noojens wrote:That thing is almost completely useless.
I disagree... It's completely useless, lol. All it does is use an arbitrary formula to calculate a weighted average of the components of one's application and compare it to arbitrary values they assigned to each program.
-Firstly, the relative importance of things like GPA, GRE, Recs, whether you have a major in the desired field, Minority Status etc... vary by school and are not published. The published data on page 2 only give some averages and cutoffs for GPA and test scores, so your "estimator" results can't really be based on more than these, which clearly aren't enough to give a sensible estimate. This thing can't prentend to evaluate your whole application with that magical weighted average "score" which it says it does.
Even the categorical questions don't make sense...
-Students aren't going to know what their recs say, and even if they've seen them, how "good" they are according to each committee. Nor will they know how good their own SOP is... what they say could be liked by some schools and disliked by others. We may make reasonable predictions about these, as we often do in our profile thread for instance, but to quantify our recs is ridiculous.
-The formula used here could only be accurate if it reflects the formula used by each school. But again these aren't published, and most schools don't even use formulas. It comes down to individual profs, on a committee which changes each year, looking at individual apps and evaluating them however they see fit.
-The formula doesn't even consider the strength of one's undergrad program (which obviously influences how GPA is interpreted).
-By twistor's experiment, it seems the VGRE and QGRE are equally important... downwright wrong for most schools.
From all this, you can't possibly estimate "chances" at a particular school. All the system seems to do is say that better stats give a higher weighted average which gives a higher chance of admission at a given school, and a stronger school requires better stats. DUHHHHH!!! Don't we already know that? The system really isn't personalized to each program's individual preferences and idiosyncrasies. It looks like it's based only on an overall prestige number for each program. The estimator bars are vague and and all you can infer from them is... likely admitted, probably, maybe, probably not, you're screwed... Is that helpful? People already know what schools are reaches and what are safeties, and for schools in the middle, is getting a "probably" vs. a "maybe" from this system going to influence anyone's decisions?
I can't go through every imaginable reason why that system is garbage, but I hope enough people here realize this so that nobody embarrasses our field by trying to create one for physics students, lol.