We have discovered a striking example of how the use of impact factors to judge journal performance can bias the whole evaluation system (Nature 465, 845; 2010, Nature 465, 864–866; 2010, Nature 465, 870–872; 2010).

A surprise in the 'all journals' category of Thomson Reuters' impact factors for 2009 is the meteoric rise to second position of the journal Acta Crystallographica A. That journal's impact factor, which has not exceeded 2.38 in the past four years, has hit a whopping 49.93. Such startling fluctuations are rare — compare, for example, the impact factors of Nature and the New England Journal of Medicine over the same period at 26.68–34.48 and 44.01–52.58, respectively.

Acta Crystallographica A had 5,966 citations in 2009 for 72 articles published in 2008, of which all — except one — received no more than three citations (data from ISI Web of Knowledge v. 4.98). However, a paper by G. M. Sheldrick entitled 'A short history of SHELX' (Acta Crystallogr. A 64, 112–122; 2008) clocked 5,624 citations. It seems that this article could be responsible for the sudden dramatic inflation of the journal's impact factor.