rollingdf.loc[:, '_Row'] = rollingdf[['PlayerID', '_Row']].groupby('PlayerID').cumsum()