I was testing some new code for the 2006 EEMC pi0 cross section, including new cosmetics for the final plot.  I was thinking about the dominant source of uncertainy being the EEMC energy scale.  I thought it'd be interesting to see what the plots look like if I took all the LW pT points and shifted by a factor and reduced the EEMC energy scale uncertainty.  Left--normal.  Right--artificial shift of the pT points by 5% and reduction of EEMC energy scale from 3% to 1%.  Maybe the analysis is fine but the calibration needs tuning or something is funny with use of dead towers???