First of all, you need a table of data. Here’s one I’ve prepared, showing Quantity, Total Cost, Average Total Cost, Average Variable Cost and Marginal Cost:-



And this is what ATC, AVC and MC would look like in a graph:-



Except that’s wrong, because the MC line should start a half-tick to the left of the other data lines. So, how do we do this? The first step is to add another column to our graph. I’m going to call it MCx. I’m then going to fill it with the x-axis values for each corresponding MC value. (For large amounts of data, you can usually make this faster by observing that it’ll be the same as “Quantity – .5″.)



Now, here comes the tricky part – applying the custom x-values to the MC curve without affecting the other data. The first step is to change the type of graph you’re dealing with. You don’t actually want a “line graph” anymore. What you want is an “X Y (Scatter)”, specifically Scatter with Smooth Lines. You can see the graph type you need in the image below.



Colours notwithstanding, it looks pretty much like the last graph. Scatter plots have one major advantage though – you can select a custom x-axis for each set of data! So whilst the ATC and AVC lines will get their x-axis values from the Q column, we can change the MC line to use the MCx column for its x-axis values. To do this, right-click the graph and go to “Select Data”. Highlight “MC” and click “Edit”.



Now you can set the x-axis values to the column we created earlier. All you need to do is delete everything in the “Series X values” box, and replace it with your own data, in this case everything from F2 to F9.



Press OK a couple of times to get out of those menus and your graph will now show the MC line shifted half a tick to the left – where it’s supposed to be. Technically this trick can be applied to give ANY data offset values on an axis. The secret really lies in using an X Y Scatter plot.