Since 2000, the lowest % the SOA has passed for Exam 3/M was 31.9% (and that was in 2000). Since 2005, every Exam M had passing %'s of 41.7% or more. Nov 2006 had 46.7% of people passing with an almost definite pass mark of 24/40. Meanwhile, the other forum had 78% of exam takers (over 100 sample size) with a score of 24+. This time on that forum (sample size 89 as of today), only 15.7% of exam takers scored 22+. This is from a forum where respondants historically do much better than the overall exam population. 25.8% got 21+, 38.2% got 20+, 44.9% got 19+, 57.3% got 18+, and 65.2% got 17+.
I really think the pass mark will be 18 or 19 (MAYBE 17)
the other forum had 78% of exam [M] takers (over 100 sample size) with a score of 24+... This time... 65.2% got 17+.
I guess I'm missing the logic behind why you really think the pass mark will be 18 or 19, or maybe 17. You must be making some judgement about the sample size this time around -- thinking perhaps that as it grows, the percentage of people scoring 17+ will approach the 78% mark (the level that implies the pass mark, based on the poll taken last year)?
If the exams were of equal difficulty, I might expect the same thing -- a passing score around 18 (meaning, that as the sample size increased the percentage of people scoring around 17/18+ would likely approach the 78% level). However, knowing that this exam was slightly more difficult supports the trend the poll is already showing -- that is, a pass mark that will be lower (on a percentage basis) than the 24/40 we saw last November. As of now, I am going to continue assuming that the sample size does a decent job of mimicking last November's poll and not speculate on how the bias from the responses so far would directionally effect a theoretically "complete" poll. That leaves it at 15 or 16...
I thought I wouldn't even pass this exam. Yet, I got my results in the mail today and I got a 9. I really can't even believe it. I'm stunned. I wonder how few questions you could answer correctly and still get a 10.