Covid 19 Part XXXV-956,720 ROI (5,952 deaths) 452,946 NI (3,002 deaths) (08/01) Read OP

Good to see the decline in hospital numbers continue. Hammers home the worthlessness of Nolan's models even more.

What previous sources? RTÉ news? independent.ie, Irish times? I know they’re not independent at all now & blatantly pushing propaganda but they’re the sources you’d look to.

I just asked a question, you didn’t know the answer.

I know there are certain posters on this thread doing their best to keep the labels flying out to anyone they don’t agree with, or to anyone who questions the logic for the current policies. Off with you.

Much as I've had my own feelings about the modelling they do serve a purpose as a possible indicator of the spread of disease. As has been stated more than once at NPHET briefings it is not an exact science. The real trouble though is governments are used to models and predictions that can inform and direct policies in the real world. Epidemiological models just don't do this, yet they are taken as gospel and are used to make wide ranging decisions that can look quite ludicrous and excessive.  

I'm not sure if there is an answer to this. The modellers, in good faith, are doing all their scenario analysis but while they only look at disease spread and not things that mitigate it they just can't tie it to any reality.

Sure except that the government doesn't only use one single model. It has access to a wide variety of data and sources.

The point is they can have any number of competing opinions to evaluate and consider. Here there is only one. The measures recommended out of this advice are an entirely different conversation but the models are used to justify them.

Bloody scientists with their maths and graphs. Back in my day a stick and a good patch of sand was all we needed to figure stuff out.

The UK has left recovery out of the new Covid pass rules so if Ireland may follow suit -

But the big difference is that in the UK a negative antigen is acceptable anywhere you need a Covid pass, so in reality it doesn't stop anyone doing anything (self tested antigen)

The science and the theory behind the modelling is perfectly sound. In fact the modelling systems are completely open-source and available for anyone to look at and find fault with.

They are supported as being solid by practically every expert in the field. And when the last models were roundly mocked, many of the experts took a bit of offence and asked that people examine it for themselves.

The problem is not the modelling itself; the maths and the systems are fine.

The problem is a very basic premise in data science and analytics - when you put bullshit into your models, you get bullsh1t back out. The model itself is fine, it does exactly what it's supposed to. But models are only ever as good as the data you start with. Models can't fix bad input.

So when the model doesn't match the reality, you usually go, "OK, we made some mistakes about our input, some assumptions were wrong" and you learn.

Except in this case, it's been plainly obvious a number of times that the input into these models was wrong. We didn't even have to wait to see if the models were accurate, it was clear that the input was wrong, so therefore models were too.

As Lumen says, when you're using these models for economic forecasting, or some architectural woo, then it's OK to have made these mistakes and then come back and "learn". When the models are being used to apply restrictions and to try and get a grip on a live pandemic, then it's not good enough. Every input into the model needs to be validated, every assumption justified and peer-reviewed. If you pluck numbers out of the air or use data that you know to be out of date, then the models are just fancy lines on a screen and nothing more.

Thanks for the information. Why is recovery left off the Cert in the U.K. I wonder? At least they have antigen for now. If Ireland take away recovery as an option and don’t even allow for testing, we’re lost as a country.

If all of this is in the name of health, why would you need a vaccine shortly after your immune system has successfully fought off Covid and would now have antibodies and T cell immunity for at least 6-9 months. And possibly much better immunity and antibodies to the Omicron variant which the vaccines can only offer limited protection against.

With hospital numbers dropping, a push for the reopening of night clubs must be on the agenda.

Well considering you can't get a vaccine/booster within 6 months of having Covid I would imagine the recovery cert will be time sensitive to allow for that going forward

They’ve just shortened it recently….

@I see sheep wrote:

I thought that couldn't possibly be true, it's too stupid, but then I looked it up. Yes, you can self-report your own negative antigen test to get a covid pass in the UK. That is the most comical, inept thing I've heard all week. Just when you think the Brits can't make any more of a balls of this, they go ahead and exceed my expectations again.

==========

570 in hospital this morning. The decline in hospital admissions continues. The booster programme making a huge difference.

Today's number is the lowest since 6th November, but if it drops below 444 this week, then statistically it'll be the lowest since mid-October.

@Lumen wrote:

Oh yeah completely. But models can be presented as "the most likely outcome from our current position". Not to mention the war, but we do it all the time with the weather, for example. Which is a system equally as chaotic and unpredictable as human behaviour, but granted it's an older science. And we issue weather warnings and warnings to take action, based on the outcome of those weather models. And often they don't land exactly right, but nowadays they're rarely way off.

So in that regard I have no issue with modelling looking at the course of the pandemic and presenting a "here's what will probably happen" and using that to consider a very macro response. But like weather models, you run the modelling continuously and react when it's time to react. When a weather model shows you a world-ending snow event 2 weeks out, you say "probably not", but keep an eye on it anyway. If the models, day after day, keep telling you that this snowstorm is coming, then you give it more and more credence, and you react to it at an appropriate time. This is what the NPHET modelling team should be doing. Tracking the course of this thing using up-to-date input and identifying things coming down the line which the models are getting more and more confident about.

At the moment what we do is the equivalent of using a weather model that tells us there's a big snow storm coming in a month, so it's time for everyone to grit the roads, bang up the heating, put on heavy clothes and sit and wait for it. And stick doggedly to that even as the temperature starts going up.

The fact is nphet have recommended restrictions based on models they know are not accurate.

They are presenting models that don't include the effect of vaccines or boosters because the models wouldn't be as scary.

The government are then implementing restrictions on models that are predicting scenarios that don't include vaccines.

Basically nphet are deliberately lying to the government to advise them to introduce restrictions.

This is in spite of the vaccine uptake and how many have adhered to restrictions and people wonder why people are fed up and not listening anymore.

I can't understand how anybody can have faith in nphet an organisation that is restricting their life's and jobs based on lies.