That approach is directly suggested in the accompanying article. The suggested problem is that square footage may not be sufficient to express the actual value preference difference between rooms. For instance, as the article suggests, what is the value of having a window?
That's, thus, exactly the neat part of this algorithm. So long as the participants are honest then it converges on an "envy-free" set of room selections and prices. In other words, by using only your own preference for room-X-at-price-Y it builds a set of compatible room-X-at-price-Y choices provided to each participant such that they've shown they would not want to trade with anyone else for their room-X-at-price-Y option.
The beauty of this is that the entire interface is done through these room-X-at-price-Y preferences and options so it abstracts over technical details like the relative value of a window. It's simply up to the participants to decide what they value.
Comments
That approach is directly suggested in the accompanying article. The suggested problem is that square footage may not be sufficient to express the actual value preference difference between rooms. For instance, as the article suggests, what is the value of having a window?
That's, thus, exactly the neat part of this algorithm. So long as the participants are honest then it converges on an "envy-free" set of room selections and prices. In other words, by using only your own preference for room-X-at-price-Y it builds a set of compatible room-X-at-price-Y choices provided to each participant such that they've shown they would not want to trade with anyone else for their room-X-at-price-Y option.
The beauty of this is that the entire interface is done through these room-X-at-price-Y preferences and options so it abstracts over technical details like the relative value of a window. It's simply up to the participants to decide what they value.