This article was created 23 days ago.

Just How DOES the New Lottery System Work?

11:59 pm - 04/21/08

Yesterday, block information was released for the 2008-2009 academic year. For the classes of 2010 and 2011, this also marked the first in-practice demonstration of the new housing lottery system.

While the new system is designed to be more fair, so it is powered by a program which removes most chance from the system. The core idea behind the program is to have every Swarthmore's students three housing lottery numbers average out to the same number.

The Rising Sophomore Lottery

This lottery is very simple—-the members of the rising sophomore class are randomly sorted and then assigned sequential numbers.

The Rising Junior Lottery

This is where things start to get complicated.

Every student's number are first adjusted based on the housing decisions made for sophomore year. Students who decided to block are penalized by the number of students in their class divided by three–this means members of the class of 2011 who blocked will be penalized roughly 122 points. New students (like transfers) are randomly assigned numbers.

This leaves the the College with an ordered list of adjusted-sophomore housing numbers. At this point, the program moves halfway down the list and assigns the middle student the top number in the Junior class. The list wraps around, so the student with the best number in their sophomore year would get the middle number their junior year.

Imagine six students. We've assigned them each a number in the sophomore housing lottery (the first number in parenthesis).

Jill (1)
Bob (2)
Tim (3)
Sue (4)
Jeff (5)
Sally (6)

If Tim and Sally blocked, their effective-sophomore numbers would move up 2, so the adjusted list is created and then Junior numbers assigned (in brackets).

Jill (1) [4]
Tim (1) [5]
Bob (2) [6]
Sue (4) [1]
Sally (4) [2]
Jeff (5) [3]

The Rising Senior Lottery

This system won't be implemented until the 2009-2010 school year.

The ranking is determined based on the sum of the student's previous numbers, as no students are penalized for blocking. Senior lottery numbers are assigned such that the student with the worst sum is given the best senior number. Returning to our example, the sum is in parenthesis and the senior lottery number is in brackets.

Jill (5) [6] - Total: 11
Sue (5) [5] - Total: 10
Tim (6) [4] - Total: 10
Sally (6) [3] - Total: 9
Bob (8) [2] - Total: 10
Jeff (8) [1] - Total: 9

28 Responses to “Just How DOES the New Lottery System Work?”

Simon
4.22.08 @ 12:04

I may be alone in this, but I still don't get it.

apparently doesnt get it
4.22.08 @ 01:04

so doesn't sally get a better number (2) having blocked than she would have had she not blocked and been in Jeff's original place (3)? and doesn't Jeff get a worse number than he would have because of sally's blocking the year before?

vivaan
4.22.08 @ 07:04

so as rising seniors this year we WILL be getting punished for previous blocks?

im 3rd cohort and scared.

Reader
4.22.08 @ 08:04

I'm tempted not to believe this because they are no explicit sources stated.

DG, please state your sources without having your readers forced to ask for them.

Miles Skorpen
4.22.08 @ 08:04

This system has no effect on rising seniors.

Your "effective number," which determines where you are in a class, is based off the number you pick on—the number you are assigned doesn't really matter much.

Bobby
4.22.08 @ 08:04

Penalized for blocking? Not fun. I get it: people who don't get their blocks and have to live in ML or PPR deserve some love, but people who block, especially those that don't land in the most desirable places on campus, seriously don't deserve to be penalized. Though only anecdotal, it seems that most students on the waitlist (with probably pretty bad numbers) end up in reasonably nice places.

Miles Skorpen
4.22.08 @ 08:04

Rising sophomores have *always* been penalized for blocking—under the old system, you'd automatically be placed in the top cohort, which could be an even bigger penalty than this system provides.

My understanding of the reasoning behind this is that students who block as sophomores regularly get into housing that they could only obtain with much higher numbers in the ordinary lottery—good locations or good rooms.

The system is designed to give every student roughly similar rooms all four years. This means that the lottery has to adjust for the fact that blocking rising sophomores get better rooms.

Mommy & Daddy didn't take care of me
4.22.08 @ 08:04

Yeah, it's really fair. That's why a freshman living in Alice Paul this year got his block to live in Alice Paul again next year. Whereas, without a rich mommy or daddy, I am screwed.

Jeff
4.22.08 @ 09:04

I don't really see what the complaining is about. This seems quite fair, and, as usual, the Daily Gazette did an excellent job of covering it (but then again, maybe I'm just a math major…). It certainly is much better than the system in place previously. Freshman year housing plays no part in the lottery and probably shouldn't, since you don't get to pick: one person's hell (like ML) may be another's paradise.

I don't know how money would play into it unless your balance wasn't settled previously. But it seems like almost everyone was on the list outside of the Housing office, not just rich persons. Unless you think Liz Derickson takes bribes…?

aaron
4.22.08 @ 11:04

I have a question about sophomore blocking. Say i ended up with a really good sophomore number in my block. Let it be an 850, with my overall block average as a 900. Next year, what number will they use after it is adjusted? 900 +/-122 (i don't understand which way it goes,) or 850 +/- 122.

Thanks

David
4.22.08 @ 12:04

So what is the tiebreaking system? Is it random or do students who have blocked lose(win?) the tiebreaker if two people end up having the same number?

Miles Skorpen
4.22.08 @ 01:04

Your number is always based on what you picked with–so, Aaron, in your case your effective sophomore year number would be 900-122 = 778. (You subtract the 122 because lower numbers = better, and blocking is considered an advantage.)

I'm not sure what the tie-breaking system is. I'll speak with George Dahl and Meggie Ladlow on the topic.

Jean
4.22.08 @ 01:04

Well, I blocked into Woolman last year with an average of 1000 so I could avoid living in PPR/ML like half my class, but I wouldn't say Woolman is exactly desirable. And yet my entire block this year got numbers in the 760's, which basically means off-campus for us again.
Whereas the sophomore block who got Alice Paul had an average 100 lower than us last year, and this year they all got in the 640's, lower than us again.
Since basically all seniors get on campus singles, it doesn't seem fair to me.

Miles Skorpen
4.22.08 @ 01:04

Jean—

Like with the old system, everyone gets a high number, low number, and medium number. It could be that you had a middle low, middle middle, and middle high number.

Bottoms Up
4.22.08 @ 02:04

Okay, so (in the example with the class of 6 people, Sally and Tim each end up with a number higher than what they WOULD have gotten if nobody in their year had blocked (they got 2 and 5, respectively, instead of 3 and 6).

In this example, Sally and Tim both benefit from the block that they had Sophomore year. But Bob, who may not have been awarded his block Sophomore year, gets a worse number than he would have had if Sally and Tim didn't block, meaning that Bob has less of a chance to get a Junior year block… am I understanding this correctly? Do rising Juniors get indirectly penalized for not being awarded their Sophomore block (and thus not getting to live with their friends Sophomore year)?

Miles Skorpen
4.22.08 @ 02:04

There is no penalty to your overall position for trying to block—only for successfully getting a block.

But yes, because Bob was pushed down on the list—though in theory, if Sue had blocked, Bob would have been pushed down one more slot and received the highest number in his class.

curious
4.22.08 @ 02:04

So is this program which powers the lottery system a single computer program? I'd be interested in the details of its function.

Miles Skorpen
4.22.08 @ 02:04

As I understand it, the program was coded in Python by two students, and runs off of a Filemaker database.

NA
4.22.08 @ 07:04

"Like with the old system, everyone gets a high number, low number, and medium number. It could be that you had a middle low, middle middle, and middle high number."
Doesn't the new system end up worse for people with mediocre numbers all four years? I feel that my number last year was at the high end of bad, and this year at the low end of bad. So, in theory this would yield a middle of the road number next year. At least under the old system, you didn't get mediocre numbers all 3 years.
Another point: The waitlist is not a godsend for rising sophomores. I know many people (myself and roommate included) who got the options of: ML, ML, ML, ML or, ML basement.
Lastly, wouldn't it be fair to somehow institute a desireability scale for generating sophomore numbers? Ie, there are clearly dorms that you are placed in as a frosh that are more desireable (a Wharton quad) than others (Willets basement), but in the lottery, the Willets basement resident is just as likely to end up in ML against their will as the Wharton resident is to end up in David Kemp. Does anyone else agree with me that this doesn't seem exactly just?

Miles Skorpen
4.22.08 @ 07:04

In the old system, it was much more random—so you were just as likely to get three low numbers in each cohort as it was to get three high numbers in each cohort. And while we might like trying to odds, in reality it ended up meaning that some students got consistently bad housing while some students got consistently good housing. The new system averages it out—a small number will always get middle, but most will get a good spread of numbers.

If you got two bad numbers, NA, you are in-line to get a good number next year. Possibly a very good number. A good way to check: where are you in your class? If you know your position in your class, you can figure out your effective number for sophomore year and, in turn, find out what number you'll get senior year.

Jeff
4.22.08 @ 09:04

From the numbers that the people I picked with this year got in the lottery, it appears that tiebreaks go by alphabetical order of last name. Since my last name starts with a W, I was #697, one of us with last name H was #696, last name F was 695 and so forth.

seth
4.22.08 @ 11:04

I agree with jean (though, because I blocked into AP and got like a 640 this year, I'm pretty much the ideal benefactor of this weird system). As Roy is explaining right now, the ideal for us is to get a good number as rising sophomores, then a middle number as a junior and a bad number as a senior because all seniors get on campus singles.

Basically the numbers should be equalized for sophomore and junior years, and then senior year should be random again. Otherwise, we get really good things 2nd year, pretty good things 3rd year, and, by default, great things senior year. Other people get shitty things 2nd year, less shitty things the next year, and then great things senior year. It should be equalized so that your sophomore year good number cancels out the next year and not over the next two.

Miles Skorpen
4.22.08 @ 11:04

Seth,

You are basically right … but it isn't significantly different from the current cohort system. There was always a chance you'd get high sophomore, middle junior, low senior. The big difference with this system is that some people won't get high numbers in all three cohorts and others low numbers in all three cohorts.

Chris Green
4.23.08 @ 12:04

I'm sure we could come up with some rough ranking system for which dorms are better than others, but on an individual level there are large differences. I know some people who lived in Willets basement and loved it. I have also known Wharton residents who were ambivalent about it. I was put in ML, loved it, and have stayed here ever since.

The housing lottery is really painful (though I've never played the lottery, I have gone very year and most of my friends played it), but I don't really see that it's more unfair than chance.

Additionally there are two consolations: even ML, way off in the boondocks, is only 15 minutes away from campus on foot. This is better than almost all mid- or large university housing. Second, if your living conditions are completely untenable the housing coordinators have been amazing about getting you something you can live with. It may not happen instantly, and it may not be that sweet Wharton or Parrish single, but it probably won't be ML if you don't want it, either.

The best words come from the Hitchhiker's Guide to the Galaxy:

Don't panic.

off campus
4.23.08 @ 12:04

If the new system only counts numbers you pick off of, how does it account for unused numbers? I won't be picking off my junior year number. Will it still count towards my average if I want to live on campus senior year?

also
4.23.08 @ 01:04

how did people who blocked into a lodge get numbers like 465? that seems soo unfair

Eric
4.23.08 @ 05:04

Given the information in Miles's 4.22.08 @ 01:04 comment about blocking, isn't the example you give incorrect? If Tim and Sally blocked with each other, their block number would be a 4.5, which adjusted by two would be 2.5, each one's effective sophomore year number for purposes of junior numbers. Then the tiebreaking mechanism (the alphabet, apparently) would go into place. I'm pretty sure this is the case, given that my current block now has six numbers in a row in the 750s, meaning that our new numbers were all assigned off the average block number from last year.