Interesting Links for 18-11-2023

[andrewducker]

1. This series of The Crown is the best yet

(tags:TV review monarchy Queen UK )

2. Finally, a solution to the Teleporter problem!

(tags:StarTrek funny death comic )

3. "Personalised Learning" is not the best approach to teaching children

(tags:children Education )

4. Criticisms of BMI have been grossly overstated

(tags:BMI weight health )

5. Half-Life gets a big update for its 25th anniversary

(tags:games valve half-life )

6. Edinburgh named as one of worst cities in Scotland for vacant homes with almost 7,000 lying empty

(tags:housing Edinburgh epicfail )

7. Apple, Disney, Comcast and Warner Brothers Discovery have all halted advertising on X amid antisemitism storm

(tags:bigotry Jews advertising twitter ElonMusk )

Comments

Criticisms of BMI have been grossly overstated

[jack]

This article seems to be muddling two different things. If we're now getting conclusive evidence that interventions to lower weight are successful at reducing serious illness and mortality, then that's significant (although I may continue to weigh the evidence because we've had convincing studies going back and forth over that a lot for a long time).

On the other hand, it says "it's fine to use BMI as a measure because it's good approximation". And like, yes, if this is the only time in their life someone is told "hey, being overweight can be a problem or a symptom of a problem, and sometimes can be be easy address" then yes, taking a quick approximation is useful. Or if it was RANDOMLY 20% accurate, and someone who goes to the doctor a few times a year is told the wrong answer every year or so, then that would probably be fine. But if it gives the wrong answer for the same 20% of the people ALL THE TIME, then perhaps instead of spending 20 seconds calculating a BMI, it would be better spend spending 40 seconds every other visit using a bit of common sense, rather than saying "OK, every three months we're telling someone to stage a medical intervention with significant risks that will only have downsides and no upsides, and shaming them if they don't agree, but that's ok, it only happens to 10% of people." Or maybe less, I don't know.

Lots of metrics are like this. A good guide to "has the situation changed" or "do we need to look deeper", but a bad guide to "should we focus 100% on improving the metric regardless of whether it makes the underlying situation better or worse".

But I know it can make me overly ranty :( I think the situation is something like, something can be useful "most" of the time, and there can still be a serious disagreement between people who are saying "this has significant problems, can we use it less [then 100% of the time]" and people saying "but this is still useful, can we keep using it [more than 0% of the time]", because both sides feel like the other extreme is eclipsing the middle they want, even if one is more right than the other...

Re: Criticisms of BMI have been grossly overstated

[andrewducker]

Oh yes, the way that doctors treat obesity and weight in general is awful. But that would happen with any weight metric.

Re: Criticisms of BMI have been grossly overstated

[jack]

Yeah, I did rather conflate the different problems. But I don't think that was just my reply, in retrospect I think that a reaction against the article implicitly doing the same thing.

If you're recommending something that's harmless to almost everyone and helpful to a big chunk of people, then it makes sense to recommend it based on a very approximate heuristic because it doesn't matter if you're wrong. I the article saying "BMI is useful" is implicitly saying, "it's useful, because it's quick AND THE COST OF GETTING IT WRONG IS LOW". But it's only low if you're confident that the advice you give people with high BMI isn't counterproductive or harmful, whereas it often can be...

[channelpenguin]

When you say "OK, every three months we're telling someone to stage a medical intervention with significant risks that will only have downsides and no upsides, and shaming them if they don't agree, but that's ok, it only happens to 10% of people." What medical intervention do you mean?

(I'm a lifelong skinny person, so I have no idea what doctors say or recommend to anyone else!)

[jack]

Sorry, I was a bit over the top. I meant trying to diet without knowing what was realistic, if it hadn't worked for you before.

Personally the advice I was given was mostly fairly reasonable, the NHS page on losing weight was pretty helpful and responsible (although I didn't do exactly that). But I know that "just eat less" doesn't really work for a big proportion of people.

(I think "just exercising more" is probably usually positive unless someone's already an athlete, although not always easy depending what their life is like. But may lead to being healthier in other ways without losing weight depending on their metabolism, or could lead to fat loss which is partially offset by muscle gain.)

[channelpenguin]

I never thought of dieting as having risks, but I suppose there's some very unsensible versions going about.

Exercise won't use up much excess food, but progressive weight training will build up muscle (slowly!) which increases metabolism and is useful besides. Plus we lose it naturally with age if we don't. Cardio is, well, good for the heart and for some people improves their mood - but we all need to be careful not to overdo it to start with, and sheer weight pressure on the joints can be a problem with the impactful sorts like running.

So yeah, I get it. Imbalanced starvation diet plus excessive running would indeed be a very bad plan that I can imagine some folks might think they should be doing... Sigh. I'm glad it seems the NHS page gives sensible advice.

[anna_wing]

The BMI categories aren't arbitrary either. They reflect the increasing statistical likelihood of the various serious medical conditions that are associated with excess body fat. Which is why it's adjusted for Asians, since sadly because of a greater tendency to accumulate visceral fat we get all the increased risk at a noticeably lower BMI than other racial groups. The actual problem with BMI is that it's too lenient and it doesn't catch slim but unfit people who have too high a percentage of visceral fat.

[channelpenguin]

Yes. One of the problems. The "skinny-fat" slip through.

[bens_dad]

2. That is fine ... until they decline to take the puppy, whether because it of the risk on/to the away-mission, or out of stubbornness. At which point they are not *voluntary* taking the puppy.

Rather than using puppies, the operator could believe that they are assisting in suicide.

[calimac]

To me, "BMI" means Broadcast Music, Inc., a US performing-rights organization.