Most testers are simply betraying themselves by saying they are testers not judges.noobpwnftw wrote: ↑Mon Feb 22, 2021 7:22 pm I don't see how it is a beneficial thing to do to make a gigantic list of entries that are irrelevant.
So here is the thing:
If you test every random SF clone that people may release, how is that different from just testing every SFdev commit like NCM? Do you do that to your "general" rating lists? I don't think so.
Now, I also have a plenty of experiments of different NNUE arch, size and they vary in performance, in fact, I probably explored more architecture variants before some other guy re-invented them, do you consider that I would be offering an "edge" to the independent testers if I just go by nicknaming all my nets and let people play around it all? I don't think so either.
So I'm quite mystified on what the standards being for them in technical terms other than having a skill set involving anus licking.
Actually they all judge too! But it seems they cannot accept this truth themselves.
Otherwise all rating lists would be full of Raubfish, JudasPro, Salt, Orca and whatever, but they are not!
So they judged and their judgement led them to decisions, which are/were doubtful or plain stupid sometimes, that's it.
To be fair, there are also 'so called' wannabe rating lists, which really don't 'judge at all', but their results are disgusting of course.
https://chessengines.blogspot.com/2021/ ... 10221.html
(you can guess by the names what was cloned or not)