The statement is mathematically nonsensical.
Unless there is insufficient data, you can estimate your probability. Different estimates may have different error variances, but if there is data, you can estimate. If you go to one of the NYC exam schools, your school is likely to have Naviance where you can get scattergrams of acceptances from your school given your SAT and GPA, and there is likely to be a huge amount of data for schools within 5 hours of NYC.
For example, for Stuyvesant
http://stuy.enschool.org/apps/pages/index.jsp?uREC_ID=126943&type=d&termREC_ID=&pREC_ID=484165
You can draw a circle around your dot and calculate the number of people who have applied and the number of people admitted. The ratio is an estimate of your probability of admission. If you’re not sure how big a circle, I would say 100 points on the SAT and 0.2 on GPA. If that doesn’t provide enough data, make it a little bigger. Since it’s not a great estimate, but an ok one, round to 1 significant figure, i.e, one of 0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7,0.8, 0.9, 1.0.
Now you rank your preferences. If your first choice is Yale, and your second choice is Princeton both with a probability of getting in at 0.2 for someone with your stats, based on your scattergram, then the probability of attending Yale is 0.2, and the probability of attending Princeton is (1-0.2)0.2 = 0.16, and the probability of not getting into your first two choices are (1-0.2)(1-0.2) = 0.64. Then go to choice 3, and calculate 0.64*prob(choice 3), etc.
You can put this in a spreadsheet to calculate if you want. Do this for 6 schools and calculate your probability of not getting into any of the 6 choices. If it’s still pretty high after the 6 schools, you need to increase your threshold for excluding a school and eliminate choices further down where the probability of attendance is smallest in favor of schools with better odds.
This will give you a pretty good way for formulating a strategy.