So, the one person Mr. Sardi quotes as saying the numbers are fixed is John Williams of Shadowstats. Let's look at his alternate numbers.
He claims he's only adding "long term discouraged" to the U6 measurement. Discouraged is defined as willing to work, could work if offered, looked for work in last 12 months but not last 4 weeks because of a belief that the individual would not find work. Williams is claiming he's only adding in the people who would be classified as discouraged except for the 12 months time limit.
The U6 is (Unemployed + Marginally Attached + Part time for economic reasons)/(Labor Force + Marginally Attached)
Marginally Attached has the same definition as discouraged, but stopped looking for any reason. It includes discouraged.
So...we can get Unemployed and Labor Force from Table A-1, and Part Time for Economic Reasons from Table A-8 (use the seasonally adjusted numbers) and Margianally Attached from Table A-16 of the Employment Situation Report.
That makes it (12,088,000 + 2,517,000 + 8,613,000)/(155,062,000 + 2,517,000) = 23,218,000/157,579,000 = 0.147 14.7%
So how many does Williams have to add to get 22.8%?
(23,218,000 + X)/(157,579,000 + X) = .228 Solve for X and you get 16,464,000 people he's adding. But if you look at Table A-38, we can see that there are only 3,278,000 people who say they want a job but haven't looked in over a year. That's including people who could not take a job if offered and people who stopped looking for reasons other than discouragement. So how the fuck does he get 16.5 million? We'll never know.
As for his claims that seasonal adjustment was different, well of course it was...August and September have different seasonal patterns. And the Unadjusted UE rate dropped from 8.2% to 7.6%, so it's odd to try to claim seasonal adjustment as a factor.
