chunkyda

Once the data is entered you can create any graph you want. Here I have compared my computed vs car's trip computer "tank average" gas mileage and a plot of cost of gas per gallon. I also keep notes such as when a tank includes a highway trip or those rare times when my wife drives (she has a lead foot). This data entry takes a few seconds every week and it allows me to determine cost to own and operate my cars, it also would indicate a problem by showing reduction in mileage over the course of time.

jayclapp

You have enough data to be able to calculate and plot plus and minus 3 standard deviations from the overall average. Then you could apply standard statistical control chart rules to see if the process is stable.

If stable, then you would know that all points that occur within the plus and minus 3 standard deviation limits are normal deviation, meaning everything is normal and nothing has gone wrong with the car. Occasional points that fall outside of the plus and minus 3 standard deviations would be attributable to special causes and you could find the reason for it and "fix" it. For instance, a long cruise with high mpg would probably fall above the upper limit, but you would know that was the cause and not be concerned. Or if you really wanted high mpg all the time you could only drive the care on long cruises. Of course that is ridiculous, but just trying to make a point.

Statistical control charts are invaluable if you have a process that is stable or can be made stable.