Treebeard's Homepage : Stumpers

Treebeard's Stumper Answer
12 November 1999

Five Fridays

I write these stumpers for the Dunn Middle School Friday Newsnote, and I keep collections of past newsnotes and stumpers on the Web. I noticed last week that there were five Friday newsnotes in the month of October 1999. I believe this is the first time that this has ever happened, though there have been other months that could have had five newsnotes if not for school vacations and other interruptions. This suggests a tricky stumper. Ignoring vacations, what are the most (and least) months with five Fridays that can occur in a single calendar year (Jan. - Dec.)? How uncommon is this event, and what hidden patterns does it reveal?

Five Fridays in October 1999.   Unlucky Strikes
98.6 in the Shade
World Party
Too Many Ancestors    
Ollie Up
  1 Oct 99
  8 Oct 99
15 Oct 99
22 Oct 99
29 Oct 99
Five Fridays in October 1999.    

Every year has at least four months with five Fridays, but this year there are five. The first four weeks of each month account for 4 x 7 x 12 = 336 days. That leaves 365 - 336 = 29 extra days that will become four five-day months for each day of the week, plus one day left over. A year is 52 weeks plus a day (or two in a leap year). So a year that begins on a Friday like 1999 (or a Thursday in a leap year) will have five five-Friday months. This result is quite general. January 1, 2000 is a Saturday, so there will be five five-Saturday months next year.


There are too many fives in this stumper!

Whew! Hyphens aren't enough for this, but brackets might help on that last one, for LISP programmers at least:
Let me explain...

I thought of this stumper when I first noticed there were five Friday Stumpers in October, but I didn't know if it would go anywhere. I checked a calendar and discovered that in 1999 there are four five-Monday weeks, four five-Tuesday weeks, four five-Wednesday weeks, and so on for every day of the week, except there are five five-Friday weeks! I knew I was on to something, so I made a bunch of copies of a perpetual calendar from an old almanac, and I started marking five-day months with a highlighter.

Five five-Friday months in 1999.

This year 1999 begins on a Friday as shown, and it's not a leap year. Everything else is determined. There are five five-Friday months (January, April, July, October, and December) shown in yellow, but only four five-Monday months (and every other day), as shown in cyan.

Leap years like the coming year 2000 are different in that they have two five-day months:

Five five-day months in 2000.

The year 2000 begins on a Saturday, and it is a leap year. (The year 1900 wasn't a leap year! Why not?) There are five five-Saturday months shown in yellow, and five five-Sunday month shown in green. There are only four five-Friday months (and every other day), as shown in cyan.

Eventually, I found this simple pattern:

Cool pattern! Then I started thinking about it, but Graybear beat me to it:

I'm kicking myself right now because I immediately tried to solve this by looking at a calendar and determining, day by day (Sunday, Monday, etc.) how many months had five of that day (That's how I had solved the Friday the 13th stumper on November 13, 1998 a year ago). Then I saw that it would have been more exciting to solve it in my head. Therefore, here is the 'cerebral solution':

Every month has either 4 or 5 Fridays - because each month has at least 28, and less than 35, days in it. If there were exactly 52 weeks in a year, eight would have 4 Fridays, and four would have 5 Fridays. Think of it this way - twelve months of 4 Fridays equals 48 Fridays, so the remaining 4 Fridays must mean that 4 months each have an extra. In other words (or numbers), ( 8 * 4 ) + ( 4 * 5 ) = 52. Since a year is actually 52 weeks plus one day, if a year starts on Friday, (Thursday or Friday in a leap year) there will be five months with five Fridays.

Exactly right, and to the point. I had to read this over and over to really understand it. There are too many fives, which makes it confusing!

The almanac's perpetual calendar is based on the fact that there are only 14 different possible calendars, depending on which day of the week the year begins, and whether it's a leap year. These different calendars repeat in a sequence of 28 years called the Solar Cycle. Each day of the week is the start of three normal years and one leap year in this 28 year sequence:

YearFirst day
of year
Months with
Five Fridays

1972SaturdayYes4Mar, Jun, Sep, Dec
1973Monday4Mar, Jun, Aug, Nov
1974Tuesday4Mar, May, Aug, Nov
1975Wednesday4Jan, May, Aug, Oct
1976ThursdayYes5Jan, Apr, Jul, Oct, Dec
1977Saturday4Apr, Jul, Sep, Dec
1978Sunday4Mar, Jun, Sep, Dec
1979Monday4Mar, Jun, Aug, Nov
1980TuesdayYes4Feb, May, Aug, Oct
1981Thursday4Jan, May, Jul, Oct
1982Friday5Jan, Apr, Jul, Oct, Dec
1983Saturday4Apr, Jul, Sep, Dec
1984SundayYes4Mar, Jun, Aug, Nov
1985Tuesday4Mar, May, Aug, Nov
1986Wednesday4Jan, May, Aug, Oct
1987Thursday4Jan, May, Jul, Oct
1988FridayYes5Jan, Apr, Jul, Sep, Dec
1989Sunday4Mar, Jun, Sep, Dec
1990Monday4Mar, Jun, Aug, Nov
1991Tuesday4Mar, May, Aug, Nov
1992WednesdayYes4Jan, May, Jul, Oct
1993Friday5Jan, Apr, Jul, Oct, Dec
1994Saturday4Apr, Jul, Sep, Dec
1995Sunday4Mar, Jun, Sep, Dec
1996MondayYes4Mar, May, Aug, Nov
1997Wednesday4Mar, Jun, Aug, Nov
1998Thursday4Jan, May, Jul, Oct
1999Friday5Jan, Apr, Jul, Oct, Dec

2000SaturdayYes4Mar, Jun, Sep, Dec
2001Monday4Mar, Jun, Aug, Nov

Count the years with five five-Friday months and you'll see that there are five years with five five-Friday months every 28 years. It's almost too much to mention that there are usually five four-Friday-month years between successive five-Friday-month years. Like I said at the start, there is a {five [five (five Friday month) year] year} cycle!

There's another unexpected pattern revealed here too. Five-Friday months occur in January, April, July, October, and December in a normal year or January, April, July, September, and December in a leap year. When any day of the week occurs five time in a month, five times in a year, it is always in those months!

The 28 year solar cycle breaks on century years that are not leap years according to the revision of the Gregorian Calendar by which century years are only leap years if they are divisible by 400. The year 2000 is a leap year, but 1900 and 2100 aren't. Does this effect the distribution of five-Friday-month years in the long term?

I tried my usual Web searches on this stumper and came up empty. Maybe this is an original?

Back to Stumper

Last modified .

Copyright © 1999 by Marc Kummel /