net with your handle. I promise it will only be used for statistical purposes if you so request. Thanks again Andy. Larry Kaufman, USSF Chairman --Boundary_(ID_8ke0pg1ZVVt1tO3V9PQQ3Q) Content-type: text/html; charset=iso-8859-1 Content-transfer-encoding: 7BIT

"Intercept 1409.449, slope 0.424628. In other words, to convert a SC24 rating to Pan-Atlantic based on that scale, multiply by .424628 and add 1409. The R-squared is around 72%. Accuracy is pretty weak, with a standard error of 96 points." Andy Lipscomb.

Thank you Andy.

Wow, that suggests that the SC24 ratings are much more spread apart
than I imagined (I guessed a .8 slope before gathering all of this data).

Could you redo the calculation with the slope set arbitrarily at .5 (I can't
believe it could really be lower than this with enough data) so we can have
a simple formula to convert (divide by 2, add ?). Thanks for your help.

(Andy did this giving an intercept of 1298, which we can round to 1300.

So based on the available data, the best simple conversion of an SC24 rating

to Pan-Atlantic is to divide by 2, then add 1300. So 2000 becomes 2300,

1600 becomes 2100, 1200 becomes 1900. 1300 would become 1950, which

is about midway between 2 Dan (1860) and 3 Dan (2000) on Pan-Atlantic scale.

So Mr. Niwata's table listing 1300 as 3 Dan is only a slight stretch, based on this.

If he listed 1400 as 3 Dan, he would be exactly correct according to this data.)

We need more data. Anyone with ratings on both scales based on at
least 30 games PA and 80 SC24 who might not have a handle known to me,
please email me at lkaufman@comcast net with your handle. I promise it will
only be used for statistical purposes if you so request.

Thanks again Andy.

Larry Kaufman, USSF Chairman

--Boundary_(ID_8ke0pg1ZVVt1tO3V9PQQ3Q)--