<p>We should all know that nobody can ‘chance’ an applicant with any degree of accuracy, and that it’s all a bit random in any case. On the other hand, many CC-ers have seen the grades and stats of many, many applicants. In light of this, instead of asking for chances of getting in, applicants should ask CC how they rank in terms of Harvard applicants: this would involve far less ill-qualified speculation</p>
<p><a href=“http://talk.collegeconfidential.com/harvard-university/1253171-official-harvard-university-2016-early-action-decisions-thread.html[/url]”>http://talk.collegeconfidential.com/harvard-university/1253171-official-harvard-university-2016-early-action-decisions-thread.html</a>
<a href=“http://talk.collegeconfidential.com/harvard-university/1305027-official-harvard-university-2016-rd-results.html[/url]”>http://talk.collegeconfidential.com/harvard-university/1305027-official-harvard-university-2016-rd-results.html</a></p>
<p>If you go through the above threads, students who were deferred, denied, or waitlisted often have identical stats to those students who were accepted. So, I’m not sure the ranking of a student’s grades and tests scores is anymore useful than the ill-qualified speculation of chance threads.</p>
<p>Direct posters to parchment.com, which uses a data-based approach to chancing. It’s nifty! :P</p>
<p>
</p>
<p>That should mean nothing more than that having those stats =/= having 100% chance of getting in. But if a large number of people are getting accepted with those stats (say, a third), then any individual with those stats can still roughly estimate that they have a 33.3% chance of getting in.</p>
<p>@gibby it isn’t a <em>good</em> measure by any means, but it is better than ‘chancing’</p>
<p>@Philovitist what parchment and other such approaches can’t do well is assign a value to ECs. CC can do that better</p>
<p>Tbh, I’m not too bothered - just putting a suggestion out there, that next time you see someone asking for his chances, simply tell him how you’d rank him in terms of Harvard applicants that you have seen. This measure involves less ‘judgement’ on the part of the ‘chancer’.</p>
<p>Parchment’s data-based results are skewed because the algorithm depends on rejected students going back and updating the program. Rejected students are sometimes too depressed or upset to bother going back, whereas accepted students are elated and happily go back, so Parchment’s results give you a much “rosier” chance that you actually have. Search the forum’s for Parchment and you’ll find many threads discussing the unreliability of the website. Here’s one I responded to: <a href=“http://talk.collegeconfidential.com/harvard-university/1392611-how-reliable-parchment-com-online-chancing.html?highlight=parchment[/url]”>http://talk.collegeconfidential.com/harvard-university/1392611-how-reliable-parchment-com-online-chancing.html?highlight=parchment</a></p>
<p>That’s true, but it’s much less arbitrary a system than our chance threads, right? If you just keep in mind the skewing (maybe predict how skewed the data is?), add a disclaimer to your reports, etc., you might get a better result.</p>
<p>But since parchment is used for applying to a lot of different schools, users seem to come back when they get accepted to any college, even if it isn’t their first choice, and report decisions on all of their colleges.</p>
<p>
</p>
<p>If your ECs are different from the common variety, maybe. Parchment does have a measure for that, too, though.</p>
<p>Oh, yea. Parchment also has a confidence level along with its percent chances, based on how accurately its skewed data has predicted chances in the past.</p>
<p>For example, after I computed one CC chancer’s stats into my account, the site says he has a 60% chance of getting into Brown, and that this guess is 77.4% accurate.</p>
<p>Parchment works to calibrate its chances despite this obvious source of skewing, and succeeds to at least some extent.</p>