Photometric redshift accuracy. Small black dots show the mock set of simulated afterglow spectra and their corresponding photo-z. Thick blue lines show the average photometric redshift after distributing the 4000 mock afterglows into redshift bins of 100 afterglows each. The lowest two panels also show in blue-shaded areas the quadratic sum of the typical difference to the input redshift and the 1σ statistical uncertainty of the photo-z analysis averaged over 100 afterglows in absolute (Δz = z_phot − z_sim) as well as relative (η = Δz/(1 + z)) terms. The statistical probability that the true value is in the interval of +/−20% around the quoted 1σ accuracy is  >95%. Grey shaded areas represent the z_phot > 0 constraint. The thin, horizontal, dotted lines in the lowest two panels represent Δz = −0.1,−0.05,0, +0.05, +0.10 and η = Δz/(1 + z) = −0.1,−0.05,0, +0.05 and +0.10. The large red dots show final UVOT/GROND photo-z measurements for real bursts where a spectroscopic redshift has been obtained (see Table 1). The green dot shows the photo-z of the flat-spectrum radio quasar PKS 0537-286 derived in a similar manner (z = 3.10; Bottacini et al. 2010; Wright et al. 1978).