How accurate is Parchment in predicting college admission chances?

Broad strokes, talking about a dozen schools overall with, broad strokes, 1/3 denials, 2/3 admits. The denials were all selective LACs with overall acceptance rates ranging from 25-35%. Kid’s test stats were in the upper end of the middle 50 for all of the denials. (Fuzzier on the GPA.)

The real ringer was the one where Parchment gave him a 96%/green light chance of admission at a school where he was still only in the upper end of the middle 50% and the school itself only takes 35% of applicants. That seemed wildly misleading to me. Though he did actually get in. I just thought that was an awfully sunny prediction under the circumstances. For most of the green light schools his stats were well above the middle 50.

Not in ANY way suggesting this is a reliable tool. Just interesting to look at in combination with a lot of other factors.

@SpikeyMike13 That seems like a pretty poorly predictive result to me. I bet I could setup a one-man shop and do better than only getting the 80% chance places right.

I agree with the other posters who said Naviance and the Common Data Set are far more predictive, if you are fortunate enough to have Naviance at your school, have enough years of data and a large enough population to have enough data points. My kids HS has 8 years of data for 350-400 graduates a year, so it was highly predictive for my S and his peers. Of course it doesn’t give you a simple color-coded predictive graph – you have to mine the data yourself.

@citivas Parchment was useful to us though a bit optimistic. We didn’t count on anything but the 99% schools, so no harm done. Our small school has only had Naviance for one year, so we had to make do…

1 Like

As an experiment, I created an account and ran my college freshman son’s data and the exact schools he applied to. As I expected, it was good for basically identifying which schools were harder and easier to get into. In that regard, it was spot on in that all the schools with the lowest probabilities were the ones he didn’t get into or were waitlisted at. But whereas @SpikeyMike13 experience was it only accurately predicted acceptance to schools with 80% or greater probability, my S got into every school it predicted a 40% chance or better at, and that included 9 schools between 40-80%. He didn’t get into one at 32% and was waitlisted at one with 24% (all RD). All the other denials were 20% or less. I guess it could be useful for broadly helping define safeties vs. matches vs. reaches.

Even Naviance which is much more specific to you still can be misleading if you have a lot of hooked applicants at your school which we do. Generally we know the recruited athletes so we can figure out those but we don’t always know other hooks like legacy, URM, etc. If our circle is in the middle of the chart we tend to think that is not a 50/50 thing that it looks like it would be.

My son’s experience was similar ; he was accepted at all schools > 60% (4 green, 1 yellow) , declined at all schools <40% (1 yellow, 3 red) . Schools were he was declined had admission rates of <8%, he was admitted to all schools with rates >8%. All applications were RD or EA. No schools he applied to were between 40-60%. Naviance also corresponded with the Parchment results at his school except for Vanderbilt, where Parchment correctly predicted admission and Naviance was uncertain.

I would guess that Parchment is most accurate for students who are average applicants for their particular stats.

I’d hesitate to use the word “accurate” but it’s one more interesting indicator. Add it to the mix as one more prism through which to view your kid’s list. If everything’s in the red zone or nothing’s in the green zone, you might want to rethink it.

(Actually, you DEFINITELY want to rethink it if everything’s in the red zone.)

Parchment seems extremely optimistic in our case. Granted, we aren’t shooting for reaches or terribly competitive schools (for merit money and stats reasons) but still. 5 green (4 @ 99 and one @ 98) and 2 yellow ( 88% and 94%) and calls for 7/7 acceptance.

I believe it’s 2 green and 5 yellow and while I do think 7/7 is possible, I wouldn’t call it with that kind of certainty to be sure. It will be interesting to see how it plays out.

My S is applying as an engineering major, which can be a lot more competitive (depending on the school) than other majors. I’ll be curious to see how Parchment’s predictions compare with his actual experience since it doesn’t seem like they take the major into consideration when determining likelihood of acceptance. Any other parents of engineering kids willing to share their experience?

Parchment accurately predicted the number of schools that would accept my son but it was hit or miss for predicting individual school results below the 80% mark.

So in our experience, Parchment was a reasonable tool to use for validating that his list as a whole was realistic. I wouldn’t use it to predict individual results.

@wvuhopeful I don’t have an engineer but I’d like to point out that a nice feature of Parchment is that you can see the profiles of other applicants. By looking at their ECs, stats and other applications it is possible many times to figure out which ones are pursuing similar interests.

I wonder if maybe Parchment has a hard time dealing with high stats applicants when it does not know if they have visited or interviewed and how demonstrated interest and yield management strategies might impact the school’s decision.

For example, looking back I see that D applied to 6 schools for which she had a 56-77% chance of acceptance according to Parchment (overall acceptance rates 21-32%). She visited many schools, but only 1 of those 6, and she was ultimately wait-listed at all 6.

Maybe they were giving her another chance to demonstrate interest - however, she rejected the wait-list spot at all 6. Love the school that loves you back…

I went back and ran the numbers for my older son and it was right but really shouldn’t have been. He was accepted to all his green and yellow schools including some in the 30% range and was not accepted to his red schools. In running it for my daughter’s not quite final list it seems to predict the same for her that she will not get into any red schools but will get into all her yellows which are still low odds. It includes 2 very lottery type school as a yellow for her though which would be nice but isn’t realistic. The 2 schools she is deciding between for ED are both red of course but parchment doesn’t consider ED. Fun to play but probably not helpful if you know your stats.

Just for fun I plugged in my D18’s info. She’s looking at Canadian universities. I know that there isn’t as much data for these schools, which is probably why it’s so wacky. They say she has 100% chance of admission at UBC. 15% chance at U Toronto, but 89% of applicants with 10-20% chance are accepted. What? 8-}

Parchment is not worth the paper it’s printed on.

I think Naviance is a good tool to estimate your chances. It doesn’t mean that if it tells you that you have a probability of getting in of 0.8 that you will get in. There is still a probability of 0.2 that you won’t. Nonetheless, it conditions out 1) your school, 2) the grading at your school, and 3) SAT scores. Therefore, it’s probably more predictive than most tools.

As post #21 points out Naviance is worth nothing unless you have a large enough sample size (data points) over multiple years. This means not only does the high school graduating class need to be large but a meaningful number must apply to a specific school in order to glean anything useful. Maybe every poster here praising Naviance has a high school where 14 kids apply every year to all the top 100 colleges in the country, but that doesn’t happen at our public Oregon high school. For me Naviance was just absolutely frustrating.

Just curious. Does anyone know if an applicant has been rejected in the green zone, specifically >98%? If so why do you think the prediction was so far off?

D16 had an 11% chance and was accepted early. University of Toronto is not accurate in parchment at all. The thing is, it should be. Except for engineering, U of T admissions is purely numbers based.

Similarly, another Canadian university, University of Waterloo, gave D16 a 97% chance. Some of the most important information that Waterloo considers, the AIF form, is not even part of the parchment database. Although D16 was admitted, there were students with higher numbers (GPA’s and SAT scores) who were not. Parchment (and other predictive sites) calculates chances based on the numbers, but there is so much more than numbers that goes into holistic admissions.

Just to show the absurdity of Parchment from the perspective of an applying senior:

GPA: 4.0 UW, 4.79 W
ACT: 34 SS

College list/ chances according to parchment:
-Brown ED: 51%
-Harvard (legacy): 26%
-Northeastern: 97%
-Pomona: 51%
-Tufts: 77%
-UCB: 70%
-UVM: 99%
-WashU: 47%
-Williams: 49%

Says that I will likely be accepted to all of my colleges. Hoping for Brown, but who knows.