LogFAQs > #934702417

LurkerFAQs, Active DB, DB1, DB2, DB3, DB4, DB5, Database 6 ( 01.01.2020-07.18.2020 ), DB7, DB8, DB9, DB10, DB11, DB12, Clear

Topic List	Page List: 1
Topic	Politics Containment Topic 271: Two Turds With One Stone
red sox 777 02/21/20 6:49:29 PM #253:	LordoftheMorons posted... 1. The people with the most relevant expertise to analyze this are economists, not doctors. 2. If everything you say is 100% true and electing Bernie is going to save tens of thousands of lives and you see people getting turned off by mean tweets, would it not logically follow that, I don't know, these people not being a dick to everybody who disagrees with them? After all, it's saving lives! I've been clear that I think there is massive risk in transitioning to M4A. It's an extremely complex task with a ton of potential for unintended consequences, and nothing that Bernie has done has convinced me that he can bring those risks down to a low enough level to justify it. LOTM, I get the feeling that you didn't study economics in college. I did. There's a big difference between how models are tested in economics and in hard sciences. In hard sciences, it's regarded as critical that the experiment that tests the model comes after the model, and is replicable. In economics, especially macroeconomics, it often isn't possible to construct experiments capable of testing models and so they are "tested" with past data. Which means there is a huge danger of overfitting. If you already know the data you need to explain, and you are trying to construct a model that makes sense intuitively and fits the data, that is a lot easier than constructing a model that can predict data you haven't seen yet. And it's even harder to make that replicable (so I can't just run a gazillion experiments until one of them fits one of my models purely from luck). --- September 1, 2003; November 4, 2007; September 2, 2013 Congratulations to DP Oblivion in the Guru Contest! <cite>red sox 777 posted [p:934702417] in Politics C... [t:78420451]...</cite> <quote>LordoftheMorons posted... </cite> <quote>1. The people with the most relevant expertise to analyze this are economists, not doctors. 2. If everything you say is 100% true and electing Bernie is going to save tens of thousands of lives and you see people getting turned off by mean tweets, would it not logically follow that, I don't know, these people not being a dick to everybody who disagrees with them? After all, it's saving lives! I've been clear that I think there is massive risk in transitioning to M4A. It's an extremely complex task with a ton of potential for unintended consequences, and nothing that Bernie has done has convinced me that he can bring those risks down to a low enough level to justify it.</quote> LOTM, I get the feeling that you didn't study economics in college. I did. There's a big difference between how models are tested in economics and in hard sciences. In hard sciences, it's regarded as critical that the experiment that tests the model comes after the model, and is replicable. In economics, especially macroeconomics, it often isn't possible to construct experiments capable of testing models and so they are "tested" with past data. Which means there is a huge danger of overfitting. If you already know the data you need to explain, and you are trying to construct a model that makes sense intuitively and fits the data, that is a lot easier than constructing a model that can predict data you haven't seen yet. And it's even harder to make that replicable (so I can't just run a gazillion experiments until one of them fits one of my models purely from luck). --- September 1, 2003; November 4, 2007; September 2, 2013 Congratulations to DP Oblivion in the Guru Contest!</quote> ... Copied to Clipboard!
Topic List	Page List: 1