AAR is weak at the baseline

[banana_peel]

Played several vit variants with the new mod on Hallanx offhand courtesy of the resident AAR enjoyer @KoS9K

Also played this kind of generic fire Sorc AAR with two items with mods for the skill, high rr%, very high %dmg but basically solo AAR to isolate the performance of the skill.

Fire AAR Sorcerer SR30-31 [1.2.1.6 pt] - 5:47

Very good layout with two short rooms and a lot of close nemesis spawn. A decent build in this patch would finish it in 4:20, 4:30 at most. Not 5:47.

AAR had several improvements including the leech mods. But the baseline dmg on it still feels very low. Here is some breakdown by points:

1. Tooltip is low:
If we compare AAR to other channeling casts, FoI and DE, AAR lands somewhere in the range of 35-40k a tick, with vit and aether pushing a little further due to stronger support. At the same time DE usually sits at 40-45k now. FoI is also somewhere in that range with chaos, RtA and aether variant pushing much higher numbers, 50-70k.

2. The skill is not compensated enough for single rr%
Arcanist has no rr. For most skills Arcanist provides good defences and %dmg opportunities that allow you to shift some power in gear into offence and compensate. But AAR is different. It has a hard ceiling on dmg output as a channeling skill with no WD part.

3. The skill is not compensated enough for unstable enemy behavior
AAR always relied on lining enemies up with the ray by correct repositioning. But now enemies constnatly bounce around and flank you which drastically reduces the effective channeling time. This problem was partly solved for ranged AA and PS gameplay with reduced Impaired Aim and Fumble sources on enemies, and for FoI - by increasing its cone and turn speed. DE just has no issues with enemy movement at all, and it’s a big advantage. AAR still suffers a lot.

4. No reason to play over PRM
After the latest buffs PRM is miles better than AAR, for every dmg type. It has higher solo dmg, higher aoe clear, not very sensitive to enemy behavior, rather it can randomly reward you with overlaping dmg from grouped enemies. Most importantly, it has the same “stand in place and cast a ranged stream of dmg gameplay” as AAR.

---

Suggestion: I think, AAR needs to be closer to 1k flat per tick at the baseline (Clairvoyant bonuses would have to be reduced).

Right now it has 781 at both hardcaps (and no mods provide huge increase for it). In comparison, converted Seal of Blight gives 464 flat + 36% WD + bog dot + inbuilt leech. If you have a big 2hander with some global flat, you can actually reach or even surpass flat value of AAR and have a skill with a better dmg pattern. One component vs 26 sp, 50/50 Arcanist skill. Not good.

@grey-maybe and @romanN1 can share their recent experience with the skill.

[Gnomish_Inquisition]

only thing i have to add is Clairvoyant AAR warlock felt absolutely gutted after the last round of changes , but that might just because i was using a non-fancy setup/regular “level and assemble” char.

Chaos AAR felt worse than FoI, but it might be accented by my slower run and facetank nature giving FoI aoe an even bigger benefit

I like AAR in terms of entertainment and enjoyment, but even to me Lee’s PRM feels stronger these days, rather stark shift with how good PRM has become (i outright used to dislike it)

and while it’s probably obvious/objectively worse than Banana’s/“suffering from gnomish inquisition approach”, my latest fire sorc Sorcerer, Level 100 (GD 1.2.1.3) - Grim Dawn Build Calculator experience was def not anything great - not even gonna consider OFF AAR mageslayer these days

on that note i’d like to mention a perception issue AAR might sometimes suffer from, atleast from beginners and maybe slightly less experienced end builders (also even if not constantly/everyday frequent posts). And that is, “why spend 28 skill points+items on AAR, when you can just use this Conflagration - Items - Grim Dawn Item Database”
Obviously their notion doesn’t factor in item mods/benefit of item modifiers, sustain potential and true overcap dmg scaling. But if nothing else it might suggest that the base dmg to some feels weirdly unattractive specially when contrasted with a freebie item skill.

other thing i’ve mulled over lately was perhaps too much of AARs dmg was baked in with and relying on crit dmg in mind?
The difference of going from regular/avg player 3050’ish OA homebrew legend setup to 3450-3500? OA on green MI/“proper top” setups feels relatively noticeable even on a the more casual front, making it feel even more mandatory to get right as skill/be done proper build

[mad_lee]

@banana_peel

The skill is not compensated enough for single rr%

Good point. Currently skills like DE or FoI usually have double rr. And overall all good points. Personally I stopped doing AAR builds long ago because it was just not worth the effort.

The only upside of AAR builds is some of them can be obscenely tanky.

[_maverick]

@mad_lee

The only upside of AAR builds is some of them can be obscenely tanky.

Elemental ones lack behind the other types. I could see a damage boost to AAR but chaos and phys would then be SR 40 builds…

Also some classes like Sorcerer (fire, aether), Spellbreaker (cold, pierce) have inherent weaknesses. Mage Hunter, Binder and Templar are better.

I would like to see less interrupred AAR gameplay. How will the new channeling skill in Berserker work with all the interruptions? This needs to be changed imo.

[KoalaeiO]

@banana_peel

Very good layout with two short rooms and a lot of close nemesis spawn. A decent build in this patch would finish it in 4:20, 4:30 at most. Not 5:47.

I find interesting to note the evolution of the language here. “Decent”. Just decent. Not even good. In my book, that should be “top tier”. 4:30 is the timing reached by the characters in the “Top 20 Softcore builds” thread.

[Gnomish_Inquisition]

From my understanding, and please correct me if wrong Lee/Banana
top 20 isn’t top20 because they have the highest performance, but because they clear all content, while being (potentially) simple to assemble, doing so under decent clear speeds and have “guaranteed” performance(/clears don’t require skilled pilots nor varies much/isn’t risky one run to the next).
There’s a bunch of builds that clear faster than top20, but they’re either more expensive to build, don’t have the same neat straight performance line, or requires more controller/pilot expertise to get that speed out of/“less fool proof”.

And while i personally agree it’s fast these days , noticing from posts the trend is indeed just that, the target frame seems to be 4-5mins
Likewise cruci has shifted too

*also while 4.30 on appearance might seem lightning fast, it doesn’t mean it translate to everyone, but what these skilled players manage to push out of the same build as others might do 60-90secs slow on avg(like me)
But even if more skilled they then still have the same/their own applied avg curve like we do

so if 75 builds is 4.30 “in banana hands”, and this supposed strong skill build is 5.47, it’s then a “big” difference in that same comparison pools

[Bas]

While I don’t agree with everything they say and I sometimes feel they sometimes complain about skill that are already ‘ok’ (just not good) while there are others that to me seem worse, banana does have a point here. I tried to build an aether version not too long ago (even if it was a bit of a silly one) and while damage seemed fine in the campaign, when I moved to SR it was just atrocious. And maybe someone can save it, but for me it was so bad I didn’t even want to try anymore.

Comparison to PRM I find a bit… eh… (I think PRM is OP since last patch). Comparison to the other channeling skills is totally fair and show’s it’s lacking. Though I’d also argue the skill is awkward to build around, since you basically need one of two items to give it leach, otherwise you’re out of luck for sustain.

[banana_peel]

@KoalaeiO

“Decent”. Just decent. Not even good. In my book, that should be “top tier”. 4:30 is the timing reached by the characters in the “Top 20 Softcore builds” thread.

The game’s balance improves rapidly. It would be much harder to make a top20 for this patch. over 50% of builds i tried this patch had this top performance. That’s why 5:47 on that layout is very noticeable.

@Gnomish_Inquisition

From my understanding, and please correct me if wrong Lee/Banana

All is correct.

SR still has a lot of timer variance in it, too much variance for my liking. Mainly it has to do with very inconisstent distribution of big packs and Nemesis spawns. You can’t even do “average” SR clear timer for the build, you’d have to make like 40 full runs with it, nobody has time for that.

So the feedback is somewhat feel/experience based hence not super reliable. I try to look at the build’s performance in typical situations, rather than at full timer. For example, it took Fire AAR 38 seconds to complete a 3 tantacle fatties room while a good build completes it in 17-25 secs depending on if it’s solo or aoe focused. Then you factor in a lot of rooms with close Nems spawn - they also take much longer than a good build would take, etc.

[AngmarDT]

I strongly object against making AAR “better” by gimping Clairvoyant bonuses to AAR.
please find another way.

[Gnomish_Inquisition]

curious, how would it matter if +25 dmg was on the skill, boosting all AARs, including Clair, vs exclusively on clairvoyant/only benefitting Clair ?(and primarily binder at that) - you lose nothing and everything gains

[AngmarDT]

If it is like you say, then yes. If clairvoyant is ripped 10-20% of damage for base AAR to get extra 5%, then it is not a fair trade off.

[Gnomish_Inquisition]

i’m also not getting the impression that’s what’s being proposed anywhere, since that would be a nerf, and clair aar warlock is already hit - and the notion is to maintain status quo by reducing clair bonus, but have it apply to all aars

[The_Mess]

@banana_peel

You can’t even do “average” SR clear timer for the build, you’d have to make like 40 full runs with it, nobody has time for that.

Well, I do But usually I go to record 10 runs and post the shortest one I manage to gumby up along with a rough estimate on average time/max time. Only if I’m motivated can I manage to record times down.

As for the actual value needed for an average, it’s 70 at the smallest, simply because SR has so many Degrees of Freedom, so you can actual calculate the sample size if you can be bothered. Which is a lot of bloody runs. Only takes about 2hrs to get, but that’s a lot of repetition and I burnt out trying it last time because of that…

Anyhow, the whole thing back in the day with AAR was massive crit damage, but with power creep and changes to balance that out, AAR no longer hits as hard. Especially compared to FoI. Probably the easiest fix would be to increase the flat values and scaling and tweak Chaos ARR to accomodate that.

[thepowerofmediocrity]

Just adding another +1 to all suggestions. To provide another data point, I decided to choose either Lightning or Cold AAR to coincide with the Pulsing Shard changes. Here’s what I came up with:

Spellbinder, Level 100 (GD 1.2.1.3) - Grim Dawn Build Calculator

This is my general approach when using a high damage, non-leeching skill: in this build I use Ill Omen as massive damage reduction and crowd control while using semi-invested Bone Harvest for leech. AAR Tooltip was around 33K which matches what banana stated (though remember that some of that tick is DoT damage which doesn’t help against tough single targets).

I’m able to do SR30-31 reliably, but I know I would absolutely perish in the speedrun scenarios other people test in - I needed to thin out the herds using Ill Omen’s confusion so that AAR could do its job in lazering the weak packs and relied heavily on Spellbinder’s huge CDR / Eternity procs and Mirror to make it through tough scenarios.

AAR has multiple problems: elemental AAR - outside of the weapon / amulet / shoulder slots - is miserable to gear for. I had to use a faction chest and an ill-fitting head piece just to get the skill points required. Casting speed was a surprising problem I had to account for. I was constantly running out of Energy when facing the boss rooms (also, I hate the new Energy potion changes, I feel like I have to chug a pot when I’m 3/4 full just so I can sustain enough energy for the pot’s cooldown). Sustain was actually one of the stronger points of the build because of its crowd-control capabilities (this is why I don’t advocate for “make everything leech,” I liked having to solve for that variable), but Lightning was a non-starter because I didn’t have anything like Bone Harvest to use for sustain.

So +1 to all of banana’s suggestions and then some. AAR needs to be the nuke of the channeling skills, it has no right having the weakest tooltip.

[Gnomish_Inquisition]

@thepowerofmediocrity

also, I hate the new Energy potion changes, I feel like I have to chug a pot when I’m 3/4 full just so I can sustain enough energy for the pot’s cooldown)

mentioned this in another thread
really seems like the major boon for this change is levelling/early game
Endgame builds that has mana issues and needed to chug, now just need to chug even more

@thepowerofmediocrity

Lightning was a non-starter

can do it with regen, Druid, Level 100 (GD 1.2.1.3) - Grim Dawn Build Calculator gets a bit scuffed, but i think Banana had a semi competent variant that didn’t make as many concessions as my approach (and yes i know mine still uses pulsing shard but i use that on everything for facetank comfort guarantee )

[_maverick]

@thepowerofmediocrity

I decided to choose either Lightning or Cold AAR to coincide with the Pulsing Shard changes.

No changes if I am not mistaken… Binder seems like a good idea for cold now but making it without leech is not a good thing… SR 30-31 is a deep meme territory these days… With pulsing shard and some other changes the Binder could perform a lot better.

I have a cold Spellbreaker that is almost an SR 36 build… Can complete it in slightly less than majority of times within the timer. Moosi is a bitch tho… Spellbreaker, Level 100 (GD 1.2.1.3) - Grim Dawn Build Calculator Just needs slightly more defense. Breaker is really hard to build… Mage Hunter would be the easiest for cold. And Binder should be better than Breaker if built differently.

@Gnomish_Inquisition

can do it with regen

Here is mine… [1.2.1.5] Lightning AAR Druid, Ravager, Callagadra, SR 36+ Needs more DPS. 330k is not that high today… Cold Breaker has 350k. 400k is ideal.

I also have a non clairvoyant aether Binder that can do SR 38 almost. Will tune it a bit and post sometimes… Unfotunately has only 300k DPS… But is a tank…

[thepowerofmediocrity]

@_maverick

No changes if I am not mistaken… Binder seems like a good idea for cold now but making it without leech is not a good thing… SR 30-31 is a deep meme territory these days… With pulsing shard and some other changes the Binder could perform a lot better.

Did I imply somewhere that particular SR range was the top of the build’s capabilities? No? So what’s the point of your message? My build was meant to be a starting point to see what kind of problems AAR is facing, and everything I stated above (gearing problems, energy consumption, bad tooltip) is going to apply no matter how deep you go into SR.

EDIT so it’s not just an angry response: I didn’t realize the Pulsing Shard conversion was a typo, in which case, yes, Pulsing Shard is BiS for nearly all AAR builds. I was testing to see if AAR without leech was even possible, much less optimal.

Made a Lightning version Druid as well for comparison. Tooltip on AAR is higher than Cold version, but really misses not having a source of damage reduction anywhere compared to Ill Omen on the Cold version. Also, Storm Shepherd belt was made for Lightning AAR, and yet the 100% Aether to Lightning conversion clashes with Star Pact, meaning I have to pick a suboptimal Exclusive.

So offensive is better, survivability was way down, but now that Pulsing Shard is an option again the leech will probably offset that.

[_maverick]

Also Spellshock Tome seems to be useless for lightning AAR because of the need to use Pulsing Shard… Something should be done for it if anyone has any ideas…

And weapon damage conduit still exists… Should be remade to something useful. Chaos PRM would be dope…

[Kurapika]

@_maverick

And weapon damage conduit still exists… Should be remade to something useful. Chaos PRM would be dope…

I second this. Either chaos or pierce, physical, bleed etc

[Gnomish_Inquisition]

i think,thiiink, not 100% sure since i apparently didn’t save it, that spellshock was part fo Banana’s better regen setup? and i also recall… Roman? using it in a setup i think (but might have been ages ago when leech/sustain requirement was much smaller like before 1.8 )