Hello all,
OGS has put together the following survey: https://www.surveymonkey.com/r/goserversurvey2015. We’d appreciate you taking a moment to fill it out. Our goals are (1) to help us here at OGS improve our existing ranks and ratings and (2) to provide information for the benefit of the go community. In light of the latter goal, we’ll release the survey results as a summary and raw answers.
If you have any questions, feel free to ask!
On point number two, is there any chance that the results could be published on sensei’s library, so that the wider community can access it from a common point? I know that a lot of people use that as a point of reference, however I believe that it’s quite out of date at present.
The results will be published here at the very least, and anyone can add them to sensei’s library.
Very initial results posted here: https://gist.github.com/thouis/0b43424e9a61dd84478f
I’m happy to do deeper analysis, based on what people think would be useful.
I think @anoek will post the raw data, at some point soon.
Hi, thanks for the nice graphs.
Can you confirm that I’m reading them the right way?
First I think the negative values are kyu levels and positive are dans?
The OGS-KGS graph has a median shift of 1.69. I seem to see and understand that on average (or rather median) people have a higher rank on KGS than on OGS. Quantitavely this would be about 1.69 higher --> 10 kyu on OGS would on median be 8.31 on KGS.
Does this sound right?
Oh, sorry, I forgot to fix the axes. 0 = 1d.
You are correct on the rest. OGS ranks are about 1.7 stones lower than KGS, so KGS 10k ~= OGS 8k.
huh. well thats interesting.
it really looks like there are two kinds of rankings. TYGEM and WBADUK are
co-linear, and everyone else is co-linear, but they are kinda cubic wrt to one another,
with a crossover point around SDK.
does anyone know if there are cultural or algorithmic reasons why those are the same as
each other and so different from everyone else?
thx 