Statistics-based college rankings

<p>We've all kind of accepted the US News and world report rankings as dogma when it comes to which colleges are best, but a way to do it that takes about five minutes and gives results very comparable to the USNWR with their exhaustive methods is just to take 1-((acceptance)(1-yield)). the highest score anyone can get is a 1, which makes for easy comparisons, and the list itself is comparable to the us news and world report list in a pinch. There are some head scratchers (BYU at 24 thanks to its monstrous yield), but who is REALLY to say that Penn is absolutely without a doubt 6th instead of 7th. anyways, it's interesting. the 1st number to the right of the college name is its score, 2nd is usnwr ranking, and third is the difference between the two.</p>

<p>1.Harvard .98068 1 0
2.Yale .9693 3 1
3.Stanford .96910 4 1
4.Princeton .96702 2 2
5.MIT .96125 5 0
6.Columbia .95654 8 2
7.Penn .9456 6 1
8.Brown .9384 16 8
9.Dartmouth .92656 11 2
10.Cal Tech .89522 7 3
11.Notre Dame .8922 18 7
12.Georgetown .88976 23 11
13.Washington and Lee .88582 12 1
14.Duke .8666 9 5
15.Cornell .88658 14 1
16.Berkeley .86253 21 5
17.UCLA .85368 25 8
18.UNC .8460 30 12
19.Johns Hopkins .83962 15 4
20.USC .83555 27 7
21.UVA: .832 24 3
22.Rice .83183 17 5
23.Northwestern .82312 13 10
24.BYU: .8298 113 89
25.Nebraska .8202 89 64
26.Tufts .81575 28 2
27.Emory .8096 19 8
28.Boston College .80617 34 6
29.Vanderbilt: .802 20 9
30.William and Mary: .7858 32 2
31.Carnegie Mellon .7844 22 9
32.Chicago: .776 10 22
33.Lehigh: .776 35 2
34.NYU: .7743 33 1
35.Pepperdine: .7655 56 21
36.Texas: .7654 47 11
37.Yeshiva: .7585 50 13
38.Brandeis: .7518 31 7
39.George Washington: .741 53 14
40.Wake Forest: .7354 29 11
41.UGA: .7354 58 17
42.University of Miami: .7226 51 9
43.Michigan: .715 26 17
44.Maryland: .7039 54 10
45.Northeastern: .7036 96 51
46.SUNY- Binghamton: .7036 77 31
47.FSU: .703 104 57
48.Ohio State: .6991 57 9
49.Clemson: .695 62 13
50.NC State: .694 87 37</p>

<p>Where is Wash U. in St. Louis?</p>

<p>If I had to guess, the low yield killed it, eric.</p>

<p>Although I’m not sure why Chicago is on the list, then, since I thought those numbers were pretty comparable.</p>

<p>In any case, this is a method that’s probably true at the top, but starts to peter out quickly.</p>

<p>actually that was just a fail on my part. washington and lee = washington in st. louis. my bad.</p>

<p>A better way to do this is:</p>

<p>rank </p>

<p>(yield)*(Avg SAT)/(admit rate)</p>

<p>The admission rate should be the one before the school takes people from the waitlist.</p>

<p>Then caltech would probably shoot up, as will Rice and MIT.</p>

<p>I think yield is overrated - can be too easily gamed. Some schools take almost 50% of their class ED which skews that data, as does aggressive waitlisting.</p>

<p>One issue with this is that acceptance rate and yield are not independent variables at all, so that in effect this is just a way of double-counting yield. Acceptance rate = slots/yield/applications. So all your formula measures is which schools get the highest yield and the most applications per available bed.</p>

<p>Kolyyu: Actually, I think W&L is a small, very conservative school in Virginia.</p>

<p>Wash U in St. Louis is in… Wait for it… St. Louis.</p>