WISQARS vs UCR: The Numberings

So the Brady's put out this little bit of trite:

Senator Jeff Sessions has repeatedly voted against expanding lifesaving Brady background checks. He has been a champion of the evil PLCAA law that strips gun violence victims of their day in court. And during his time representing Alabama in the Senate, gun homicides have risen 29 percent in his home state.

So I hop over to the Alabama UCR and 'Disaster Center' and note that while there was a little bit of difference between the two, both showed a decrease in overall homicides since '97 when Sessions was first elected. Neither, however, showed 'gun homicides' separately.

So where did the Brady's get their info from?  I figure it's from the 'GunPolicy' site, their usual source, and they utilize the CDC's WISQARS.

Note the differences in the UCR compared to WISQARS on 'Homicides'.
2011: 311 to 392
2012: 325 to 405
2013: 342 to 417
2014: 270 to 374
2015: 353 to 473

and it goes all the way back actually showing an increase in homicides from '99 (the year WISQARS started and two years after Sessions took office, near the end of a drop in homicides from their peak years earlier).  A difference of 20-30% higher numbers than the state and FBI provide. Not a small error.

So whom to believe?

Update: Discrepancy likely found. Medical examiners determine 'Homicide' as " the death of a person comes at the hand of another person."  iow including justifiable, police, and accidental whereas the UCR's use the definition of "violent acts subject to criminal prosecution"  . The exact reason the CPD tries to redefine them as 'accidents' etc because fewer people use the CDC numbers.