When looking at the success rate or duration statistics for builds by day/month, if no builds occured for a certain day/month the success rate is shown at 0% for success and <1mil for durations. Could days/months where no builds have been done be skipped instead so that a direct line is drawn between the two surrounding days where builds actually occured? As it currently stands when looking at the success rate it's hard to tell bad build days from say weekends where no builds occured.