分享一个牛贴：marketing phd录取概率

benxu

转载自gradcafe作者：possible_phd
非藤校 (but strong enough) marketing phd

Sobefore anyone points out the obvious, this analysis is of course verybasic and subject to tons of assumptions. haha With that being said, Ican't help but run some numbers through my head with regard to mychances of getting in somewhere this year. I'm thinking about this froma marketing admissions point of view, but it should apply to theothers, as well.

Assume, on average, a 10% chance of admission at any given program ifyou apply across the top 20 or 50 or something. The logic being thatyeah, maybe only 3% get admitted to the very top programs, but figurethat some portion of applicants are not actually competitive becausethey have sub-par stats, don't understand research and want to go toindustry, or for various other reasons. So maybe that bumps theeffective acceptance rate up to 5% or so for applicants that areactually competitive. Apply that same logic to lower ranked schools,and the effect is even more dramatic -- maybe you're so competitiverelative to this pool that you actually have a 20% chance of admissionhere. So on average, 10% is easy enough to work with and might not betoo unreasonable.

If you were to apply to 20 schools, the odds of being rejected by ALLof these programs is then only 12%. That doesn't sound too depressing,right? haha But let's think about this in terms of how we would stackup if we were to rank order applicants and compare ourselves. Let's beconsultants and make even more assumptions!

If you figure that an average of 2 spots are available per program in,say, the top 30, then there are 60 spots total. So really, you onlyhave to be in the top 60 of applicants that year to get a spot in a top30 PhD program! So how many applicants are there?

Let's assume an average of 350 total applicants per program per year.From a quick look at the breakdowns for Emory and Northwestern, itseems like 15-17% of total PhD applicants are marketing applicants, solet's assume a value of 16%, on average. That works out to an averageof 56 marketing applicants per program per year. For the top 30, thatis a total of 1,680 applications.

However, we all know how PhD admissions work -- people apply EVERYWHERE! So let's take that into account. If we assume the average applicant applies to 9 programs, that meansthe 1,680 applications really only amount to ~187 applicants. So, beingin the top 60 out of 187 applicants only requires you to be in the top32% of these applicants. If you have the requisite GMAT and GPA, goodSOP, and strong LORs, you're very likely to be in that upper third, somaybe we shouldn't freak out so much about the 3-5% acceptance rates,because the odds of us being able to get into a top school are actuallynot too bad. We've just got to be smart about where we apply, sincemost of us will not send out applications to every school in the top 30!

Like I said, I know this logic is very simplified. It assumes the samepool of applicants applies everywhere, that applicants apply to everysingle school in the top 30, etc. But hey, this is just a back of theenvelope calculation, alright?

他的计算太复杂. 概括一下,想法是这个样子的:

Assume
1. 前30的学校,每个申请者申9个学校;
2. 20% 的申请者背景,分数和能力不qualify， 属于盲申.
3. 前30的学校官方平均录取率为3%
p=3%/(1-20%) *9=34%
4. RE, SOP, LOR等优势可以让你站到前33.3%.

推论:
只要条件还不错 (不用完美), 运气还可以, 申请到前30的学校是很容易的. 概率大于34%.

事实上, 他的这个结论只适用于20-30名的学校. 对于前10名 (10-20名)的学校,修改假设如下:
1. 前10 (10-20)的学校,每个申请者拒绝offer的概率为50% (80%);
2.  前10 (10-20)的学校 的申请者有10%的人, 背景,分数和能力不qualify， 属于盲申.
3. 前10 (20) 的学校官方平均录取率为2% (3%)
4. 拥有RE, SOP, LOR等闪光点, 可以排到前50%.

修正:
前10名: Adjusted p=10%
前20名: Adjusted p=32%

现在考虑这样的学生A和B, A 申8所前十的学校; B 申9所10-20的学校

设同批次学校录取概率完全独立, 固知

A 被所有学校所有轮次据的概率: 0.9^8=0.4
B 被所有学校所有轮次据的概率: 0.68^9=0.03

在有闪光点的前提下, 被至少一所前10名的学校录取需要40% 以上的运气 (或者double, triple 闪光点的强度); 被前20名的学校录取只要申的足够多就行

All factors controlled.

以上计算完全从possible_phd大牛同学的推出. All errors are his            

补充说明: (看来不得不补充)
1. Just for fun. possible_phd的帖子是为了证明被录取的几率远大于统计上的几率.
2. 计算误差大是因为假设过于简单. 并非学术讨论. 请勿曲解

galluplinc

amazing things happen every single day. so do the shit ones. if you are good enough and in possession of strong recommendation letters, just be positive that good results will eventually emerge.

it's probably non-sense thinking about the probabilities.....

haohaosun

哈哈有趣。我之前也一直在计算这个概率, just for fun :p
的确按照多数30－50学校的官网上所说，商学院平均录取率在3％左右，假设it applies to every single program. 再算上20％盲申的人话，申9所学校的话，的确有29％的概率至少被一所学校的某program录取，条件是每个申请人equally competitive。1－（1－3％／80％)^9 = 29%

但问题在于，不是每个人都一样的competitive, 而对于这样3％左右的小概率事件而言，实力的不均可能造成很大的差异，况且你还不知道自己处于哪个实力区间。。。最极端的情况是，你申的那9所学校你的所有竞争者也都申了（你的竞争者是指跟你申同类型排名学校的人，不是指比你牛出很多的人），假设你有100个竞争者的话，那你至少排第27位才可能被一所学校录取。但你也可能很悲催的排在28名，那这样的情况下你一所学校都没有。。。。更悲催的是，比你牛的很多人很可能不被top 10之类的牛校录取，只好转投你的竞争领域，这样的话你排第27都没啥可能被录了。。。。

当然你排第几位不是概率决定的，所以人算不如天算，还是洗洗睡吧，我对自己说。。。。

benxu

Nothing is for certain. So probability always exist, and has been or will be considered by all applicants. The only problem is whether we have an accurate model to estimate it. Because of the variety of uncertainty, a perfect or certain model for the estimation may not exist by definition. It is not a problem here, because the discussion I posted is simply for fun, as any one with basic reading skills can clearly tell.

You misunderstood my point and misinterpreted the whole picture. When you say "good results will eventually emerge," the likelihood of such event is ambiguous and deliberately ignored; When you say "just be positive,"  you misled the readers that faith is a more accurate estimator.  Although faith is indeed more important, and sometimes, the only thing we can rely upon, your argument that "it is nonsense thinking about the probabilities" is irrelevant to our discussion and unsubstantiated by your own reasoning.

amazing things happen every single day. so do the shit ones. if you are good enough and in possession of strong recommendation letters, just be positive that good results will eventually emerge.

it's probably non-sense thinking about the probabilities.....

-- by 会员 galluplinc (2011/1/29 5:01:35)

kndx5

这孙子是kellogg的，不是藤校的。

benxu

多谢...

我马上改.

benxu

是啊

对每一个学校来说, 3% 就好像在抓阄一样. 在申请者都各有优势的时候, 只能看各个学校自己的偏好了.

哈哈有趣。我之前也一直在计算这个概率, just for fun :p
的确按照多数30－50学校的官网上所说，商学院平均录取率在3％左右，假设it applies to every single program. 再算上20％盲申的人话，申9所学校的话，的确有29％的概率至少被一所学校的某program录取，条件是每个申请人equally competitive。1－（1－3％／80％)^9 = 29%

但问题在于，不是每个人都一样的competitive, 而对于这样3％左右的小概率事件而言，实力的不均可能造成很大的差异，况且你还不知道自己处于哪个实力区间。。。最极端的情况是，你申的那9所学校你的所有竞争者也都申了（你的竞争者是指跟你申同类型排名学校的人，不是指比你牛出很多的人），假设你有100个竞争者的话，那你至少排第27位才可能被一所学校录取。但你也可能很悲催的排在28名，那这样的情况下你一所学校都没有。。。。更悲催的是，比你牛的很多人很可能不被top 10之类的牛校录取，只好转投你的竞争领域，这样的话你排第27都没啥可能被录了。。。。

当然你排第几位不是概率决定的，所以人算不如天算，还是洗洗睡吧，我对自己说。。。。

-- by 会员 haohaosun (2011/1/29 5:06:18)

haohaosun

我乱说的阿~ 说不定可以弄个simulation model然后把比如top100的学校的录取概率放进去, 然后变量是你对自己的定位。就好比simulate一个队列，排第一的话肯定所有学校100%要你，然后概率依次递减，说不定可以算出个最优申请学校数，given你的排名位子，你要申多少所学校你被topX录取的概率是50％以上。

说不定Sheldon算的出来，不过他不用算，他反正排第一了，只有偶这种排到门外的才每天惶惶不可终日的乱算

galluplinc

Nothing is for certain. So probability always exist, and has been or will be considered by all applicants. The only problem is whether we have an accurate model to estimate it. Because of the variety of uncertainty, a perfect or certain model for the estimation may not exist by definition. It is not a problem here, because the discussion I posted is simply for fun, as any one with basic reading skills can clearly tell.

You misunderstood my point and misinterpreted the whole picture. When you say "good results will eventually emerge," the likelihood of such event is ambiguous and deliberately ignored; When you say "just be positive,"  you misled the readers that faith is a more accurate estimator.  Although faith is indeed more important, and sometimes, the only thing we can rely upon, your argument that "it is nonsense thinking about the probabilities" is irrelevant to our discussion and unsubstantiated by your own reasoning.

amazing things happen every single day. so do the shit ones. if you are good enough and in possession of strong recommendation letters, just be positive that good results will eventually emerge.

it's probably non-sense thinking about the probabilities.....

-- by 会员 galluplinc (2011/1/29 5:01:35)

-- by 会员 benxu (2011/1/29 5:28:44)

i was not trying to criticize your argument. ur counter-argument made u look like an idiot.


btw, ur last quote on my words was inaccurate. i never said "it is nonsense ..."

benxu

I am sorry for your attitude and your thinking pattern. Given the nature of your remarks and the signals from your school's placement, there is no point for me to take your words seriously or to quote them right.

If you want to use strong language to start a fight, fine. But if you are a loser, then even winning a flight cannot possibly change the fact.

That's my final comment to this dispute.  





i was not trying to criticize your argument. ur counter-argument made u look like an idiot.


btw, ur last quote on my words was inaccurate. i never said "it is nonsense ..."

-- by 会员 galluplinc (2011/1/29 8:12:35)

[/quote]