Cranky Nick, you really need to join a church. > Now, what most people wanted to know from Nate Silver is whether Clinton was > going to win the election. Nate constantly says that making such predictions > is, strictly speaking, not his job. As long as what happens falls within the > error of his prediction, he feels justified in having made it. He will say > things like, "actually we were right." I would prefer him to say, "Actually > we were wrong, but I would make the same prediction under the same > circumstances the next time.” In other words, the right procedure produced, > on this occasion, a wrong result.
The thing you say here that “most people want to know” of course, you know full-well, doesn’t exist. So you need to join a church because they are the ones who will tell you they are giving it to you, when at least you, and maybe even they, know it doesn’t exist. What Nate gave you is a sample estimator for a probability distribution (each of those words means something specific; they are not an evocative construction within common vernacular). He didn’t even give you the “actual” probability distribution for the underlying process, because, as Pierce saith both rightly and interestingly, the “actual” probability distribution is something we don’t have access to. What we have, and all we ever have, are sample estimators to probability distributions. Nate’s estimator includes biases. Some of these, like method biases in polling, are things he can also try to estimate and correct for. Others, like systematic biases in the relation between sampling and underlying correlations — as in the really interesting and exactly relevant link Marcus sent — are things Nate (et al. of course) haven’t identified. The acknowledgement of those, too, was in the advertising. So, the sample estimator for a probability distribution, with known biases described and correction methods listed, and unknown biases acknowledged, is what Nate gave you, and in the only sense that “right” can be applied — which is an accurate rendering of methods — it was right. If someone gives me a revolver with two filled chambers, and in the afterlife I protest that I didn’t pull one of the empty ones, well, we know what we think of my judgment, and we don’t spend a lot of time on this list putting that out as a philosophical problem. I don’t actually write this note to be nasty -- because of course I know you know all this as well as your interlocutors do — but to be colorful to make a different point. It has to do with liking the fact that learning is not most interesting when one accretes an acquaintance with new facts, but when one realizes new ways of using words are necessary as a vehicle to taking on new frames of mind. The claim that “right/wrong” are only allowed to be applied to certain and definite values, and are _not_ allowed to be applied to more composite deliverables such as sample estimators for probability distributions, is where terminology nazis close off conversation by insisting on a language in which terms that are needed to express the pertinent ideas are disallowed. We see it in every field. Stanley Miller ruled out metabolism as being a concept that could be presaged in geochemistry by “defining” metabolism as chemical reactions catalyzed by enzymes within a cell. Historical linguists did it for a century insisting that absolutely regular sound correspondences (none of which ever actually exist) were the only signatures of genetic relatedness among languages, and probabilistic fingerprints had no interpretation. The Stochastic Thermodynamics cabal do it when the say that thermodynamic laws for non-equilibrium processes that don’t come from Boltzmann/Gibbs free energies have “no physical meaning”, thereby scoping “physical” to refer to equilibrium thermodynamic states, the narrowest of special cases. And Dave did it in his post of long questions some weeks ago — which at the time I didn’t want to respond to because my responses are sort fo dull and unhelpful — when he said most physicists are realists but quantum physicists are anti-realists. What the quantum physicists say is that the old classical assumption that “observables” and “states” are the same kind of thing turned out to be wrong. They are different kinds fo things. States can be real, and can even evolve deterministically, but may not be associated with any definite values for observables, because observables, when formalized and fully expressed through the formalization, are different kinds of things (they are a kind of operator, which one can think of as a rule for making a mapping) than states or than particular numbers that the observables can yield as their output from some states. So to claim that the quantum physicists are anti-realists is to scope “real” as coextensive with interpreting “observables” not as operators but as simple definite numbers. That is, to adopt the frame of classical mechanics. So Dave’s “anti-realist” actually means “anti-classical-mechanics-assumptionist”, which of course is exactly right, but never the scope I would use for the word “real”. Anyone who insists that is the only way it is allowed to be used has just dictated rules for conversation in which there is no way I can engage and still work for sense-making. Anyway, the whole tenor of the discussion is fine. I enjoy all the parts of it, including your stubbornness for its own sake. Wittgenstein was reportedly impossible in that way, though I forget the reference and source. Some fellow-philosopher complaining that “it was impossible to get Wittgenstein to admit there was not a rhinoceros in the room." Eric > > That’s all, > > Nick > > Nicholas Thompson > Emeritus Professor of Ethology and Psychology > Clark University > [email protected] <mailto:[email protected]> > https://wordpress.clarku.edu/nthompson/ > <https://wordpress.clarku.edu/nthompson/> > > > -----Original Message----- > From: Friam <[email protected] <mailto:[email protected]>> On > Behalf Of u?l? ? > Sent: Friday, April 17, 2020 4:45 PM > To: FriAM <[email protected] <mailto:[email protected]>> > Subject: Re: [FRIAM] anthropological observations > > Again, though, you seem to be allowing your metaphor to run away with you. > When someone who does quantitative modeling says "expected value", they do > NOT mean what the layperson means when they say "I expect X". We can pick > apart your statement and accuse you of an ambiguity fallacy if we want. > > Your first use of "expected value" relies on the jargonal definition. Then > you switcheroo on us and your 2nd use of "I expect that" relies on the > vernacular concept. Up to this point, we can give you the benefit of the > doubt. We all munge things a bit when talking/thinking. But *then*, on your > 3rd use of "what he expected", you explicitly switched the meaning from > jargon to vernacular. > > I don't think you do this on purpose. (If you do, I laud you as a fellow > troll! >8^) I think it's an artifact of your being a "metaphorical thinker", > whatever that means. > > FWIW, I only had to pull a little on the Sabine Hossenfelder thread to find > that she tweeted this, as well: > > Embracing the Uncertainties > While the unknowns about coronavirus abound, a new study finds we ‘can handle > the truth.’ > https://www.nytimes.com/2020/04/07/science/coronavirus-uncertainty-scientific-trust.html?smid=tw-share > > <https://www.nytimes.com/2020/04/07/science/coronavirus-uncertainty-scientific-trust.html?smid=tw-share> > > The effects of communicating uncertainty on public trust in facts and numbers > https://www.pnas.org/content/117/14/7672.abstract > <https://www.pnas.org/content/117/14/7672.abstract> > > If they're right, then the right-leaning local media might band together with > the clickbaity national media and give it to us straight ... or they might > simply skew their "expected value" reporting to continue serving their > politics. Pfft. > > > On 4/17/20 2:58 PM, [email protected] <mailto:[email protected]> > wrote: > > If expert X tells me that the expected value of variable A is K, then, > > when it's all over and the data are in, and A did not equal K, I expect > > that expert to admit that /what he expected did not happen./ Only after > > that confession has been made, should a conversation begin about whether > > the expert’s prediction process was faulted or not. It seems to me that > > the shaded area is part of that second conversation. > > -- > ☣ uǝlƃ > > .-. .- -. -.. --- -- -..-. -.. --- - ... -..-. .- -. -.. -..-. -.. .- ... > .... . ... > FRIAM Applied Complexity Group listserv > Zoom Fridays 9:30a-12p Mtn GMT-6 bit.ly/virtualfriam > <http://bit.ly/virtualfriam> unsubscribe > http://redfish.com/mailman/listinfo/friam_redfish.com > <http://redfish.com/mailman/listinfo/friam_redfish.com> > archives: http://friam.471366.n2.nabble.com/ > <http://friam.471366.n2.nabble.com/> > FRIAM-COMIC http://friam-comic.blogspot.com/ > <http://friam-comic.blogspot.com/>.-. .- -. -.. --- -- -..-. -.. --- - ... > -..-. .- -. -.. -..-. -.. .- ... .... . ... > FRIAM Applied Complexity Group listserv > Zoom Fridays 9:30a-12p Mtn GMT-6 bit.ly/virtualfriam > <http://bit.ly/virtualfriam> > unsubscribe http://redfish.com/mailman/listinfo/friam_redfish.com > <http://redfish.com/mailman/listinfo/friam_redfish.com> > archives: http://friam.471366.n2.nabble.com/ > <http://friam.471366.n2.nabble.com/> > FRIAM-COMIC http://friam-comic.blogspot.com/ > <http://friam-comic.blogspot.com/>
.-. .- -. -.. --- -- -..-. -.. --- - ... -..-. .- -. -.. -..-. -.. .- ... .... . ... FRIAM Applied Complexity Group listserv Zoom Fridays 9:30a-12p Mtn GMT-6 bit.ly/virtualfriam unsubscribe http://redfish.com/mailman/listinfo/friam_redfish.com archives: http://friam.471366.n2.nabble.com/ FRIAM-COMIC http://friam-comic.blogspot.com/
