Barrage build possible?

So is there any conceivable way to make a viable (if sub-par) build that uses Barrage as its main skill and DOESN'T rely on CoC?

I love the mechanics of the skill, and I always used the rapid fire range attacks in Torchlight, but Barrage's damage just seems.. unusable, even with a bunch of pure-DPS support gems and a decent weapon. Am I just missing something?

I'm definitely looking forward to the Ranger Ascendancy classes, because that +1 chain node with no damage reduction might just make a Ricochet Barrage build viable.

I suspect that the tooltip DPS is simply lying. I'm trying to compare lv. 12 Barrage vs lv 18 Frenzy right now (single-shot frenzy vs full barrage hitting 1 target) and Barrage seems to be doing just about as much damage if not more to an Aurate Statue EDIT: even though Barrage tooltip is about 1000 while Frenzy's is about 3500.

May report back if my testing shows more conclusive results.

EDIT: Went to Merc Ledge A1 instead (lawl you can see how super op my bow is that I have to go to such a low lvl area :P) and I'm 100% sure of it. Barrage-all-hit-same-target deals more damage per volley than a single Frenzy shot. The tooltip is definitely lying.

As for ideas on how to make it a "viable main attack skill," because of the way GMP/LMP interact with Barrage, I would definitely wager that 100% Pierce chance is necessary to take full advantage of Barrage's abilities. Even in Merc Ledge Barrage's crowd clearing was pretty pitiful on multiple targets because each projectile deals so little damage. However, each enemy would take roughly 300% more damage if I had 100% full pierce chance, because then most enemies could take damage from most projectiles.

That's a good recommendation, thanks. I'll try that. I definitely agree, the tooltip isn't accurate.

Since I first made this post I did some minor testing on Merc Kaom's Dream with Chain instead of Pierce, and even with only 3.5k tooltip (it was only like a lvl 12 Barrage gem), the crowd clearing was actually quite adequate. Not great, but at least viable even with under-leveled gems. I tried Pierce for a moment and it was fiddly to get the mobs lined up, but every Pierce-based build is like that so I can't complain too much (and I don't have a Drillneck or anything, which would be a huge DPS boost). Chain just fit my playstyle better.

Conclusion: Ricochet Ascendancy node will definitely make a Barrage build fully viable, since I'd be able to swap Chain gem for a DPS/multi projectile gem and still have essentially the same crowd-control functionality. Not really practical until then, sadly.

I had the skill tree for a dual puncture totem build that then used Barrage-GMP-Chain-ChanceToFlee on a four link. looked fun but I never got around to making it. Act four makes leveling so long!!!

If you are looking to use it as damage I recommend trying to use Wands, I believe passable damage could be obtained an elemental crit wand build with chain if you got max crit an tons of muli. (Favorite wand set up is still still Elemental Hit in a +3 melee gem shield)

If looking for a bow specificly they did leak a low level bow with +1 chain on it. Might be able to stack a bunch of eli damage on that.

If you want strait up phys bow Barrage or bust try Chin Sol Point Blank and become melee Ranger without shield.

Thanks for the input. Any recommendation on an actual tree for a Barrage build? Stuff to focus on besides crit?

P.S. please post that Chance to Flee build. I've been wanting one for ages.

A tip - replace Chance to Flee with Knockback gem.CtF doesn't work against unique and rare monsters,Knockack does.

"

MasterBLB wrote:

A tip - replace Chance to Flee with Knockback gem.CtF doesn't work against unique and rare monsters,Knockack does.

I was messing with Barrage and Knockback today too lol, modeled after my old Torchlight 2 character. Even in PoE it's surprisingly effective in terms of defense with a decent aspd.

Just did some more testing (swapping between GMP/LMP/Faster Proj + Pierce) and I can confirm beyond a doubt that the tooltip is SUPER inaccurate. It claims a much higher DPS with LMP than GMP, yet swapping to GMP is a massively noticeable increase in actual real-world DPS.

Is there some formula out there for calculating true Barrage DPS?

"

MaraWuti wrote:

Is there some formula out there for calculating true Barrage DPS?

Not that I know of, it's a funny skill

"

Mark_GGG wrote:

Adding extra projectiles to barrage will make it take longer than your attack time, as each one takes time to fire.

"

Mark_GGG wrote:

The reason that can't happen is that Barrage (like other attack skills) uses the attack time of the attack to know how long to take - the initial "windup" phase takes 40% (from memory, might be slightly wrong) of the attack time, and each projectile takes a set percentage of the attack time for itself.
Since Barrage's actual total time taken is necessarily a function of the attack time value, it's not possible to have it match in the lots of projectiles case.
Long term, we'd ideally be able to introduce a separate stat for barrage to show how long the full barrage will take, but for now the actual stat we have is the attack time stat of the skill, which the skill uses to time the animations.

And viperesque did some testing which have shown that:
1. Barrage does not scale in a linear fashion with attack speed.
2. Barrage's attack time is not dependent on the number of projectiles fired, in contradiction to apparent developer intent.
3. That said, LMP or GMP each seem randomly capable of inducing a 4-frame delay in the attack time, which remains until you restart the game. I assume this is just client-side, and this is just my best interpretation of what's going on with some confusing data that I don't want to put hours more into figuring out.

Mark answered with this at the time

"

Mark_GGG wrote:

The issues with barrage at high attack speed are relatively well understood and being worked on.
There is no difference in calculation as such on the client and server - there is a difference in how long a frame takes to execute, which can cause noticeably different timing results between the two detecting when a very short animation has ended, since there's no granularity below a single frame. In the case of barrage, which plays several very short animations (each triggered when the previous one ends) this is exasperated further.
I am currently working on changes to the animation system that should allow the client and server to keep "series" of animations like these better in sync despite having different frame times. I cannot say at this stage when such changes might be ready for deployment.

So, if it's a sync issue it got fixed when they introduced lockstep, probably in predictive too. Haven't really been doing any testing of Barrage myself since.

So what you're saying is..

1. Prioritize number of bullets over aspd

2. Prioritize crit multi and pierce over base damage?