Lockdown!

Posted on by

Share your experience, tips, advice, questions…

As it seems most communities world-wide are going into voluntary or enforced quarantine that involves staying at home and avoiding physical contact as much as possible, I thought we could have a thread where we could try a bit of mutual support by cheering each other up over the next few days, weeks, months… Who knows?

I don’t know: suggestions on films to watch, books to read, gardening tips, exercise ideas

Usual rules apply plus a guideline. Let’s be kind and supportive to each other.

932 thoughts on “Lockdown!”

  1. Kantian Naturalist: There’s actually an interesting question about whether evolutionary theory has any predictive utility, and if predictive utility is the most important thing to look for in a scientific theory.

    I would look on predictive utility as a useful litmus test, and not the “most important thing to look for”.
    I also understand that many of the predictions of evolutionary theory may seem underwhelming in their utility…
    {where we will find which fossils} one could argue that this of utility only to fossil hunters…
    {every single time someone sequences a previously unsequenced fragment of naturally occurring DNA} in itself, it’s really only a test of evolutionary theory — there’s no external utility.
    {the way that bacteria, viruses, and cancers will ‘escape’ treatments, if possible} extremely high utility, but you could argue that this is a truism, and is unrelated to ‘history of life’-style evolutionary theory.
    {the ability to design robust assays for DNA from specific species} it may sound underwhelming, but it’s undeniable.

  2. DNA_Jock,

    I would argue that there is great utility in determining which species are more and which less closely related to another. Medical sciences test treatments on model organisms — without that they’d be in bad shape. Creationists like to argue that MDs don’t use evolution in practice. They don’t think they do, but they forget the model organisms. Knowing that a mouse is better for testing medicines than, say, a carrot is an inference from evolutionary biology.

  3. I agree with you, Joe. But creationists somehow argue that the broad-strokes degrees of similarity are “obvious”; think Linnaeus. So according to their limited understanding, evolution doesn’t help.

  4. Alan Fox: Can you find anything to confirm that assertion?

    Well known fact.

    Kantian Naturalist: There’s actually an interesting question about whether evolutionary theory has any predictive utility, and if predictive utility is the most important thing to look for in a scientific theory.

    No question. It’s settled if even Alan Fox agrees. No process, no theory either. “Evolution” is just bullshit end to end.

    Allan Miller: Creationism evades that how?

    Creationism is a belief… just like “evolution”. Doesn’t pretend to be “scientific” and “theory”. You should agree “evolution” is just like creationism as you imply.

    ID is another matter. And for that, there’s this pure science: http://theskepticalzone.com/wp/intelligent-design-detection/

    Allan Miller: It is a simple mathematical fact that if something increases in frequency everything else in that collection – whatever it is – decreases in frequency. Because all frequencies must add up to 100%.

    False and stupid. Learn math. A [6-face] die outcome collection can be 1,2,3 only or 2,4,6 only. And then probability adds up to 50% ONLY.

    Flint: Well, evolutionary theory predicts which past organisms will be found in which strata, and where to find those strata. My understanding is that a good many fossils have been found ONLY because evolutionary theory predicted where they’d be found in some detail.

    False. It’s just your regular interpolation (if you know what that is). Has nothing to do with “evolution”.

  5. Joe Felsenstein: Knowing that a mouse is better for testing medicines than, say, a carrot is an inference from evolutionary biology.

    False. Has nothing to do with “evolution”. Just with observing a mouse is more like a human than a carrot. Zero value in “evolution” other than entertaining locked up desperados.

    DNA_Jock: I agree with you, Joe. But creationists somehow argue that the broad-strokes degrees of similarity are “obvious”; think Linnaeus.

    Exactly. And true.

  6. Nonlin.org: Alan Fox:

    Can you find anything to confirm that assertion?

    Well known fact.

    Which parses to?

    non-lin has no evidence for its unsupported claim.

  7. Nonlin.org,

    False and stupid. Learn math. A [6-face] die outcome collection can be 1,2,3 only or 2,4,6 only. And then probability adds up to 50% ONLY.

    Huh? The die has 6 faces. If you look at 3 of them, you’ve looked at 50%. That leaves 50% unexamined, ie 100% total.

  8. Nonlin.org,

    ID is another matter. And for that, there’s this pure science: 

    Did you miss some scare quotes? How’s design detection not post hoc?

  9. Allan Miller: Huh? The die has 6 faces. If you look at 3 of them, you’ve looked at 50%. That leaves 50% unexamined, ie 100% total.

    Are you sure this is right? I used to think that 100% meant 1 (one, unity,) but maybe non-lin is doing something non-linear. He might explain. You never know.

  10. Alan Fox: Are you sure this is right? I used to think that 100% meant 1 (one, unity,) but maybe non-lin is doing something non-linear. He might explain. You never know.

    As far as frequency’s concerned, yes, I think it’s right. When we talk of the frequency of an allele, we’re talking the percentage of the population that has it. It must add up to 100% across all alleles at a locus of interest. 100% of the pie is the pie, however sliced.

  14. Alan Fox,

    She is always excellent. I’ve largely refrained from political commentary at this time – it’s not certain anyone’s doing better in the immediate vicinity – but, strictly entre nous, I wonder if we are reaping the rewards of restricting the talent pool to committed Brexiters.

    Just as I wrote the above, Channel 4 were announcing that later today they’ll be showing the 1940’s Vera Lynn favourite “We’ll Meet Again”. I’m sure tongues are somewhat in cheek, but I did splutter!

  15. Some interesting preliminary results from the Covid tracker app, for the nerds among us. First is that the best predictors of incipient infection aren’t necessarily dry cough/fever as per Government advice, but anosmia, fatigue and breathlessness. Second is that the data indicate some degree of optimism, since reported symptoms are dropping. Unfortunately, there’s been a million drop in users as the novelty wears off, which may affect the reliability of the data.

    But, if real, this is useful, because the daily horrors of the reported Worldomoter stats (mindful of my daughter working in ICU) are very biased towards the tail of the distribution, since testing (and death, of course) is largely among those ill enough to be hospitalised, 3 weeks shifted from the actual arrival of new infections.

    One thing puzzling me is that France is continuing to pull away, despite locking down a week ahead of the UK, and being more draconian. Somewhat pointlessly draconian, IMO, in terms of effect on spread. The threatened equivalent tightening of the rules in the UK seems designed to make the job of the state easier rather than necessarily for solid epidemiologic reasons. It’s popular: people are buying wholesale into the memetic ‘stay home, protect the NHS’, resistant to more nuanced arguments. A friend I’ve been jousting with has just announced his departure from Facebook with a valedictory “obey the rules even if you don’t agree with them”. Which leads us into a whole new philosophical thicket!

  16. Allan Miller,

    It seems that the effort needs to move from educating Nonlin on a theory with a vast amount of facets to consider to educating Nonlin on principles of basic mathematics. I suppose it’s not surprising, given nonlin’s OP on regression to the mean, but it is rather tedious, especially when someone denies such basic mathematical principles as the percent sign.

  18. Schizophora: I believe the proper word is “Brexiteers,” but I could be completely off base.

    Both are used. But there is an increasing mood among the rest of us to avoid that one, as it conjures up images of bold swashbucklers, and they themselves adopt it with pride.

  20. There’s a statistical analysis showing that having CV doubles your risk of dying in the 12 month period in which you have it. Which doesn’t explain anything, but it describes the risk. It perfectly reflects the risks associated with age and medical condition. You can die at any age, but the odds are easy to calculate.

  23. Allan Miller: False and stupid. Learn math. A [6-face] die outcome collection can be 1,2,3 only or 2,4,6 only. And then probability adds up to 50% ONLY.

    Huh? The die has 6 faces. If you look at 3 of them, you’ve looked at 50%. That leaves 50% unexamined, ie 100% total.

    Doesn’t matter. Your collection is whatever you want. Go learn math. As such it needs not add up to 100% of another collection.

    Allan Miller: How’s design detection not post hoc?

    Post hoc to what? It’s ongoing. I explained in that OP.

  24. Nonlin.org: Doesn’t matter. Your collection is whatever you want. Go learn math. As such it needs not add up to 100% of another collection.

    Not sure what that word salad means. If you have all of something (an entire population, say) you have 100% of it. The frequencies of any different alleles at a given locus in that population cannot help but sum to that global 100%.

    If the strings make up (say) 40% of an orchestra, are you genuinely dumbstruck as to the percentage the other instruments must sum to? Is it 42?

    Post hoc to what?

    Post hoc in exactly the same sense evolution is. If coming along after the fact is a failing, ID does that. Design detection inevitably requires that the design be there to detect.

  25. Schizophora:
    New update: on top of Corona, there are thunderstorms and tornadoes here.

    Way ahead of you. We had floods in early March (I wrote my car off in one, a place I have never seen flood before). It hasn’t rained since …

    War and Famine to go, we’ll have the set. 2020 vision …

  27. Nonlin.org: Go learn math.

    Physician heal thyself. Time to put away childish things and start speaking as an adult. (Unless you are a child, of course, in which case my apologies.)

  28. Allan Miller: If you have all of something (an entire population, say) you have 100% of it. The frequencies of any different alleles at a given locus in that population cannot help but sum to that global 100%.

    I already explained very clearly. Go read again and/or think harder.

    Allan Miller: Post hoc in exactly the same sense evolution is. If coming along after the fact is a failing, ID does that. Design detection inevitably requires that the design be there to detect.

    Design is ALWAYS there. Follow the diagram: http://nonlin.org/intelligent-design/

  29. Nonlin.org:
    Design is ALWAYS there. Follow the diagram: http://nonlin.org/intelligent-design/

    I looked at your flow chart, and I could not get past the initial question in the circle – “Designed or not.” The answer to this question MUST be presumed a priori.

    Consider that archaeologists often debate as to whether some stone was or was not deliberately shaped by some ancient ancestor for some purpose. Often enough, functionally identical stones are commonly found around the world. Now, sometimes there is a context – the stone was found in close proximity to other evidences of human habitation. But without the context, without a knowledge that ancient peoples DID this sort of thing, and some solid understanding of WHY they did it, the design is ambiguous.

    This gets us to the distinction between the process of design (deliberate, done to achieve some purpose) and the result of design. Used as a noun, a “design” is anything whatsoever that has form, shape, color, texture, etc. We can work forwards from the designer to his process to his product. We can NOT work backwards from the product to the process to the designer, unless we simply presume. Which cannot be based on evidence. The evidence absolutely requires a context, which we must arbitrarily impose if we don’t already know it.

    Without that context, our presumptions are circular. We see something, we presume your god did it. It’s his handiwork. And how do we know your god even exists? Why, because we see his handiwork all around us! And how do we know it’s all his handiwork? Because we know he did it! And how do we know? Because we see “design” in everything! All your flowchart accomplishes is to verify the correctness of the original assumption

  30. Allan Miller,

    That’s quite something. To be fair to you, I live in the United States in the southeast, so thunderstorms and tornadoes often take a backseat to hurricanes here.

  31. Schizophora:
    Allan Miller,
    That’s quite something. To be fair to you, I live in the United States in the southeast, so thunderstorms and tornadoes often take a backseat to hurricanes here.

    Sure, everything’s bigger over there!

  32. Flint: I looked at your flow chart, and I could not get past the initial question in the circle – “Designed or not.” The answer to this question MUST be presumed a priori.

    You don’t know how to read flowcharts? Oval is not a question, it’s a “do statement” – “start test” in this case (first oval). Instead, rhombus is a question. So first question is “Reject ‘random’ null hypothesis?”

  33. Allan Miller: The frequencies of any different alleles at a given locus in that population cannot help but sum to that global 100%.

    My “Go read” didn’t do it and this is important. Let’s try something else: given the typical gene size average of 10–15 kbp,
    a) what makes an allele? and
    b) how many alleles does that typical gene have? and
    c) of these alleles, what percentage are actually known at any time so we can do “sums and differences to 100%”?

  35. Nonlin.org: You don’t know how to read flowcharts? Oval is not a question, it’s a “do statement” – “start test” in this case (first oval). Instead, rhombus is a question. So first question is “Reject ‘random’ null hypothesis?”

    And the most important path to take from the first question is “insufficient data”. Stuff we KNOW was designed by people is one path, stuff we KNOW was NOT designed by people is another path, stuff we observe being designed by other organisms (like spider webs, bird nests, termite mounds) is another path. These paths are all trivial.

    That leaves a very important path, the path where we simply do not know and have no reliable way to tell. Some things can be redirected off this path with research – watching to see the process that produced something. Watching provides us a context that the something, by itself, cannot tell us.

    Your actual question “smuggles in” your preferred results already. Wind designs some fascinating formations, which are far from random but certainly not conscious or deliberate. What path do we take for dunes, waves, watersheds? Do we create a “nonrandom null hypothesis”? This gets important when we look at living organisms. Absent any context, I guess we’d say these are random? But we have extremely extensive context. So do we say FOR SURE these are random? They are clearly not designed by any known organism. I suppose we could say that they are partially designed by environmental constraints, and partially designed by statistical drift. But actual real-life nuances like this don’t fit your flow diagram.

  36. Nonlin.org: My “Go read” didn’t do it and this is important. Let’s try something else: given the typical gene size average of 10–15 kbp,
    a) what makes an allele? and
    b) how many alleles does that typical gene have? and
    c) of these alleles, what percentage are actually known at any time so we can do “sums and differences to 100%”?

    Why go through that exercise. All of everything is, by definition, 100% of it. Presuming alleles exist at all, than all of them together total 100% of all alleles INCDLUDING alleles which are unknown or even unsuspected. Since there can be unknown alleles, we cannot precisely calculate the percentage of the total any particular allele represents. We can, however, SAMPLE genes and get a percentage of our sample. As the sample size grows, the probability of having a correct percentage approaches 1.

  37. Nonlin.org: My “Go read” didn’t do it and this is important. Let’s try something else: given the typical gene size average of 10–15 kbp,
    a) what makes an allele? and
    b) how many alleles does that typical gene have? and
    c) of these alleles, what percentage are actually known at any time so we can do “sums and differences to 100%”?

    If the character state is measured in base pairs, an allele can be any length desired – it does not have to map to a coding gene. And you don’t need to do any of this to know that the frequencies of all such alleles in a population – all variants of a particular segment – must add up to 100%.

    You don’t need to do anything practical to demonstrate that the entirety of a set is 100% of it.

  38. Flint: And the most important path to take from the first question is “insufficient data”.

    Nonsense. Data is plentiful to establish non-randomness in biology.

    Did you read: “A quick glance at biological systems show extreme precision repeated over and over again and indicating essentially zero probability of system-level randomness. Kidneys and all other organs are not random, reproduction is not random, cell structure is not random, behavior is not random, etc. “?

    Flint: Wind designs some fascinating formations, which are far from random but certainly not conscious or deliberate.

    Nonsense. How would you know that’s not “deliberate”?

    Flint: Why go through that exercise.

    Just to see you demonstrate your ignorance. Is that exercise too difficult for you? Of course it is.

    Allan Miller: If the character state is measured in base pairs, an allele can be any length desired – it does not have to map to a coding gene. And you don’t need to do any of this to know that the frequencies of all such alleles in a population – all variants of a particular segment – must add up to 100%.

    So you can’t answer anything? Anything at all? Then why comment about things you just don’t understand?

    Have it your way:
    – What’s “character state”?
    – give an examples of a collection of allele of “any length desired”
    – explain how you know for sure that (whatever) collection of alleles is for sure 100% of all possibilities

    Allan Miller: You don’t need to do anything practical to demonstrate that the entirety of a set is 100% of it.

    You DO need to do something practical to demonstrate you understand something. Anything.

    Will you pass the practical exam? Or will you reply with more foggy bullshit?

  40. New York Times has an article on antibody testing which gives a bit of background to the UK being “days away” from being able to test the population for resistance.

  45. Allan Miller,

    Statisticians and computer scientists dominating genetics is like accountants dominating company boards. Looking at trees can get you lost in the woods!

  46. Interesting brief reference in The Guardian‘s never-ending Covid Blog that the Dutch have gone for ‘intelligent lockdown’. I’ve been advocating similar (among my vast audience of 50 or so followers). Our own lockdown is a midpoint between the extremes, but much of it seems based on ease of management rather than science and trust. France’s earlier and more extreme measures are not more obviously successful than our own, despite a wide belief that our government failed, by being too slow and not harsh enough. The more rule-oriented members of our society are getting increasingly shrill about people doing things that frankly carry little or no transmission risk.

    The Dutch are gambling on an ‘intelligent lockdown’ to beat coronavirus, AFP reports. Shops are open and families cycle along in the sunny spring weather in the Netherlands, which has opted for what it calls an “intelligent lockdown” to curb the coronavirus pandemic.

    In contrast to most other European countries, where people are virtually housebound, the Dutch authorities have merely advised people to stay home and to keep 1.5 metres (five feet) of social distance.

    While restaurants, bars, museums and its infamous sex clubs remain shut, and the famed cannabis ‘coffee shops’ are open for takeaway only, the outdoors-loving Dutch are otherwise allowed to leave home when they want. Schools meanwhile start to reopen from 11 May.

    The Netherlands has the 14th-highest number of confirmed infections worldwide, with 38,998, and 4,727 people have lost their lives in the country so far.

    Dutch Prime Minister Mark Rutte – who came up with the phrase “intelligent lockdown” – has been clear on the policy. “We don’t work like that in the Netherlands, where the government says ‘you have to do this, you have to do that,” Rutte told a press conference at the end of March.

    The authorities have admonished the public when the country’s beaches have become too crowded, but while police have closed car parks to stop crowds flocking there at the weekends, they remain open.

    The Dutch position – very similar to Sweden’s – also reflects a wider philosophical split in both Europe and the world on how to balance the need to curb the disease against the catastrophic economic damage caused by harsh lockdowns.

  47. Allan Miller,

    Makes so much sense. Also trusting your circle of people who you could interact with socially and professionally while distancing from unknowns.

  49. We had a minor rebellion inviting a (very) few friends for aperitifs in the garden. It was wicked! 🙂

  50. Allan Miller,

    Does it really matter though? I mean, looks like some people in the Netherlands are irresponsible enough to crowd the beaches, just like everywhere else. Here they would have been fined, there they get a reprimand instead. But it seems to me that countries that have been successful at containing the initial outbreak are the ones which stocked up early on tests and protective gear, and also those with a strong health care system with a relatively large number of intensive care units. I’m not saying lockouts don’t matter, just that it may not matter all that much how harsh or lenient governments were at implementing them.

    That of course would support your idea that harsh lockouts are unnecessary, but in the grand scheme of things, when so many people are dying and economies are taking a huge hit, who cares if we can’t go out running or biking for a couple of months?

    Just to be clear, I agree that it’s much better to rely on mutual confidence between people and government and viceversa, rather than coercion. It’s just that I don’t see that affecting the outcome much in this case. But what do I know.

Leave a Reply