The behaviour or these functions in VDDT was always confusing to users and so it was modified in ST-Sim to make it consistent and simpler to use. Best to forget about VDDT syntax and look at the online help at:
http://wiki.syncrosim.com/index.php?title=Probabilistic_transitions
Neither feature ever “trumps” the other – they do slightly different things and so now work together. Age Reset if calculated first, followed by the Age Shift.
First thing is to determine the effect of the Age Reset – this says whether or not to reset the age of a cell after a transition. Default is yes (which is what happens when it is left blank). If yes then the new age of a cell is set to the start age of the post-transition state class; otherwise the cell maintains its pre-transition age after the transition.
The age shift is then applied to whatever age is first calculated using the Age Reset for a transition. Default is 0 (i.e. no shift if left blank). An age shift of +20 would add 20 to whatever age was first determined using the Age Reset.
Note that the age of a cell can never be beyond the bounds of its state class min and max ages.
So you are correct in that the -15 for Age Shift is redundant. In fact the Age Reset= Yes is also redundant. Leaving both fields blank invokes the default behaviours for both: resetting the age (in this case to 0) and having no shift.