Hey everyone,
in the recent weeks I've been testing the new DH builds made possible by the reworked Hellcat Waistguard and also dived a little bit more into the belt's mechanics as well. I did some tests on the PTR to find out more about how everything works, which is probably worth knowing for everyone who wants to take his grenade builds seriously. Since we are quite late in the PTR cycle and DH seems to be in a okay spot, I doubt we will see many, if any, changes here anymore.
I did a total of 5 different tests. First, nailing down the damage behavior of the procs, second, recording the average hit chances of bouncing grenades on a stationary target, third, validating proc chances by skill, fourth, counting some kill times with different belt rolls, and fifth, exploring the Buriza mechanics with grenades and the belt.
Here is a screenshot of my sheet for the first test:
My base damage is 56 without the belt equipped, using Grenade - Tinkerer. If I equip the belt, my "base" attacks will go up by 50% to 84 already, meaning the initial attack aready counts as a bounce. The following bounces each add 50%, and the last one will deal 800% total, not 100% + 800% = 900% as the wording on the belt could suggest ("... dealing an additional 50% damage on each bounce / This bonus is increased to 800% on the final bounce"). This means the total damage with all grenades hitting are 1700% (5 affix), 1400% (4 affix) and 1150% (3 affix), respectively.
On my next test, I took a stationary target (Hellbringers in A3) and did 200 casts with each possible belt affix, counting the procs as well as how many times each separate bounce would hit the target. This test was done with the Grenadier passive equipped. Here is a screenshot:
It seems likely that the differences in proc chances are only due to statistical errors, and the real proc chance is probably precisely 50% (also tested more thoroughly below).
It is worth noting that Hellbringers have a rather large hitbox, which rivals that of only the biggest monsters (not counting RGs). Interestingly, the damage share of the initial attack is rather high, around 20-27%. But as you can see on the numbers, especially for the belts with 4 and 5 affix, hitting a target more than twice is rather uncommon. The third, fourth and fifth hits all have average hit chances of below 50%, going as low as 28%. Obviously the numbers go down as the count in bounces increases, as the grenades have a higher chance of jumping somewhere outside of the initial attack area.
Looking at the numbers for only the last and most important bounce for each affix, you can see that the "big" explosion at the end doesn't even do that much for the 5-belt, with only 32% of the total damage done (as opposed to the possible 47% if we had 100% hit chance). Same for the 4-belt, where it is 40% (possible 57%), and only for the 3-belt we have some good numbers here, with 59% of the total damage (69% if we had 100% hit chance).
This result is very interesting because it suggests that the "bad" rolls on the bad are actually not that bad, and might in fact be better than a good roll in some cases. Of course the potential is much higher with a 5-belt, however the average chances of hitting the right targets (or anything at all) are very low for the strong, late explosions, while you get those much more consistently with a low affix. More importantly, bouncing grenades disappear as soon as they hit an obstacle or a wall. This test was done in an open area with no obstacles around, however in a GR, even in good map layouts (excluding most of arreat crater, perhaps), we have walls and obstacles everywhere, and grenades will disappear often before they can fully finish their bounces, favoring a lower affix on the belt. On top of that, you have to consider that for high GR pushing, the ideal playstyle is to surround elites with a lot of trash in order to nuke them down with area damage. With the high-affix belts, the grenades can be expected to do the opposite, i.e. jumping out of the center and killing all the small targets outside, leaving you with nothing but the high hp targets and no area damage to kill them. Based on my limited GR testing on the PTR it seems like there is a chance that people will prioritize lower rolls on their belts for certain builds. For speed farm (torment) builds specifically, you might want to run with a low belt anyway, simply because it takes a long time for all the bounces to happen in the first place.
This incosistent bouncing behavior is generally not so bad for skills like Cluster Arrow - Cluster Bombs or Grenade - Cluster Grenades, because those don't jump so far or just hit a large area in general. However, for Rain of Vengeance - Anathema, Strafe - Demolition or Grenade - Grenade Cache the difference is definitely noticeable and you can see that very often you won't hit your targets with your last explosions. I think overall higher belt affixes will still win out (especially when we consider GR fishing for big density), however probably not by as much as one could think without looking more closely.
My next test was just a confirmation of proc chances by skill. I equipped all the relevant grenade skills and counted how many times I saw a proc. It was rather difficult to do for RoV, so I only did 5 casts, watching them in ultra slow motion a couple times and counting the hits / nonhits until I felt I caught most of them. It seems like the belt has 50% proc chance across the board, independent of skill and belt affix roll:
Based on my findings about the low affix rolls, I went on to a field-test killing a big target in a rather open area (Ghom) and a small target in a very confined area (skinny a1 zombie near Adria's hut), counting the seconds for a kill.
It's only a small sample size, but it indicates that indeed the lower belts might not be so bad, especially on small targets and in areas where grenades tend to bounce against obstacles. Note that this test was done with Grenade - Grenade Cache, which probably has the highest bounciness of all grenades. It's also interesting that despite the highest average kill time, the fastest kill by far was done with the lower belt, too (on zombie target).
Right now I would say that the (5) belt is probably going to be the best on average if you intend to use it for fishing good maps and big density, and a (3) belt is probably best used for fast farm runs (torments and low GRs), where we have to fight in all kinds of small corridors and also want to deal the big damage as fast as possible (however only if you intend to basically oneshot stuff with a proc). Unfortunately it is rather difficult to acquire reliable data on this, so I can only go with my gut feeling here.
My last test was done with Buriza cubed, to see its behavior with grenades:
This screenshot doesn't show much more than that the extra grenades are also working with the belt, as expected. What it doesn't show is that the Buriza is actually quite difficult to proc, and it has something to do with how many targets are present and how they are distributed locally. It seems you need at least two targets and get the best results if you have a second (or more) target in a straight line right behind the first mob you hit with your grenades, which will make Buriza proc for additional damage.
Here is a little indication of what I mean:
This is only really relevant for the UE generator build, though, and keep in mind that Buriza still has an ICD of ~33 frames (so you can only get a proc on every second attack).
----
TL;DR: Hellcat Waistguard gives us many new builds, especially for solo DH. The proc chance seems to be 50% across the board. Some of the grenade skills we have are quite bouncy, making the juicy late bounces more difficult to unfold their potential when they jump out of a pull or against an obstacle and disappear. A low roll on the belt does not necessarily mean it's bad, especially for RoV - Anathema, Strafe - Demolition or Grenade - Grenade Cache and speedfarm builds in general. Buriza works for grenades but is difficult to proc, because it depends on your and your target's positioning.
----
Now that we're through with all of this, I also want to present you my guides for next patch. They are all aimed at pushing solo, and with this new belt we actually have a couple very competitive builds now. Personally, I will probably play Marauder's for pushing next season. *I will add some numbers to each build to give you an idea of what I would expect this to clear with ~1000-1400 paragon, good item rolls, augments and maybe 100-300 keys.*
UE generator (video / D3planner / D3planner torment version) [GR 92-95]
Fire Natalya with Shi Mizu (video / D3planner) [GR 95-97, builds without Shi Mizu or including FoK are also 90+]
Cluster Bombs Marauder (video / D3planner) [GR 96-97]
New concise version of my LoN FoK + Shi Mizu guide (video / D3planner) [GR 96-98]
LoN Strafe for T11-13 (video / D3planner)
I also plan on adding short written versions when Diablofans resets its build section with the new patch.
----
I hope you enjoyed the read and / or one of my guides, and maybe you're also looking forward to pushing solo with a different build next season.
wudijo
4