ROC - Receiver Operating Curve

An excellent resource is http://www.anaesthetist.com/mnm/stats/roc/Findex.htm.
The common misunderstanding I find is the following
- ROC is not same as traditional lift curve that marketing analysts use, though both look the same
- (1-Specificity) vs. Sensitivity is the plot of ROC where as Decile vs. Sensitivity is the lift curve
- The area under ROC is not same as area under lift curve
- Hosmer-Lemeshow test may not be the one marketing analysts will use to see the goodness of fit of the model; they may use area under ROC curve and KS statistic.
- The statistic "c" in SAS output is area under ROC curve
- KS just tells you what maximum differences occur in the ROC between cumulative response (sensitivity) vs. cumulative non-response (Specificity).