Right - working on the back log here:
Mike,
I'm glad to that Vanguard popped in to underscore my point: That the website were clearly explains what I do and do not use the data for. So you're butchering the straw man when you say people are so stupid to think that I used any data for anything.
I use data to figure out which teams are unbalanced (my explicit definition) under CRP rules. As I explicitly state on the site.
I do not use data to figure out which exact shape these team specific changes should then have. Which I also explicitly state.
You're welcome to put in the work and make better changes. That would be awesome.
But this is what I'm prepared to do for now. And there's no secrecy about it.
I'm presuming they don't understand statistics enough to understand why they're relevant
I have no way of divining what it is you specifically think I don't understand. I've picked up some basics since we first met, and I stick to that. You're right I'm not a statistician.
But the last time I got confused, it was because you ranted that people (only name-dropping me) don't use inferential data. So I mistakenly assumed that you actually knew what you were talking about and had looked at the website - so I did half a rewrite only to have Dode tell me that the stats I do use are in fact inferential. Thanks by the way.
So I do - at the basic level - understand mean and CI, and how to calculate them. And I did.
You didn't go back, do analyses, and then build your rules around what the numbers showed.
I did actually do the analyses, and both removed stuff that was unsupported by the data, and added stuff supported by data, that I had not seen before. So yes, I did make stringent changes as to which teams were nerfed/buffed. But as explicitly stated, I did not use the data to decide on the specific shape of said nerf/buff.
You went back and looked for numbers that would help support what you'd already decided was right.
No.
Which leads my back to the implication that I sinisterly use some data in order to mislead people to think that I'm actually using other data for other things.
Trip Down memory lane, shall we?
I had my original rules, based on mainly discussion and unscientifically eyeballing some stats.
I was quite happy with those rules.
But than you above all others (but also Shteve0 and Dode) felt like I was pretending to have proper scientific backing, and demanded that I started working with the existing piles of data that was available.
And so I did. I analysed to the best of my ability the existing CRP data that you spoke of. And I made changes to my rules accordingly. And presented that on the website.
So now you and Dode claim that this use of CRP data, which you specifically asked my to do, is a sinister ruse to mislead people into thinking that I'm actually using CRP+ data to sell my snake oil. Even though you had me do it, and even though I explicitly spell it out on the website.
That. Is. Just. So. Rich.
Sigh.
In fact, very close to the top of the page on my site that presents my roster changes it states that:
"To do this, I’m using inferential statistics to examine which teams have been overperforming or underperforming under CRP rules. For further details read on here [Link]. It is worth noting, that I’m not using the same definition of the tiers as the BBRC – most notably I use 30-40 point TV bands rather than lifetime performance, I don’t include mirror matches, I raise the bottom of tier 1 and I move up tier 2 and 3. The result is a narrower definition of the tiers which in turn leads to changes to teams that do not seem imbalanced under the BBRC’s definition of balance.
The results is changes to 10 of the 24 official rosters indicated by the inferential data.
As for the particular changes made, they were based on online polls, personal preference and feedback from players. They were not based on match-level statistics, and I do not pretend that these are the only ways that these rosters could have been fixed. In short, the changes are intended to nerf or buff particular teams, but whether they work as intended remains to be seen."
So, I think it is very clear.
And I object to the idea that you or Dode get to tell me what to do with my time.
Cheers
Martin
(PS: Edited a few confusing spelling errors)