Comments on: Models Didn’t Bring Down Wall Street; People Brought Down Wall Street http://blog.sentrana.com/2009/05/12/models-didnt-bring-down-wall-street-people-brought-down-wall-street/ Turning complexity into competitive advantage Wed, 20 Jan 2010 23:15:41 -0500 http://wordpress.org/?v=2.9.1 hourly 1 By: Greg Hill http://blog.sentrana.com/2009/05/12/models-didnt-bring-down-wall-street-people-brought-down-wall-street/comment-page-1/#comment-124 Greg Hill Tue, 26 May 2009 18:32:44 +0000 http://blog.sentrana.com/?p=185#comment-124 Katrina, Isn't the real problem that people make decisions which, unless you're a determinist, are autonomous? Unlike the tides, human decision making is not subject to causal laws that remain constant over time. Katrina,

Isn’t the real problem that people make decisions which, unless you’re a determinist, are autonomous? Unlike the tides, human decision making is not subject to causal laws that remain constant over time.

]]> By: Staufer http://blog.sentrana.com/2009/05/12/models-didnt-bring-down-wall-street-people-brought-down-wall-street/comment-page-1/#comment-123 Staufer Tue, 26 May 2009 05:58:37 +0000 http://blog.sentrana.com/?p=185#comment-123 Have you read "The Black Swan" by N.N. Taleb? I find his critique of all predictive models very persuasive. There are simply too many unknowable factors that can have a decisive impact on the future. You can only predict on the basis of historical data, but new data might pop up. You don't even have all old data, or you have them in a not very reliable form. Take a closer look how government collects data and you will see the point. So you have an incomplete set of somewhat dubious data, you have no way of incorporating future data that will pop up outside of your present variables, and you think fancy mathematics will turn this dog's dinner into a a five-star French diner? Get real. Have you read “The Black Swan” by N.N. Taleb? I find his critique of all predictive models very persuasive. There are simply too many unknowable factors that can have a decisive impact on the future. You can only predict on the basis of historical data, but new data might pop up. You don’t even have all old data, or you have them in a not very reliable form. Take a closer look how government collects data and you will see the point. So you have an incomplete set of somewhat dubious data, you have no way of incorporating future data that will pop up outside of your present variables, and you think fancy mathematics will turn this dog’s dinner into a a five-star French diner? Get real.

]]>