Reply to thread

Message: <blockquote data-quote="Ovinomancer" data-source="post: 7474498" data-attributes="member: 16814">I'm not trying to be abusive.  I actually like and generally respect your posts -- you're one of the posters I perk up when I see the handle, as you usually have an interesting take on whatever's being discussed.  I usually try to avoid the fisking style of discussion, as it usually leads to gross mischaractization of points and gets into <img src="https://cdn.jsdelivr.net/joypixels/assets/8.0/png/unicode/64/1f642.png" class="smilie smilie--emoji" loading="lazy" width="64" height="64" alt=":)" title="Smile    :)"  data-smilie="1"data-shortname=":)" /><img src="https://cdn.jsdelivr.net/joypixels/assets/8.0/png/unicode/64/1f642.png" class="smilie smilie--emoji" loading="lazy" width="64" height="64" alt=":)" title="Smile    :)"  data-smilie="1"data-shortname=":)" /><img src="https://cdn.jsdelivr.net/joypixels/assets/8.0/png/unicode/64/1f642.png" class="smilie smilie--emoji" loading="lazy" width="64" height="64" alt=":)" title="Smile    :)"  data-smilie="1"data-shortname=":)" /> for tat, but failed here.  I also apologize for my contributions to the tone.Moving forward, process oriented is where the process is the important part. If the process isn't followed, the results are usually diminished or useless (luck happens sometimes).  Outcome oriented, on the other hand, really is about results over process -- it doesn't matter how you did it if it worked.  A good example of this is sports.  It doesn't matter how you score, the score is the thing.  Bringing this around to science, there's no amount of being right that justifies not following the method.  The method is what allows science to be ultimately self-correcting* and to be shown to others so that they can believe the outcome rather than have to accept it on faith.  Being correct in a prediction doesn't require science, so science cannot be just being correct in a prediction (and, arguably, some of the most important science done has been showing predictions to be wrong).This is why I say outcomes cannot be the definition of science.  Ultimately, science is not just the result but also the completely reviewable chain of evidence that led to that result.  And, to again say it, statistics is subjective in all cases -- you choose a model to use with the associated parameters because you think it will provide the best fit.  That choice, the choice of model, is subjective, and each choice provides differing answers.  Further, if you select data input to your model based on how you think they did rather than a rule (and Silver evaluates polls individually), then that is also subjective.  This isn't to say that the subjective choice cannot be informed or subject to expertise, but, by nature, it isn't dealing with clean data, and anything you do past the initial step is working with modeled data rather than real data.  Silver is exceeding good, and I certainly do not recommend dismissing him, but what he does is fundamentally not science in the same way that sports betting isn't science.  * Although, there's likely a case where any tool used by humans cannot self-correct at a certain point -- a hammer cannot exceed the skill of the carpenter.</blockquote>

Verification