For the tick marks, either the wavelength/x-axis calibration is incorrect, or the lattice-parameters/space group of your sample is incorrect. If you are trying to refine peakshape with the tick marks in the wrong place, they will just smear themselves out to be broad jumps like you describe. I would fix the peak shape to some sensible value, and then work out why the tick marks are not where you expect first. You can then refine the peakshape afterwards.