dcmom wrote:I am PP who talked extensively to the DME about this.

The tracking number is based I believe on when you initially registered for my school DC. It is not your lottery number. But PP is correct--all applicants for one grade are sorted in order and that order is used for all schools. So, say there are 10,000 students applying for PK3; the students' tracking numbers are randomly ordered, and that random ordering--what people here are referring to as a single lottery number--is used for each process in the iteration. That number is used to break ties among people with the same level of preference. So if I have a sibling attend a charter school, then my applicant will get in over all other applicants who don't have a sibling. However, if there are more applicants with siblings than there are spots, then the random lottery number will be the tiebreaker, and those with the best lottery numbers get in. Everyone else is waitlisted in the order of the random lottery number.

Not everyone applies to all schools, and you are dropped from waitlists of schools you rank as less desirable from the school you get into, so your actual position on waitlists will vary based on whether you qualify for a preference category (in-boundary for PK3/4, proximity, sibling, founder for charter, etc.), who else applied to the school, and where the other applicants get in.

Remember, the tracking number is not the same as your lottery number. Just because you are # 20,000 doesn't mean you are #20,000 in the PK3 lottery. The computer will create the lottery order randomly after the lottery application closes.

Anonymous wrote:This is a question:

Let's pretend I am applying with no special preferances for PreK4.

I am applying to schools A, B and C.

My lottery # for PreK4 is 280.
so for schools A, B and C - anyone who has a lottery # less than 280 will across the board have priority in front of me?

Anonymous wrote:This is a question:

Let's pretend I am applying with no special preferances for PreK4.

I am applying to schools A, B and C.

My lottery # for PreK4 is 280.
so for schools A, B and C - anyone who has a lottery # less than 280 will across the board have priority in front of me?

Anonymous wrote:I submitted a few days after it opened and have numbers in the low 2000s.

Anonymous wrote:This is a question:

Let's pretend I am applying with no special preferances for PreK4.

I am applying to schools A, B and C.

My lottery # for PreK4 is 280.
so for schools A, B and C - anyone who has a lottery # less than 280 will across the board have priority in front of me?

dcmom wrote:

Anonymous wrote:This is a question:

Let's pretend I am applying with no special preferances for PreK4.

I am applying to schools A, B and C.

My lottery # for PreK4 is 280.
so for schools A, B and C - anyone who has a lottery # less than 280 will across the board have priority in front of me?

The answer as anything with the lottery seems to be "it's complicated." Here's how the DME office explained it to me:

The computer has everyone request a seat from their first choice school. So in round one, the computer has you request school A, and let's just say it's a public non-dual language school. School A say has 50 people request seats from them, but they only have 30 seats. Of the 50 who apply in round one, let's say 5 people apply who are in-boundary with siblings, 10 people apply who are in boundary, 2 people apply out of boundary with siblings, and everyone else is just out of boundary. They take everyone they can from the preference category -- 17 people -- and then they have 33 applicants left for only 13 out of boundary spots. Of the 33 applicants, they take the first 13 lottery numbers. So if you are 280, then it depends on whether there were 13 people out of boundary who applied to your #1 school in round 1 of the lottery. Let's say there aren't, there are only 12, then you get the 13th spot. But it's only a temporary assignment.

In round 2, the computer has everyone who did not get a spot in round 1 request a seat from their second-choice school. Let's say someone else put school A as their second-choice school and they are in-boundary. All the people who got temporarily assigned to the school in round 1, along with all the people who bid in round 2, are now compared by the computer. Again, the seats first go to those with preferences. So even though you were temporarily assigned a seat in round 1, now in round 2 someone who is in-boundary gets a seat and knocks you onto the waitlist.

The process continues to the next round. At the end of round 2, anyone not assigned (which now includes you) bids on their next-lower choice. For you, it is choice B. The same process continues on and on until it has iterated through everything.

The DME office told me that the Nobel Prize people who came up with it said the outcomes are much better when only one lottery number is used, which is why they are going with that. This seemed counter-intuitive to me, but I tried it out with my own example and they are absolutely right--more people are matched with a higher-ranked option with just one lottery number, and you always get into the highest-ranked school with availability when there is just one number. When I ran it with a unique lottery number for each school, there were people who should be "traded"--i.e., someone who preferred A but got into B, and someone who preferred B but got into A. Using a single lottery number in the example I ran made it so this never happened.

It is a very different algorithm from what was used last year, which essentially had each DCPS running its own lottery but then letting people only be assigned to one school.

Hope this helps. I'm a total data geek, which is why I asked them. Obviously, understanding the algorithm makes no difference to how it actually plays out. Main thing I took away from my discussion is that:

1) I should have at least one "safety" school, since I may have number 10,000/10,000 PK3 applicants and I want to make sure my kid goes somewhere. That's why I explored my in-boundary school and decided to apply. It is not a popular school and historically all in-boundary applicants get in during the initial lottery. I also added a few other schools that are convenient and traditionally go through their entire waitlist.

2) Once I choose the school I'm applying to, I should rank them in the order that I actually want my kid to go. There is no benefit of being strategic in rankings.

Anonymous wrote:I don't see how the following statement can be accurate if the process described above is correct: "Once I choose the school I'm applying to, I should rank them in the order that I actually want my kid to go. There is no benefit of being strategic in rankings."

If "The computer has everyone request a seat from their first choice school . . ." and then "the computer has everyone who did not get a spot in round 1 request a seat from their second-choice school," then it matters in what order you rank them.

To use Hill schools as an example: if I'm IB for Brent PK3 but slightly prefer the Reggio program at SWS, and don't like any other options, then I have a "strategic" decision to make. Let's say I have a 40% shot of getting a spot at Brent if I list it first but only have a 10% chance of getting into SWS by listing it first. In that case, I'd clearly list Brent over SWS, despite my preference, because by the second round Brent would be filled with other IB students. I'd basically be choosing between a 40% or 10% chance of getting into a school I find acceptable.

dcmom wrote:

Anonymous wrote:This is a question:

Let's pretend I am applying with no special preferances for PreK4.

I am applying to schools A, B and C.

My lottery # for PreK4 is 280.
so for schools A, B and C - anyone who has a lottery # less than 280 will across the board have priority in front of me?

The answer as anything with the lottery seems to be "it's complicated." Here's how the DME office explained it to me:

The computer has everyone request a seat from their first choice school. So in round one, the computer has you request school A, and let's just say it's a public non-dual language school. School A say has 50 people request seats from them, but they only have 30 seats. Of the 50 who apply in round one, let's say 5 people apply who are in-boundary with siblings, 10 people apply who are in boundary, 2 people apply out of boundary with siblings, and everyone else is just out of boundary. They take everyone they can from the preference category -- 17 people -- and then they have 33 applicants left for only 13 out of boundary spots. Of the 33 applicants, they take the first 13 lottery numbers. So if you are 280, then it depends on whether there were 13 people out of boundary who applied to your #1 school in round 1 of the lottery. Let's say there aren't, there are only 12, then you get the 13th spot. But it's only a temporary assignment.

In round 2, the computer has everyone who did not get a spot in round 1 request a seat from their second-choice school. Let's say someone else put school A as their second-choice school and they are in-boundary. All the people who got temporarily assigned to the school in round 1, along with all the people who bid in round 2, are now compared by the computer. Again, the seats first go to those with preferences. So even though you were temporarily assigned a seat in round 1, now in round 2 someone who is in-boundary gets a seat and knocks you onto the waitlist.

The process continues to the next round. At the end of round 2, anyone not assigned (which now includes you) bids on their next-lower choice. For you, it is choice B. The same process continues on and on until it has iterated through everything.

The DME office told me that the Nobel Prize people who came up with it said the outcomes are much better when only one lottery number is used, which is why they are going with that. This seemed counter-intuitive to me, but I tried it out with my own example and they are absolutely right--more people are matched with a higher-ranked option with just one lottery number, and you always get into the highest-ranked school with availability when there is just one number. When I ran it with a unique lottery number for each school, there were people who should be "traded"--i.e., someone who preferred A but got into B, and someone who preferred B but got into A. Using a single lottery number in the example I ran made it so this never happened.

It is a very different algorithm from what was used last year, which essentially had each DCPS running its own lottery but then letting people only be assigned to one school.

Hope this helps. I'm a total data geek, which is why I asked them. Obviously, understanding the algorithm makes no difference to how it actually plays out. Main thing I took away from my discussion is that:

1) I should have at least one "safety" school, since I may have number 10,000/10,000 PK3 applicants and I want to make sure my kid goes somewhere. That's why I explored my in-boundary school and decided to apply. It is not a popular school and historically all in-boundary applicants get in during the initial lottery. I also added a few other schools that are convenient and traditionally go through their entire waitlist.

2) Once I choose the school I'm applying to, I should rank them in the order that I actually want my kid to go. There is no benefit of being strategic in rankings.

Anonymous wrote:

dcmom wrote:

Anonymous wrote:This is a question:

Let's pretend I am applying with no special preferances for PreK4.

I am applying to schools A, B and C.

My lottery # for PreK4 is 280.
so for schools A, B and C - anyone who has a lottery # less than 280 will across the board have priority in front of me?

The answer as anything with the lottery seems to be "it's complicated." Here's how the DME office explained it to me:

The computer has everyone request a seat from their first choice school. So in round one, the computer has you request school A, and let's just say it's a public non-dual language school. School A say has 50 people request seats from them, but they only have 30 seats. Of the 50 who apply in round one, let's say 5 people apply who are in-boundary with siblings, 10 people apply who are in boundary, 2 people apply out of boundary with siblings, and everyone else is just out of boundary. They take everyone they can from the preference category -- 17 people -- and then they have 33 applicants left for only 13 out of boundary spots. Of the 33 applicants, they take the first 13 lottery numbers. So if you are 280, then it depends on whether there were 13 people out of boundary who applied to your #1 school in round 1 of the lottery. Let's say there aren't, there are only 12, then you get the 13th spot. But it's only a temporary assignment.

In round 2, the computer has everyone who did not get a spot in round 1 request a seat from their second-choice school. Let's say someone else put school A as their second-choice school and they are in-boundary. All the people who got temporarily assigned to the school in round 1, along with all the people who bid in round 2, are now compared by the computer. Again, the seats first go to those with preferences. So even though you were temporarily assigned a seat in round 1, now in round 2 someone who is in-boundary gets a seat and knocks you onto the waitlist.

The process continues to the next round. At the end of round 2, anyone not assigned (which now includes you) bids on their next-lower choice. For you, it is choice B. The same process continues on and on until it has iterated through everything.

The DME office told me that the Nobel Prize people who came up with it said the outcomes are much better when only one lottery number is used, which is why they are going with that. This seemed counter-intuitive to me, but I tried it out with my own example and they are absolutely right--more people are matched with a higher-ranked option with just one lottery number, and you always get into the highest-ranked school with availability when there is just one number. When I ran it with a unique lottery number for each school, there were people who should be "traded"--i.e., someone who preferred A but got into B, and someone who preferred B but got into A. Using a single lottery number in the example I ran made it so this never happened.

It is a very different algorithm from what was used last year, which essentially had each DCPS running its own lottery but then letting people only be assigned to one school.

Hope this helps. I'm a total data geek, which is why I asked them. Obviously, understanding the algorithm makes no difference to how it actually plays out. Main thing I took away from my discussion is that:

1) I should have at least one "safety" school, since I may have number 10,000/10,000 PK3 applicants and I want to make sure my kid goes somewhere. That's why I explored my in-boundary school and decided to apply. It is not a popular school and historically all in-boundary applicants get in during the initial lottery. I also added a few other schools that are convenient and traditionally go through their entire waitlist.

2) Once I choose the school I'm applying to, I should rank them in the order that I actually want my kid to go. There is no benefit of being strategic in rankings.

This may be how it will actually work, but I think it sucks and I liked it better when there was a separate lottery (within the common lottery) for each school. In that system, everyone who applied for school A is given a random lottery number just for school A, and then everything works as the PP above said, but just for that school. That way, if you get #840 for a popular school that is your first choice, when the computer goes to your 2nd choice, school B, you get a whole new number within the lottery for school B. So hopefully you'll do better than 840 in at least ONE of the lotteries for your top 3 or 4 schools, since that random number will also determine the waitlist for schools you're waitlisted into, won't it?

One single randomly assigned number sucks, as it means that if you get the last possible number in the whole lottery (apparently number 30,000 or so), then you are last on every list. Isn't that the case? Or am I missing something?

dcmom wrote:

Anonymous wrote:This is a question:

Let's pretend I am applying with no special preferances for PreK4.

I am applying to schools A, B and C.

My lottery # for PreK4 is 280.
so for schools A, B and C - anyone who has a lottery # less than 280 will across the board have priority in front of me?

The answer as anything with the lottery seems to be "it's complicated." Here's how the DME office explained it to me:

The computer has everyone request a seat from their first choice school. So in round one, the computer has you request school A, and let's just say it's a public non-dual language school. School A say has 50 people request seats from them, but they only have 30 seats. Of the 50 who apply in round one, let's say 5 people apply who are in-boundary with siblings, 10 people apply who are in boundary, 2 people apply out of boundary with siblings, and everyone else is just out of boundary. They take everyone they can from the preference category -- 17 people -- and then they have 33 applicants left for only 13 out of boundary spots. Of the 33 applicants, they take the first 13 lottery numbers. So if you are 280, then it depends on whether there were 13 people out of boundary who applied to your #1 school in round 1 of the lottery. Let's say there aren't, there are only 12, then you get the 13th spot. But it's only a temporary assignment.

In round 2, the computer has everyone who did not get a spot in round 1 request a seat from their second-choice school. Let's say someone else put school A as their second-choice school and they are in-boundary. All the people who got temporarily assigned to the school in round 1, along with all the people who bid in round 2, are now compared by the computer. Again, the seats first go to those with preferences. So even though you were temporarily assigned a seat in round 1, now in round 2 someone who is in-boundary gets a seat and knocks you onto the waitlist.

The process continues to the next round. At the end of round 2, anyone not assigned (which now includes you) bids on their next-lower choice. For you, it is choice B. The same process continues on and on until it has iterated through everything.

The DME office told me that the Nobel Prize people who came up with it said the outcomes are much better when only one lottery number is used, which is why they are going with that. This seemed counter-intuitive to me, but I tried it out with my own example and they are absolutely right--more people are matched with a higher-ranked option with just one lottery number, and you always get into the highest-ranked school with availability when there is just one number. When I ran it with a unique lottery number for each school, there were people who should be "traded"--i.e., someone who preferred A but got into B, and someone who preferred B but got into A. Using a single lottery number in the example I ran made it so this never happened.

It is a very different algorithm from what was used last year, which essentially had each DCPS running its own lottery but then letting people only be assigned to one school.

Hope this helps. I'm a total data geek, which is why I asked them. Obviously, understanding the algorithm makes no difference to how it actually plays out. Main thing I took away from my discussion is that:

1) I should have at least one "safety" school, since I may have number 10,000/10,000 PK3 applicants and I want to make sure my kid goes somewhere. That's why I explored my in-boundary school and decided to apply. It is not a popular school and historically all in-boundary applicants get in during the initial lottery. I also added a few other schools that are convenient and traditionally go through their entire waitlist.

2) Once I choose the school I'm applying to, I should rank them in the order that I actually want my kid to go. There is no benefit of being strategic in rankings.