Nice! NFTC has outdone even the venerable FIFA in establishing new awards for 2025.
I'm considering creating an annual award for best new award -- but ran into a problem: I wish to make ineligible any award that is nominated for itself. Will my new award be eligible for itself? It will be only if it isn't. To resolve this, I'm currently plowing through Principia Mathmatica and after 400 pages or so I have the proof that 1 + 1 = 2. I anticipate soon proving that two plus two equals four, and hope with the assistance of chat AIs to have the definitive list of nominees by the end of this century.
Hank Green is the millennial Bill Nye, but better. He is like me, a hobbyist and dabbler and lover of too many things to fit cleanly into one category, but he is excellent at explaining pretty much anything.
Great choices, Mike! I am a fan of all three! Been watching Anton for years (Hello, wonderful person!). I discovered Hank and Vlad more recently, but they are all the best things on YouTube in their respective niches. Food for the brain!
:Often results are preliminary and need verification—something a lot of science communicators use as a territory of grift. Research studies that have statistically-significant findings at low P-values—a signal that you should do more experiments because there could be something there—are often dramatized into game-changing breakthroughs, apocalypse fear-mongering, etc.
Anton is very careful to guard his readers away from these flights of fancy. He is epistemically responsible! And for this, he is my first recipient of the NFTC Award for Epistemic Excellence in Science."
I would suggest that you are being epistemically irresponsible in the way you have presented the above. In essence, you are not providing the whole story, and in doing so you may be tossing out the baby with the bathwater.
The editorial effectively throws out the baby (the concept that a low p-value is the necessary starting point for statistical inference) by failing to explain that the problem is not the p-value itself, but its interpretation without context. This misleads the non-scientific reader. It connotes or at least unsatisfactorily addresses the issue of "necessary and sufficient."
Risk of Misinformation: By not mentioning Effect Size, the confidence interval (CI), and the quality of the study, the editorial risks teaching the reader to simply distrust any low p-value, which is a harmful oversimplification. A medical professional knows a statistically significant finding, even a preliminary one, may possess high clinical value when the effect size is large, and/or the confidence intervals are narrow and the quality of the study is excellent. These nuances are not cited in your commentary discussing possible scientific grift. You've gone too far.
The most epistemically responsible writing on this topic would acknowledge the low p-value as a required threshold and immediately pivot to the metrics of estimation (Effect Size and CI) to measure the clinical validity. By failing to complete this pivot, the editorial demonstrates a lack of intellectual thoroughness in its own argument, (i.e., the commentary is epistemically irresponsible).
I have not read any of Anton Petrov's writings. For sure, I have in my realm of expertise (prostate cancer) seen published studies where the clinical benefit of Regimen A vs Regimen B is a matter of a few weeks. Added to that, the study was multi-institutional and does not account for the wide variation in medical care from one MD to the next. Possibly, the differences could be due to a bunch of mediocre docs at San Quentin State Medical Center vs much more attentive docs at Ain't She Beautiful Health Center. I know of one study where the side effects of the preferred regimen are far more serious and detrimental to the quality of life yet the extra two weeks and less frequent visit of the patient to the oncologist pushed it into the preferred treatment category. So, sure, there are poorly done studies.
Your comment on this particular subject does deserve clarification.
I did go to the Vlad Vexler site and do find his site interesting.... but. I see comments to one of his posts that are glib, typically the type one sees on what I consider superficial social media sites and which I allude to as "Have a Nice Day" level of commentary. In other words, I think your readers' comments are indicative of far more reflection and thought. I believe the site I went to was per your link to the "Main" YouTube channel. Maybe my sampling was far too superficial, but I think your audience is more learned. His site feels disorganized also and I cannot navigate to recent chats. I see "two weeks ago" and many links to the same general categories (e.g., Ukraine, Russian aggression, Putin). It may be my inability to grasp the structure of his site.
Damn, I totally forgot about vlad. It's been years since I watched his videos but I recall being taken by surprise at some of his insights. Will definitely revisit that one.
Nice! NFTC has outdone even the venerable FIFA in establishing new awards for 2025.
I'm considering creating an annual award for best new award -- but ran into a problem: I wish to make ineligible any award that is nominated for itself. Will my new award be eligible for itself? It will be only if it isn't. To resolve this, I'm currently plowing through Principia Mathmatica and after 400 pages or so I have the proof that 1 + 1 = 2. I anticipate soon proving that two plus two equals four, and hope with the assistance of chat AIs to have the definitive list of nominees by the end of this century.
Sharp wryness. Much respect.
Hank Green is the millennial Bill Nye, but better. He is like me, a hobbyist and dabbler and lover of too many things to fit cleanly into one category, but he is excellent at explaining pretty much anything.
Very much in agreement with Vlad Vexler. I'll have to check out Hank Green.
Great choices, Mike! I am a fan of all three! Been watching Anton for years (Hello, wonderful person!). I discovered Hank and Vlad more recently, but they are all the best things on YouTube in their respective niches. Food for the brain!
Thanks for the recommendations!
Subscribed to all three
Mike, you wrote:
:Often results are preliminary and need verification—something a lot of science communicators use as a territory of grift. Research studies that have statistically-significant findings at low P-values—a signal that you should do more experiments because there could be something there—are often dramatized into game-changing breakthroughs, apocalypse fear-mongering, etc.
Anton is very careful to guard his readers away from these flights of fancy. He is epistemically responsible! And for this, he is my first recipient of the NFTC Award for Epistemic Excellence in Science."
I would suggest that you are being epistemically irresponsible in the way you have presented the above. In essence, you are not providing the whole story, and in doing so you may be tossing out the baby with the bathwater.
The editorial effectively throws out the baby (the concept that a low p-value is the necessary starting point for statistical inference) by failing to explain that the problem is not the p-value itself, but its interpretation without context. This misleads the non-scientific reader. It connotes or at least unsatisfactorily addresses the issue of "necessary and sufficient."
Risk of Misinformation: By not mentioning Effect Size, the confidence interval (CI), and the quality of the study, the editorial risks teaching the reader to simply distrust any low p-value, which is a harmful oversimplification. A medical professional knows a statistically significant finding, even a preliminary one, may possess high clinical value when the effect size is large, and/or the confidence intervals are narrow and the quality of the study is excellent. These nuances are not cited in your commentary discussing possible scientific grift. You've gone too far.
The most epistemically responsible writing on this topic would acknowledge the low p-value as a required threshold and immediately pivot to the metrics of estimation (Effect Size and CI) to measure the clinical validity. By failing to complete this pivot, the editorial demonstrates a lack of intellectual thoroughness in its own argument, (i.e., the commentary is epistemically irresponsible).
I have not read any of Anton Petrov's writings. For sure, I have in my realm of expertise (prostate cancer) seen published studies where the clinical benefit of Regimen A vs Regimen B is a matter of a few weeks. Added to that, the study was multi-institutional and does not account for the wide variation in medical care from one MD to the next. Possibly, the differences could be due to a bunch of mediocre docs at San Quentin State Medical Center vs much more attentive docs at Ain't She Beautiful Health Center. I know of one study where the side effects of the preferred regimen are far more serious and detrimental to the quality of life yet the extra two weeks and less frequent visit of the patient to the oncologist pushed it into the preferred treatment category. So, sure, there are poorly done studies.
Your comment on this particular subject does deserve clarification.
I did go to the Vlad Vexler site and do find his site interesting.... but. I see comments to one of his posts that are glib, typically the type one sees on what I consider superficial social media sites and which I allude to as "Have a Nice Day" level of commentary. In other words, I think your readers' comments are indicative of far more reflection and thought. I believe the site I went to was per your link to the "Main" YouTube channel. Maybe my sampling was far too superficial, but I think your audience is more learned. His site feels disorganized also and I cannot navigate to recent chats. I see "two weeks ago" and many links to the same general categories (e.g., Ukraine, Russian aggression, Putin). It may be my inability to grasp the structure of his site.
I have an award system on Bluesky but it is to be unexpected for the damned. I value your award platform.
Damn, I totally forgot about vlad. It's been years since I watched his videos but I recall being taken by surprise at some of his insights. Will definitely revisit that one.
This is awesome. Great idea.
Thank You for Sharing.
Contemplating “what’s real”.