Skip to main content

Using game theory to fine-tune architectural diplomacy

The attached article is about one English architecture firm’s use of Game Theory when designing new buildings, and interviews the firm’s founder about the applications of Game Theory in this field.  This particular firm, Mzo Tarr Architects, specializes in apartment building, usually in urban areas.  This leads to each building having a number of tenants, as well as the usual variety of clients, neighbors, officials and investors.  Each of these many parties has a variety of needs, all of which must be addressed by the architect.  Using Game Theory, Mzo Tarr can look at all these demands and quantify them into ranked priorities (known as Utility Curves).  From there the firm can see where to sacrifice some preferences and by how much to get a better return elsewhere.

An interesting challenge the firm used Game Theory for was in regard to the amount of garden space in a building as compared to the size  and quality of each flat. From these two desires the firm modeled a game that let the 5 flat tenants per floor to create an amount of garden space in the middle of their floors.  Each flat could give as little or as much floor space (coming from their flat) as they wanted in order to create this garden. To create more of an incentive, the firm would create 2 meters of public garden space for every 1 meter of private space given up.  This encouraged the flat members to work together to all give up a little space for a nicer garden.  Some flat owners were selfish and hoped to use the public space created by other tenants giving up some of their space, however the one person being selfish would lead to the other floor members to be less generous, creating a less nice garden space.  The best results were obtained when all the flats on a floor collaborated and gave up some space, demonstrating that in this case better collaboration led to a better outcome for all.


A second example of the firms use of Game Theory was when juggling the requirements need to get approval for a building in Essex. This particular area where the building was being constructed had a number of constraints and planning policies.  These constraints came from a variety of parties, each of course with varying interests. Mzo Tarr constructed a model that factored in all the groups interests and allowed a predicted approval decision to be predicted for each player.  Using this the firm was able to design the building, and for each change that was made in the design, the change in response from the consulted parties could be predicted.  This allowed the architects to design a building that would be likely approved by all parties.  As of the time of the article being written, the official response has not been released, but initial responses seem positive.


Leave a Reply

Blogging Calendar

September 2016
« Aug   Oct »