The attached .pdf shows versus day in run-9 left-right and top-bottom asymmetries as measured online by the BBC with the algorithm originally developed by Joanna KIryluk.  The symbols alternate between fills between open and closed.  The first transverse period is easily recognized, as it has the expected large left-right asymmetries.

The averages of the asymmetries (_not_ analyzing powers; indeed, this could be refined) are:

Yellow Left-Right 0.004089 +/- 0.000032 (transverse)
Yellow Left-Right 0.000340 +/- 0.000005 (longitudinal)
Yellow Top-Bottom 0.000457 +/- 0.000041 (transverse)
Yellow Top-Bottom 0.000221 +/- 0.000006 (longitudinal)

Blue Left-Right 0.003763 +/- 0.000033 (transverse)
Blue Left-Right 0.000678 +/- 0.000005 (longitudinal)
Blue Top-Bottom 0.000177 +/- 0.000041 (transverse)
Blue Top-Bottom 0.000586 +/- 0.000006 (longitudinal)

and served as input to the tuning of the rotator current for the Blue beam that was (finally) done on Monday June 22, a week before the end of run. This tuning reduced the residual LR asymmetry by a factor of about 4.

A related presentation by Chuck Whitten at the June 19 RHIC Spin Collaboration meeting may be found here.