<p>Would any current students be willing to talk with me about their experiences with Columbia's lottery? I'm doing a theoretical study of assignment systems and wanted some general ideas of how people approach them in real life. I just have a few questions now, but might send you a couple more as work progresses, if you don't mind.</p>
<p>Just PM me if you're interested!</p>
<p>(apologies in advance if I'm slow to respond; lots of irons in the fire this time of year)</p>
<p>Talk to Denzera.</p>
<p>The Lord of Housing</p>
<p>There are generally two types of selection times: Group selection and general selection. Group selection is with more than 1 person, general is if you want to go alone. The lottery begins at the end of freshmen year, where you can choose either to go into group or general. The important thing is that group selection always chooses first before general, so most people pair up. Furthermore, you can always drop down into individual general selections if you want to during group, but you can't go into groups after you get your general selection number.</p>
<p>You get two numbers, one is the class number (10 for incoming sophomores, 20 for juniors, 30 for seniors) and one is the actual number (varies between 1-3000). Each person will have two numbers (10/300... 30/2425). The higher first number, the senior in this case, will always choose before the lower first number. In addition, some groups will have a first number of 15 (a sopohmore and junior), some 21.666 etc (groups vary from 2-8). </p>
<p>If you choose to drop down into general selection, in most cases everyone will hold that same number. In group selection you can only choose your own suite size (if you entered as 8, you must pick a suite size of 8), or you can drop down into individual groups. </p>
<p>Most upcoming sophomores want singles in Furnald, however they will go into group selection as doubles. If they get a bad number (10/2000), they will still be able to get a decent double, but if they get a great number (10/320), they will split into general selection, and each will then be able to select singles during general selection (remember, singles can't be taken in group selection).</p>
<p>More information is posted on the housing website, but I hope this summary helped. </p>
<p>Yeah, some people get shafted by the lottery, some people don't. It's all luck. But do try to go into group selection unless you absolutely are sure you want a single.</p>
<p>Thanks, that's really helpful!</p>
<p>(anyone else can feel free to post whatever they think of the system, too. even if it DOESN'T help me, it'll probably help someone else here eventually)</p>
<p>I replied via PM as requested, but can you tell us all -</p>
<p>are you a student at columbia? if so, what year? if not, what sort of study are you doing that you're asking this?</p>
<p>Give us some background here.</p>
<p>No, I'm a Barnard alumna. The "house allocation problem" is a pretty common one for people studying markets; I wanted to get a sense of some of the different tweaks various schools have done to the standard mechanism, ideally get some data to test predictions, and maybe make some statements about how different rules affect strategic group formation. I'm minimally familiar with Barnard and Columbia's rules, so I've started there.</p>
<p>Oh, and does anyone know when Columbia instituted the senior regroup option? I'd love some before and after outcomes on any of these rules.</p>
<p>But yeah, thanks for all your help, everyone. :)</p>
<p>I believe the first year of senior regroup was the lottery for the 0405 year, meaning it took place in spring 04. I remember it well, because somehow a Hogan 5-person suite made it to regroup and half the senior class regrouped to groups of 5 to try for it (irrational!). Groups of 6 and 2 made out like bandits that year.</p>
<p>I have a document - a research paper, actually - that you may find interesting, PF. PM me your email and i'll fire it over.</p>
<p>-D</p>
<p>Ooh, good to know it was so recent! That's exactly the sort of thing I'm interested in.</p>