It was raised in the last jet meeting, that for my data the average track pT had dropped fro ~0.9 GeV (plots in Novemer or earlier) GeV to ~0.75 GeV (plots after November).  In that time period I changed from reading the MuDst's directly into histograms, to reading the MuDst's into an analysis tree.  When coding the new method to save everything into a tree, I found a bug and wasn't saving all my tracks(or towers) into the tree.  This has been corrected and as shown below the two methods agree.

Histogram taken from the tree(left), Histogram made using the method that reads the MuDst and directly saves the histogram (right). 

Now both histograms have the same entries, mean, and RMS tho the binning isn't exactly the same.  Also since I wasn't saving all my tracks for both data and embedding, it could be the cause for the track pT discrepancy.  If I was saving lower pT tracks in data and higher pTs track in embedding; this would lead to the embedding being much harder than the data.