The peculiar methodology of the research assessment exercise was always likely to produce strange league tables ("Reviewers raise concerns about RAE gameplaying", 8 January). Take the composition of the sub-panels: 15 or so senior UK academics, some of whom were nominated by subject associations, ploughing through hundreds of outputs.
No one could claim that all those outputs were RAE-assessed with the expertise brought to them by their publishers' readers. That community - which really does represent world opinion - would be better served by journal rankings.
Without disrespect to the people concerned, it would require unusual fortitude in a subject association nominee not to feel that he or she was there to ensure that the picture looked as rosy as possible when shown to the funding council purse-holders. On both counts, the secrecy of the panel meetings is scandalous.
And who on earth dreamt up the idea that you can judge the research environment of a department or university without knowing what proportion of staff take part in it?
David Roberts, Newman University College.