Jump to content

Talk:Goodhart's law

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

This is an old revision of this page, as edited by Belteshazzar Mouse (talk | contribs) at 14:24, 27 February 2024 (Concern on the often used quote, which is not from Goodhart: Strathern). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.


Margaret Thatcher note

I believe this should be removed from Formulation for being too specific for this section, or moved elsewhere in the article. — Preceding unsigned comment added by 141.226.240.109 (talk) 16:52, 29 June 2020 (UTC)[reply]

Absolitely. This is Alex McVean AKA Joey Plus currently from bluelight.org. The WORST program ever devised! Johntitor93 (talk) 00:09, 25 July 2023 (UTC)[reply]

Concern on the often used quote, which is not from Goodhart

The quote "When a measure becomes a target, it ceases to be a good measure" presented in the first line of the article as a formulation of Goodhart's law is in fact a citation from Marilyn Strathern paper, as stated under Generalization. The first line of the article does not explain this, posting the quote as an alternative version of the Goodhart's law. To provide correct attribution and prevent misleading the reader about the author of the quote, shouldn't the page clearly present that the often used quote is not in fact the Goodhart's law but it's generalization?

In addition, Strathern's quote talks about measures, while Goodhart's quote talks about analysis (observed statistical regularity). Measures are not bad, they are simply measures. We should not complain that a scale reads more pounds than we want or that an inch is too long when we are not as tall as we wish. If someone tells us that we are overweight or not tall enough, they are engaging in analysis. The measures, pounds and inches, do not change or become bad. Measures are only signs that lead us toward analysis. Goodhart specifically calls out statistical regularity, which is analysis, and notes that human beings will try to game the system when that statistic is observed and shared with others as an indicator.