<p>cranberriez:</p>
<p>
[quote]
the system works.
[/quote]
</p>
<p>No, it doesn't. Read on.</p>
<p>
[quote]
I didn't say that s/he was definitely in or out of anywhere.
[/quote]
</p>
<p>Well, of course not, since %s don't say whether the person will get in or not; it tells the persons chances.</p>
<p>
[quote]
S/he could get into every place or rejected from every place (or any combination of the two), and it would still be within my predicted range.
[/quote]
</p>
<p>That's true, but it's also true if he/she were rejected if I gave him/her a 99% chance, or accepted with a 1% chance. They're all "within range" because none of them actually predict an outcome; rather, they predict the odds of a possibility.</p>
<p>
[quote]
A 63% chance indicates that you have a slightly higher than average chance of getting accepted, whereas a 6% chance indicates that it is highly unlikely, not impossible, but not likely.
[/quote]
</p>
<p>It's not easy to be that precise about this. That's why, on CC, people generally go by 10-percent intervals. If you go into the ones place, you're increasing the precision tenfold. Here's generally how the scale works:</p>
<p>100% - guaranteed/open admission
90+% - safety
80% - safe match
70% - match
60% - high match
50% - slight reach (could go either way)
40% - reach
30% - reach
20% - big reach
10% - nearly impossible</p>
<p>That, to me, seems precise enough.</p>
<p>
[quote]
my chance predictors are not solely based on percents admitted to each university.
[/quote]
</p>
<p>That's good. They shouldn't be. But you can gauge your chances with those stats in mind. Many people don't seem to understand (and you do) that % admitted can not tell you the % chance. So Berkeley, for example, admits 50-60% of its ELC applicants; that doesn't mean an ELC applicant has a 50-60% chance. In order for you to say that, admissions would have to be random events; they're not (basic statistics). But here's the counterargument: when the stats are at extremes, you can make safer judgments on one's chances. So, for example, UCSC admits 99.2% of its ELC applicants; you can justifiably say that if someone is an ELC student, then UCSC is a safety (as safety is defined as >90%).</p>
<p>
[quote]
The above should explain why ^ Perhaps only 10% of the total admits are out of state, but this person's chance may be greater than other out of state applicants based on their information, statistics, ECs, recs, test scores, etc.
[/quote]
</p>
<p>To continue the point above: when the statistic--10%--is at an extreme, you can make judgments about it. You can justifiably say it's probably a reach.</p>
<p>But that's just a small argument. I based my judgment on my own past experience and knowledge: the OOS applicant pool is very self-selective, yet only has a 20% or so acceptance rate. Generally, you have to be extremely qualified to get in (as I've seen in many, many cases). UC has an obligation to take in-state students first, as that's its main purpose; in-state admission is extremely competitive (40,000+ applicants), so it's difficult enough to get in. As such OOS is even more difficult to get in. For the majority of OOS (and in-state, for that matter) applicants, it's a reach. Much of the same can be said of other publics, too, as they have the same obligations.</p>