Do you know a kid who was screwed in the college process in last few years?

[Anonymous]

Personally, I think that using AI to read essays is a terrible development. Too subjective.

Setting that aside, I still think that the problem with AI is the post-review shaping of the class algorithm from the college's enrollment management consultant. Intuitively, I think their models are likely crap. They think certain variables mean things that they don't, they weight them wrong, etc.

Ultimately, it doesn't matter that the algorithm is unfair if they end up with the class that they want. And thus, we have OP's title, that certain kids are getting screwed by those algorithms. A hard truth to swallow.

I don't remember if it was this thread or a different one that talked about leaking. One day that will happen, or someone will figure it out, though I suspect that some of the key variables will not be items that can be easily changed/gamed.

[Anonymous]

The thing is if your kid is the top of the class/top stats, you can’t really just wing it. Once you meet the baseline for grades and stats it doesn’t really mean much.

So if your kid is planning on applying to many T20 schools, they really should be starting the research on each of the supplemental for each school over the summer. Continue to refine. Dig deep. Sign up to the school newspapers, go to the departmental websites, find events at the university that might interest them.

The problem is these kids try to do all of it in a 10 day period in late December and truthfully it’s just painfully obvious how shallow those supplemental essays are. They never work. I have yet to see a really good supplemental essay for a top school that was done in a very short period of time.

Why? Because they don’t go deep. The best essays are soaked in the fabric of the school, the intimate details that take a lot of time and research to develop.

[Anonymous]

Anonymous wrote:Personally, I think that using AI to read essays is a terrible development. Too subjective.

Setting that aside, I still think that the problem with AI is the post-review shaping of the class algorithm from the college's enrollment management consultant. Intuitively, I think their models are likely crap. They think certain variables mean things that they don't, they weight them wrong, etc.

Ultimately, it doesn't matter that the algorithm is unfair if they end up with the class that they want. And thus, we have OP's title, that certain kids are getting screwed by those algorithms. A hard truth to swallow.

I don't remember if it was this thread or a different one that talked about leaking. One day that will happen, or someone will figure it out, though I suspect that some of the key variables will not be items that can be easily changed/gamed.

I think it depends on how it’s used. I do think certain kids will be institutionally disadvantaged by these new developments. Likely kids in oversubscribed majors and from large public high schools where there are dozens and dozens of applicants a year.

[Anonymous]

Anonymous wrote:

Anonymous wrote:Personally, I think that using AI to read essays is a terrible development. Too subjective.

Setting that aside, I still think that the problem with AI is the post-review shaping of the class algorithm from the college's enrollment management consultant. Intuitively, I think their models are likely crap. They think certain variables mean things that they don't, they weight them wrong, etc.

Ultimately, it doesn't matter that the algorithm is unfair if they end up with the class that they want. And thus, we have OP's title, that certain kids are getting screwed by those algorithms. A hard truth to swallow.

I don't remember if it was this thread or a different one that talked about leaking. One day that will happen, or someone will figure it out, though I suspect that some of the key variables will not be items that can be easily changed/gamed.

I think it depends on how it’s used. I do think certain kids will be institutionally disadvantaged by these new developments. Likely kids in oversubscribed majors and from large public high schools where there are dozens and dozens of applicants a year.

Wanting to study a popular major makes you disadvantaged? Attending a large public high School. Makes you disadvantaged? I think that is a very silly way to look at this. No one is disadvantaged if they don't get into Harvard.

There are not a lot of spots and competition to get them is fierce. but it is more than possible to pursue whatever it is you want to pursue without getting admittance to Harvard. And good thing because the vast majority of students are not going to Harvard.

[Anonymous]

So with AI, the major becomes even more important not less? And they will be doing metadata searches for keywords and metrics that show national-level achievement and impact? I assume AI will give them a score for “authentic alignment with major” based on the other applicants in the field?

So they’ll be able to give everyone a score relative to each other, both at the regional level and by major? If so, I imagine that it distorts and overvalues certain achievement over others?

Do I have this right?
Very troubling.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Personally, I think that using AI to read essays is a terrible development. Too subjective.

Setting that aside, I still think that the problem with AI is the post-review shaping of the class algorithm from the college's enrollment management consultant. Intuitively, I think their models are likely crap. They think certain variables mean things that they don't, they weight them wrong, etc.

Ultimately, it doesn't matter that the algorithm is unfair if they end up with the class that they want. And thus, we have OP's title, that certain kids are getting screwed by those algorithms. A hard truth to swallow.

I don't remember if it was this thread or a different one that talked about leaking. One day that will happen, or someone will figure it out, though I suspect that some of the key variables will not be items that can be easily changed/gamed.

I think it depends on how it’s used. I do think certain kids will be institutionally disadvantaged by these new developments. Likely kids in oversubscribed majors and from large public high schools where there are dozens and dozens of applicants a year.

Wanting to study a popular major makes you disadvantaged? Attending a large public high School. Makes you disadvantaged? I think that is a very silly way to look at this. No one is disadvantaged if they don't get into Harvard.

There are not a lot of spots and competition to get them is fierce. but it is more than possible to pursue whatever it is you want to pursue without getting admittance to Harvard. And good thing because the vast majority of students are not going to Harvard.

I also think it can disadvantage your application to private T20 schools. Its not about Harvard.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Personally, I think that using AI to read essays is a terrible development. Too subjective.

Setting that aside, I still think that the problem with AI is the post-review shaping of the class algorithm from the college's enrollment management consultant. Intuitively, I think their models are likely crap. They think certain variables mean things that they don't, they weight them wrong, etc.

Ultimately, it doesn't matter that the algorithm is unfair if they end up with the class that they want. And thus, we have OP's title, that certain kids are getting screwed by those algorithms. A hard truth to swallow.

I don't remember if it was this thread or a different one that talked about leaking. One day that will happen, or someone will figure it out, though I suspect that some of the key variables will not be items that can be easily changed/gamed.

I think it depends on how it’s used. I do think certain kids will be institutionally disadvantaged by these new developments. Likely kids in oversubscribed majors and from large public high schools where there are dozens and dozens of applicants a year.

Wanting to study a popular major makes you disadvantaged? Attending a large public high School. Makes you disadvantaged? I think that is a very silly way to look at this. No one is disadvantaged if they don't get into Harvard.

There are not a lot of spots and competition to get them is fierce. but it is more than possible to pursue whatever it is you want to pursue without getting admittance to Harvard. And good thing because the vast majority of students are not going to Harvard.

I also think it can disadvantage your application to private T20 schools. Its not about Harvard.

If it can disadvantage your application, it can also advantage your application. This is a game and it always has been and will be. Understand the game and apply accordingly.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Personally, I think that using AI to read essays is a terrible development. Too subjective.

Setting that aside, I still think that the problem with AI is the post-review shaping of the class algorithm from the college's enrollment management consultant. Intuitively, I think their models are likely crap. They think certain variables mean things that they don't, they weight them wrong, etc.

Ultimately, it doesn't matter that the algorithm is unfair if they end up with the class that they want. And thus, we have OP's title, that certain kids are getting screwed by those algorithms. A hard truth to swallow.

I don't remember if it was this thread or a different one that talked about leaking. One day that will happen, or someone will figure it out, though I suspect that some of the key variables will not be items that can be easily changed/gamed.

I think it depends on how it’s used. I do think certain kids will be institutionally disadvantaged by these new developments. Likely kids in oversubscribed majors and from large public high schools where there are dozens and dozens of applicants a year.

Wanting to study a popular major makes you disadvantaged? Attending a large public high School. Makes you disadvantaged? I think that is a very silly way to look at this. No one is disadvantaged if they don't get into Harvard.

There are not a lot of spots and competition to get them is fierce. but it is more than possible to pursue whatever it is you want to pursue without getting admittance to Harvard. And good thing because the vast majority of students are not going to Harvard.

I also think it can disadvantage your application to private T20 schools. Its not about Harvard.

If it can disadvantage your application, it can also advantage your application. This is a game and it always has been and will be. Understand the game and apply accordingly.

Explain how it (what is it) can advantage your application? AI?
So now you pepper your application with the keywords they are looking for?

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Personally, I think that using AI to read essays is a terrible development. Too subjective.

Setting that aside, I still think that the problem with AI is the post-review shaping of the class algorithm from the college's enrollment management consultant. Intuitively, I think their models are likely crap. They think certain variables mean things that they don't, they weight them wrong, etc.

Ultimately, it doesn't matter that the algorithm is unfair if they end up with the class that they want. And thus, we have OP's title, that certain kids are getting screwed by those algorithms. A hard truth to swallow.

I don't remember if it was this thread or a different one that talked about leaking. One day that will happen, or someone will figure it out, though I suspect that some of the key variables will not be items that can be easily changed/gamed.

I think it depends on how it’s used. I do think certain kids will be institutionally disadvantaged by these new developments. Likely kids in oversubscribed majors and from large public high schools where there are dozens and dozens of applicants a year.

Wanting to study a popular major makes you disadvantaged? Attending a large public high School. Makes you disadvantaged? I think that is a very silly way to look at this. No one is disadvantaged if they don't get into Harvard.

There are not a lot of spots and competition to get them is fierce. but it is more than possible to pursue whatever it is you want to pursue without getting admittance to Harvard. And good thing because the vast majority of students are not going to Harvard.

I also think it can disadvantage your application to private T20 schools. Its not about Harvard.

If it can disadvantage your application, it can also advantage your application. This is a game and it always has been and will be. Understand the game and apply accordingly.

Explain how it (what is it) can advantage your application? AI?
So now you pepper your application with the keywords they are looking for?

I do not understand the exact process that ai is performing on these applications. But whatever it is, understand it is and use it. if you need to have keywords in there for AI put the keywords in there for AI.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Personally, I think that using AI to read essays is a terrible development. Too subjective.

Setting that aside, I still think that the problem with AI is the post-review shaping of the class algorithm from the college's enrollment management consultant. Intuitively, I think their models are likely crap. They think certain variables mean things that they don't, they weight them wrong, etc.

Ultimately, it doesn't matter that the algorithm is unfair if they end up with the class that they want. And thus, we have OP's title, that certain kids are getting screwed by those algorithms. A hard truth to swallow.

I don't remember if it was this thread or a different one that talked about leaking. One day that will happen, or someone will figure it out, though I suspect that some of the key variables will not be items that can be easily changed/gamed.

I think it depends on how it’s used. I do think certain kids will be institutionally disadvantaged by these new developments. Likely kids in oversubscribed majors and from large public high schools where there are dozens and dozens of applicants a year.

Wanting to study a popular major makes you disadvantaged? Attending a large public high School. Makes you disadvantaged? I think that is a very silly way to look at this. No one is disadvantaged if they don't get into Harvard.

There are not a lot of spots and competition to get them is fierce. but it is more than possible to pursue whatever it is you want to pursue without getting admittance to Harvard. And good thing because the vast majority of students are not going to Harvard.

I also think it can disadvantage your application to private T20 schools. Its not about Harvard.

If it can disadvantage your application, it can also advantage your application. This is a game and it always has been and will be. Understand the game and apply accordingly.

Explain how it (what is it) can advantage your application? AI?
So now you pepper your application with the keywords they are looking for?

I do not understand the exact process that ai is performing on these applications. But whatever it is, understand it is and use it. if you need to have keywords in there for AI put the keywords in there for AI.

NP and differing POV here:
I think the people most influential and important in college admissions in this next cycle will be former college counselors who just left T20 admissions offices - so their contacts are "fresh" and they have a giant roster of people to connect/call - and can get real-time advice in this coming cycle. We are not going through this next cycle (just went through it - for the 2nd time) and not going again for another 2 years, but I see how/why having up-to-date info will be really helpful and differentiating. And we didn't use a counselor for this last time, but its just too much changing real time, and information asymmetry will be real.

[Anonymous]

Anonymous wrote:The thing is if your kid is the top of the class/top stats, you can’t really just wing it. Once you meet the baseline for grades and stats it doesn’t really mean much.

So if your kid is planning on applying to many T20 schools, they really should be starting the research on each of the supplemental for each school over the summer. Continue to refine. Dig deep. Sign up to the school newspapers, go to the departmental websites, find events at the university that might interest them.

The problem is these kids try to do all of it in a 10 day period in late December and truthfully it’s just painfully obvious how shallow those supplemental essays are. They never work. I have yet to see a really good supplemental essay for a top school that was done in a very short period of time.

Why? Because they don’t go deep. The best essays are soaked in the fabric of the school, the intimate details that take a lot of time and research to develop.

That is my kid. Did not even finish the common app essay in summer. Finished the common app essay just in time for the EA deadline.

For a few supplements a few topics were selected but the writing the most intensive was the 1 week prior to the EA deadline for the publics and the 2 weeks during winter break for RD.

DC got into 5 T20s including two of HYPSM. That is where I disagree with the gist of your post. That it takes some deep metaphysical understanding of colleges. Going in, that is what I believed too as that seems to be common wisdom.

I am not talking that the kid was screwed, the results were far beyond my expectations at least. There were 6 colleges that we thought DC had a good chance in and 5 that were probably less likely. Got into 3 of these less likely colleges, where DC did not spend much time as the focus was mostly on the 6 that we thought offered better chances.

For one T20 private, DC was offered $200k in merit. It is a last minute add, because questions were similar to another and changes were no more than a few word replacements for obvious name changes and one new sentence. This is definitely the college that took the very least amount of time. It took more time for safety college.

I thus think there is way more randomness that much of what paid consultants peddle is primarily directed at their self interest rather than students.

[Anonymous]

That’s amazing for your kid and highly atypical.

[Anonymous]

If you read stories of college admissions (books or newspaper articles), many will say that there is a layer of applicants at the top schools where the "yes" decision was simple...like they were immediately in the yes pile within a week of their application coming in the door.

I am not talking about athletes or donors or whatever, but just incredible stats and I guess they satisfied whatever institutional priorities these schools needed that year.

That's why it's more likely to see someone get accepted to multiple Top 10s then it is to see someone accepted to only one Top 10.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Personally, I think that using AI to read essays is a terrible development. Too subjective.

Setting that aside, I still think that the problem with AI is the post-review shaping of the class algorithm from the college's enrollment management consultant. Intuitively, I think their models are likely crap. They think certain variables mean things that they don't, they weight them wrong, etc.

Ultimately, it doesn't matter that the algorithm is unfair if they end up with the class that they want. And thus, we have OP's title, that certain kids are getting screwed by those algorithms. A hard truth to swallow.

I don't remember if it was this thread or a different one that talked about leaking. One day that will happen, or someone will figure it out, though I suspect that some of the key variables will not be items that can be easily changed/gamed.

I think it depends on how it’s used. I do think certain kids will be institutionally disadvantaged by these new developments. Likely kids in oversubscribed majors and from large public high schools where there are dozens and dozens of applicants a year.

Wanting to study a popular major makes you disadvantaged? Attending a large public high School. Makes you disadvantaged? I think that is a very silly way to look at this. No one is disadvantaged if they don't get into Harvard.

There are not a lot of spots and competition to get them is fierce. but it is more than possible to pursue whatever it is you want to pursue without getting admittance to Harvard. And good thing because the vast majority of students are not going to Harvard.

I also think it can disadvantage your application to private T20 schools. Its not about Harvard.

If it can disadvantage your application, it can also advantage your application. This is a game and it always has been and will be. Understand the game and apply accordingly.

Explain how it (what is it) can advantage your application? AI?
So now you pepper your application with the keywords they are looking for?

I do not understand the exact process that ai is performing on these applications. But whatever it is, understand it is and use it. if you need to have keywords in there for AI put the keywords in there for AI.

NP and differing POV here:
I think the people most influential and important in college admissions in this next cycle will be former college counselors who just left T20 admissions offices - so their contacts are "fresh" and they have a giant roster of people to connect/call - and can get real-time advice in this coming cycle. We are not going through this next cycle (just went through it - for the 2nd time) and not going again for another 2 years, but I see how/why having up-to-date info will be really helpful and differentiating. And we didn't use a counselor for this last time, but its just too much changing real time, and information asymmetry will be real.

In two years..maybe..but the power of former and current AOs will be diminishing over time. AOs themselves do not set or influence who the campus wants, this happens with the Deans and senior admin. AOs are lower leverage admins tasked with carrying out what university leadership wants. As systems with predictive data decisions and AI take over, the AO role will drastically change and diminish.

[Anonymous]

I really think if schools went back to having class rank, college decisions would make a lot more sense to parents. I have a 2023 and a 2026 at two single sex private schools, not in DC. Familiar with their classes and a year or two around them. Since schools have cum laude and give grade based awards, contrary to dcum belief, it’s easy to see that admissions actually do very closely track weighted gpa and test scores unless a kid totally whiffs on extracurriculars. This is supported by scattergrams. Sure, there may be a kid or two with a rare talent that does better than stats might predict, but it isn’t that common.

I think there are a lot of affluent schools with a significant amount of grade inflation where it is either hard for colleges to differentiate between kids leading to what seems to be random results or parents don’t realize kid isn’t even in top 20 percent of class because there is so much grade inflation.