I included the five weeks before the December downturn begins to establish a bit of a baseline. Aside from Thanksgiving week (Week 10), both years are settled at a relatively consistent "True PUT"* of about 35.0 or a bit lower. The first week of the holiday downturn is typically the second full week of December, Week 13. It bottoms out in Week 14 (week ending 12/25/11 or 12/26/10) at around 31.0. Even though broadcast barely tries at all in the period between Christmas and New Year's (week 15), there is typically a relative viewing recovery in that period to about the week 10/week 13 level. Then in week 16, midseason stuff and the NFL playoffs arrive and we begin the most-viewed period of the year.
How does the Week 14 viewing compare with other low-viewed times of the year? Week 14 last season (31.5 TPUT) still had more viewing than 14 of the first 15 weeks of the summer. The week ending 7/3/11 (27.9) was lowest, but most summer weeks are somewhere in the 29s or 30s.
To break it down by day, I'll do a chart of the PUT changes week-to-week. This is better than just breaking down the PUT by day because it lessens the effect of the day of the week.
The holiday downturn doesn't start right at the beginning of Week 13, but at least by Wednesday of Week 13, viewing is consistently on the decline, and it remains that way pretty much through the end of Week 14. (They vary a little more wildly at the end of Week 14 because Christmas Day falls on a different day of the week.) I don't have my Week 15 estimates for this year, but last year saw an about even week-to-week Monday, then a pretty consistent upswing through the rest of the week (New Year's Eve excepted). So the viewing is on a continual decline for about two weeks, from early week 13 through early week 15 (depending on the day of the week on which the holidays fall).
*- "True PUT" is the old methodology Persons Using TV I use in the True Strength calculations. After 3/27/11 (so for the 2011 portions of these charts), that includes the Methodology Adjustment.