On the basis of a brief perusal of that post and the Wikipedia article it links to, I’d say it looks ad hoc and overcomplicated. It’s entirely possible that people working with it have had useful insights, but if the approach turns out to be anything like the best way of doing decision theory under conditions of extreme uncertainty then I’ll be extremely surprised.
(I also looked at the website made by the theory’s chief proponent. It seems to have scarcely any actual mathematical content, and to exist only for publicity; that doesn’t seem like a good sign.)
Seeking comments on Info-Gap Decision Theory:
http://www.stat.columbia.edu/~cook/movabletype/archives/2009/04/what_is_info-ga.html
On the basis of a brief perusal of that post and the Wikipedia article it links to, I’d say it looks ad hoc and overcomplicated. It’s entirely possible that people working with it have had useful insights, but if the approach turns out to be anything like the best way of doing decision theory under conditions of extreme uncertainty then I’ll be extremely surprised.
(I also looked at the website made by the theory’s chief proponent. It seems to have scarcely any actual mathematical content, and to exist only for publicity; that doesn’t seem like a good sign.)