Covid 19 Business Funding

[Borin' Baggie]

I won't apologise for being a rational optimist in a message board of hysterical pessimists.

Whatever floats your boat fellas.

Rational? You were repeating conspiratorial criticisms against the Imperial modelling with no grounding in reality a couple of weeks ago as a means to justify your criticisms of the lockdown, that was not rational.

[Protheroe]

I won't apologise for being a rational optimist in a message board of hysterical pessimists.

Whatever floats your boat fellas.

Rational? You were repeating conspiratorial criticisms against the Imperial modelling with no grounding in reality a couple of weeks ago as a means to justify your criticisms of the lockdown, that was not rational.

The Imperial modelling was wrong. Even Neil Ferguson has admitted it.

[Borin' Baggie]

I won't apologise for being a rational optimist in a message board of hysterical pessimists.

Whatever floats your boat fellas.

Rational? You were repeating conspiratorial criticisms against the Imperial modelling with no grounding in reality a couple of weeks ago as a means to justify your criticisms of the lockdown, that was not rational.

The Imperial modelling was wrong. Even Neil Ferguson has admitted it.

Yeah, he admitted that some assumptions were being underestimated due to the available data hence the estimated death toll was below the actual. That is a) not what you were complaining about and b) the opposite of what your argument relied on.

[baggy1]

I won't apologise for being a rational optimist in a message board of hysterical pessimists.

Whatever floats your boat fellas.

Rational? You were repeating conspiratorial criticisms against the Imperial modelling with no grounding in reality a couple of weeks ago as a means to justify your criticisms of the lockdown, that was not rational.

The Imperial modelling was wrong. Even Neil Ferguson has admitted it.

Yeah, he admitted that some assumptions were being underestimated due to the available data hence the estimated death toll was below the actual. That is a) not what you were complaining about and b) the opposite of what your argument relied on.

Nice one BB - that really does sum up Proths need to argue the toss even in the face of overwhelming evidence.

[Protheroe]

Yeah, he admitted that some assumptions were being underestimated due to the available data hence the estimated death toll was below the actual. That is a) not what you were complaining about and b) the opposite of what your argument relied on.

And we don't know that he's right now.

Given that he's been wrong about every other bloody disease threatening the country over the years I, for one, don't expect him to break his duck here.

There is no "overwhelming evidence", just an overwhelming amount of often contradictory data. And I get accused of confirmation bias 

[baggy1]

No, you were pulled up on using info to support your view when in fact that info contradicted it.

[Borin' Baggie]

Yeah, he admitted that some assumptions were being underestimated due to the available data hence the estimated death toll was below the actual. That is a) not what you were complaining about and b) the opposite of what your argument relied on.

And we don't know that he's right now.

Given that he's been wrong about every other bloody disease threatening the country over the years I, for one, don't expect him to break his duck here.

There is no "overwhelming evidence", just an overwhelming amount of often contradictory data. And I get accused of confirmation bias 

This response is so disingenuous. I have explained to you how modelling works before and you respond with this turgid nonsense?


First point, because of the use of assumptions when you look at models you look at trends and not absolute values. This is a very basic data science concept. You judge your response by comparing the various relative comparisons between scenarios within the isolated model to mitigate any errors in the assumptions affecting the outcome. This is why you look at x scenario causing a y percentage rise over z scenario and not x scenario is y amount higher in absolute terms than z scenario. In this instance, you look at how many deaths would occur if you implement various strict restrictions compared with lighter restrictions and no restrictions within the model.

Second point, Ferguson has been involved in modelling three potential domestic epidemics. In all three cases, actions were made by the governments at the time ranging from culling animals to limit the spread of foot and mouth, screening and travel restrictions to limit the spread of Swine flu and the current Coronavirus situation. In all three situations actions were made which would obviously impact the outcomes of the models, no shit. That's basic common sense. This criticism is so flawed that a five year old could tear it apart. If something happens that the models didn't initially account for then of course the outcomes are going to be different, that's why you don't model based on absolute outcomes.

Third point, the model used by Imperial has been verified to provide reproducible and parallel results by external, independent teams. The Github page also shows edits have been made over time to adjust the model as more data has been available to use which is more secure than using the dodgy Chinese state published crap that initial assessments had to rely on.

Fourth point, ICL are not the only institution providing models that are being looked at. A quick look at SAGE shows that epidemiologists and statisticians from various institutes sit on it.

Fifth point, to get back to my point your criticism related to the age of the code (irrelevant as the code was only based on previous models as it's stupid to not reuse common assets and libraries like population interaction and entry points, the model itself is new and the code being used has been constantly updated) and the results were wrong (incorrect as the results have been shown to be reproducible and parallel) which have literally been used by conspiracy websites in an effort to undermine scientists and push the narrative that the virus is less infectious than it actually is.


Sixth point, what contradictory data? Sweden has implemented a light touch solution and they've done worse than comparable countries like Denmark, Finland and Norway in terms of deaths. Brazil have done fuck all and they've reported a tenth of all global confirmed cases.

You've constantly criticised the epidemiologists and statisticians for doing their job, i.e. the potential impact on health, because you think the lockdown is causing too much economic damage relative to the health risks. Why should the epidemiologists look at the economic modelling? That's the job of the economic advisors and then it's up to the government to plan their response.

[Brentbaggie]

Yeah, he admitted that some assumptions were being underestimated due to the available data hence the estimated death toll was below the actual. That is a) not what you were complaining about and b) the opposite of what your argument relied on.

And we don't know that he's right now.

Given that he's been wrong about every other bloody disease threatening the country over the years I, for one, don't expect him to break his duck here.

There is no "overwhelming evidence", just an overwhelming amount of often contradictory data. And I get accused of confirmation bias 

This response is so disingenuous. I have explained to you how modelling works before and you respond with this turgid nonsense?


First point, because of the use of assumptions when you look at models you look at trends and not absolute values. This is a very basic data science concept. You judge your response by comparing the various relative comparisons between scenarios within the isolated model to mitigate any errors in the assumptions affecting the outcome. This is why you look at x scenario causing a y percentage rise over z scenario and not x scenario is y amount higher in absolute terms than z scenario. In this instance, you look at how many deaths would occur if you implement various strict restrictions compared with lighter restrictions and no restrictions within the model.

Second point, Ferguson has been involved in modelling three potential domestic epidemics. In all three cases, actions were made by the governments at the time ranging from culling animals to limit the spread of foot and mouth, screening and travel restrictions to limit the spread of Swine flu and the current Coronavirus situation. In all three situations actions were made which would obviously impact the outcomes of the models, no shit. That's basic common sense. This criticism is so flawed that a five year old could tear it apart. If something happens that the models didn't initially account for then of course the outcomes are going to be different, that's why you don't model based on absolute outcomes.

Third point, the model used by Imperial has been verified to provide reproducible and parallel results by external, independent teams. The Github page also shows edits have been made over time to adjust the model as more data has been available to use which is more secure than using the dodgy Chinese state published crap that initial assessments had to rely on.

Fourth point, ICL are not the only institution providing models that are being looked at. A quick look at SAGE shows that epidemiologists and statisticians from various institutes sit on it.

Fifth point, to get back to my point your criticism related to the age of the code (irrelevant as the code was only based on previous models as it's stupid to not reuse common assets and libraries like population interaction and entry points, the model itself is new and the code being used has been constantly updated) and the results were wrong (incorrect as the results have been shown to be reproducible and parallel) which have literally been used by conspiracy websites in an effort to undermine scientists and push the narrative that the virus is less infectious than it actually is.


Sixth point, what contradictory data? Sweden has implemented a light touch solution and they've done worse than comparable countries like Denmark, Finland and Norway in terms of deaths. Brazil have done fuck all and they've reported a tenth of all global confirmed cases.

You've constantly criticised the epidemiologists and statisticians for doing their job, i.e. the potential impact on health, because you think the lockdown is causing too much economic damage relative to the health risks. Why should the epidemiologists look at the economic modelling? That's the job of the economic advisors and then it's up to the government to plan their response.

Ouch.