Ivy League Admissions Difficulty is Exaggerated

<p>I agrree that's easier than most ppl think, but some of you guys are making it sound too easy. It's hard to imagine that >2300 and even valed and decent everything else will get you into an ivy almost certainly. I think you're putting too much weight on SAT's and grades.
but i'm going to say I think the chances at different schools ARE independent. The P(A and B)=P(A)P(B) holds true I think, and a decision at one school will not impact the decision at another school</p>

<p>
[quote]
but i'm going to say I think the chances at different schools ARE independent.

[/quote]
</p>

<p>I had better quote a statistics teacher again. I invite any onlooker to this thread to show the various posts to an experience AP or college statistics teacher, to ask the teacher what the teacher thinks about the statistical reasoning here. </p>

<p>


</p>

<p>The chances of admission at different Ivies are not independent. Indeed, I'm quite certain that if you took two groups of STATISTICALLY identical students (same scores & GPA/rank) and measured their success across all Ivies, you would find that there was some consistency in acceptances (i.e., some applicants would get rejected everywhere, some would get a good number of acceptances, etc.) The key thing is that statistically identical studens aren't identical applicants - some will have demonstrable passion, leadership, etc. while others will appear to be boring bookworms. Many will fall between the two extremes.</p>

<p>Still, there is no doubt at all that chances can be improved by applying to more schools. Even for a student likely to be rejected at most schools, it's possible that one school may need a decent oboe player or that one application reader may find something in the application particularly appealing. There is certainly SOME indepence between decisions.</p>

<p>Here's how I'd look at it...
1) Students who meet the stats levels described by the OP will all survive the "stats cut" where applicants with marginal academics get whacked.
2) These applicants will also not lose many head to head academic comparisons with other applicants when various candidates are being discussed.
3) Since the defined level of academic achievement is outstanding, it shouldn't take as much on the EC side to tip the balance at one or more schools.
4) Since even superbly qualified candidates are still subject to random elements in the process (which eliminates many superbly qualified candidates at each school), more apps DOES increase the chances of success.
5) A candidate with great stats but few outside accomplishments shouldn't assume that playing the numbers game will result in success.</p>

<p>roger you are using statistically pertaining only to scores and GPA. when I say statistically equal I mean they are equal in all facets (scores, ECs, etc) and they have the same statistical chance of getting into any particular school. I realize being exactly statistically equal is not possible, and I realize there are schools that cater to a particular area (MIT and math, Chicago and essays).
ok I understand the not independent argument now. But still, I think we can approximate using the multiplier rule, even if it breaks the rules of statistics, it's just an approximation, and it still demonsstrates a good point.</p>

<p>i can however think of contradictions to the stats teacher's explanation.</p>

<p>"What is independence? It means that when you learn about the outcome of one event, it has no influence on your guess about the probability of success in another event."</p>

<p>so I have a coin and I tell you it's unfair. I flip and you call tails, but it is heads. Is that going to change your decision what you'll cal next. If I flip the unfair coin 5 times and get 5 heads, what are you going to call next? Probably not tails again (although I want to point out that it is still POSSIBLE to get tails). Did learning about the outcome of those five events influence your decision? So does that mean that the flips are NOT independent? I think not. We all know that fflips of a coin, whether fair or not, are independent of each other. Admissions decisions for an applicant are like unfair coins. Heads and tails are not equally likely, and admit/deny are not equally likely for an applicant.
Somebody (nicely) correct me if I'm making a humungous statistcs blunder</p>

<p>The way this thread uses statistics is pretty pointless, but it's (sort of, mildly) entertaining nonetheless. Reminds me of a certain cliche/quote: [Noun/pronoun] use statistics like a drunk uses a lamp post - for support rather than illumination...</p>

<p>
[quote]
We all know that fflips of a coin, whether fair or not, are independent of each other.

[/quote]
</p>

<p>Ask the best statistics teacher you know what the reasoning should be if one has </p>

<p>a) a fair coin, defined as a coin equally likely to land heads or tails, </p>

<p>and </p>

<p>b) a biased coin, defined as one known to land more often either tails or heads than the other way. </p>

<p>Then request the teacher to kindly read this thread, and to tell you what he or she thinks about the reasoning here. You're not paying me to teach you statistics, while the fine taxpayers of some jurisdiction are paying somebody in your town to teach statistics, so I'll simply say that the high school student who "reasons" that he has a 99-out-of-100 chance to get into an Ivy League college on the basis indicated early in this thread has a mistaken understanding of statistics. He may still get into one or more of the least selective Ivy League colleges, but that doesn't make the reasoning correct. </p>

<p>The 2008 U.S. News college guidebook reports that the eight colleges in the Ivy League enrolled 13,737 distinct students in the fall of 2006 (4,227 for HYP alone), and the College Board reports that the cumulative number of class of 2007 students who scored 2240 or higher on the SAT Reasoning Test was 12,909, so mathematically some student with a score below 2240 (or no SAT score at all) has to get into some Ivy League college. (Mathematicians call the principle that underlies this proof the Pigeonhole principle.) So, yeah, I agree with the thread-opening assertion that if </p>

<ol>
<li>You scored over 2300 on the SAT</li>
<li>You have a 3.9+ GPA with a challenging courseload</li>
<li>Have 700+ (or 750 for you asian-gunners) on SAT II's.</li>
<li>Your teachers don't hate you with a passion and you can write english in a grammatically correct fashion.</li>
</ol>

<p>then you have a decent shot at being admitted to some lower-tier Ivy League college to which you apply, because students who meet only the first condition are much too rare to fill all eight of the Ivy League colleges, and students who combine characteristic 1. with characteristics 2. through 4. are rarer still. But it's still possible in principle for a student with all those characteristics to be rejected by all Ivy League colleges, not to mention the most selective H, Y, and P.</p>

<p>The original post said 2300+</p>

<p>What if I got a 2290? ;) JK</p>

<p>I actually did though btw, 730-760-800 1 sitting</p>

<p>I can't believe I just read that entire thing.</p>

<p>It was light outside when I started.</p>

<p>^ lol</p>

<p>it's interesting though</p>

<p>i hope it's true</p>

<p>LOL, gollygoshkins... It's actually an interesting thread, as it contrasts the two most common misconceptions:
1) Stats alone aren't enough to get you into an Ivy (truth: really stellar stats create a pretty good chance to get into at least one or two).
2) Stats alone WILL get you into an Ivy (truth: students DO get rejected who have very good stats).</p>

<p>This, of course, begs the question: should "getting into an Ivy" be a student's objective? That concept ignores the very real differences between the eight schools, and the fact schools outside that group may more closely resemble some of its members than the other members themselves.</p>

<p>If someone gets a 2300+, has a 3.9 GPA and writes a reasonably good personal statement - they have a very good chance of getting into a single Ivy. </p>

<p>This is news? Do people simply not have a quantitative understanding of how the world works?</p>

<p>
[quote]
If someone gets a 2300+, has a 3.9 GPA and writes a reasonably good personal statement - they have a very good chance of getting into a single Ivy.</p>

<p>This is news?

[/quote]
</p>

<p>I know a lot of people in real life, and encounter many students here on CC, who think that a single B is a disqualification from the school of their dreams, and they aren't always dreaming about HYP. (And in fact, it is possible to get into any one of HYP with a B here or there on the transcript, but it's wise to have a transcript with a lot of hard courses on it.) Some high school students are way too fixated at having perfect grade-point averages, even at the cost of taking very wimpy courses.</p>

<p>I love the mathematical perspective of this thread. Although as a student who is applying next year, I find it neither encouraging nor clarifying. The admissions game is always a myth. So long as the myth isn't debunked, we will still be getting intriguing threads like this one.</p>

<p>tokenadult, while what the stats teachers said is true, I do not believe they discredit that "if an applicant has a 50% chance at each of the 9 schools he applies to, he has less than 1% chance to get rejected by all." (NOTE: this was not the idea behind the original calculations of the OP, but I want to point out the accuracy of THIS statement). I think the actual problem lies in the fact that we are taking somebody based on their stats and GPA, not their actual chance of admittance.</p>

<p>Let me clarify. Let's say I take Bob and somehow manage to run a simulation at a college an arbitrarily large number of times. By simulation, I mean I have him go through the selection committee, each time with slightly different order, readers, etc. Now, if he gets accepted in half of those simulations, then he has a 50% chance of getting in. Now, if the simulation showed a 50% chance at each school, the statement above is perfectly applicable. </p>

<p>Of course 50% is arbitrary and can be replaced by any number.</p>

<p>ml2,</p>

<p>The problem with the simulation, as I'm sure you're aware, is the controls needed. That's why I rarely like to use the percent chance of admission method to give people an idea of their chances. It's far too specific a figure for a very "fudgy" probability given. I think your best bet is to see where you fall within last year's cohort and ask if you are near the mean admit. If you are, great. If not, then how far are you?</p>

<p>Someone who is within one SD of the mean admit (even accounting for ECs) is likely to be admitted to at least ONE school in the Ivy League/Ivy Caliber.</p>

<p>I find this extremely truthful in a mathematical sense. A girl at my school, ranked 2nd, with 2200+ SATS, applied to nearly every Ivy and was accepted to just Brown. She had a decent shot at all of the Ivies, but if you assume her chances were 25 at each and she applied to 6, that's a 72% chance she'd get into atleast one. That doesn't mean she definitely will, but a pretty good chance she will. And as it is shown in real life, that "good chance" landed her in Brown.</p>

<p>It's mathematical data that should be proven with actual data. I'd love to see if students that apply to more top schools are more likely to actually get into more top schools. CC should collect some data.</p>

<p>At my childs high school last year, 2,900 students, 875 graduating seniors, affluent suburb, 8 kids got into Ivy's. Seven were athletes, 1 was legacy. 2 of them also the stats cited in the op. Just too many students with stellar stats and without a hook of some kind the odds stated in the original assumption are propably more like a 20% chance at a given school IMO. You would also assume the odds are not constant with the chance at HPY being less than other Ivy's. So maybe the need for two separate categories.</p>

<p>yay MIT =(</p>

<p>"4) Since even superbly qualified candidates are still subject to random elements in the process (which eliminates many superbly qualified candidates at each school), more apps DOES increase the chances of success."</p>

<p>I think Roger's summary is good...I would just quibble with one thing--I would replace "random" in the point above with "apparently random." If you get into a school because they are looking for somebody from Idaho, that isn't random--it just appears random from the outside of the process.</p>