I've been looking at this in The Linking Ring VI August Number VI (1927)

How To Find Any Day Of The Week On Which Any Particular Days Falls by W.W. Durbin

I really like the simplicity of the explanation and have looked on the net at the Doomsday method, Scott Cram's and other's etc. My questiion is am I OVERSIMLIFYING this presentation or am I DISCOUNTING the other methods because I think they are much harder to effect...

Here's the Durbin presentation:
1800-1899
Jan/Tue Feb/Fri March/Fri Apr/Mon May/Wed Jun/Sat July/Mon Aug/Thur Sep/Sun Oct/Tue Nov/Fri Dec/Sun (The next century is a day off of above).
(I think this is the only hard memorization part in the entire method).
To find the day of the week on which any particular date fell, proceed as follows: Take the preceeding leap year (for instance 1860) cut off the first two figures (18) leaving 60, divide by 2 and then subrtacrt from the quotient the number of years that the year given is removed from the leap year, divide the remainder by 7, the number of days in the week, and whatever the remainder is, this date will fall on the day of the month according to the table above. If there is no remainder then it will mean that the 7th day of the month fell according to the table given above.

For instance, take this example. On what day of the week did June 30, 1877 fall? Take the precedeing leap year, 1876, cut off the 18, this leaves 76, divide by 2 and you have 38. 1877 is removed one year beyond 1876so you subtract 1 from 38, leaving 37, divide 37 by 7 and it goes 5 times with a remainder of 2, therefore the 2nd Day of June fell on Saturday.

Isn't that easy? Or as my original questions asks am I dicounting these other methods as too difficult. Or have I possibly missed an entirely easier method?

Thanks again,

Jay

I am not quite clear how you derive Saturday from '2' by using the table above.

I think I can explain.

The calculation involving the year yielded a result of 2.
(1877 = 1876 + 1
76/2=38
38 -1 =37
37 divided by 7 leaves a remainder of 2).

The month/day-of-week conversion table tells us that in 1877
Jan 2 was a Tue.
Feb 2 was a Fri.
Mar 2 was a Fri.
Apr 2 was a Mon.
May 2 was a Wed.
Jun 2 was a Sat.
Jul 2 was a Mon.
Aug 2 was a Thu.
Sep 2 was a Sun.
Oct 2 was a Tue.
Nov 2 was a Fri.
Sat 2 was a Sun.

Since Jun 2 was a Sat., so was Jun 9, Jun 16, Jun 23, and our target date,
Jun 30.

---------------------------------------------------------------------------

Another (1800-1899) example:

What day of the week was Oct. 15, 1845?
1845 wasn't a leap year so we go back to 1844.
44 divided by 2 is 22.
Since 1845 is 1 year after 1844, we subtract 1 from 22 to get 21.
When you divide 21 by 7, there is no remainder so the 7th day of each
month falls on the days given in the table.
This tells us that in the year 1845,
Jan 7 (Jan 14, Jan 21, Jan 28) was a Tue.
Feb 7 (Feb 14, Feb 21, Feb 28) was a Fri.
Mar 7 (Mar 14, Mar 21, Mar 28) was a Fri.
etc.

For October,
Oct 7 (Oct 14, Oct 21, Oct 28) was a Tue.

Therefore Oct 15 was a Wed.

------------------------------------------------------------------------

One more example: Nov 19, 1863 (Gettysburg Address)

1863 is not a leap year. The previous leap year was 1860.
60 divided by 2 is 30.
Since 1863 is 3 years after a leap year, we subtract 3 from 30 to
get 27.
27 divided by 7 leave a remainder of 6.

By month/day-of-week conversion table, we know
Jan 6 (Jan 13, Jan 20, Jan 27) was a Tue.
Feb 6 (Feb 13, Feb 20, Feb 27) was a Fri.
etc.

Nov 6 (Nov 13, Nov 20, Nov 27) was a Friday.
Therefore, Nov 19 (one day earlier than Nov 20) was a Thursday.

--------------------------------------------------------------------------

One mistake and one omission in the Durbin write-up:

For years in the range 1900-1999, the above will place you two
days off (ahead.) You must go back two days:

Example:
Nov 22, 1963

Previous leap year: 1960.
60/2 = 30
30 - 3 =27
27 divided by 7 leaves a remainder of 6.
Since the 1800-1899 table matches Nov to Fri,
Nov 6, 1863 was a Fri.
Going back two days for 1900-1999 tells us
Nov 6, 1963 was a Wed.
So Nov 13 and Nov 20 were also Wednesdays.
So Nov 22 was a Friday.

That's the error. The omission: If the date falls in Jan or Feb of
a leap year, you must go back one day from the day given
by the above algorithm.

For years in the range 2000-2099, you must go back three days.
(So for Jan/Feb dates in a leap year in this range, you must go back
four days...three for 20xx and one more for JanFeb/leap.)

Today is Dec 19, 2005.
The previous leap year was 2004.
4/2=2
2-1=1
So in 1805,
Jan 1 was a Tue.
Feb 1 was a Fri.
Mar 1 was a Fri.
...
Dec 1 was a Sun.

We don't want 1805. We want 2005, so we go back three days:
(Sat, Fri, Thu...or count forward four (7 minus 3)
days: Mon, Tue, Wed, Thu.)
Dec 1, 2005 was a Thu.
Thus Dec 8 and Dec 15 were also Thurdays.
Dec 19 is four days later. (Fri, Sat, Sun, Mon).

Thus Dec 19, 2005 is a Monday.

---------------------------------------------------------------------------
For my purposes, I'd rather use 1900-1999 as my base.
(And go forward two days for 1800-1899 dates, and backwards one day
for 2000-2005 dates.)
Adjusting the table gives
Jan--Sun
Feb--Wed
Mar--Wed
Apr--Sat
May--Mon
Jun--Thu
Jul--Sat
Aug--Tue
Sep--Fri
Oct--Sun
Nov--Wed
Dec--Fri

Stanalger: thanks for that, the penny dropped!
The corrections and the rebasing of the algorithm are welcome too. I had been looking for a method that did not involve much maths and this one looks like a winner.

Stan's the man!


Jack Shalom

That's what I was seeking! Happy Holidays!

Great, very usefull method, thanks a lot

Yes, I also liked it a lot.

Too bad more of us don't spend more time reading the older literature.

The easiest way to do the any day for any date effect is with Jack Dean's "You were There" which does all the calculations and work for you.

PSIncerely Yours,
Paul Alberstat

Thoughtreader,
Where can I find mention of Jack Dean's "You were There?"

Bob, you can find Jack Dean's "You Were There" on Paul's own site (click and scroll down), as well as on the Center Tear Site.

I discuss a few excellent references for the day for any date feat on my March 23rd blog entry, including my own new version, which allows you to quickly and easily determine the day of the week for any day in 2006 (I'll have a 2007 version of this calendar out later this year).

Take a look at the calculation on:

http://www.quincunx.org/calendar/index.htm

Any thoughts/criticisms/complaints/compliments about it?

I personally like the Doomsday method but I'm intruiged by Durbin's method.It seems much easier - I wonder why it isn't as well known as Doomsday.

JPL

This looks challenging and fun, but I have a serious question (as a beginner). How do you make this entertaining? It seems that there would have to be some way for specs to check your answer in order for it to be valid, otherwise you could just lie, saying march 7, 1826 was a wednesday. How do you prove it is true? I like Scott's 2006 calendar (both because it can be given to a spec- I assume- and they can check your answer, and because it contains a key to help the calculation). When people do this effect, do they usually hand out old calendars, or what?

Excellent question! Here are a couple of tips that can help make it entertaining:

• Make sure you use a day special to the spectator, such as a birthday or an anniversary
  
• Set it up as a challenge, framed as your mind vs. the calendar (not you vs. the spectator)
  
• Do it as part of another effect. Mention the day of the week as part of your post-peek revelation.
  
• Do it as a throw-away. Act confused at their reaction, as if you thought everyone was able to do this.
  

Strong Magic has several ideas that could be applied to this feat to help you make it entertaining.



Usually, when doing this effect, you should have a calendar to verify the date. With the Day For Any Date Calendar, the spectator is checking the calendar themselves (it can be used opened for their asking or closed, as in my race suggestion above!).

With You Were There, a 250-year perpetual calendar is included, so people can verify the date. With the Train Your Brain and Entertain software, in which I teach the legitimate Day For Any Date feat, I include an 8400-year perpetual calendar (1600-9999) that you can print. I've printed my own version on cardstock, with a clear plastic sheet for a front and back cover, so that it lasts a long time.

I think you can find a method in Corinda´s 13 steps...

One of the best references on this that I've recently come across is "Day For Any Date" by Sam Schwartz (published by Karl Fulves). He not only explains all the adjustments of the calculation method, but the reasons behind them. He also gives some excellent presentations, including ones that help build your confidence as you start performing the routine.

My favorite idea from the Sam Schwartz book is to hand out 7 calendars from recent years, one to each of 7 spectators. You mentally number the spectators from 1-7, left to right. Spectator 1 gets a calendar with a year whose key number is 1, spectator 2 gets a calendar with a year whose key number is 2, and so on up to spectator 7 who gets a calendar whose key year is 0 (7 mod 7). Important: None of the calendars are leap years, so you don't have to make any adjustments for that.

You ask an audience member who doesn't hold a calendar to call out any day of the week, and then turn to your calendar-holding volunteers and ask who has a calendar on which January 1st falls on that day of the week. For example, the audience member says, "Wednesday!" You would then ask each of the calendar holders to look at their calendar and raise their hand if theirs has a January 1st that falls on a Wednesday.

One person will raise their hand. This does 2 things: It establishes to the audience that all the calendars are truly different, and keys you into the year code automatically and secretly. Spectator 4 raised his hand? The year code is 4! You ask what year his calendar is (or look at the cover), as if you need to get an idea of the year. You then ask that person to give any date. You do the standard calculations, and you get the day of the week quickly! If the person says "May 13th", you add 4 (for the year) + 1 (for May) + 13 (for the day), getting 18, with 18 mod 7 being 4, you quickly know that May 13th on his calendar is on a Thursday (4)!


I've been toying around with another approach, too. When I started, I memorized the key numbers for the years 1900-1999, for the classic calculation method. Playing around with the Doomsday Algorithm, I noted that, whatever the year, you could simply add 3 to the year code and get the code for the day of the week that is the doomsday! For example, 2007 is a 0 year, so you simply add 3, which gives you 3. 3 is the code for Wednesday, so Wednesday is the Doomsday for 2007!

Does this work in leap years? Yes, and surprisingly with no adjustment! The year code for 2008 is 2. Adding 2+3, we get 5, thus giving Friday (5) as the doomsday for 2008!

What about 2011? The year code is 5, to which we add 3. This gives us 8, or rather 8 mod 7, which is 1. The doomsday for 2011 is Monday (1)!

You're probably wondering, "why don't we have to make adjustments for leap years"? Simple - you're subtracting 1 from the year code to adjust for January and February in a leap year, but you're also adding 1 day to February for the leap year (the doomsday is always the last day of February, regardless of whether it's the 28th or 29th). Since you're adding 1 and subtracting 1, this has the same effect as adding 0.


I apologize for this stream-of-consciousness post, but sometimes you just get inspired!