I remember a time when a lot of people flat out called for nerfs to barbs. I don’t ever remember a time of reading people “dog whistling” because that is not a term I use to describe peoples posts.
It is a type of coded language that apparently only people like you seem to be able to pick up. Although, I really think it is a type of attack against people whose message you don’t agree with.
You act like you 100% know he was calling for nerfs to barbs even though he never once asked for it. The part that confuses me is that you tell me I don’t know what it is and then point to a post describing something that wasn’t dog whistling.
You have a lot of merit to argue his numbers and his methodology about his post which is why I get confused as to why you would then go the route of adding on about dog whistling. Don’t stoop to such underhanded tactics it just makes you look weak.
Micro, it has been mentioned to you quite a few times already and you are still not seeing your error in your approach. Simply averaging some numbers do not qualify for a representative data. What you are doing does neither approximate nor describe the behaviour of the data. You simply average top 10s.
Imagine a formula 1 race with 20 racers. And you are taking the top speed of the top 5 racers. Then you average these speeds and call those numbers as the average speed of the racers for the whole race. This is what you are doing.(Imagine Wizards are in F1 league, Necro in F2 and Barbs are racing at F10 league). You cannot compare a boxer from heavy weight to a boxer from light weight.
Yes you are suggesting a 10gr worth of buff, which may or may not be correct, but your way of thought reaching to that conclusion is flawed.
It is not a robust way of thought/analysis.
I think you have quite a long thinking that you should carry out before posting suggestions or presenting “analyses”
Currently I am on holiday and I dont have access to a fast internet. However, I see no use doing a 2.6.5 analysis at this stage. However, once 2.6.6 is out, I will definitely do carry out my approach, which considers low-average-high parts of the data spectrum generated by players, which makes it more evident what level of buffs will benefit low-mid-high tiers of players.
However, do note that this type of a general buff only elevates the gr clesring potential at the same clearing efficiency. We actually achieved this back from S11 to S12. Afterwards, the real problem of barb class had to be addressed through legendaries and items and this did not happen.
Also I cannot pinpoint specific posts from past regarding your questions as I dont have to proove myself to you. We have been very active in these forums so these are our observations made in the last 5 years. We kind of know what works for devs or not.
Given this fact, I would argue that it is incorrect to call it my approach. My approach was based entirely on Free’s. I think that it would be better to say: “The approach used by Free and duplicated later by you has problems.” My response to this statement is “I fully agree but his/my approach has limitations/problems. I feel that is good enough with acknowledged caveats.”
I am not convinced that Free can argue that the merits of his methodology to calculate these numbers is superior to mine since we used the same methodology. These numbers and approach are from the 2.6.6 barb buff proposal website linked above. Free and Rage are the ones who made that website.
I do not have a problem with Free’s method. It is good enough. I fully recognize that the method is far from perfect.
My problem is simple. The first person in that thread that Free directed others to that disagreed with the methodology was Free himself (see post #10). Ironically, he complained that the methodology used was invalid. He failed to recognize the fact that he did the exact same thing on his website (i.e., averaged the top 10 solo greater rift clears per region). That irony was not lost on me.
Some posters claim that Free’s numbers/methods are good/appropriate. When I post Free’s numbers directly from his website/use his methodology (without saying that these are from Free), I get eviscerated by some of the same people who tell me to go to his website to see the right way. I dislike hypocrisy.
I look forward to this future post. The prior analyses that you have done were quite good. Ideally to have the best estimate of class power, you need to consider paragon levels, legendary gem levels, augments, hours played, and how many competitive vs. casuals are playing each class. Unfortunately, we players do not have access to all these variables to make a perfect comparison.
I think one thing that you need to show is that there is a very good correlation between paragon levels and two key variables: augments and legendary gem levels. For people who play primarily non-meta classes, one would anticipate that their average augments and paragon levels would be lower.
Also, you have talked about class power relative to the average Joe. I do not think that you or anyone else has really captured that. I base this on the premise that the average joe is not on the leaderboard/top 2000 per class.
Anyone can argue anything at anytime that they feel like. I am all for a lively debate about how things should be or why they should be. I am not for underhanded tactics like calling someone a dog whistler.
I get it man, I really do. I still think there is always room for debate and discussion and really only wanted to make a point in not using shady tactics to shut down these discussions. I do think using those numbers is the quick and easy way to get to the same conclusion that barbs (and other classes) still need buffs/tweaks. I like Pro’s method a bit better as it go way more into detail and allows for more nuanced changes to happen.
Aliens? Robots? People who have come into contact with a dictionary? Fascinating!
Ah, no. Please read my posts carefully:
Not quite.
The data on your GD thread is not the same as the data from Rage’s and my site. We looked at the top 10 non-Season GR clear for every region, then averaged it solely to prove that Barbs (and Monks) clear significantly lower than all other classes in non-Season.
There’s a reason we emphasize non-Season, because our concerns is with the top-end power potential of the class. Top-end power potential is not necessarily reflective of low-end or mid-end, and though we didn’t publish it, we did crunch the numbers on those clears as well. Season data had to be excluded due to the instability created by Season themes and exclusive bonuses, but also because the presence of botters is felt more heavily in Seasons; by the time you have 4-6k Paragon, all botting is doing is fetching you keys and mats.
Our philosophy has always been: Work from the top down. Look at what we’re capable of on the high-end, buff accordingly, and make special exceptions for builds that are severely underperforming (primary skills, IK HOTA, etc). Our analysis looks at global averages only to examine solo clear potential in non-Season. After that, we get into the specifics of wall-charging and so forth. In other words, even in our proposal, the numbers are not the true tell.
Where people are taking major umbrage with your work is A) the emphasis on group play, and B) the argument that Barbs haven’t been the worst for a very long time.
My response to both is that A) that’s a separate issue, so leave it out, and B) in eras where we cleared higher than some classes, it was due to broken sets and exploits and we still haven’t been addressed in other ways.
In other words, the numbers alone are not the final tell. My Measured Response thread spells that out plain as day.
Context and details make numbers significant. We’ve asked you over and over, so here’s me, once again, asking nicely: Leave Barbs to Barb experts.
Negative, mah dude. You don’t understand how Barb builds work.
+10 GRs = zero for WW in fishing, rolls, frustration, front-loaded damage, etc.
+10 GRs = zero for Frenzy or Bash or Cleave. It means nothing for wall-charging. It means nothing except slightly larger base numbers.
Boulder Toss will still be garbage. Slam builds will still be 10-15 GRs behind HOTA. See where this is going?
Global buffs are not the answer. I can’t say it any more plainly.
Buffs need to come from targeted supporting legendaries. Barbs have received nothing for a very long time (so help me, if someone says Mortick’s is something, you’re fired). Look at my Measured Response thread. It spells it out in detail.
One more time, Micro: Leave Barbs to Barb experts. Stop making threads about this stuff. Stop with the suggestions. I’m happy to discuss things with you on the forums or answer questions, but we’re all better off with less arguing.
I am primarily focused on cross class parity. I also have made a thread on having better parity for solo versus groups that is of interest to me. I think that my desire to have class parity is a hypersensitive subject because barbs are the worst solo class currently in terms of higher greater rift clears.
On ESPN, they often have the segment called blind resume to compare players analytically. I am curious about your philosophy on class and build parity.
Edit: For clarity, a two fold increase in power level “in D3 terms” is the ability to clear a greater rift that is ~4.5 GRs higher.
Which would be the best of 1-5 in your mind?
Which would be the best of 6–8 in your mind?
Power Goal = 3000
Original Build Power
New Power of Scenario 1
New Power of Scenario 2
New Power of Scenario 3
New Power of Scenario 4
New Power of Scenario 5
Best Build
1000
3000
1000
1000
1000
3000
Build 2
500
500
3000
500
500
1500
Build 3
400
400
400
3000
400
1200
Worst Build
250
250
250
250
3000
750
Power Goal = 3000
Original Build Power
New Power of Scenario 6
New Power of Scenario 7
New Power of Scenario 8
Best Build
1000
3000
3000
1000
Build 2
500
1500
3000
500
Build 3
400
1200
400
3000
Worst Build
250
750
250
3000
P.S.
Point of clarification: My OP had the data for seasons and eras. I copied your table verbatim at around post 200 and was told that that table was “bull” also.
This seems to be one of your biggest issues. The “It” = wall charging bug began “before 2.4.1”. If the barb community has been advocating for this to be fixed for years without any luck, I am not sure why would expect it to change now.
I’m entitled to my opinion even if I am not a barb expert. Remember what Nev said. Nevalistis Community Manager
This is simply a question about preferences not about any class in particular.
The numbers gives a builds functional power in arbitrary units (the higher the number, the more powerful). The goal for functional power is 3000.
A build whose functional power is 1000 is twice as powerful as a build with 500 power, and four times that of 250.
Column 2 refers to the functional power of 4 builds for an underperforming class prior to buffing).
Column 3 -the end refers the new power of each build following a potential buff. These column represent different scenarios of buffing to achieve at least 1 build of 3000 power.
Rows 2-5 refer to 4 builds (most powerful to least powerful within a given class).
I hope that helps.
For scenarios 1-5, my preference is for 5 then 4.
For scenarios 6-8, it is more difficult to choose. I prefer 8 most then 6 then 7 the least.
I really don’t understand why you are going through these mental gymnastics to try to prove some esoteric point. Let’s look at some actual almost-real-time data from today. I finished a 106 gr with my seasonal wastes barb. That earned me rank 230 on the barb leaderboard (fwiw, highest I’ve ever been on a seasonal board…yay me…ahem…). I was fairly impressed with myself until I looked at the other leaderboards and discovered (ok–confirmed what I already knew) that I wouldn’t have made the top 1000 on any of them. You want my preference? I want my rank on the barb leaderboard to be +/- 50 of the rank on any other leaderboard. I know this isn’t going to happen, but that’s my preference. I’d also prefer we try to get back to the point at hand, but I suspect that’s not going to happen either.
What does any of this mean? Why are we using abstract numbers and nonsense terms like “functional power?” What does that mean? What are you talking about?
When my undergraduates present me with someone like this, I call it “gross convolution.” Then I make them re-write the essay until there’s a coherent dang thesis.
I have no response to this because it’s meaningless, abstract nonsense. Nothing really refers to anything else. Nothing has a tangible referent.
You want to talk about tangible referents?
There’s one.
Here’s another: Barbs are, globally, 4.1 GR tiers behind other classes in solo GR potential.
In conclusion, The Grapes of Wrath is a really good book, and look into their eyes, ma, you’ll see me, you’ll see me, you’ll see me, you’ll see me
Crap–here I thought I was the only professor who still played Diablo. I’m going to have to re-examine my priorities…
Oh–and as for a thesis, I’m a fan of the old “assume you are right–tell me what we should do.” I like to throw it back on the students to try to make sense of their own gibberish. That wouldn’t be completely out of line here.
Excellent–good luck with that. Back in the day I focused on early modernists, but I like the later American stuff, too. You can’t go wrong with Malcolm Lowry (Edit: of course, after I posted this I remembered he’s English–still love Under the Volcano).
Lol I am an engineering professor too. Different from Free, I already have a PhD. Good luck with the tunnel with a light at the end. I would not dare doing another PhD
I have a saying: If you get 10 academics in a room, you may end up with 11 opinions.
From our discussions, it was clear to me that Free is coming from a non-STEM background that is consistent with him being an English professor. His writing style has a certain eloquence that mine lacks. His arguments still have room for improvement with respect to nuanced data interpretation and analytical reasoning. My self-assessment clearly highlights the fact that I need to improve in multiple areas. Educators definitely value lifelong learning as a worthwhile endeavor.
For Prokhan, I thought that he had a STEM background, but his terminal degree was in engineering/technology rather than science/math. It makes sense to me that he is an engineering professor. I based my guess on his educational background on two things: 1) The data analyses that he present and 2) Once he said that augments and legendary gem levels are 100% correlated with paragon level. No professional statistician/mathematician would make this claim. A statistician’s determination would be more refined and presumably dumbed down for this forum because of lack of knowledge about the educational background of forum members. A professional statistician would claim that it was highly correlated and if addressing a more educated crowd, might talk about R value or other statistical parameters. Needless to say, lectures/presentations are very different if given to undergraduates, graduate students, postdocs, or colleagues in the field at national meetings. I appreciate the fact that Prokhan’s assessment of Free and my calculations is that both are insufficient to assess completely accurately differential class power. I disagree with his conclusions and think that there are a “reasonable” approximation with noted caveats and limitations. When I was criticized for my statistics/calculations, I responded with this:
This statement would not be coming from an uneducated imbecile. I admire Free’s passion but you may want to ratchet it down as illustrated by your OP in “Hold on. Pause. Excuse me?”, the creation of this thread, and other comments that you have posted. Creation of this thread also is walking a line where it might be construed/misconstrued as targeting another poster for potential harassment in a different thread.
My OP presents how the data was acquired and how the data was analyzed. Anyone could look at what I did and then draw their own conclusions, including questioning the data and its analysis. Of all the posts in that thread, the one that is particularly astute in my mind is Rashiels’ who questions sampling bias since we do not know the number of people who play each class and of that number, how many are playing competitively to push the top build. When I looked at Prohkan’s analysis of season 11 (I think), I noticed that for certain classes, the slope of the curve increased at the top end relative to high end clears. This likely tells us something about that dataset relatively to high end player number (i.e. that these classes have less high end players). Franky, I think Rashiel may be right in that there is inherent sampling bias that may be a significant confounder, because we don’t know how many people play each class in general and at a “competitive level”.
Lets start from common ground: We three agree that barbs are severely underperforming currently and need a buff. The conversation about dog whistling, ulterior motives, and precise methodology that we used to reach the conclusion that barbs are bad is not core to the issue that we all want addressed in that underperforming classes should be buff. Free has his preferred method of buffing support legendaries for barbs while I am primarily focused on buffs to achieve cross class equity relative to each class’s top build.
I think that we may be missing the forest for the trees. If Blizzard implemented all of Free’s changes for barbs that would be best. If they did this, all the other classes would want the same thing and they would be making their own websites. They would also want changes to shake up the meta.
The game is 7 years old. The best case scenario in my mind is that they buff 1 or 2 builds per class occasionally. D3 has a skeleton crew/classic games running the show and we already paid for the game. Developmental resources are being dedicated to new Diablo releases as mentioned in the 2nd quarter ATVI investor report. Kotick claimed that we would be able to experience these Diablo projects soon.TM
Back to my question about preference for scenario 1-5 and preference for scenarios 6-8. The reason I want to talk about an abstraction and what Free considers “gross convolution” is to remove some noise from the discussion.
I just got to ask, why would you throw out a chart of numbers that only have some kind of abstract meaning only to you and expect that anybody else would understand where you are coming from?
Honestly Free, I think you must have the patience of a saint.
Between the continued fight to get changes for Barbs, and dealing with the trolls, idiots, and people who just don’t get it, I am surprised you haven’t flipped out and told them all to jam it, or ended up in counseling by now.
