to say the least.

With that in mind, my original order of preference for a healing proc was:

  1. Lifesteal from Lightning Tethers themselves. Unfortunately impossible, as item modifiers can’t be applied to skill modifiers.
    This would have been a great fit for Allagast playstyle, as well as representing a different take on lifesteal; one that rewarded targeting, positioning and movement for a change. (I’ll admit, however, that balancing would be finicky: in my brief testing, going from 1% to 3% meant going from “just alright” to “eep too strong”.)
  2. Leeching tethers on autocast. Also not possible, it would seem.
    Losing control of targeting makes this solution less interesting tactically, but it would still be nice, as well as easier to balance.
  3. Summons with leeching tethers.
    They work well enough in testing and add a bit of color. Easy enough to tune. Reasonable fit thematically.
  4. Plain “n% health restored every x seconds”

With the first two options being off the table, let’s consider tethering summons.

Summoning mode

The wisps would work best as a permanent invulnerable pet à la blade spirit, but I’ve never seen permanent invincible pets granted by items. So, short-lived wisps it is, which is not ideal: tethers are broken when a wisp is replaced, and I couldn’t find a saturating autosummon controller, i.e. one that would only summon a new wisp if there’s less than their maximum number alive.

Adjusting wisp count, lifetime and cooldown is a simple way to mitigate the issue. In testing, 2 wisps with a lifetime of 8 seconds (with 6 seconds of tether time) worked well on a 2.5 seconds cooldown.

Pet behavior and parameters

I settled on mobile summons because I felt that they were a better fit for Allagast’s kiting style, as well as being a bit more fun to play with. Tethering totems could work – and they’d give Kaisan a taste of his own medicine :P – but the soft requirement for autocast would prevent us from controlling their position, which might prove troublesome; even so, they might be worth testing, perhaps by spawning them on the player’s location.

With wisps, I originally intended to retain the ability to control tethers indirectly by having pets maintain a near-fixed position relative to the player. Unfortunately, the available pet controllers don’t seem to allow for that. Similarly, pet AI has no notion of how tethers work, which means that the wisps can’t use positioning and movement to maximize their effect.

So the wisps merrily wobble and roam their way around the battlefield like unwary puppies on an electrified leash. The results aren’t half bad, although tinkering is probably needed to find the right values for leash length, aggression, and so on – something I’ll happily leave to people who actually know how pet AI works in Grim Dawn. :P One thing to note is that wisps work best if they’re allowed to flee and move away from targets, thus extending their tethers; they seem to snuggle up otherwise.

Tether parameters for pets

Although AI limitations factor in my preference for multiple weaker wisps over a single powerful one (more tethers mean a better chance of randomly landing one across multiple enemies), the primary reason for it is consistency and ease of balance.

Two wisps with one tether can each target the same enemy, which means that their yield doesn’t vary quite as much between mobs and individual foes. (Two wisps with two tethers each are mighty fun, but start feeling a bit gratuitous. :P) From there, it’s easier to find the desired baseline without worrying about bursts getting out of hand, which is a problem with Bat. Tethers do burst against mobs, but less than you might expect, since monsters generally cross them for a few ticks’ worth and no more.

Closing notes

This post more or less concludes my active contribution to a potential Allagast revision.

Tuning is always warranted, alternatives can be explored, other ideas considered, and so on, but on the whole, I reckon that these changes represent a significant and perceptible improvement to Allagast as a set based on Storm Box and TSS. They feel good in testing and play very well. Whether they fit the designers’ concept for Allagast is of course another matter.

Also of course, it’s perfectly possible that I missed something and my changes are totally broken and OP. So, if one or more of our valiant Praetorians enjoys Allagast gameplay and feels like taking the test mod for a spin, I’ll personally stand them a virtual drink of their choice. :P

Update: the revised Allagast can now be tested in Crucible with a nearly plug-and-play mod, to be found here. Detailed installation instructions have been added to the dedicated post.

These suggestions all look very reasonable to me. Summonable tether-wisps also look great and very fitting thematically, nice work! Hoping that you’ll be in the business of modding after Crate stops updating the game. :slight_smile:

A few things were broken by the migration to Discourse, notably internal forum links, attachments, and some formatting. They should be fixed now.

My finalized recommendations can be found in the test mod or the TLDR in the OP, skipping the posts themselves.