1.110683What we want to be comparing is <sin(\Psi_1 - \phi_{proton}^*)> which is (in words) the average of the sine of the difference between the first order event plane, \Psi_1, and the phi component of the protons momentum in the Lambda frame, \phi_{proton}^* (the * denotes Lambda rest frame). Of course \phi_{proton}^* is really a proxy of \phi_{spin}^*. This sine is equation (3) in this paper: http://arxiv.org/pdf/0705.1691v2.pdf, without the 8/(pi alpha) scaling. This comparison is focused on 19.6GeV. The results should not be corrected for resolution or any other factor. The data presented should also be for 0-80% (minbias) centrality. Finally lets post only on mass peak results.

Hopefully this does add confusion(! ) but maybe we can put our results in this document for comparison and update it when necessary. For clarity I'll put my stuff in a section with my name as a header.

-Isaac Upsal

Isaac Upsal:

I could paste a graphic here but since we're running just one energy (19.6GeV) I'll just quote some numbers instead. I prefer using Latex style symbol typing, if this is annoying and inscrutable let me know and I'll do something different.

The following data is using a convention for the BBC which is my own. If adc which corresponds to two shared tiles I take that to be an effective tile which is in the middle of where the two physical tiles are. For example tiles 7 and 9 correspond to adc channel 7 so I assign all of the adc of channel 7 to pi/2, which is midway between tiles 7 and 9. This is a sort of greying out scheme. The more typical way of doing this is to assign the entire adc to either 7 or 9, that choice is made randomly with a sort of coin flip. In the past I noted that this doesn't make a big difference. I will have to rerun to make a more fair comparison.

For "my" set of cuts - which is really Alex's set of cuts (maximized for S/sqrt(S+B)) - I get:
<sin(\Psi_1 - \phi_{proton}^*)>_{\Lambda} = 0.0004091 \pm 0.000132019
<sin(\Psi_1 - \phi_{proton}^*)>_{\bar{\Lambda}} = 0.000610146 \pm 0.000332158

For Shusu and Xu's set of cuts - which I copied from their code - I get:
<sin(\Psi_1 - \phi_{proton}^*)>_{\Lambda} = 0.0003831 \pm 0.0001963
<sin(\Psi_1 - \phi_{proton}^*)>_{\bar{\Lambda}} = 0.0007427 \pm 0.0004513

The following data is using the standard convention for the BBC and my cuts. If adc which corresponds to two shared tiles it is randomly assigned to one XOR the other. Also I do not do the recentering correction
<sin(\Psi_1 - \phi_{proton}^*)>_{\Lambda} = 0.0003744 \pm 0.0001220
<sin(\Psi_1 - \phi_{proton}^*)>_{\bar{\Lambda}} = 0.0004534 \pm 0.0003009

The following data is using the standard convention for the BBC and Shusu/Xu cuts. If adc which corresponds to two shared tiles it is randomly assigned to one XOR the other. Also I do not do the recentering correction
<sin(\Psi_1 - \phi_{proton}^*)>_{\Lambda} = 0.001064 \pm 0.0002631
<sin(\Psi_1 - \phi_{proton}^*)>_{\bar{\Lambda}} =  0.001228 \pm 0.0005811

*note, extremely minor difference. I forgot to change the mass range so this is using my nominal mass range of 1.110683 - 1.120683 instead of Shusu and Xu's nominal mass range of 1.110-1.120 which is an extremely minor difference