I have one final question.
It's been suggested, and I don't know how seriously, that the more people you have in your data bank, the more people can be exonerated by a no-match sample, so we haven't found the criminal but we've exonerated 156,000 people every time we run a match. Do you think that has any value as a statement of principle in terms of the value of a data bank of this nature?
Anybody care to comment on that?