Receiver operating characteristics (ROC) curves provide critical performance insights for the evaluation of an authentication algorithm. However, there are two different versions of it in literature I'm not sure which one to follow.
ROC curves mostly appear to have the FAR or FPR on the $x$-axis and the sensitivity or verification rate along the $y$-axis as shown below.
Is there a significant difference between the two or is it just a matter of preference?
Is one more recommended over the other?
The latter is also referred in some articles as a decision error trade-off (DET) curve and accompanied by a dotted $y=x$ line for EER. Is it advisable to do the same in the former representation of ROC with a $y=1-x$ line?