Model A serial number confusion

I have three serial number charts. The first three years are a bit off; one list shows 8 experimental tractors in 1933. From 1936 to 1941 they are in agreement.
Then it all goes nuts!
The chart from TractorData.com does not agree with my "Dunning" JD Tractor Data Book or Yesterday's Tractor chart.
The tractor Data Book agrees for the most part with the chart here at YT.
EXCEPT in the year I'm trying to decipher which coincides with the ending of WWII. There are three tractors in question around the year 1946; Dad's 558312, My 563648, and Uncle Forest's 575930.
From TractorData.com, Dad's and mine are 1946, Forest's is 6320 tractors into 1947.
From the JD Tractor Data Book, all are 1946; Early, Mid, and Late.
Using Yesterday's Tractor chart, Dad's is a late 1945 (which we've always called it; it was sold to him as a 1945). Mine and Forest's are 1946.

The yearly production totals are all over the board, however on my spread sheet the total # of tractors agree on all three!
Here is a screenshot of the spread sheet I made to chart them. Note the color coding.
Let me explain the comparisons on the right side:
Yellow: If charts 1 and 2 agree, Yes or No.
Blue: If charts 2 and 3 agree, Yes or No.
Green: If all three charts agree.


The very bottom boxes indicate how far into the serial numbers for that year that the tractor is.


So, which chart is correct, or does anyone really know?

Well first off, I don't pretend to know. I have the same problem with the B serial numbers charts. My B is a 37 on some charts, a 38 on others. Someone here once explained that Mother Deere changed production year beginning and end dates from Nov 1/ Oct 31 to some other months (that I don't recall,) but then changed back again sometime after WWII and it has remained the same ever since. Hence, this may be the origin of the confusion. Maybe someone else with better knowledge than I will chime in here. PS Two Cylinder Club research says mine was shipped out on Sept.28, 1937. So I believe it to be a late '37.

The "year" of one's tractor can be a complicated issue.

When Deere first published beginning serial numbers, they used the beginning of their fiscal year (Nov. 1 to Oct. 31) as the basis. This data was printed and duplicated all over the place, including those little wallet cards initially furnished by Two-Cylinder, and it was copied by many others. Later, Deere published revised data as better research showed that the Production year (usually considered the "model" year) was the appropriate basis for determining the beginning serial number. Unfortunately, much of the copied data which is incorrect still exists.

Initially, the Production year did coincide with the fiscal year, but in June 1938, the Production year began to diverge from the fiscal year and model years started to change to the same time the factories had their summer shutdown, usually late June to early August. This means a tractor built in (say) August 1942 would fall into the 1943 production (model) year. Regardless, there are always exceptions. The Serial Number Registers, until the early 1950's, are clearly marked to show when the production (model) year changed, and the most accurate data for this information is probably Mr. Thinker's Almanac published by Green Magazine, taken from first-hand research of the Serial Number Registers by the late J.R. Hobbs. Data from Yesterday's Tractor is incorrect as their info is a virtual copy of Deere's early data, but YT made some typos in doing it - for example, Deere originally stated 588817 was the first 1946 "A", but that number is also incorrect as it clearly is out of sequence. So Yesterday's Tractor assumed somebody made a mistake and they assumed the number should be 558817 - and that is also a mistake as the authorities all agree that 555334 is the correct beginning S/N for the 1947 "A". Another problem with the YT data is for 1950 which they show as 646530. That number is unique with YT and it's unclear where it came from.

Anyway, all this concern about the "year" of one's tractor isn't too important. The tractor's serial number, which will establish what parts were used to build it, can be used to determine the date of its build (or more accurately, the warehouse date which was the date it was entered into inventory) since that information is recorded and available. And because many tractors received mid-model year changes (and some models were changed at an arbitrary time), the "year" of a tractor isn't necessarily the best identification, in my opinion.

Thanks, Dan! Figured it was something like that.
Uncle Forest died in 1978, Dad died in 2017 (at 102!). We have both of their tractors. Mine was a freak buy in 2005. A friend in Colorado Springs was looking for a "B". He sent me a link to an "A" asking my opinion. That one was 40 miles from me!!! Got it for $1500 with new rubber on the back. I've since put new rubber on front.

For some reason, it didn't look "right", but I couldn't put my finger on it. One day back on the home place I was looking our '45 over and it hit me: The cultivator brackets!! So I pulled them off and put them on the '46. Will never have a cultivator on it, but now it "looks right"!

The serial numbers listed on TractorData are the same as in the Almanac and on the current printing of the Two-Cylinder Club card.

Not all serial numbers ended up as production tractors, so the "number produced" column isn't quite accurate to actual totals.

(quoted from post at 15:17:28 02/04/20) The serial numbers listed on TractorData are the same as in the Almanac and on the current printing of the Two-Cylinder Club card.

Not all serial numbers ended up as production tractors, so the "number produced" column isn't quite accurate to actual totals.

Click to expand...

Actually, the "Number Produced" is a result of a formula in my spreadsheet, not a number from a chart.

(quoted from post at 15:17:28 02/04/20) The serial numbers listed on TractorData are the same as in the Almanac and on the current printing of the Two-Cylinder Club card.

Not all serial numbers ended up as production tractors, so the "number produced" column isn't quite accurate to actual totals.

Click to expand...

Actually, the "Number Produced" is a result of a formula in my spreadsheet, not a number from a chart.