<p>Yeah, I know what you mean, but if you get into Harvard, you do have a better chance of getting into Yale (because if you got into Harvard you probably have great stats, essays etc.), don’t you?</p>
<p>^ Not because you got into Harvard. You would still have had great, Yale-caliber credentials even if you had applied nowhere else besides Bemidji State. Your probability of admission at Yale would be the same regardless of the Harvard decision.</p>
<p>I’m being overly-technical because the mathematical definition of independence is critical to the discussion at hand. OP’s error was in assuming that he could estimate his probability of admission, and in doing so poorly by substituting the overall acceptance rate.</p>
<p>^Applying math in everyday situations. Haha.</p>
<p>Reason why statistics is so important.</p>
<p>The reach schools may all think that you are not the right fit for their school, and because they have the ‘privilege’ of being selective, they could reject you.</p>
<p>You can’t just add probabilities, you have to do binomial</p>
<p>Assuming it is straight probability, the chance of getting rejected to all 8 with a 10% chance is ~40%, 27% for 15% chance, but your “chance” may be lower</p>
<p>In my opinion, your chances at a college are either 100% or 0% and the admissions office decides which one it is. I don’t understand how you could have a 50% chance. If you got rejected, an identical applicant would also get rejected. Therefore, your application would have a 0% chance of getting accepted. If you apply to a million 0% schools, you still don’t get in.</p>
<p>I think that ^^ has some truth to it. I am no statistics whiz, but it seems to me that you fail to take into account that top universities tend to look for the same things. Therefore, the idea that if you are accepted into Harvard, you have a higher chance of being accepted to Yale and Princeton than someone who was rejected probably has some truth to it because an individual who was accepted to Harvard must have some kind of admirable trait that other top universities would probably also find attractive.</p>
<p>^ No… see posts #20 and #22</p>
<p>About how much variance do you guys think there is in the mood of an admissions officer from day to day, and how strong do you guys think is the correlation between mood of the admissions officer and a student’s acceptance?</p>
<p>I mean I’m pretty sure that it would be a very low correlation. But I also imagine that the admissions officers are, in general, pretty stressed so their moods might swing a lot. Hmmmm…</p>
<p>I also now want to have data tracking every admissions officer’s emotions (it would be according to the officer, though so that’s kind of eh for gathering accurate data, but whatever, this is a huge research project, tons of data, idgaf) throughout the admissions process.</p>
<p>But then you also need a control and you’re going to need to track their emotions all the time. So then they’re stuck writing stupid little diary entries about their emotions all year.</p>
<p>Would the survey have to track more than one year, do you guys think?</p>
<p>
Just my guesses:</p>
<p>Fairly high variance, we all have good and bad times.
The correlation is probably low on its own, but if you compare a group of similar applicants near the border for admission I imagine it might become quite significant.</p>
<p>
</p>
<p>Yesh. OP’s probability skills are lacking.</p>
<p>First, they aren’t independent. Given that you get accepted to Harvard, your chances of getting into other perceived “reach” schools is much higher.</p>
<p>^Totally wrong. If you got into Harvard, chances are that your application is stronger which leaves you in a better position for Yale, etc. You don’t have a better chance simply because you got into Harvard, you have a better chance because you are a strong applicant.</p>
<p>Okay, from a probability perspective they definitely ARE NOT independent variables. Consider this thought experiment. Suppose there is a 10% chance of getting into Yale and a 50% chance of getting into UConn. Suppose we have a student who only applies to UConn and we know nothing else about him. We would reasonably conclude that there is a 50% chance of him getting accepted. You should then roughly break even if you were betting on him getting in with 2:1 odds. </p>
<p>Suppose we have a student who has gotten into Yale but has not yet heard from UConn. Would you bet on him with 2:1 odds? Definitely, because his probability of getting into UConn is now very close to 100%. If they were independent variables then you wouldn’t have any higher expectation that the student accepted into Yale would get into UConn then an applicant who only applies to UConn. </p>
<p>Suppose acceptances are independent variables. There are about 15 schools which admit ~10% of applicants (super-selective). Then a random applicant has a 1 - .9^15 chance of getting at least one acceptance. This is 80%, which is clearly way too high. What the actual correlation is, I don’t know.</p>
<p>I know but you have to realize they really aren’t connected at all. Look at it this way: Getting into Harvard does not equal get into Yale. Being extremely smart = getting into Harvard and being extremely smart = getting into Yale.</p>
<p>^Yeah, it’s not because a student gets into Yale that his chances of getting into UConn increase. He had the higher chance of being accepted there due to his credentials in the first place. The Yale acceptance just exaggerates his original good probability.</p>
<p>
‘Totally wrong’ seems to be excessive. The main thing seems to be his using the word ‘independent’.</p>
<p>
Interesting, but wrong. Let’s say that the actual probability P(UConn = admit) = 0.99 for this student. P(UConn = admit | Yale = admit) still = 0.99. P(UConn = admit | Yale = reject) still = 0.99. The condition of being admitted to Yale does not change the actual probability of admission to UConn; it just changes our best guess of the probability of admission to UConn. The actual probability is not affected in any way.</p>
<p>Of course, this doesn’t do anyone hoping to apply the OP’s strategy much good because we can’t actually estimate the probability of admission.
If the student actually had a 10% chance, that would be completely accurate. Your error is the assumption that a random applicant actually has a 0.1 probability. They don’t. Their probability is either much higher or much lower depending on their stats.</p>
<p>As an aside, the relationship between SAT score and UConn admission would not be independent.</p>
<p>So as a summary, I think it would be safe to say your chances of getting in a college would only be between 1% and 99% if you are a borderline applicant.</p>
<p>Otherwise if your stats and essays are clearly below-par for a college, your chances are < 1%, and if your stats and essays are above-par, your chances are >99%.</p>
<p>This is assuming you’ve already completed your standardized testing and application essays.</p>
<p>I guess that means matches are actually schools that are between 1% and 99% chance for applicants, and reaches are schools with <1% chance for an applicant to be admitted to. In this case, it would not help much to apply to a ton of reaches.</p>