Compare the W^- distribution based on the two embedding versions the new  SL16g and the old SL16f:

• The number of events in the new sample is larger than the old one, checked with Devika she said in the subEmbeddingTemplate.sh for W^- if [ $sample == Wme -o $sample == Wmt ]then norm=0.34 it was (0.32) in the old sample, and this is why we have more events now, the ratio of the events in the new to the old is 1.06 with more events in the new version.
• Since the number of events are different, I did not take the ratio as in W⁺, I am showing side by side the distribution for the same quantity from SL16g and SL16f.

   Also I attached a  copy of the W- QA plots for both the new and the old version