Take this BBC article as a prime example. The, probably, eminant doctor says something we can all identify with. Tiredness. Especially tiredness when the clocks go back 1 hour (in the U.K. at least this occurance happened early hours of today. 28-102007 at 2am) but then he spoils it all by saying "six out of 10 people" are failing to relax before sleeping and that this leads to "semisomnia" and that this change in time can take 3 days to recover properly from.
It is not what he says overall that is annoying, just the obvious figures out of thin air bit. I mean, everyone can identify with tiredness, especially feeling just that little bit more tired after the clocks are turned back 1 hour and for a couple of days afterwards we all can identify with feeling that little bit more tired than usual. But where exactly did those figures, "six out of 10", come from? I would hazard a guess that he looked through either his own log book or that of the hospital he works in and then pulled out all instances of tiredness in the 3 days following the clock change then calculated that figure against the population count as a whole. Not exactly rocket science nor does it make those figures correct. However, "6 out of 10" looks very impressive and is sure to trigger something in the minds of the readers but here, all it does is provide evermore proof that this type of portrayal of figures is nothing more than an ineffectual method that annoys.
This pulling figures out of thin air to backup or bolster whatever point is being made is quite simply ineffectual to all but the lowest of the low on intelligence. It is annoying. It is almost an affectation in so much as the person talking seems to have a driving need to back up that whih they are saying by pulling out outlandish figures to back up their claims.