<p>For al6200: </p>
<p>Maybe it would make more sense to break the SAT I down into sections? I was under the impression that MIT doesn’t even consider the writing section. I would also imagine that a relatively low Math I score would not be a positive. If you scored CR 800, W 800, and M 650, I imagine the admissions committee might worry a bit about your ability to succeed at MIT.</p>
<p>and you know what? no matter how high or low your “chance” is… in the end it’s only whether you are in or not… a 99% chance or even 100% on this thing can’t guarantee you anything. </p>
<p>stop stressing over chances since it doesn’t do anything at all! Enjoy the rest of your high school!</p>
<p>I should add that from everyone I’ve talked to–on the campus, at my interview, at info sessions–the match is more important than scores and GPA and whatnot after you’re past a certain threshold. </p>
<p>Formulas are interesting, but not very useful. Everyone wants to know whether or not they’ll get in, but the only definitive answer is the one that comes out online (or in the mail) in the spring.</p>
<p>according to the old method I had 158% chance…deferred. I saw the new method but only after i had already done the old one and I’m too lazy to redo it now haha.</p>
<p>.1330772466</p>
<p>err… that was by mistake… Neways I have 85% chance of being accepted (YAY!!) using first method… Though I didn’t include GPA and AP factor as I am an international…
ps: you should also include domestic-international factor(dat’s an imp one)…</p>
<p>Yeah but it’s still nice to be able to get an indication of whether MIT is even possible. So I think the formula isn’t completely worthless.</p>
<p>.1337 is 13.37%, which is pretty solid. </p>
<p>I’m thinking about integrating geography / ethnicity into my data. Then I might pull some more tricky statistical stuffs. This is making my brain hurt. </p>
<p>But I’m seriously thinking about sending this formula in to MIT as a supplement to my application.</p>
<p>It’ll will be fun… You should send it definitely… but isn’t the deadline anyways dead…
ps- .1337 was just a part of it… 85% sounds much better than 13.37%… ;-)</p>
<p>lolz, 1337.</p>
<p>Yeah, it still over-estimates your chances, because it places so much weight on the subject tests. </p>
<p>It is after the deadline, but since its a supplement (not required), I think it will be okay. Besides, I’ll overnight it, and I don’t think they make the decision until like February or later.</p>
<p>I think I’ll link to this thread somewhere in my letter so that if they google the formula, they’ll know its mine and not stolen.</p>
<p>
This is correct – supplemental information can be sent through the end of January, approximately.</p>
<p>woot, really? Sweet. I have the perfect ending too. </p>
<p>I seriously hope no one has done this before…</p>
<p>50.44%.
Accepted EA.</p>
<p>My results:
1st eqn: 90.4%
2nd eqn: 64.8%
3rd eqn: 36.3% [My GPA (3.79) and my 8 AP 5s or 4s dragged me down 9%]</p>
<p>Nice work al, but probably nowhere near accurate…</p>
<p>did you get in TapedDuck? </p>
<p>Also, 30% seems like a reasonable number. </p>
<p>The 1st one was kinda flawed since the data set was way too limited. The second data said was far more accurate, since it took into account more than 3 times the number of people as the previous one. </p>
<p>The third data set was standardized so that it better matched the data that MIT publishes. </p>
<p>But is 30% unreasonable? I don’t know, seems sensible to me. I’ll try testing it out on my data set to get an accuracy percentage.</p>
<p>The AP stats actually were best suited to a quadratic regression (although it was still very limited), where students who took about 5 APs did the best. </p>
<p>But thanks for the compliments.</p>
<p>50.1%. I already got deferred EA, so by the law of averages ;), I should be good RD.</p>
<p>^^ 
I like that thought.</p>
<p>I’m creating a new system for weighting SATs and GPA that will lead to much better results. Currently in the works…</p>