Crosswords0 min ago
Why Is Covid-19 Less Survivable In The Uk?
According to a study by the Institute for Health Metrics and Evaluation (IHME) in Seattle released yesterday, the UK could be the European country worst hit by the coronavirus pandemic, suffering a death toll over 66,000 by August.
The IHME, which produces the Global Burden of Disease study, believes that deaths in the UK will peak with an estimated 2,932 deaths on 17 April, and predicts 66,314 total deaths in the country by 4 August.
In the EU, Italy, Spain and France are forecast to suffer the highest death tolls, at 20,200, 19,209 and 15,058 respectively by 4 August.
https:/
The IHME, which produces the Global Burden of Disease study, believes that deaths in the UK will peak with an estimated 2,932 deaths on 17 April, and predicts 66,314 total deaths in the country by 4 August.
In the EU, Italy, Spain and France are forecast to suffer the highest death tolls, at 20,200, 19,209 and 15,058 respectively by 4 August.
https:/
Answers
Best Answer
No best answer has yet been selected by Gromit. Once a best answer has been selected, it will be shown here.
For more on marking an answer as the "Best Answer", please visit our FAQ.
//Some models are more accurate than others but I doubt a PM would get away with high taxes because he had "an inkling" his plans for the next five years required them.//
I didn't say they were inappropriate for everything. I simply believe they are not appropriate for this. And any forecast which has a range of -60% up to +212% is frankly not worth taking the trouble to produce in any detail. Simply look at the trend and extrapolate it as far as you want to (which, in the case of the virus, is not worth doing much beyond a few days because it's going to be wrong and the trend will change when new actuals are available). My forecast for today will be just as "correct" (and about as useless) as anything else produced. The government is trying to fool people into believing they know what's going to happen. And they don't.
I didn't say they were inappropriate for everything. I simply believe they are not appropriate for this. And any forecast which has a range of -60% up to +212% is frankly not worth taking the trouble to produce in any detail. Simply look at the trend and extrapolate it as far as you want to (which, in the case of the virus, is not worth doing much beyond a few days because it's going to be wrong and the trend will change when new actuals are available). My forecast for today will be just as "correct" (and about as useless) as anything else produced. The government is trying to fool people into believing they know what's going to happen. And they don't.
Then tell me how you can take an objective view (and objective action) when a forecaster says "My forecast is 100. But it could turn out to be 40 or it might turn out at 212"? Absolutely useless. The forecaster's salary would be better spent treating one of the 100 (or 40 or 212) who happen to fall sick.
I'd agree that models are not necessary for predicting short term things like the deaths over the next few days and the forecasts need not have a big error margin but become useful for forecasting things like what would happen to hospital admissions or deaths over the next month under different assumptions about lockdown measures (and the resultant transmission rates)
//Easy. Error analysis is part of the objective approach.//
If you don't mind me saying so, Jim, a glib answer to what I thought was a sensible question. Let me put it another way: how do you react to a forecast that has a range of error of -60% to +212%? It is not like +/- 10%. It is no better than useless. It is no better than my forecast for today (which actually proved reasonably accurate but was plucked from thin air). The work being done to produce such figures is utterly, utterly pointless.
Meanwhile, despite the lockdown, our rate of new cases continues to rise at a considerably higher rate than that of Sweden (where, despite what I've heard, there are still no signs of them introducing stricter measures). Since 1st April their increase in new cases has been 140%, ours has been 234%. They, of course, will have nothing like the problems we shall have in developing an exit strategy. Their measures have seen most of their economy still functioning; ours has been virtually brought to a halt. All we have been told is "it would have been much worse". I'd like to know how that claim can be substantiated.
If you don't mind me saying so, Jim, a glib answer to what I thought was a sensible question. Let me put it another way: how do you react to a forecast that has a range of error of -60% to +212%? It is not like +/- 10%. It is no better than useless. It is no better than my forecast for today (which actually proved reasonably accurate but was plucked from thin air). The work being done to produce such figures is utterly, utterly pointless.
Meanwhile, despite the lockdown, our rate of new cases continues to rise at a considerably higher rate than that of Sweden (where, despite what I've heard, there are still no signs of them introducing stricter measures). Since 1st April their increase in new cases has been 140%, ours has been 234%. They, of course, will have nothing like the problems we shall have in developing an exit strategy. Their measures have seen most of their economy still functioning; ours has been virtually brought to a halt. All we have been told is "it would have been much worse". I'd like to know how that claim can be substantiated.
I've pointed out to you the various countries whose responses you can use to reach that conclusion, and I've pointed out to you why your continued references to Sweden are inappropriate and taken out of context. Never mind the fact that the "weekend effect" in reporting is even more stark in Sweden than many places. Compare Sweden to Norway, not to to us, and you'll see the failures in their approach; which, I've also explained have been acknowledged as mistaken by their Prime Minister, by their scientists, and by the international community. By contrast, the only criticism of the lockdown here is that it came too late for many.
At this point it's bloody-mindedness in the face of overwhelming empirical evidence, coupled with a clear refusal to recognise what common sense should tell you: that if we aren't in contact with people as often or in as large numbers then a disease doesn't spread so fast.
At this point it's bloody-mindedness in the face of overwhelming empirical evidence, coupled with a clear refusal to recognise what common sense should tell you: that if we aren't in contact with people as often or in as large numbers then a disease doesn't spread so fast.
With regard to the error discussion, it's a fair point, but still there's something defeatist about it. Even a vague indication of the future has its merits if the limitations are understood. Sometimes it's the best we can do. Indeed, if anything, you'd be surprised the other way round. How frankly absurd it would be if every estimate for the future turned out to be spot on. But, as ff suggested, more important is more the longer-term trend, rather than fluctuations in day-to-day figures which may be sensitive to other factors outside your control. Today's announcement of 761 new deaths included at least 11 from March that had been overlooked.
There is a balance to be struck between acknowledging the limitations of models and abandoning them altogether. By definition and by necessity, we clearly have to try our best in understanding the potential consequences of different approaches if we're obliged to choose between them. Just winging it won't do.
There is a balance to be struck between acknowledging the limitations of models and abandoning them altogether. By definition and by necessity, we clearly have to try our best in understanding the potential consequences of different approaches if we're obliged to choose between them. Just winging it won't do.
// Easy. Error analysis is part of the objective approach.//
Jim there is no AB group idea on variance or confidence intervals - dont even go there.
I was gonna say - extrapolate up and it always goes up
and extrapolate a down curve and it always goes down
so - - - extrapolation is useless for detecting a turning point ( or saddle ) or top of the hill
and I thought Even That was an Idea too far for the average innumerate ABer
Jim there is no AB group idea on variance or confidence intervals - dont even go there.
I was gonna say - extrapolate up and it always goes up
and extrapolate a down curve and it always goes down
so - - - extrapolation is useless for detecting a turning point ( or saddle ) or top of the hill
and I thought Even That was an Idea too far for the average innumerate ABer
>Since 1st April their [Sweden's, no lockdown] increase in new cases has been 140%, ours [with lockdown] has been 234%.
Putting aside the issue of where they are in terms of time, what we don't know is what it would have been in Sweden if they had had lockdown. It's been an interesting discussion between you, New Judge, and jim, and I can see a case on both sides. But I'm puzzled by the comparison with Sweden. I'd be interested to know whether you had a view on what Sweden's figures would have been if they'd had a lockdown like ours- would they have been higher or lower? If higher, then could that be explained as it seems counterintuitive since the chance for transmission and infection is clearly reduced isn't it? If lower, how much lower- marginally lower or significantly lower.
And how do we know the differences weren't due to different factors rather than the presence of a lockdown?
Putting aside the issue of where they are in terms of time, what we don't know is what it would have been in Sweden if they had had lockdown. It's been an interesting discussion between you, New Judge, and jim, and I can see a case on both sides. But I'm puzzled by the comparison with Sweden. I'd be interested to know whether you had a view on what Sweden's figures would have been if they'd had a lockdown like ours- would they have been higher or lower? If higher, then could that be explained as it seems counterintuitive since the chance for transmission and infection is clearly reduced isn't it? If lower, how much lower- marginally lower or significantly lower.
And how do we know the differences weren't due to different factors rather than the presence of a lockdown?
Latest Posts
ChatterBank2 mins ago
Jobs & Education4 mins ago
Jokes4 mins ago
ChatterBank6 mins ago
News7 mins ago
Related Questions
Sorry, we can't find any related questions. Try using the search bar at the top of the page to search for some keywords, or choose a topic and submit your own question.