According to SKY analysis http://www.skysports.com/football/n...ague-one-and-two-clubs-ranked-for-improvement I suppose this is fair enough as a basic analysis, but it only compares clubs who remained in the same division and doesn't take account of league position, which of course will be affected by clubs that are promoted and relegated into the division. To give perhaps a rather extreme example, unless Manchester City score more than 100 points next season their performance by this measure will have declined.
I can understand your reservations about the method used, however, using a percentage points improvement/deficit seems as good a way as any to measure improvement. The Manchester City example isn't particularly extreme, this is a measure of "improvement" so even if City finished the same next year they wouldn't have improved.
