A new number-generating program mistakenly issued duplicate Draw numbers to 81 groups this year, adding additional worries to an already complex process for the 135 affected students.

After consulting with the ASSU Undergraduate Housing Advisory Committee, Housing Assignments opted to notify students who received duplicate numbers by email last Friday, after the deadline for listing housing preferences had passed. Students were assured that their housing assignments would not be affected by the mistake.

“This situation occurred because of some enhancements we made this year to the computer process that issues the numbers,” said Director of Housing Assignments Sue Nunan, in an email to The Daily. “Unfortunately, our careful testing of the enhancements did not reveal any anomaly until after actual numbers had been issued.”

Only one of the affected students contacted Student Housing with additional questions, which Nunan said indicated that most students seemed unconcerned by the news.

Scott Schonfeld ‘08, one of the students assigned a duplicate number, said he felt that Housing was too quick to brush away students’ worries.

“It seemed nebulous because they basically said, ‘Don’t worry about it,’” Schonfeld said. “The email stressed that this wouldn’t have any bearing on my number but the reality is that it has to. It bumps up my number one space or two spaces.”

“If there was a mistake in the program that assigns random numbers, who knows what mistakes there are in the Draw program itself,” said Rafe Kinsey ‘08, a student who has worked with Philosophy Prof. Marc Pauly to study the Draw. “I think this shows the importance of having experts working on making sure that these resources work properly.”

Pauly said that a mistake in the number generator could have an impact on the fairness of the Draw if the numbers generated were non-random.

“If there are biases introduced at that level the fairness of the overall system will be impacted,” Pauly said. “The only question is whether that happened.”

Housing emphasized that it has used statistical analysis to double check that no such bias occurred.

“The randomness of the Draw numbers was not affected by this situation,” Nunan said. “The completely separate Draw program and process that assigns students to housing was also not affected in any way by the random number generator process.”

Pauly also said that the mistake raises worries about the opaqueness of the system.

“Historically, what happened is similar to something like the tax system,” Pauly said. “[The Draw] was implemented in a simple form and then in the course of its history people came to add more features.”

“For me the main lesson from this whole error in the Draw numbers,” he said, “is that it is just another argument for really simplifying the process.”

For Schonfeld, worries about fairness after the mistake center primarily on his strategy in selecting housing preferences.

“I structured all my rankings to reflect the confidence that I had in my number,” he said. “We weren’t notified early enough in the process, and I could have been really screwed.”