"They were independent barons," said Young. "I think some of the stereotypes in the freshman mythology date back to the days of masters' choice."

But master's choice eventually faded out as the arbiter for housing assignments.

By the late 1950s, Dean of the College L. Watson created a review committee to "add diversity" to the houses by taking 30 percent of house assignment away from the masters.

"Although it's a great deal of work, it is far preferable to the IBM method used at Yale," said Watson in a 1959 reference to the New Haven university's policy of random housing.

Student choice was on the rise, however. In 1959, 66 percent of freshmen gained entrance to their top choice. In 1965, the last year of this process, 70 percent received their first choice.

Social Engineering

But the turmoil of the 1960s youth movement soon swept over Harvard. Searching after a "fairer" ideal, a 1966 House Assignment Committee revolutionized student life after the years of masters' dominance.

Through careful planning--reminiscent of social engineering--the new process aimed to please students while organizing houses into a thorough mix of the student body.

Qualities such as student personality and secondary school background weighed heavily in the system. Students whose characters ranged from "outstanding" to "innocuous" to "possible problem" were blended into the houses. Meanwhile, each of four categories of high schools--"traditional preparatory," Exeter and Andover, "other" private schools and public high schools--had to be represented in each house.

"Our first aim is to please the students, but we have to stay within certain limits," Watson said in 1968.

The limits were breached more drastically than Watson imagined they would be. In 1972, coeducation came to Harvard, adding a degree of complication to housing that led to computerized selection.

A 1973 computer system, programmed to diversify the Houses by evenly distributing students according to field of concentration, rank group, high school background and master's preference, resulted in fewer than half of all freshmen receiving their first choice. Fifteen percent were denied all their top five choices.

Outraged students petitioned and protested. A lottery was proposed, but the new Dean of the College Charles P. Whitlock said such a process "doesn't fit into the Harvard way of doing things."

Protest changed Whitlock's mind in 1974. The Committee on Housing and Undergraduate Life created a partial lottery system, from which the current housing lottery process has evolved.

Under the 1974 plan, students ranked their preference of the 12 houses. Houses were lotteried only if more students listed them as first choices than rooms were available.