Free and Prokahn,
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.
Of 1-5, I told you that I prefer 5 then 4.
Of 6-8, I prefer 8 then 6.
What say ye?