Friday, April 1, 2011

Can someone 'splain me the math?

The March 2011 Student Assignment Report has the following explanation for the Density Tiebreaker (Appendix C, page 26):

Elementary (K5) students who live in attendance areas that do not have enough space to accommodate all the students living in the attendance area receive a density tiebreaker.

For each attendance area we calculated the number of kindergarten applicants who live in the attendance area (regardless of the choices they listed on their application form) as a percent of seats in the attendance area school. While 100% of kindergarten applicants live in an attendance area only 84% of kindergarten seats are in attendance area schools; the remaining 16% of seats are in city-wide schools. Therefore, if the percent of applicants in an attendance area was equal to or greater than 116% of capacity all elementary students (K5) who live in the attendance area got the density tiebreaker for all of their requests.


This passage seems to imply that city-wide seats were allocated to each attendance area at a fixed amount of 16%. Hum?

Is it correct to add 16% to 100% for the tiebreaker (116%) if only 84% of seats are in attendance area schools?

Can someone explain the assumptions and the math?


  1. The passage doesn't "imply" anything of the sort. 100 percent of students live in an attendance area. 84 percent of seats are attendance area seats, while 16 percent of seats are citywide. The number of seats has no relationship to the number of students.

    In any given attendance area, if the TOTAL NUMBER of students submitting a K application to SFUSD = 116% of capacity at the attendance area school, then all of those students got a density tiebreaker for all of their requests.

    It might be simpler to explain it as a concrete example: School A and School B each have 66 K seats available. 116% of 66 seats = 77 (with rounding). Only 70 children who live in School A's attendance area submitted K application forms for any SFUSD school, so those children WILL NOT receive a density tiebreaker. However, 77 children who live in School B's attendance area submitted a K application form for any SFUSD school, so all of those children WILL receive the tiebreaker.

  2. Thanks for the explanation. I also read and re-read that section. Then I printed off a District map and colored in the attendance areas with the density preference. (Coloring is a skill I've picked up in the last few years...)

    But the map looks like nothing I expected. The NE has no schools with the density preference. Some schools in the SE have it, some don't. Same story in the West.

    Going into the lottery I thought that the density preference would be based on families choosing their own attendance area. If it's over-filled then it's dense. That's what I thought it would be and I also thought that in some ways it duplicated the purpose of the neighborhood preference. I pictured the entire West-side getting it. But that didn't happened. Still it doesn't make much sense.

    Bottom line, I'm not a fan of the density preference because the map looks random, half the City got it and half didn't, and as it's worded it doesn't seem to promote equity.

  3. I'm not sure how I feel about the density tiebreaker either. But if we're just talking about why it works the way it does, it looks to me that the district is taking the capacity of the system as a whole into account when it awards the density tiebreakers.

    If you are someone who lives in the attendance area of a very popular school, you don't really like the district's approach because it means you don't get a density tiebreaker until applications from your attendance area reach 116% of your school's capacity (the assumption being that many people from the attendance area of a popular school will choose that school and make the attendance area tiebreaker less valuable for everyone).

    On the other hand, from the perspective of someone who lives in an area with lots of children and low-performing schools (e.g., Mission and Bayview), the district's approach is reasonable. The assumption is that people in high-density, low-performance areas will a)not all be able to fit in their local schools and b)want other options. The density tiebreaker, as currently defined, gives these families (especially those who don't live in a CTIP census tract) a slight edge.

    From the district-wide perspective, systemwide K capacity (and the density tiebreaker applied only to K this year, since middle and high school had no attendance areas) is still above demand for seats, though it seems to be tightening up every year. Already, the density tiebreaker applied to something like 60 percent of K applicants this year -- why would you make it *easier* to qualify for a density tiebreaker in some high-demand areas, since that lowers the value of the density tiebreaker for everyone?Therefore, it makes sense to factor in the citywide seats as a pressure reliever for all schools. Therefore, since citywide seats are 16 percent of total seats, it makes sense from the systemwide perspective to award tiebreakers based on the system capacity and not individual school capacity.

  4. This is yet another example of the extreme ridiculousness of this district's ever evolving Grand Social Experiment.

  5. Should the District base the density tiebreaker algorithm on the total number of available K seats per school or on the actual number of K seats per school after placement of non-attendance area siblings?

    Let’s take the School A example above. Parents in this attendance area didn’t get the tiebreaker because 70 children is less than 116% of capacity. Now, let’s assume that 20 seats were taken by non-attendance area siblings. That would leave only 46 seats for the 70 attendance area children who submitted K applications. The new tiebreaker is 116% of 46 seats = 53 seats (with rounding), and School A attendance area qualifies for the density tiebreaker.

  6. The simplest way to get into the school of your choice is through sibling priority. This strategy is used all the time by families in Daly City who have common last names shared by friends and relatives in the city whose children already attend desirable schools (some as "siblings" of other kids already at the school).

  7. Well, as 10:51 illustrates, reasonable people can differ about the algorithm the results as shown by the map of the attendance areas with > 116% density is incoherent.

  8. At a minimum, the density preference should not apply to City-wide programs. After sibling, CTIP1 and CDC/Pre-K preferences are considered, everyone else should have an equal chance at City-wide schools and programs.

  9. I agree with 11:58. As someone with no density tie breaker, and an undesirable neighborhood school, I am completely shut out of pretty much everything when you consider that 60% of K received a density tie breaker.

  10. If,as SFUSD has claimed, this student assignment system is simpler than the previous one, why are there so many threads needed on the SF Kifiles to clarify all the various aspects of it?

  11. I think SFUSD should leave it as simple as possible. Neighborhood school priority, yes. Sibling priority, yes. CTIP1 should be need based, for low income families only. Then, the District should poor $ into low performing, South East schools with great teacher incentives to draw the best teachers/administrators there. It's not rocket science. Forget about the rest of the priorities. In my mind, random chance would distribute people more fairly than the way they are working it now.

  12. "CTIP1 should be need based, for low income families only."

    That's fine, as long as neighborhood preference is needs based too.

    I know, everyone's steamed about the mythical reams of families moving into the Mission to get a preference.

    But why's it OK for someone to move into e.g. Grattan attendance area to get a ticket in but not OK for someone to move into CTIP1?

  13. 9:01: because people are stupid.

  14. I agree that the density tiebreaker is a farce. We were surprised to learn that we live in a "high density" area, which in turn helped us get one of our top choices (#4 actually). Like others, I thought that "density" was defined as more students in the attendance area CHOOSING the school than there are spaces to accommodate them. We live in a neighborhood where there are a lot of kids, but our attendance area school is not good and we would never enroll there. The result is that we got into an immersion program because we live in a dense attendance area even through our designated school still has plenty of slots open. We are happy with the results, but that doesn't make it right or equitable in any way.