Once in our CS department (not CMU), we had given PhD acceptance to a wrong student (call him Student2) since he carried the same name+surname with the candidate we intended to admit (call him Student1).

Instead of rescinding the acceptance, we let Student2 join. Of course we also admitted Student1 to the department as well.

It turned out that Student2 (the candidate we admitted by mistake) performed much better than Student1. :-)

Regehr notes a similar observation. http://blog.regehr.org/archives/147

I remember when I was applying to graduate schools I was waiting to go to my last (fourth) recruitment interview, but I was pretty sure I wanted to go to the third school I had visited. Well I wasn't back for more than a day from the last recruitment trip when I called up that third school to ask them when I should show up and what paperwork I needed to fill out... it wasn't until about two years later that I realized I couldn't remember receiving an official acceptance letter or phone call from them...

As an aside -- I have always found this type of sentence formulation, common with software engineers, so mystifying. It would be just as easy to understand without the sentence variables -- it doesn't add clarity but it does add length.

The alternative:

> Once in our CS department (not CMU), we had given PhD acceptance to a wrong student, since he carried the same name+surname with the candidate we intended to admit.

> Instead of rescinding the acceptance, we let both join. It turned out that the candidate we admitted by mistake performed much better than the other. :-)

My personal impression:

Formulation 1 (with variables, short and concise sentences): easy to understand.

Formulation 2 (without variables, long sentence): difficult to understand.

==> I prefer Formulation 1.

One doesn't always know whether the structure of the prose they are about to produce will permit discernment of its subjects by context and description alone. That is to say, in informal communications, an intuitive notation such as "[class][index]" can easily map between an abstract thought and its textual representation, without a critical re-reading on the part of the author to ensure the subjects are properly differentiated, potentially followed by subtle sentence refactoring.

In support of the variables, the comments seem to leverage Student1 and Student2 to describe the students the commenters are referring too, which is a lot easier than having to reexplain who you're referring to. (edit: DRY principle)

It's a style you also see quite frequently in legal documents where you first introduce the parties, give them a shorthand name and then use that name consistently throughout the document. I personally prefer that style over endless repetition of a description (with possible inconsistencies between the repetitions).

To some of you reading this who haven't been to graduate school, this mistake is not as large as you might think. A lot of graduate schools have the policy of overadmitting and then pruning out during the oral exam (which usually 1-2 years into the program). I do not know if CMU is one of those schools (I attended one which did not have this policy; also, this policy can vary by department within the same school).

This reminds me of the novel _Small World_ where the main character is a young researcher who got a position thanks to a mistake like this. https://en.wikipedia.org/wiki/Small_World:_An_Academic_Roman...

In that story, however, Student1 didn't get the position.

Also reminds me of the movie Orange County, where main character Shaun Brumder does not get admitted to Stanford because they mistakenly admit classmate and stoner Shane Brainard instead. Hilarity ensues.

Wasn't there once some sort of experiment of admitting a random sample from a college application pool? Couldn't find it on the spot, but I think it had some similar results to the PhD admission experience. It also does make sense from a armchair psychologist standpoint: The one being admitted on weak grounds will probably put in a lot more effort to keep up with the class/competitors than the one having always some other, similar options to fall back on.

Did Student2 know he was given a pass due to the mistake?

No, as far as I know. Only some faculty knew, until it was divulged at a graduate review meeting.

It would be interesting to look back and see what indicators might have indicated Student2's success that was not seen by the department.

A third degree friend of mine resigned from her job as soon as she received acceptance mail. She received rejection mail in the morning & was devasted. As depressing as it could get.

Might want to consult with a lawyer.

http://en.wikipedia.org/wiki/Estoppel

> Estoppel occurs when a party "reasonably relies on the promise of another party, and because of the reliance is injured or damaged"

I thought that was just called, er, detrimental reliance?

The first thing I'd do after being accepted for a Fall 2015 Masters program would be to quite my job.

Oh. My. That's going to be a mess. Some people will have quit their jobs, turned in notice, etc. CMU might want to be prepared to reaccept some of those people.

My response exactly.

wasn't the error corrected quickly though?

I am sure that someone got the email, yelled loudly, and then wrote their "I quit" email to their fine employer.

Or perhaps turned down other graduate school offers...

Don't burn your bridges until the chickens are hatched or something.

You can't reasonably expect to be rejected /after/ being accepted though. Resigning when you apply, chickens not hatched yet. Get a letter from the program you applied to informing you of your acceptance, that's a pretty hatched chicken.

Perhaps I should have just said don't burn your bridges? It's not strictly necessary to quit in a way that can't be unquit immediately upon receipt of an acceptance letter.

can't reply to tedunangst but it doesn't matter how you do it, you can't "unquit" once your employer knows you've been applying elsewhere, no matter if it's for school or another job.

Keyword being immediately. Waiting 24 hours from acceptance notification to quitting doesn't seem entirely unreasonable. One may not expect to be unaccepted, but a little prudence rarely hurts in the event of the unexpected.

That said, it certainly is possible to unquit at many employers provided you didn't quit by taking off your pants and running out the door shouting "adios, motherfuckers."

I don't know why you think you can't in general. If you're on good terms and quit on good terms I would expect most places you could undo it once you explained the situation. I know for a fact that I could at my employer.

Ugh, this apology seems lamely antiseptic. "We are reviewing our notification procedures" is not the level of self-abasement I would expect after a mistake this serious.

It's CMU. I wouldn't expect them to admit a mistake at all, so this is pretty good.

why is that ?

They're the top ranked in CS, and also highly ranked in general.

If you're the best or close to, you're allowed to make mistakes (See Linus' toxic remarks). God have mercy on you if you aren't.

At least it wasn't for a software testing-focused master's degree.

I thought the discipline was ironic too, but it was probably a secretary with a massive spreadsheet that messed up a formula or a mail merge.

I can't imagine what those applicants are going through.

Both sad and ironic!

Aren't those emails somehow legally binding?

No, you usually need an offer on letterhead from the dean of the graduate school (or similar).

If the email comes from a school employee then that may be enough to make it legally binding. But even if it isn't, if you threaten to sue the school will almost always let you in. This sort of thing happens all the time, e.g. from coaches who tell athletes they are accepted before the admissions committee officially makes a decision, and the reason you never hear about lawsuits is because the schools almost always let the candidates in if they complain in order to prevent bad publicity.

I don't understand, wouldn't the law account for honest mistakes if you corrected them in a timely fashion (e.g. before anyone relied upon the notification, resulting in actual harm)? Why would a mistake like this be legally binding?

EDIT: Looks like someone further up the thread is in exactly this situation: https://news.ycombinator.com/item?id=9066942

Good ol' Robert'); DROP TABLE Rejects;--

I will put money that they use the incorrect email template (acceptance vs reject)

This scary, I got my admit to UW last week. I just hope they don't rescind it.

Was it the PhD program? I know they usually notify people who were accepted in a phone call, so this kind of mistake is less likely to happen :).

for grad school, or undergrad? Either way congrats! I've spoken with a lot of the CS students and they're super bright.

Note that "UW" is a very overloaded term:

- University of Washington

- University of Waterloo

- University of Wisconsin

- University of Wyoming

etc.

But we all know which UW is the right one :)

Can't tell if this is clever sarcasm or you mean Waterloo...

Americans will probably think of Washington first.

This is exactly what I'm talking about. No one knows what each other is talking about when saying "UW". Looking at chris_va's profile, it appears he was talking about the University of Washington. Most of the US probably thinks of UW that way, though in Silicon Valley I've met a lot more people from the University of Waterloo than from the University of Washington.

Ah, CMU.. it's certainly no Stanford.

Flagged. This, of course, is totally fake.