The monopoly which is Nielsen is now admitting that they made an error when putting out the February ratings for the Boston market. 

Nearly a month after February sweeps results were announced, Nielsen finally admits the numbers were bogus and point to a single household. 

Nielsen was forced to take a serious look at the irregularities of the February book after WBZ, WCVB, WHDH, and WBTS (NBC Boston) all filed formal complaints.

The ratings issue arose toward the end of the February book when WFXT FOX 25 saw a sudden surge in their ratings. The ratings increase did not go unnoticed at the station, and station management started celebrating the ratings wins - FOX 25 general manager Tom Raponi grew a "ratings beard" and the station threw the staff a pizza party.

Before the sudden rise in the ratings, FOX 25 had been losing to WHDH at 10pm for the first time - and was also losing to WHDH in the mornings.

WBZ was also on track to win at 11pm. The rise in the ratings skewed the monthly average in such a way that FOX 25 was able to pull out the wins in the morning and at 10pm against WHDH, and at 11pm against WBZ.

Nielsen's investigation found that the ratings anomaly was tied to one household in the sample.

New England One reports that the particular household did not watch FOX 25 regularly, then all of a sudden started watching only FOX 25. The ratings bump was also due to visitors of this particular household being punched into the people meter to count them as additional viewers.

The Nielsen investigation has concluded that no one in this household works at or is affiliated with FOX 25, but Nielsen cannot say who the visitors are.

Really? 

It appears that the greatest jumps were during the 10pm and 11pm newscasts, but 7am and 6pm also saw some irregularities.

In the adults 25-54 demographic for example, FOX 25 registered big bumps between February 21 and March 2.

Nielsen sent this memo to 

Amazing that the Boston stations pay Nielsen thousands of dollars every month and the numbers can be skewed so badly by a single meter. 

Kind of shows how small the sample really is, doesn't it?

H/T New England One