That is not quite correct - it would be more correct to say:
The "rows between" syntax is often left to the default, which is everything up to the current row, based on the order provided. Thanks for promoting analytic functions, Scott, always great to see :-) Nice, simple way of providing the user information that can be easier to read & interpret. Note this report looks a lot cleaner with display setting 'Marker' using 'None', as opposed to 'Circle'. In this case it explicitly specifies the 'window' of rows to average to be only nearby data points. The "rows between" syntax is often left to the default, which is everything up to the current value (see post comments), based on the order provided. The third column uses an analytical function to calculate the average of the 6 surrounding counts at any given point on the x-axis. ,round(avg(count(*)) over (order by diff rows It's fairly easy from a SQL point of view - in this case it's another column in the original chart query
This stabilises the results, like looking at climate vs weather. What if you would like a smoother version of that line - a moving average, if you will (red line). Consider a chart with a trend that might be quite jagged across data points (blue line).