GMAT difficulty Algorithm - GMATPrep vs Actual Exam

The short answer: the GMAT Club difficulty ratings aren't necessarily accurate! The difficulty of actual, official GMAT questions is determined empirically by GMAC (that's exactly why unscored "experimental" questions exist), but if you're looking at the GMAT Club difficulties, those are a completely different thing. The difficulty levels here at GMAT Club are determined automatically by the results of the "timers" on each page. Basically, if GMAT Club users frequently miss a given question or take a long time to answer it, then it'll be labeled as "high" difficulty.

Of course, that has nothing at all to do with the actual difficulty levels determined by GMAC. GMAT Club users aren't necessarily representative of test-takers as a whole, and we may or not behave the same way on the "timers" as we would on actual exams.

Put another way: if you're getting a 49-50 on the actual exam, then (by definition!) you're seeing plenty of questions that are officially "difficult."

Aditim12
Premises : GMAT prep mock test 1 and 2 are VERY old, I guess 2010. They (GMAC) have not changed the questions.
Conclusion : Don’t determine difficulty level from those mocks.
Assumption : GMAT Questions from the previous years/decades were considerably easier than the ones being presented these days.

Flawed Assumption?

I wanted to know if this level of difficulty is in line with the actual GMAT exam as well. Has anyone else done this sort of analysis?
Also, Can someone tell me how GMATclub rates the percentage of Official question difficulty and how accurate is it as per official data.