Anonymous wrote:

Anonymous wrote:A mix. Some people do Basis #1. Some people might have a sibling at Latin Cooper so that for them went ahead of 2nd Street. Some people prefer a Hardy/Deal feeder, SWWFS, or another school because its close to them.

If you have sibling preference at Latin Cooper, it's idiotic to list it #1 above Latin 2nd. You would list Latin 1 first, so that you could match at Latin Cooper and stay on the waitlist for Latin 1.

Eh, I wouldn't do this. If I had a kid at Latin Cooper, I'd stay there and not pull my older kid to Latin 1 just because the younger kid got in there. I'd just stick with Latin Cooper rather than force my poor older kid to switch schools after settling in.

Anonymous wrote:A mix. Some people do Basis #1. Some people might have a sibling at Latin Cooper so that for them went ahead of 2nd Street. Some people prefer a Hardy/Deal feeder, SWWFS, or another school because its close to them.

By the results, it would be silly for anyone to bother wasting a lottery spot trying for a Deal/Hardy feeder or SWWFS for 5th.

0 matches for 5th grade for any Deal/Hardy feeder school or SWWFS other than 4 spots at Eaton ES. Though plenty tried because there are anywhere from 20-50 waitlist spots for 5th at each school.

Anonymous wrote:A mix. Some people do Basis #1. Some people might have a sibling at Latin Cooper so that for them went ahead of 2nd Street. Some people prefer a Hardy/Deal feeder, SWWFS, or another school because its close to them.

If you have sibling preference at Latin Cooper, it's idiotic to list it #1 above Latin 2nd. You would list Latin 1 first, so that you could match at Latin Cooper and stay on the waitlist for Latin 1.

Anonymous wrote:Kids with a sibling at another school (BASIS, WOTP school, DCI feeder) maybe??? It does still seem high.

If you have sibling preference at another school, you're going to match at that school over matching with Latin, so why list Latin below? That's stupid.

A mix. Some people do Basis #1. Some people might have a sibling at Latin Cooper so that for them went ahead of 2nd Street. Some people prefer a Hardy/Deal feeder, SWWFS, or another school because its close to them.

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:These numbers seem incorrect …

My guess is you think the Latin number should be lower. Maybe this helps.

Of all Latin I applicants on match day:
20% matched at a school they ranked higher
7% matched at Latin with no preference
5% matched at Latin with preference
68% were still on the waitlist

In the past we didn't have data on applicants who matched at a school they ranked higher. So people traditionally calculate the likelihood of getting in as matches/(matches + waitlist). If you do that you get:
9% matched with no preference
6% matched with preference
85% were on the waitlist

It's just another way of thinking about the data. I wouldn't advocate for it to replace the traditional way, but it can potentially provide a bit more insight on school preference.

That part doesn't make sense to me. What are people listing higher than Latin?

We listed schools higher than Latin (BASIS plus DCPS schools with better feeders). I know it wasn't totally rational because if we got a high enough number for BASIS, the latins would drop off the list. But that was our true preference. Maybe there are others.

Anonymous wrote:Kids with a sibling at another school (BASIS, WOTP school, DCI feeder) maybe??? It does still seem high.

Exactly. The piece not accounted for when looking at the # that got a higher match is that they might have had a sibling at their higher match, but still listed additional schools. Per above post, people aren't exactly rational - we did this even when we do we got sibling preference at our school and were guaranteed a spot. It could be that the majority do this for all we know.

So it doesn't quite give true odds for someone without any preference at all.

Kids with a sibling at another school (BASIS, WOTP school, DCI feeder) maybe??? It does still seem high.

Anonymous wrote:

Anonymous wrote:These numbers seem incorrect …

My guess is you think the Latin number should be lower. Maybe this helps.

Of all Latin I applicants on match day:
20% matched at a school they ranked higher
7% matched at Latin with no preference
5% matched at Latin with preference
68% were still on the waitlist

In the past we didn't have data on applicants who matched at a school they ranked higher. So people traditionally calculate the likelihood of getting in as matches/(matches + waitlist). If you do that you get:
9% matched with no preference
6% matched with preference
85% were on the waitlist

It's just another way of thinking about the data. I wouldn't advocate for it to replace the traditional way, but it can potentially provide a bit more insight on school preference.

That part doesn't make sense to me. What are people listing higher than Latin?

Anonymous wrote:Where are you seeing data for the applicants who matched at a school they listed higher? I am admittedly not a numbers person

You have to calculate it.

Students who matched at a school they ranked higher = total applicants - matches - waitlist

Where are you seeing data for the applicants who matched at a school they listed higher? I am admittedly not a numbers person

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:I find it hard to make sense of as even those who got into a higher ranked school might have been majority sibling preference applicants for all we know.

If sibling preference at Basis, I'd still list Latin I and II on my application...

Unless you can fully subtract out all preference applicants, you still can't quite calculate the odds.

This is true generally, but for middle school entry grades sibling preference is a guaranteed spot at every school and that makes it more straightforward.

If you rank school A first and you have sibling preference there, you are definitely getting in. There is no reason to list any other school. You'll be reflected in the matched numbers at school A.

If you rank school B first and don't get in, you'll still get into school A where you have sibling preference. You'll be reflected in the matched numbers at school A. You'll only be reflected in the waitlist numbers at school B, since you didn't get into a higher ranked school.

But would I also be listed as an applicant of school B?

That's where I struggle with coming up with overall odds for an applicant with no preference

When we applied for our younger kid with sibling preference, we still listed additional schools as a just in case.

For 5th and 6th grades this makes no sense. Maybe people do it anyway but it is entirely irrational.

Different story for someone applying to something like PK3 at Maury or 9th at Latin where sibling preference doesn't necessarily guarantee a seat.

Study after study has demonstrated people don't behave rationally when it comes to school choice lottery lists. And, it likely doesn't hurt them. They put down extra schools, they match at their top choice still.

Kids can and do have sibling preferance at multiple schools also so this slightly bizarre odds calculation exercise will always have some noise.

Anonymous wrote:Just realized total applications were already available for all DCPS schools (but not DCPCS) all the way back to SY14-15 from the downloadable data files here: https://enrolldcps.dc.gov/node/61

For fun:

SY14-15 SWW had 1097 applicants, Banneker 754, McKinley 897

SY19-20 SWW 1213, Banneker 796, McKinley 780

SY24-25 SWW 1527, Banneker 1291, McKinley 1271

Thanks for pulling this. A 50% increase in applicants over 10 years is a lot. No wonder it feels like these schools are getting more competitive.

Anonymous wrote:These numbers seem incorrect …

My guess is you think the Latin number should be lower. Maybe this helps.

Of all Latin I applicants on match day:
20% matched at a school they ranked higher
7% matched at Latin with no preference
5% matched at Latin with preference
68% were still on the waitlist

In the past we didn't have data on applicants who matched at a school they ranked higher. So people traditionally calculate the likelihood of getting in as matches/(matches + waitlist). If you do that you get:
9% matched with no preference
6% matched with preference
85% were on the waitlist

It's just another way of thinking about the data. I wouldn't advocate for it to replace the traditional way, but it can potentially provide a bit more insight on school preference.

*guaranteed not guarding