VERIFYING FROM PUBLISHED DATA -- 7 processors, 524 spectator readings

  1. Weakly-detuned spectators report a LONGER look (published: -0.17 to -0.56, all seven)
     aachen           spearman -0.333   in range
     boston           spearman -0.192   in range
     fez              spearman -0.201   in range
     kingston         spearman -0.302   in range
     marrakesh        spearman -0.170   in range
     phoenix          spearman -0.558   in range
     pittsburgh       spearman -0.338   in range

  2. Reported / stated, spectators above 4 deg/us
     chip          recomputed published       n   
     aachen            1.0172    1.0172      23   OK  
     boston            1.0256    1.0256      27   OK  
     fez               1.0022    1.0022      54   OK  
     kingston          1.0457    1.0457      43   OK  
     marrakesh         1.0282    1.0282      31   OK  
     phoenix           1.0166    1.0166      63   OK  
     pittsburgh        1.0266    1.0266      43   OK  

  3. One constant across all seven
     pooled           1.0207 +- 0.0092   published 1.0207 +- 0.0092   OK  
     chi2/dof         0.52                 published 0.52           OK  
     p(one constant)  0.79                 published 0.79           OK  

  4. Against no effect (ratio 1): 2.2 sigma -- not established

  5. Without the cut, the same arithmetic
     pooled 1.0405   chi2/dof 2.22   p = 0.0384
     -> one constant does NOT fit until the ill-conditioned spectators are removed.
        The cut is stated in advance and applied to every processor equally.

  6. The 30 spectators that report a NEGATIVE elapsed time
     30 of 284 above the cut  OK   / OK  
     These come from a noisy estimator near its conditioning limit. The cut below was
     chosen AFTER the data were seen, so it is shown for transparency only. Dropping ONLY these:
     pooled           1.0331 +- 0.0083   published 1.0331 +- 0.0083   OK  
     chi2/dof         0.88                 published 0.88           OK  
     vs 1 (no effect) 4.0 sigma   (it is 2.2 sigma with the negatives left in)
     -> post-hoc, so not reported.
        The pre-specified reading (2.2 sigma) is the one reported.