So the point I am suggesting is it would be very helpful to segment that by qualifications. Like, presumably not everyone at your secondary school has something close to a 3.83UW and also something close to a 1510 superscore. And what happens with materially dissimilar applicants is not necessarily all that useful for assessing your own individual chances.
If you don’t have much data in your own personal range you can interpolate a bit, but you have to be careful. Like you mentioned yield protection, and in some cases that can make it harder to draw a straight line through a data cloud because you could actually have more cases of rejections/waitlistings toward the very top of the range.