3 ms·
I prefer Campbell’s Law — the more a metric is used in social decision-making, the more likely it is to be manipulated. You’ve stated the widely-accepted formu
by blueboo 2y ago
I prefer Campbell’s Law — the more a metric is used in social decision-making, the more likely it is to be manipulated.
You’ve stated the widely-accepted formulation of Goodhart’s, but it can be interesting to note Charles Goodhart’s original statement was that “any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes.” (The implication is people doubly go wrong presuming the regularity’s existence and placing pressure on it!)
- friendzis 2y agoI'm being humorous here. I kinda disagree with tour takeaway. I have expanded on this idea elsewhere in this thread, however, the main point is that a measure is a byproduct of an existing process. Making the measure a target, i.e. process output, changes (possibly inadvertently) the process itself. It does not matter if the regularity truly existed or was wrongly presumed to exist, as putting pressure on it invalidates the causal relationship it had before.
- zmgsabst 2y agoI’m not sure I’ve heard it summarized that way before: that good process causes good metrics, but good metrics don’t cause good process. People are indeed prone to affirming the consequent. https://en.wikipedia.org/wiki/Affirming_the_consequent https://en.wikipedia.org/wiki/Affirming_the_consequent