Introduction
(this section will repeat between my build overviews; you can skip to Build Concept if you’ve read one of my overviews before and are familiar with my rating system and test criteria)
Patch 9.8 brought, among other things, major changes to the Shattered Realm. Given that SR is where I do my endgame build testing, with the release of the patch I decided to redo all the testing for all my builds. And since I’d be putting in the time and noting down all my results anyway, I thought I might as well post the builds and their scores publicly for others. Who knows, maybe you’ve been thinking of trying something similar and are curious if it even works. Or maybe you’re just morbidly curious and like that tickly feeling in your brainhole when you see someone play something utterly stupid.
Now, what exactly does this testing entail? Each character has to do 10 SR75-80 runs (I used to do 5 runs but I’m expanding it to 10 for this second wave). Every run completed within timer counts as a success. It takes 6 successes (over 50 % success rate) for me to consider a build complete. If a build keeps failing to meet that quota, I keep improving it till it works. Ergo, no build I post here will have a lower score than 6/10, because if it doesn’t have over 50 % success rate, I’m not done improving it and it doesn’t get posted.
Three rules were followed to make the results more representative:
- no consumables other than healing and energy elixirs can be used;
- no shrines can be taken; if a build has bad resists, it’ll have to make do with bad resists, Rattosh isn’t gonna be saving its ass;
- no mutator hunting; if I get shitty mutators, I’ll have to succeed with shitty mutators.
Other than that, it doesn’t matter in what manner the runs are completed. If a build takes twice as long to finish a run than others do, that’s perfectly fine as long as it successfully finishes. Doesn’t matter whether the build has to dodge Nemeses in shards like the plague or whether it hunts them for sport, doesn’t matter if it has to kite like crazy or stands its ground with ease, doesn’t matter if it completes the runs deathless or not. Only thing that affects the score is completion within timer. Generally, a weaker build that is slow, fragile, hard to play etc. will end up with a lower score regardless because its shortcomings will affect its reliability across the 10 runs.
Now, two final things before we get to the build itself:
-
As mentioned in the title, this is a build overview, not a build guide. What’s the difference? Well, a build guide says “this is the way you should build this type of character”. What I’m saying is rather “this is what I did, here’s why I did it this way, here’s how it turned out”. It’s more of a documentary, a post-mortem, rather than an example to be followed. “Wait, that’s just a fancy way of saying you make shit builds,” you might say. No, not exactly. I try to make a strong character without deviating from the concept of the build. But the concept of the build might not necessarily be something one should even be doing in the first place
While my build concept and the in-game support will often align to create a reasonably predictable, almost cookie cutter build, just as often I’ll just be doing something abundantly demented for my own reasons. I’ll always endeavour to explain in the Build Concept section why I chose to do a thing a certain way, but I don’t want you to get the impression that what I’m presenting is a thing you should necessarily be doing. -
I’m also by no means hyperfocused on optimising the crap out of a build once it’s in a workable state. There will almost always be things to optimise on my characters, but frankly, I prefer spending an hour theorycrafting a new character rather than shuffling an existing character’s devo tree, gear, component and augment setup just to squeeze 5 % more damage out of it. I don’t care that much.
With that out of the way, let’s get to the build itself.
Build Concept
Aight, it’s writey-writey time again. This build was an attempt to combine together two loosely related concepts I’ve been a fan of for a long time but have rarely seen implemented in RPGs.
Ever since my first ARPG, Diablo 1, I’ve had a soft spot for archer characters. The Rogue was my favourite class in D1, as she seemed like a nice middle ground between the OPness of the Sorcerer and the bundle of sheer suffering that was the Warrior. But as time went on and ARPGs have tried to close the gap between melee and ranged, shooter characters kinda lost their lustre in my eyes. When you’re playing an ARPG or even a cRPG and you’re presented with a range of characters, what are you going to choose? A Warrior/Barbarian so physically strong they can wrestle giants to the ground and their blows create seismic tremors across the screen? A Druid/Wizard able to conjure devastating hurricanes or warp reality itself? Or some dude with a bent stick tied together with string, that plucks away pointy pieces of wood? However well the game might be balanced it just doesn’t feel like archers are operating at a comparable level of power to the others. The approach that renewed my interest in archer characters was when I made an Arcane Archer in Pillars of Eternity 2 (a cRPG loosely based on DnD). The fantasy of a mage who uses magic to enhance ranged combat appeals to me a lot. Yet very few games actually give you the option to make an archer feel truly wizard-like.
The second concept was that of an enchanter. Whenever I hear the term Battle Mage tossed around in games I always have to sigh over what could have been if people were a bit more creative. Because most of the time a Battle Mage is just a normal Mage. Do they use their magic in battle? Well, duh. They can cast Fireballs, what else are they gonna do with them? Gardening? All Mages are pretty much Battle Mages. How cool would it be if a Battle Mage was actually a melee combatant. but instead of using physical prowess and training with all sorts of weapons to overcome foes, they would instead utilise mundane martial techniques but enchant themselves before combat to enhance their physical capabilities (speed, strength, resilience) or conjure whatever weaponry they need? They enter an arena with just a plain shirt and pants and with just a few incantations, BAM! They’re now wearing armor that enhances their speed, carrying a shield that reflects arrows and wielding a sword that burns so hot it melts through steel as it swings. Now that’s a Battle Mage!
In GD terms, I wasn’t necessarily married to the idea that the character has to bear the class name of Battle Mage but it wound up that way anyway, because the core flavour concepts were captured pretty damn well by the combination. You don’t get more martial combat-y than the Soldier and the transmuter on Cadence really gives off that feel of an enchanted Warrior. Plus Soldier is not exactly the mastery you’d think of when it comes to Tri-Elemental so the transmuter was very intriguing from the theorycrafting perspective as well. With that Tri-Elemental foundation set, Arcanist seemed like an easy choice for its enchanting capabilities. Iskandra’s Elemental Exchange in particular really felt like the ideal realisation of that Enchanter concept, and Mirror and Sphere definitely didn’t hurt either. On the Arcane Archer side, Elemental Cadence also fit the bill, and from the practical perspective the built-in passthrough would ensure the build’s most powerful attack would come bundled with excellent AoE coverage.
Setup
Having a neat flavour concept is one thing, but having it make sense in the realities of the game is a whole nother kettle of fish. The core issue is this: by combining Soldier and Arcanist for the concept, the build is entirely locked out of all sources of on-tree resistance reduction (RR). In addition, there is no source of -% RR available for Tri-Elemental on rings or on the devo tree. So all the build would be left with is unstackable flat RR from stuff like Elemental Storm and then similarly unstackable % RR, like that found on Viper. Now, when I was starting this character, I was aware of this problem, but I was undeterred. The theory was that since I was technically dealing 3 types of damage, enemies might only resist one truly heavily and the majority of the rest would go through, so the lack of RR wouldn’t be such a problem. In addition, I would be compensating for the lack of RR by having more raw damage, as I had a 10 % TDM plus extra flat on Cadence from the Discord transmuter and I would have a fully juiced Iskandra’s Elemental Exchange (IEE) contributing its fair share of oomph to my attacks. Lastly, since there was no -%RR preceding it, the % RR from Viper might also end up doing some work for once in its life and improve that situation.
Regardless, the top priority was to get the build’s damage up as much as possible. Overcome lack of RR with sheer flat. Unfortunately, there is very little gear support for Tri-Elemental cadence specifically and even IEE doesn’t really have that much gear for it. As a result, I had to resort to mainly just stacking additional flat wherever I could either through points to the Cadence and IEE lines, or through pure on-gear flat, cause useful item skill modifiers were in short supply. The items I was set on from the very beginning were: Dawnshard Gaze and Dawnshard gloves for the flat. Living rings for the IEE points, as that would provide more flat damage than the only Tri-Elemental flat damage ring, Malmouth Arcane Seal, and I could roll them with more attack speed too. Shrould of Illusion and Leggings of Arcane Currents because of the Tri-Elemental Battlemage support. Boots of Primordial Rage for Cadence points. Ugdenbog Arcaneweaver as the only remotely viable Tri-Elemental ranged weapon, which fortunately also happens to support IEE, though we get no juice for Cadence. With so much flat damage from the skills and gear, I should be able to overcome enemy resistances and make my dream come true.
That was the theory. The practice was very, very different. This is the only build that I have ever been genuinely so ready to give up on that I had to post a thread about it to see if anybody can come up with something to save this thing. Because elemental resistances were everywhere. And not just, like, supercharged Fire or supercharged Cold. Supercharged everything. If you met a human hero it would be a miracle if they weren’t getting at least 60 % elemental resist just from their gear. It wasn’t unheard of for them to sport something more like 160 % elemental resist…overcap. In fact even that is fairly tame, I’ve met enemies with over 240 % elemental res overcap on just their gear more than once. As resist sources have been gradually shifting from single-elemental to tri-elemental for the player’s benefit, enemies were taking advantage of that very same change. The build’s damage might have been good in theory but in practice so little of it was getting through so often, it wasn’t really making up for the lacking RR the way I was hoping. Signs of issues didn’t really appear in the campaign, except for those hyper-resistant heroes, but for endgame it was a damn disaster. I could make 20 attempts at SR75-80 and make it to SR80 maybe once if I was lucky. The thing that wound up saving the character from deletion was, one, encouragement from folks like @tqFan and @Bananow9393 in the aforelinked thread. And two, Fleshwarped Casque. The Dawnshard Gaze I was using for the flat simply wasn’t doing the job and with the setup the way it was the build had survivability gaps due to War Cry downtime. And if your damage is poor, you’re gonna have trouble sustaining because you’re leeching off of so little. If you’re sustaining poorly, having tankiness gaps in addition to that was just a bad time. Fleshwarped Casque improved that dramatically, even though it leans towards Physical Battlemage support and not elemental. From the build saver thread also came the suggestion to use Haunted Steel as the weapon component instead of Seal of Resonance for flat elemental damage and phys-to-ele conversion. That inclusion then improved the lacking sustain and added an extra Oh Shit button in the form of Bloodthirster. With that, I finally managed to get the build to work in 9.6 to a success rate of 3/5. For this second round of testing, I made some further adjustments to components and attributes, landing on this:
1.2 UPDATE: Relinked the 9.8 version of the build, as the 1.2 version of GT removed the Magi’s prefixes from the helm and amulet and left them blank. In 1.2 the build’s lot hasn’t really improved. No extra skill points, defenses or damage gained, no improvements to the gear. The addition of the evade button makes it easier to line up Cadence passthrough. The leech nerf hurt the sustain, enough so that I decided to switch to Mark of the Dreadblade on medal to get more going. That incidentally also improves DPS, as even though the % damage and flat damage are all wrong, the chest piece converts a bit of it to Elemental and the total speed helps cap attack speed on the build. Going above SR80 still not advised. The build’s main strength is survivability, higher shards take that one strength away, leaving just a slow build behind. Fingers crossed that some Epic item upgrades in FoA give this build more tools to work with. Upgraded Nar’s Arcane Destroyer would be just the thing this build needs.
Before I move on to how this finalised setup performs now, I feel I should address a question that’s probably crossed your mind over the course of reading this. If I was so focused on making an Enchanter-style character, and the ranged setup I was working with was failing so miserably, why didn’t I just give up on the Arcane Archer part of the concept and use Scion of Arcane Force? It’s literally custom-made for Discord Cadence + IEE Tri-Elemental Battle Mage, after all. Well, firstly, I created this character back in Ashes of Malmouth. Scion wasn’t added until like a year after Forgotten Gods was released. By then I was pretty set on the ranged playstyle cause the character existed with it for so long. I’d rather scrap the build altogether as a fail than abandon the core of it at that point. Second…I still tried it anyway Just out of curiosity, to see if it was just my approach that wasn’t working, or if Tri-Elemental Battlemage is just a trainwreck in general.
Admittedly, my testing with that setup was fairly limited. I simply kept the gear I had (which, at the time, the ranged version was also failing with), swapped the weapon, made some changes to attributes (since I no longer needed the Cunning for the weapon) and devos (dropping Hydra) and I took it out for a run in SR75. Didn’t even make it out of the first chunk. Died there 3 times to just normalass hero groups. Obviously with a bit more practice on that setup or some of the improvements I later implemented on the ranged version it would perform better, but the feel of it wasn’t very encouraging. Being suddenly forced into full on melee on a character that’s short on damage and sustain just felt like serving myself up on a silver platter. One of the few good things I could say about the failed setup with Dawnshard Gaze was that the ranged playstyle offered it a ton of flexibility in terms of how it approaches engagements and how much it exposes itself to damage. With Scion that all went out the window, I would have to facetank everything all the time. In addition, by going melee I was now limited in how many enemies I could hit per Cadence attack, while ranged Cadence could easily pass through 10+ enemies as they noodled their way to me through a doorway or a narrow corridor. The AoE coverage and the sustain in hordes was therefore considerably better on the ranged setup. And finally, the damage. The damage from Scion was better, absolutely. But it wasn’t that much better. Even though Scion has a lot more base weapon damage, more flat, more % ele damage than my rolled Arcaneweaver, more points to Cadence, added flat to Cadence through the modifier, more RR than Elemental Storm, and more flat on the IEE modifier, there’s also things Scion lacks. My Arcaneweaver had a good bit more attack speed. And Scion lacks Phys-to-Ele conversion. For the weapon damage that doesn’t matter as that’s all pure Elemental. What’s not elemental is the flat phys/trauma on the Soldier WPS and the physical on Deadly Momentum. Converting that contributes a lot of damage, and between the Arcaneweaver and the transmuter on IEE we get most of the way there. With Scion most of Deadly Momentum goes to waste. In terms of raw damage it’s still the better option, but where you might expect something like 30-40 % more sheet damage, you get closer to 10 %. That does not factor in the increased RR gained from the weapon, of course, but it also doesn’t factor in how much weaker the WPS become without the phys conversion. And as mentioned, the safety of engagement and the AoE between the two weapon setups is night and day. Scion is much more exposed to damage, less flexible and offers less AoE than the ranged option.
But enough waffling on about weapons I’m not using, how does what I AM using actually perform…
What in the actual? This can’t be right, when this build was wearing a different helmet it was a failure and now it’s this? Hell, when this build was using the same helmet it’s using now it was a 3 out of 5 build. What happened? The number might look good but there’s a reason I don’t just settle for a number and I explain how the build feels and how it plays. The numbers don’t tell the whole story.
Now, a little bit of behind the scenes info. When I switch over to another build to test for these overviews, I usually give myself 3-4 mulligan runs where I have the option to not count the run if I failed due to my own fault. Because when I switch to a new build, I’m all fingers. Depending on what rotations a given build is using the movement skill might be placed somewhere else than it was on the character I just tested, the survivability options and Oh Shit buttons may be moved over and I’m pressing all the wrong things all the time. I’m also just feeling out the damage and survivability of the build again after not playing it for a long time so there will be situations where I overstay and get myself killed, or overestimate my damage and don’t kite an Arcane enemy when I should, etc. I also use these first few runs to see if maybe there are some final improvements I should make before I commit to doing a full 10 scored runs for a version of the build. In the first 4 runs I did with this character, I felt like I was back in the “1 success out of 20” era. I was getting hammered into the dirt before I reached the SR77 boss room. The one run I did where I made it further, I was dealing with increased enemy elemental res, which hurts this character quite a bit given the awful RR, and yet, I was doing fairly well. Then I got to the SR78 boss room, ran into Gargabol, and even though I had 5 and a half minutes on the clock when I started the fight, I failed to kill him in time. By like a minute. It was performing awful top to bottom. And yet here I am presenting this as a 9/10 character. What changed?
I learned some lessons in those 4 runs. The Grava fight was ass, so I changed the chest and medal component to fix it. I also changed boot component from Mark of Mogdrogen to give myself a little bit more slow res, since the build does need to kite now and then and it also can’t really afford to have its damage tanked heavily by reduced attack speed. I then shuffled around a few augments and shifted a couple of skill points to get a point in Veterancy and get a couple of extra points in Military Conditioning so I can move more stats from Physique into Spirit for more damage. I also finally learned where the fuck my mobility rune button is so I’m not charging into shit instead of applying War Cry. That helps.
With just these few changes and a better feel for the character, things changed dramatically and a floundering character became a successful one. So successful I almost got it to 10/10. The one run I failed was because in the SR79 boss room I started a fight with Moosilauke without luring him from his spawn first. Me hitting one of his crystals aggroed Rashalga on the neighbouring spawn and between Moose and Rashalga I simply didn’t have the kiting space I needed to handle that fight safely. An avoidable issue.
But I’m not gonna sit here and pretend like this is an excellent build worth replicating. Cause it’s not, whatever the numbers might indicate. With patch 9.8 the build’s lot has improved considerably, as there are no longer those hyper-resistant heroes running around in SR. Tri-Elemental res has just gone down a lot globally with the change to monster equipment. Also, with the nerf to healers, a previously major issue for this build has all but disappeared. Not only will human healer hero clusters be taking a ton more damage, they’ll also not be able to undo the build’s damage faster than it applies it anymore. Healer heroes went from a rage inducing experience to barely a speed bump for this build, which is massive for its success rate. Despite all that however… Tell me, have you ever heard of the term “speed”? No not the drug,the physical quantity. Cause this build is just barely grasping the concept. It’s not fast at anything. Single target, hordes, it all takes a while. Unless, that is, enemies really line up for your Cadence passthrough (or you line them up with your mobility skill); then you can actually have some very respectable clear speed and get good timer bonuses in the process to give you wiggle room for the bosses.
While the character might be living proof that RR is optional (if a certain individual feels called out here, that’s cause they are) having some -%RR sure wouldn’t hurt, cause it’s still a massive damage multiplier. Without it, the build can be sporting all the flat Elemental damage it wants but it’s just not gonna be very fast. And with low-ish speed automatically comes the issue of margin for error. If you die you’re gonna have a bad time. Either it will end the run outright, or it’ll be some damn sweaty work to recover and boss/chunk RNG will play a big role in whether or not you pull that off. There is, however, the caveat that you need to die first for this to be a problem. And once you have a feel for the character that’s not all that likely to happen. Say what you want about Battlemage but you can’t argue that it’s fragile. Between the copious amounts of health, solid armor, DA, %DR from perma War Cry and massive % absorb from Maiven, the utility of Nulli plus an Oh Shit button in Mirror, the build has all the tools it needs to stay alive, provided you know the build’s limits and don’t overcommit. So while you might end up with a relatively slow character, it’s also a character that is fairly steady in its performance, once you have a feel for it. I’ve posted 9/10 or even 10/10 characters that have died 4-5 times across the 10 runs. This character has died once. In the one run it failed.
That still makes it a difficult build to recommend. Because there are other steady, very reliable characters, and most of them aren’t saddled with poor speed. Most of them will also not demand that you farm 6 greens. Admittedly, the affixes aren’t all that strict anymore. Previously, Fleshwarped Casque only gave -2 s cooldown to War Cry, so full uptime also required some additional CDR, which I was getting through the Magi affix on the helmet and amulet. That is no longer needed, any % elemental affix will do. Having extra health on some of these pieces is needed though to avoid surprise one-shots from guys like Theodin or IM even under Afflicted or Cruel mutators. Then you’ll just want as much % ele damage as you can get, and as much attack speed on the rings as you can get. The biggest bitch to get by far will be the weapon, cause not only does it drop very unreliably and unpredictably (the one boss that can drop it moves between two dungeons), it also needs to roll at least one affix with flat ele damage (and the associated phys to ele conversion) and with attack speed. The one I have isn’t quite ideal, you’d want Enchanter’s of the Elements, but this is the best I got. And you’d be doing all that to wind up with a reliable, but by no means stellar, build that will do content slower than less gear-demanding alternatives.
If you want to replicate this, for whatever reason, and aren’t necessarily fixed on the idea of making an Enchanter and therefore juicing IEE, do yourself a favour and do this on a Tactician. Not only will you get full WPS coverage and easier access to attack speed through Ranged Expertise, you’ll also actually have a useful exclusive skill with RR bundled in. You will miss out on the phys-to-elemental conversion on the IEE transmuter, the % absorb, Nulli and Mirror, which are all useful, but I dare say, the tankiness the build has on Battlemage might be a tad excessive. Having more damage would make that tankiness slightly less necessary as there would be more sustain from the improved leech. I think it’s a trade worth considering, at least. With Battlemage, you get your reliable, steady and kinda lazy performance (there’s no buffs or heals to keep up, just War Cry to apply), it just won’t wow you the same way a more damage-juiced build might. But if you don’t care about the cost to benefit ratio of the build or were ever curious to try a build without any -%RR and see how it works, this one is definitely a viable option.
Here are the usual SR75-76 and SR80 sections of one of my runs. This is about the speed you can expect from the build with one mutator messing with its damage output, which is about average. Sometimes you’ll have more, sometimes you’ll have none, sometimes you’ll even have some extra damage. But at least one negative damage mutator will usually be around. This is also with the one survivability mutator the build cares about (Afflicted) so the survivability is usually better than this, In fact Reaper got WAY closer to killing me in the SR80 boss room here than he ever has in any of the other runs. He’s usually a very comfortable fight thanks to Nulli, Mirror and the skill disrupt on War Cry. It’s usually Grava that’s the one making things messy. I also haven’t fought Gargabol since that one disastrous mulligan run, so I can’t comment on how much the build has improved in that encounter. It has, however, killed pretty much every other scary boss you can think of across the 10 runs. Rashalga, Morgoneth, Theodin, Korvaak, Kymon, IM, Fabius, Grava, you name it, this kills it. After a bit.
