I've been wanting to do this analysis ever since I saw the site, but I've been far to lazy to type up all the values into excel and run my figures. Thanks for doing it.
I'd also draw your attention to:
https://www.admissions411.com/results.asp
According to that, GSB has a 40% accept rate. That seems high, but when you consider that the vast majority of data points are well into the 700 range, if not the 750 range, then it seems more reasonable to assume this is somewhat in the ballpark.
Now, I'll share what I did get to for GSB... this is totally back of the envelope stuff, but I had to do the math
760 1 accept
740 1 ding
730 2 accept
710 2 accept
700 1 accept
650 2 accept
640 2 ding
640 3 accept
610 3 ding
First value is gmat, second is number of alumni recs, and third is accept/ding. Based on these figures, were talking 66% accept. Thats high. What I wonder (for obvious reasons) is whether or not its too high because the sample is small, or if it's right?
If we hit R2
770 1 unknown
770 1 unknown
750 2 accept
750 1 unknown
740 1 accept
730 1 unknown
720 1 accept
720 1 ding
720 1 ding
710 1 unknown
680 1 ding
670 3 unknown
650 2 accept
unk 1 accept
unk 2 unknown
unk 3 unknown
unk 3 accept
That's only 35%. Not that high, but if we take R1 and R2 together and toss out anything under the 650 mark - we get a 47% accept rate.
Now just to try and fool myself thats its statisitically valid... lets see the percentages for those with more than 1 alumni rec and a score of 650 or above. Well, its high. Nearing 85%. The sample size though is tiny, so it's not that helpful.
Interesting anyway, especially if, like me you have alumni recs.
And anyone that enjoys stats, this is fun to read:
https://www.admissions411.com/bboard/top ... OPIC_ID=67
I especially like the guy who says "you are either a trolling or a fool". Hah.