They would have a rotating 7 year schedule, but it’s messed up by leap years. You have the seven calendars you’re thinking of and 1-2 leap year calendars mixed into those 7 years. It would have to be somewhere between 1 in 8 and 1 in 9, wouldn’t it?
There are 97 leap days every 400 years, then the calendar repeats. So you have 303/400 chance of not having a leap year, and in those years, you get a 1/7 chance of having this calendar. Thus 303/2800.
This is counterintuitive to me, because 303/2800 is .108, which is between 1/9 and 1/10. But 97 out of 400 is less than 1 out of 4, so it shouldn’t be able to interfere more than twice in a 7 year cycle, on average. But your math looks correct. I must be missing something.
1 in 7 chance [if you sample from infinite years]
the first day of the month moves forward one weekday each year except mar-dec on a leap year which moves forward two weekdays
That can’t be correct, can it?
They would have a rotating 7 year schedule, but it’s messed up by leap years. You have the seven calendars you’re thinking of and 1-2 leap year calendars mixed into those 7 years. It would have to be somewhere between 1 in 8 and 1 in 9, wouldn’t it?
I think it’s more like 303/2800 chance.
There are 97 leap days every 400 years, then the calendar repeats. So you have 303/400 chance of not having a leap year, and in those years, you get a 1/7 chance of having this calendar. Thus 303/2800.
This is counterintuitive to me, because 303/2800 is .108, which is between 1/9 and 1/10. But 97 out of 400 is less than 1 out of 4, so it shouldn’t be able to interfere more than twice in a 7 year cycle, on average. But your math looks correct. I must be missing something.
No, since there’s only 7 different possibilities, then over a sufficiently large sample, the probabilities would all still balance out to 1 in 7.
There’s 14 different possibilities because of leap years.
Oh yeah, you’re right. I was focusing on just where the first day of the week lands, not the full month calendar.