Is the micro analysis of the stats by amateurs and media actually helpful for anyone? It was in the early days when trying to get a feel for the scale of the problem. But we now know the order of magnitude that will die in this wave in the UK. It will be very sad and significant whether the final number is 17,388 or 34,203. I am far from convinced that a different policy response or different view of our government is needed depending on the total within the plausible ranges.
I think you're in danger of not being able to see the wood from the trees, in your search for the elixir of absolute statistical purity.
Certainly some governments - the French for example - have started to report deaths in establishments other than hospitals alongside their hospital deaths statistic, and they seem able to do so transparently as part of their headline reporting rather than burying it in the back of an ONS statistical annex. The Worldometer site suggests that the French move to do so at the start of April was complying with "international standards of correct inclusion".
The bottom line is that UK meanwhile is focusing on the measure which minimises the count. So the rule of thumb when making international comparisons seems to be that the UK figures are not going to compare any more favourably with other countries than might be suggested by a superficial comparison, and that a more in depth comparison might (or might not) make the UK figures look worse.
We are broadly speaking order-of-magnitude-ish stuff anyway. If you're really looking for the curvature of a trend drawn on a log-scale then so long as the % of deaths happening outside hospital is roughly constant then using hospital-only figures doesn't make any difference, because it just creates a vertical offset on the scale. The thing that matters more if you're trying to detect a turning point is to have the most up-to-date figures, ie the ones that can be collected with minimal lag. And as @Endillion points out, if you're going to do international comparisons, all the countries have different lags and different case definitions anyway. As you say yourself, superficial comparisons are ... superficial.
If your basic concern is that it's fundamentally "dishonest" to quote a lowball number, rather than what's statistically correct, then I have a degree of sympathy with you. I do think the government could do a better job of communicating the more inclusive figures, but then, it's not like they've been actively suppressed either, there has been a lot of hard work put in to producing and validating them, and there are decent reasons the lowball numbers are being used - they are the most recent, give the best indication of trend, and it's a matter of consistency with previous figures. People know the final toll will be higher, people are repeatedly told that by the government too and by now it should be widely known and understood.
Some disturbing chatter from Japanese work colleagues on Japanese thinking about how we've handled the crisis here. ... 2. Our laissez-faire approach to mass gatherings was an error and should be recognised as such. ...
(2) is correct and was a huge, stupid, catastrophic error.
I think the optics of that decision looked pretty poor at the time and far worse in retrospect. Really did give the impression of not treating COVID-19 seriously, both to the UK population and to those overseas!
Whether it was "huge", "stupid" or "catastrophic" are different questions though. The eggheads had various reasons to believe that large events, especially if held outdoors, weren't especially risky compared to restaurants or pubs. The fact there are thousands of people crowded together doesn't mean, with a virus largely transmitted by (relatively) close contact, that one person can infect thousands. It might initiate a new cluster of cases, but if the disease is already inevitably spreading around anyway, that might not be regarded as such a big deal - the effect of thousands of people mixing at a large event is far less than the effect of millions of people mixing at work or on public transport, even though there's a greater chance that one might later identify a set of transmissions that occurred at one large event rather than one particular office or train carriage.
I think by that stage the government knew that it could only conceivably delay the spread. I doubt that closing the big public events would have bought even one day of time (though may have affected things at a local level). There's a similar reason for why they didn't shut the air routes in, they had crunched the numbers and realised once the epidemic had been seeded and was spreading, a few more people with the virus landing in Britain simply made no substantial difference to its course.
If you look at the numbers large events are a small fraction of daily public transport before you factor in other offices, bars, restaurants, schools, universities etc. It is extremely unlikely that the scale of any possible mistake on that was catastrophic.
In terms of timing, the last set of games to be played in the big european countries were the UEFA games and they were only cancelled in Spain and Italy. PSG were behind closed doors but had thousands outside the ground in far closer proximity than they would have been inside the ground. The idea we were out of step with most of Europe, rather than behind Italy and to an extent Spain is just history being imagined. We were behind Italy and Spain because the virus was spreading faster there!
Even in Italy it was a mess the weekend before, with games in some areas cancelled, others behind closed doors, and one match delayed five minutes before kick off with the minister of sport phoning the referee.
Yes, but I do think this is one that could well go down in the nation's collective memory banks as a mistake, firstly because it seems so weird to have gone so fast from "we still hold concerts and sporting events" to "you can't even go clothes shopping" and so tweets or news clips of those events now look so alien and out of place, secondly because when the context gets lost it's very easy to end up as a kind of "false memory" as you point out (it now seems to be the consensus we were well behind what everyone else was doing, which simply isn't factually correct), and thirdly because the point about close contacts is not well understood. Just because thousands of people are at an event doesn't mean that one person can infect them all; if it worked like that then try getting your head around the fact 1200 people can be crammed into a tube train! The whole country would be infected in no time. Yet even intelligent, well-educated people seem to be able to look at a crowd at an event and think "wow, imagine if they all got infected".
That link says that it is unsafe to open on my ipad, by the way.
Looking at that graph, it’s way off in a number of areas, I think. France is probably ahead of the UK by a number of days, Germany is likely further down and close to France, through its suppression so far, Singapore should be bounced way back down, as they are almost starting again. Thailand and Malaysia are probably much closer to Brazil and the US is way too far up. The latter because the US is so large that different ‘waves’ are overlapping keeping them from rising up the curve. Too optimistic for most but I’d probably move Italy and Spain in the opposite direction, to just over the crown of the curve,
That link says that it is unsafe to open on my ipad, by the way.
Looking at that graph, it’s way off in a number of areas, I think. France is probably ahead of the UK by a number of days, Germany is likely further down and close to France, through its suppression so far, Singapore should be bounced way back down, as they are almost starting again. Thailand and Malaysia are probably much closer to Brazil and the US is way too far up. The latter because the US is so large that different ‘waves’ are overlapping keeping them from rising up the curve. Too optimistic for most but I’d probably move Italy and Spain in the opposite direction, to just over the crown of the curve,
A valiant attempt to simplify a complex problem into a single, ordinal, metric.
The US is simultaneously at all points on the left hand side. China could be anywhere in the rightmost 75% of the curve (although probably on the right hand side).
I think the UK and France are quite close. It's just hard to tell because the French keep muddying the waters by randomly announcing a slew of non-hospital deaths which skew all the figures. But medium term trend excluding that is still just upwards, like the UK (assuming they have similar reporting delays).
Some disturbing chatter from Japanese work colleagues on Japanese thinking about how we've handled the crisis here. ... 2. Our laissez-faire approach to mass gatherings was an error and should be recognised as such. ...
(2) is correct and was a huge, stupid, catastrophic error.
I think the optics of that decision looked pretty poor at the time and far worse in retrospect. Really did give the impression of not treating COVID-19 seriously, both to the UK population and to those overseas!
Whether it was "huge", "stupid" or "catastrophic" are different questions though. The eggheads had various reasons to believe that large events, especially if held outdoors, weren't especially risky compared to restaurants or pubs. The fact there are thousands of people crowded together doesn't mean, with a virus largely transmitted by (relatively) close contact, that one person can infect thousands. It might initiate a new cluster of cases, but if the disease is already inevitably spreading around anyway, that might not be regarded as such a big deal - the effect of thousands of people mixing at a large event is far less than the effect of millions of people mixing at work or on public transport, even though there's a greater chance that one might later identify a set of transmissions that occurred at one large event rather than one particular office or train carriage.
I think by that stage the government knew that it could only conceivably delay the spread. I doubt that closing the big public events would have bought even one day of time (though may have affected things at a local level). There's a similar reason for why they didn't shut the air routes in, they had crunched the numbers and realised once the epidemic had been seeded and was spreading, a few more people with the virus landing in Britain simply made no substantial difference to its course.
If you look at the numbers large events are a small fraction of daily public transport before you factor in other offices, bars, restaurants, schools, universities etc. It is extremely unlikely that the scale of any possible mistake on that was catastrophic.
In terms of timing, the last set of games to be played in the big european countries were the UEFA games and they were only cancelled in Spain and Italy. PSG were behind closed doors but had thousands outside the ground in far closer proximity than they would have been inside the ground. The idea we were out of step with most of Europe, rather than behind Italy and to an extent Spain is just history being imagined. We were behind Italy and Spain because the virus was spreading faster there!
Even in Italy it was a mess the weekend before, with games in some areas cancelled, others behind closed doors, and one match delayed five minutes before kick off with the minister of sport phoning the referee.
Yes, but I do think this is one that could well go down in the nation's collective memory banks as a mistake, firstly because it seems so weird to have gone so fast from "we still hold concerts and sporting events" to "you can't even go clothes shopping" and so tweets or news clips of those events now look so alien and out of place, secondly because when the context gets lost it's very easy to end up as a kind of "false memory" as you point out (it now seems to be the consensus we were well behind what everyone else was doing, which simply isn't factually correct), and thirdly because the point about close contacts is not well understood. Just because thousands of people are at an event doesn't mean that one person can infect them all; if it worked like that then try getting your head around the fact 1200 people can be crammed into a tube train! The whole country would be infected in no time. Yet even intelligent, well-educated people seem to be able to look at a crowd at an event and think "wow, imagine if they all got infected".
I think the point that's missed is that you're not yourself immediately infectious within seconds or minutes of becoming infected. If you were, then yes one infectious person at Cheltenham could easily infect almost all the rest, via (say) six degrees of contact throughout the day.
The other key point is that not every interaction results in transmission - I think 5% has been mentioned. But it's easy to imagine, given the coverage, that it's damn near 100%.
Some disturbing chatter from Japanese work colleagues on Japanese thinking about how we've handled the crisis here. ... 2. Our laissez-faire approach to mass gatherings was an error and should be recognised as such. ...
(2) is correct and was a huge, stupid, catastrophic error.
I think the optics of that decision looked pretty poor at the time and far worse in retrospect. Really did give the impression of not treating COVID-19 seriously, both to the UK population and to those overseas!
Whether it was "huge", "stupid" or "catastrophic" are different questions though. The eggheads had various reasons to believe that large events, especially if held outdoors, weren't especially risky compared to restaurants or pubs. The fact there are thousands of people crowded together doesn't mean, with a virus largely transmitted by (relatively) close contact, that one person can infect thousands. It might initiate a new cluster of cases, but if the disease is already inevitably spreading around anyway, that might not be regarded as such a big deal - the effect of thousands of people mixing at a large event is far less than the effect of millions of people mixing at work or on public transport, even though there's a greater chance that one might later identify a set of transmissions that occurred at one large event rather than one particular office or train carriage.
I think by that stage the government knew that it could only conceivably delay the spread. I doubt that closing the big public events would have bought even one day of time (though may have affected things at a local level). There's a similar reason for why they didn't shut the air routes in, they had crunched the numbers and realised once the epidemic had been seeded and was spreading, a few more people with the virus landing in Britain simply made no substantial difference to its course.
If you look at the numbers large events are a small fraction of daily public transport before you factor in other offices, bars, restaurants, schools, universities etc. It is extremely unlikely that the scale of any possible mistake on that was catastrophic.
In terms of timing, the last set of games to be played in the big european countries were the UEFA games and they were only cancelled in Spain and Italy. PSG were behind closed doors but had thousands outside the ground in far closer proximity than they would have been inside the ground. The idea we were out of step with most of Europe, rather than behind Italy and to an extent Spain is just history being imagined. We were behind Italy and Spain because the virus was spreading faster there!
Even in Italy it was a mess the weekend before, with games in some areas cancelled, others behind closed doors, and one match delayed five minutes before kick off with the minister of sport phoning the referee.
Yes, but I do think this is one that could well go down in the nation's collective memory banks as a mistake, firstly because it seems so weird to have gone so fast from "we still hold concerts and sporting events" to "you can't even go clothes shopping" and so tweets or news clips of those events now look so alien and out of place, secondly because when the context gets lost it's very easy to end up as a kind of "false memory" as you point out (it now seems to be the consensus we were well behind what everyone else was doing, which simply isn't factually correct), and thirdly because the point about close contacts is not well understood. Just because thousands of people are at an event doesn't mean that one person can infect them all; if it worked like that then try getting your head around the fact 1200 people can be crammed into a tube train! The whole country would be infected in no time. Yet even intelligent, well-educated people seem to be able to look at a crowd at an event and think "wow, imagine if they all got infected".
I think the point that's missed is that you're not yourself immediately infectious within seconds or minutes of becoming infected. If you were, then yes one infectious person at Cheltenham could easily infect almost all the rest, via (say) six degrees of contact throughout the day.
The other key point is that not every interaction results in transmission - I think 5% has been mentioned. But it's easy to imagine, given the coverage, that it's damn near 100%.
With this virus, it appears that the best way of spreading it is to get a lot of people gathering together who then open their mouths (and noses) to spread it, and who disperse back to their own area, thus maximising its transmittance. All pretty much to be expected but the kicker being that it bides its time before revealing itself. Does being outdoors change that? Has there been any study on this?
Until we can either test everyone and then track and trace everyone (and it really is useless if voluntary), then the best we can do is to stop gatherings of more than a few people at a time and keep people within a small circle of contacts (be it family or friends or whatever).
Comments
If your basic concern is that it's fundamentally "dishonest" to quote a lowball number, rather than what's statistically correct, then I have a degree of sympathy with you. I do think the government could do a better job of communicating the more inclusive figures, but then, it's not like they've been actively suppressed either, there has been a lot of hard work put in to producing and validating them, and there are decent reasons the lowball numbers are being used - they are the most recent, give the best indication of trend, and it's a matter of consistency with previous figures. People know the final toll will be higher, people are repeatedly told that by the government too and by now it should be widely known and understood.
Looking at that graph, it’s way off in a number of areas, I think. France is probably ahead of the UK by a number of days, Germany is likely further down and close to France, through its suppression so far, Singapore should be bounced way back down, as they are almost starting again. Thailand and Malaysia are probably much closer to Brazil and the US is way too far up. The latter because the US is so large that different ‘waves’ are overlapping keeping them from rising up the curve. Too optimistic for most but I’d probably move Italy and Spain in the opposite direction, to just over the crown of the curve,
The US is simultaneously at all points on the left hand side. China could be anywhere in the rightmost 75% of the curve (although probably on the right hand side).
I think the UK and France are quite close. It's just hard to tell because the French keep muddying the waters by randomly announcing a slew of non-hospital deaths which skew all the figures. But medium term trend excluding that is still just upwards, like the UK (assuming they have similar reporting delays).
The other key point is that not every interaction results in transmission - I think 5% has been mentioned. But it's easy to imagine, given the coverage, that it's damn near 100%.
Until we can either test everyone and then track and trace everyone (and it really is useless if voluntary), then the best we can do is to stop gatherings of more than a few people at a time and keep people within a small circle of contacts (be it family or friends or whatever).