An interesting thread!!
As the hackneyed phrase has it, Stats don’t lie!
An excellent analysis by vinayR....
Points, which will make this analysis even more interesting....
1. Nowadays most of the IITians and freshers from top engineering colleges join 1 coaching institute or the other, which drastically reduces the number of people managing a call without any mocks. (Keeping in mind most people working with TCS, INFY, CTS who sits for CAT also sits for mocks).
2. Now based on the above facts it is easy to assume that most of the people who are serious contenders for top slots sit for 1 of the mocks (be it from IMS,CL,PT,TIME or any other institute for that matter). Now adding all that people who generally appear for mocks would be around 65 thousand or thereabout which leaves around another 60 thousand junta of which most people are not serious(counting those ppl also who are registered with coaching institutes but harldly sit for mocks) about the exam at all and taking it just for the sake of it.
3. Based on the above assumption it’s easy to assume that your percentiles on the D-day are bound to improve. But the catch is the higher your percentiles the lesser the improvement. We can safely assume that junta scoring in 80s (percentile) in mocks would score around late 80s or early 90s. People scoring in 90-95 bracket consistently would score 95-97.5 on the D-day. But beyond that the chances of errors are more. I can bet my life on the fact that most of the junta who scored 98%+ in cat2k3 would have scored on an average higher than 95% in majority of the papers. And if this holds true for a skewed paper like cat2k3 (with such a huge bias towards VA-RC) then it would hold true for most
standard papers
Now a note of caution: If CAT-04 comes up with some shock factor like an entire section on LR or some compulsory questions or questions having different wts. then the above logic won't hold. But even then most of the 98%+ junta would be comprised of people who have scored 90%+ in most of their mocks......
Most important thing which summarizes the entire topic of discussion:
A higher probability of an event happening doesn't necessarily mean that the event would happen
Hope the message is clear