Yeah what about them? With 135-140 gems instead of 150, I think you lose less than a GR of damage. And it’s going to affect all builds. The benefit of groups is mostly in XP farming I think, and that is controlled for. These numbers are quite uncertain already, like they are all ± 3GRs at least I think, so adding in a flat -0.5GR across the baord for SSF isn’t going to change anything.
1 Like
Speaking of underperformers, I’d like to see some recognition for off-brand builds. That is, skills buffed by sets, but aren’t the main damage dealers. These don’t tend to show up on the boards because the stronger variant overwrites the weaker one (or they’re just so weak barely anyone plays them).
Here are some I can think of:
Monk - Raiment - Dashing Strike (buff Fleshrake blizz pls)
Monk - Sunwuko - Lashing Tail Kick
Crusader - Roland - Shield Glare/Bash
Crusader - SotL - Falling Sword
Crusader - Akkhan - Phalanx
Barbarian - MotE - Seismic Slam
Barbarian - IK6R4 - Vile Charge
Some love for the lesser-seen and underserved would be great.
2 Likes
Could UE get a 13 tier buff, thanks.
2 Likes
This is why I like the ethereral approach to buffing weak builds. You can target specific weapon types, specific skills. It would be a million times easier than tying to rework or rebalance all the bad sets. Eventually, you would love to see them all be “fixed” before the end of time, but this would be a great band-aid. Plus, just being able to use ethererals would be a huge draw. Example: I would actually play seeker of light, Tal Rasha, UE, etc. with shiny big magic OP weapons.
Nope, that’s all wrong.
Let’s walk through this, and I think you’ll see where it stops making any sense.
You want Stricken not to be taken into account. That’s fine, we’ll set aside the boss fight, and just pretend it takes 4 minutes, regardless of tier, mainstat, gems, everything.
And, your picture of a Conduit is basically a button that, when you press it, pushes you forward one minute on the timer, but adds 30% progression to your bar. That’s a somewhat oversimplified picture, but fine for our purposes here.
So you’re going to do a GR: Tier “X”, and it’s going to take exactly 15 minutes. We enter the rift, and click a Conduit. Now we’ve got 30% progression, and the timer is at 1:00.
We’ve now got 10 minutes to gain 70% progression, or as you put it, to do 70% damage. Let’s just call it “70 damage” that we need to do, in order to kill those mobs, and extract their progression. We do that, in exactly 10 minutes, spawn the boss at 11:00, and then kill him in 4 minutes, finishing exactly at the 15:00 mark.
Ok, so now we’re going to replay that rift, exactly the same, except one tier higher, “X+1”. Again, we enter the rift, and click the Conduit. Again, we’ve got 30% progression, and the timer is at 1:00.
But now, instead of needing to deal “70 damage” in 10 minutes, the mobs all have 17% more life, so we need to deal “81.9” damage in 10 minutes.
If we’ve also gained 17% mainstat, then our “damage per 10 minutes”, instead of being 70, will be 81.9, and we’ll finish killing the mobs in exactly 10 minutes. If instead, we’ve only gained 11.9% mainstat, the figure you suggest is needed, our “damage per 10 minutes” will be 78.33, and it’ll take us (81.9 / 78.33) * 10 = 10.46 minutes, aka about 10:28 to kill the mobs, putting us at 11:28. Then it’ll take us another 4 minutes to kill the boss, and we’ll finish at 15:28, i.e. we fail.
Or, if we had gained no mainstat at all, our “damage per 10 minutes” would still be 70, and dealing the 81.9 damage we need will take (81.9 / 70) * 10 = 11.7 minutes.
In real life, of course, you might make up a little bit of that extra time vs the boss- if you are using Stricken- since this limits the time growth to the √ of the change. Or, if not using Stricken, you would fail by even more, since the boss fight time would also grow by 17%.
Anyway, suffice it to say that the Conduit is not having the effect on the overall time that you think it is. The issue is that you are actually responsible for 100% of the progression that takes a variable amount of time. The Conduit, the time it occupies, and the progression it extracts, can all essentially be set aside.
Well, I’m 100% SSF myself, and I’m on this board, so let’s take a look at the lost damage potential.
For my 138, I used Stricken rank 136, Trapped rank 136, Zei rank 133. With this build (Leapquake), you really don’t get full power out of Zei, so let’s just use the “midway mark” for that gem, or a max 1.4x multiplier. Trapped max is 1.6.
So for those two “direct damage” gems, we’ve got (1.6 / 1.558) * (1.4 / 1.366) = 1.0525, i.e. 5.25% “missing” damage.
And for the Stricken, the ability being at 2.16%/stack vs 2.3%/stack represents about a 3% increase in the kill time, i.e. a boss who would have taken 3 minutes with a max rank Stricken would instead have taken 3:05.4, 5.4 extra seconds. Or if you also factor in/out the extra Trapped / Zei damage there, it would be about a 5.8% increase, and would take 3:10.5, 10.5 extra seconds.
So, yeah: not an earth-shattering difference.
Edit: and, if I’d had max-rank augs on all my gear, I think this would have added about 2000 mainstat, taking me from about 28k to about 30k, a 7.14% gain of damage. So, in terms of flat damage, and assuming my paragon was the same, being “non-SSF” would have gotten me about 1.0525 * 1.0714 = 1.1276, i.e. 12.76% extra damage, worth about 0.75 GRs.
2 Likes
Sounds nice in theory, but Ethereals aren’t a big enough boost for some builds. They vary in power across classes and are at most what 10GR or so extra power?
I did a Raiment DS push attempt last year and best I could do was GR85 (P1650). No Ethereal could save that build, ha.
We can give rainment both shards are well, lol.
It was before the start of season 25, I don’t remember.
Excellent thread. Seems to me the point here is not to attempt to come up with the one “right” number for each build’s GR ceiling; it is to find a rating for each build based on a consistent set of criteria so that we can get an approximate sense of the number of tiers separating any one build from the center of the distribution. Even if all the clear numbers in the table are biased downward by (say) 2 tiers, the ordering and the tier differences are unaffected. Not that I’m arguing against adjustments to the “clear number” that can be made in a clear and quantitative way across all builds, especially when those adjustments may be worth several tiers (paragon, clear time). These are a good idea because they improve the consistency of the tier number. But adjusting for something less consistent (like use of Stricken) seems hazardous. And I’m not sure that we gain much from trying to adjust the clear number based on how hard the clearer fished. That might be one tier? Maybe 2? How strongly would that affect the conclusions we draw based on the data anyway?
On to the conclusions I think we can reasonably draw from this data.
- The Necromancer as a class needs some serious attention. It has only two builds that aren’t below the median tier (137 in one version of the table): Rathma and LoD CE. Both are only 1 tier above median. Moreover, 3 of its sets are the 3 “worst” builds! We should also notice that all three of these are Corpse-skill builds that require heavy CDR investment for LotD.
The fact that these three builds all end up doing roughly the same thing (LotD —> spam corpse skill —> wait/gather a pull while doing no damage—> repeat) suggests that at least two of the sets need a redesign. This playstyle strikes me as boring anyway (probably why I don’t play much Necro), so IMO that is an additional reason to make major changes. The fact that all three sets do this badly (12+tiers below median) suggests that all of them need significant buffs in power. The fact that all three of them are (more or less) crappier versions of a build that is already being used with LoD further bolsters these points.
Brief Suggestions: Inarius should be adjusted to specifically be a generator set (see below Re: Inarius Scythe). Most other classes have a dedicated generator set so there’s strong precedent for this. Pestilence should remain a Corpse Lance set, removing empowered bone spears (These are a relic of pre-MotBC days, and this should have been addressed a long time ago). Give the set some kind of mechanism by which the player can accelerate LotD cooldown and its playstyle would be much less monotonous. Trag could be altered to buff only life-consuming spenders (Nova, Spear, Mage) and have some tie-in with Siphon Blood, a little-used skill that seems thematically appropriate.
-
Raiment monk needs serious buffs. It is one of few sets of any class with no mitigation or toughness buff built into the set itself, which might be acceptable if there was a compensation in the form of higher-tier DPS (promoting a high-risk/high-reward playstyle with heavy demands on the player’s ability to create pulls, maneuver to avoid damage spikes, and manage defensive cooldowns). But instead of being a glass cannon it’s a glass pea-shooter.
-
Recent changes for Barbarian seem to have worked quite well to make it more balanced relative to other classes. Multiple builds > 140 is pretty solid.
-
I actually think it makes sense for each class to have a set that is less powerful but more beginner-friendly, or more flexible in terms of what can be done with it, or more a showcase of certain “flashy” or “signature” skills of the class. Some of the mid- and low-tier builds seem to fit this description to some extent, e.g. Helltooth, Seeker of the Light, Natalya, IK, and Delsere all fit this description to some degree.
In my view 5 - 6 tiers below median is a reasonable place for at least a few builds.12+ tiers is not. And given that the bottom 4 builds have identifiable issues besides low DPS as well, I’d say we know where the work needs to be done.
I would add that builds that require the player to understand and apply more complicated mechanics, or are more demanding on the player’s skill and execution, should not be significantly below-median in terms of power. That just creates a situation where certain sets are completely ignored by almost everyone: “Why would I play this? It’s hard to play and it’s crap at killing stuff.”
At least if a set that is relatively underpowered also gives an easy, relaxed play experience there will be some players that will find that option attractive. sometimes that’s part of what I look for early in seasons. I don’t mind being off-META if the gearing/leveling grind is more relaxed and provides some variety.
Different note - a couple builds that I’d be interested in seeing added to the table:
LoD Blessed Shield Sader
Inarius Generator (Scythe) Necro
UE Grenades DH
Maybe these are now considered outdated but I for one like playing these builds. Curious how they come out in comparison to others.
3 Likes
I wanted to give a (relatively) quick reply to your excellent post, I’ll leave anything longer and more authoritative to dmkt since this is his thread and not mine…
A few notes/comments:
Yes, absolutely. I did this for Barb (in post #26), but now we need to do it for all the others on the list as well.
It’s more or less impossible to know how hard they actually fished, unless they’ve told us, and we believe them. But “so and so told us such and such” is not a reliable system to apply to this whole table of clears.
I think that what we can know is how fishy some builds are, overall, compared to others. Yes, all builds rely on, and benefit from, fishing/rng, to some degree. But it’s quite uneven from one build to another. For instance, I think that builds that rely on huge density to deal damage (usually via a ton of AD), and have poor ability to deal damage to single targets, are often extremely fishy.
Anyway, a “high fishing build” and a “low fishing build” might have the same clear on this table, but in terms of an average clear, the “low fishing build” will be much less far below the peak than the “high fishing build”.
Real life example: Frenzy (low fishing) vs Leapquake (high fishing). Across a wide variety of metrics, Leapquake has proven to be an overall stronger build than Frenzy, in terms of top clears. And, I’m better at Leapquake than I am at Frenzy. My top clear with Leapquake is 138, and with Frenzy, 133. But with Frenzy, I can do 100% of 130’s, without any failures. Whereas with Leapquake, I can still easily fail a 130, if even a couple of things go wrong.
Here are the refinements I’d really like to see applied to this table:
1: Where possible, replace any clears from the PTR with clears from Live. The result for LoD HOTA, for instance, should be KP’s Live 142 with 4800 paragon rather than his PTR 144 with 4200. There’s just too much uncertainty about how many tiers the GR changes are adding. Yes, for some builds (Raekor, Invoker, Inna, etc) the PTR info is the only info that counts, and we should use that info. But those numbers should get an asterisk.
2: Adjust all these “bottom up” clears for time and paragon.
3: We’ve got to do something with the “top down” info (also adjusted for time/paragon). Leaving it out would ultimately result in enormous balance problems for top-end clears. We should either include both “bottom up” and “top down” numbers on the sheet, or average the two together. So for instance, for Wastes/Rend, the adjusted clear “top down” was 146.8, and “bottom up”, 142.7. We could include both those numbers, or the average would be 144.75.
4: FOR SURE- Rate each build with a “fishing factor”. If we’re feeling confident, we could try to fine tune this to fit each build (i.e. "for build X, average will be -6 tiers below peak, and for build Y, average will be -10 tiers below peak). But, I think a better, safer, and certainly easier way is just to list each build as either “High” or “Low” fishing, with the understanding that each implies a range of tiers you might see as a loss when not pushing “to the utmost”.
5: MAYBE- Rate each build with a “difficulty of play factor”. The issue here, of course, is that people don’t agree on what is or isn’t difficult. During several of the different Barb Buff threads that Free and I created, we went around and around on which builds were hard and which weren’t. And that was just him and me- when other people got involved, they had completely different feelings on the matter too. So, maybe it’s safer if we DON’T do this, and just let people decide for themselves what is or isn’t difficult.
1 Like
I’ve said this before… this season, unmasked by any seasonal theme that gives extra power like an extra cube slot or ethereal, should give plenty of balancing data to work with. Hopefully everyone who enjoys playing this game will participate and grind the paragon to give Blizz a clear picture of how all sets are doing. Then they will be able to balance the classes as the game moves into its twilight years.
The Echoing Nightmares giving oodles of XP should aid players getting to the 3-5k paragon quickly and generate a butt-tonne* of data.
Fab thread. Thanks to you all.
* I work in metric. Don’t @ me.
I whipped up this GR Clear Adjuster to make this stuff go faster. You just type in the level of the clear, the paragon of the player, and the time of the clear (in seconds), and the spreadsheet does the rest.
1 Like
This season, as there is no powercreep, we will have the mirror of power for all classes transposed to the improvement of Greater Rift.
Adjustments.
Like I said, I’m not sure which of these are from the PTR. We should redo any that could be replaced with results from Live, and asterisk the rest.
Adjusted: 144.7
Adjusted: 145.5
Adjusted: 144.3
Adjusted: 142.4
Adjusted: 142.3
Adjusted: 141.9
Adjusted: 142.7
Adjusted: 140.7
Adjusted: 142.9
Adjusted: 140.0
Adjusted: 140.5
Adjusted: 138.9
Adjusted: 140.5
Adjusted: 139.3
Adjusted: 138.8
Adjusted: 138.1
Adjusted: 139.4
Adjusted: 139.6
Adjusted: 138.3
Adjusted: 136.9
Adjusted: 137.3
Adjusted: 136.1
Adjusted: 134.7
Adjusted: 133.7
Adjusted: 134.0
Adjusted: 134.9
Adjusted: 134.2
Adjusted: 132.4
Adjusted: 132.2
Adjusted: 132.1
Adjusted: 133.3
Adjusted: 132.0
Adjusted: 132.6
Adjusted: 131.8
Adjusted: 129.2
Adjusted: 127.6
Adjusted: 127.2
Adjusted: 127.3
Adjusted: 127.7
Adjusted: 125.0
1 Like
The bottom of that list is depressing, although it is inevitable that something would be at the bottom.
I’ll get to this very soon – work has a firm grip on my soul atm.
1 Like
No worries, man- priorities are priorities.
Looking forward to getting back to this. How’s your schedule look this week?
Also, might want to port the thread over to GD before it gets locked.
1 Like
Where’s the FoH Valor clear?
Barbarian|H90|Frenzy|136|13:50|4048
Adjusted: 137.4
2 Likes