Matt, the SFL appear to be using different figures to calculate the averages than the ones which appear in the match reports. For example Livingston's lowest home gate according to the averages is 1,355. Yet the match against Queen of the South on Oct 27th is given as 721. Similarly when St Johnstone played Dundee on Aug 18th they recorded a season's best of 6,079 according to the match report but the highest according to the averages is 6,279.
The first assumption I made is that the SFL are taking press figures for the match results but using the figures returned to them for compiling the averages.
Unfortunately there is no way of knowing for certain as they only show the average, highest and lowest. There is no link for figures per club other than looking at individual match results.
And there is no guarantee my assumption is correct. I have a figure of 186 for Albion Rovers v Elgin City on Dec 1st taken from the Rovers website. That is the lowest in the SFL so far this season. The SFL have a figure of 268 for the same game.
Comparing their figures with my own only Dundee, Alloa and Forfar are absolutely identical. However a lot of the others are broadly similar. I have no difference greater than thirty in the Third Division or fifty in the Second (barring Queen's Park where the difference is 92).
You are right that it is the First Division where the 'discrepancies' lie. Other than Dundee where the figures are the same the closest difference between their averages and mine is 43 and the biggest is 142 (there are two others at 140). It's too time-consuming for me to go through them all so I looked at the closest - Stirling Albion.
Their website is well maintained and also has a page breaking down home and away support so their own figures can be cross-referenced. Their figures
addup to the same as the home and away fans
They also add up exactly to the figures I have and the individual match tallies are identical. Yet these show an average of 43 less than the SFL.
I considered two other possibilities. First I added Stirling's League Cup game v Hearts into the equation but that put the average above the SFL figure.
Then I noticed that the SFL still had a season low of 945 for Stirling Albion (should be 946 according to the official club website) not the 883 for the game v QOS on Dec 15th. So I took the home games played by Stirling Albion on Dec 15th and Dec 26th out of the equation and- eureka! - the average is identical to that of the SPL.
As the latest Dundee home game before yesterday was Dec 8th it is apparent that the figures in the 'Insider' are running two to three weeks behind the date of publication.
However that doesn't resolve the question of the differing figures mentioned above.