But as you can see, no one from ARRS is reading this forum. Asking many times serves no purpose.
However the answer to many questions are at their website. The only claim attributable to ARRS is that the results are "statistically valid". Rather than guessing what that means, and over-reacting to wrong conclusions, all you need to interpret it is explained on the home page: http://www.arrs.net/ under the notes.
This homepage answers two questions:
"What does 'statistically valid' really mean?"
"Does 'statistically valid' mean it should be a world record?"
You look silly concluding that they think it should be a world record, especially when they never said it, and what they mean is clearly explained on their homepage.
You don't need the algorithm for most questions. You do need the algorithm if you doubt that the times are 1:37 fast for men, 1:42 fast for women, and 13.5s slow on average. But everyone can do it for themselves independently, and compare their results. The data is publicly available. Everyone can compute expected times, and compare them with actual times, defining whatever filters for outliers they want, and compare the results.
There is actually quite a lot of information about which data is included, and criteria for excluding data, and how the data is used.
Did everyone asking the questions read the website? What question do you have that actually needs the algorithm?
A Better Question wrote:
How about ARRS being more specific? How many times do people have to ask for this? Actual algorithm and actual data is what is required to allow others to critique what ARRS is supplying. Without this, it amounts to nothing more than "Hey, just trust us and shut up". Not very convincing.