Admissions Chances

College Vine regularly updates its chancing engine from time to time with enhancements to its machine learning probability algorithm and by augmenting the functionality of the interface itself. In this recent update we were informed that the main goals were to ensure that the probabilities more accurately reflected the over all lower admissions rates from the most recent admissions cycle and that an enhancement to delineating the benefits of applying early decision or early action were to be incorporated into the output results.

I'm not a software engineer so I don't know why this particular roll out was so choppy and resulted in clearly wrong outputs from the get go. Certain schools like CalTech returned results of 1% and other schools like Hampshire college returned 95%+ results. Also there were many instances of inconsistent school chances as compared to the expected outcomes just from looking at various well establish ranking lists and other college portals that annually publish data on the difficulty or ease of getting into certain school.

What I think it indisputable is that CV had some of the most knowledgeable in house experts who create video content streams, participate in blog discussions and offer private consultations to high school students and their families. Their methods are thorough, detailed and their advice is sound. Why this CV chancing engine is problematic is that it is shined a light on an area of CV which is inconsistent with the rest of the services it provides. As long as the Chancing Engine spits out results which are not consistent with how their own in house experts feel rates should be and inconsistent with the market norms for chancing probabilities, this remains an ongoing problem.

I'm certain the CV is aware that the CV is not correct and is working diligently on it.

My personal pet peeve with the CV changing engine is the following:

-First and foremost almost all the published admissions rates are dated, some 5, 6, 7 years old. I do not understand the point of continuing to ignore pleas from CV to correct this. If a 9th grader can look up an admissions rate by google or bing searching at that rate is most likely 1 cycle off. For example if you google the Harvard acceptance rate, 99% of the results are going to return either 3.19% from this cycle or 3.43% from last cycle. These are not made up numbers, these both have been published by Harvard on their press releases and official marketing documents. Therefore, why would CV insist on keeping an outdated 5% acceptance rate. If you examine each and every college on your "my list" you will se an incorrect published rate. Since CV only cover 650 schools, correcting this would only take 1 person about 1 week to complete. This should be a priority maintenance project done every cycle just like making sure other stats are accurate like "yield", "enrollment", "net cost".

-Second, I don't understand why most of the liberal arts colleges admission rates have improved their positions against Ivys and Elites. On my list Colby, Pomona, Colorado College, Swarthmore and Bowdoin are significantly harder to get into than #1 and #2 Amherst and Williams. In actuality, this is not the case. Also, on my list, these named LACs are in the same % league as Stanford, MIT, Duke, Columbia, and UChicago. There is not way that Colby 20% has the same chances as Stanford 20%, MIT 20%, no way. And no way that Colorado College 27% is harder to get into that UPenn 29%, Dartmouth 29%, Rice 30%, NorthWestern 30% JHU 31%, Cornell 33%. It's true that LACs have lower admit rate these past 2 cycles but their rates have not surpassed the lower rates of the Ivy league who now have a 8 school average acceptance rate of 4.96%. In fact the hardest LAC is only 7% and most of them are 8%. Another weird thing to me is that I would consider Hamilton a better school than say Colorado college but the CV rate is 52% for Hamilton and CC at 27% which makes no sense. Also I think there shouldn't be a huge spread between Barnard and Wellesley. 35% vs 56% seem too wide.

-Third I don't understand how US Military academies get such high status in terms of CV chancing rates when they are first of all a different kind of college entirely. I personally do not believe you can lump US military academies into the mix because 1.) the are self selecting and 95% of high achieving students would never apply to a US military academy for the following reasons, a.) they are academically a lower tier than Ivys, Elites and LACs, b.) no one wants to do 8 years of required service, c.) unless you have a personal conviction to serve as a military officer or go into the defense contractor business, most HS student want to have more flexibility, freedom and autonomy over their professional careers. 2.) all US Military standards are lower including GPA, ECs, Test Scores and course rigor. 3.) There list of majors are limited to mostly STEM majors. You would never join a military academy as a serious humanities major. If you want to study Classics or Ancient Languages, you go to Harvard, Yale, or Columbia not Westpoint or Annapolis. On my list WestPoint is listed just as hard to get into as UPenn, Dartmouth, Rice, JHU and Cornell, Williams. But on my list last year or the year before, Westpoint was like 70% and Annapolis, Airforce, Coast Guard even higher. What is a huge mis-step is to think just because admit rates are low at military academy they deserve to be ranked in terms of difficulty to get into for high achieving students just like other low admit rate schools. This is a wrong assumption for the reasons I outlined. Very few high achieving students have these schools in mind as top choices, less than 5% of HS students, perhaps as low as 1%.

-Fourth, I think Music colleges like Berklee, Julliard, Peabody, etc should not be mixed into the 650 colleges that college vine covers unless you pre-qualify as a potential candidate on the CV profile form.

-Fifth, HBCs or Historically Black Colleges should not be mixed into the 650 colleges on "my list" unless you select "I am considering applying to a HBCs" Why do I mention this? Because if you are White, Asian or non-Black, there is a 99% chance rate you are not going to apply to these colleges in the first place so what the point of having a chancing result of XX % to get into Spellman or Howard or Stillman when you are never going to apply there because of a racial, cultural mismatch. Anything CV can do to make "my list" less confusing and more accurate should be done.

The bottom line is that if the published rates were more accurate and expected rate was more accurate, CV members would feel better about their odds.

If Amherst published rate shows the accurate 6.9% rates an your CV rate is 28%, then you would know that your have a 4X better chance than the published rate. That's better than looking at "my list" and see the current published rate at 12% and comparing that to 28%. Why? because it looks like the bump I have is only 2.2X not 4X. Same for Columbia, 3.7% vs 22% looks better than 7% vs 22%, because that is the difference between 6X versus 3X.

Please check out Matt Kaye's explanation for the recent changes to chancing here: https://www.collegevine.com/questions/29908/may-2022-chancing-updates.

There was a Collegevine update that basically factored in the decreasing acceptance rates from this/previous admissions cycles to give a more accurate representation of students’ chances. It caused everyone’s rates to drop lol. My chances for Duke went from 28 to 12.

I think they’re still working out bugs tho because it’s not completely accurate for certain schools. You can contact Collegevine support also