This Dataset is created from the data in this post in The Guardian Datablog. All the periods are combined in one dataset and the scores for the previous period is added to every score.
