Covid-19 Thread

[DieselDaisy]

1 hour ago, SoulMonster said:

I have to cancel a planned trip because my conference suddenly became completely "virtual". Now this is getting serious!!!

The environment just breathed a sigh of relief. 

[SoulMonster]

31 minutes ago, action said:

So @SoulMonster is wrong to claim that the infections in china are slowing down.

What I have been saying is that the epidemic has been slowing down in China. And it has. That's a fact. Then in my latest post on this issue I stated that it could have be increasing again, but that it is too early to say. We need to wait a few days to be able to be certain.

This has all to do with statistical uncertainty and I don't think I should spend a lot of time trying to teach you statistics. You will just have to take my word for it.

But let's look at the numbers of new cases in China over the last 10 days:

New cases reported in China:

Feb 25: 518

Feb 26: 412

Feb 27: 439

Feb 28: 331

Feb 29: 435

Mar 01: 579

Mar 02: 206

Mar 03: 130

Mar 04: 120

Mar 05: 143

Can you see that the trend is that the number of new cases is going down? It can be a bit difficult to see because it isn't a steady decrease, it goes a bit up and down, but generally speaking the trend is downward. Do you see this?

I have plotted it as a graph for you:

Perhaps it is easier to see now? This is a simple graph but let me know if I need to explain it.

Here is the graph again, but now I have added a trend line that shows the trend much better:

The trend line is the dotted line crossing the bars. Can you see it is sloping downwards? That means that the trend is that less people are getting infected now in china than at the beginning of these selected dates. There is a small increase now at the end of the graph, but if you see around the dates Feb 29- Mar 1, we had a significant increase there, too. But that's how real case data often behave, they don't confirm perfectly to trends. In fact, a trend is what you get when you cancel out all such local deviations. 

 

Can you think of any reasons why the data isn't completely linear (straight line?), and here a more challenging question for you, can you understand why, as the numbers gets smaller, that coincidences will tend to have a larger effect on the data?

 

 

Edited March 6, 2020 by SoulMonster

[SoulMonster]

9 minutes ago, DieselDaisy said:

The environment just breathed a sigh of relief. 

I will probably go there anyway. I mean the flight ticket and hotel is non-refundable and Paris is nice in late March. Haven't been there in many months now. 

[DieselDaisy]

4 minutes ago, SoulMonster said:

I will probably go there anyway. I mean the flight ticket and hotel is non-refundable and Paris is nice in late March. Haven't been there in many months now. 

Oh well. The environment hoped at least.

[SoulMonster]

33 minutes ago, action said:

Unlike a common flu epidemy, where you see a spike and then a decline, here we had a spike, a decline and a rise again.

But this is completely wrong. Read up on propagated epidemic curves. No wait, please don't. It will just result in a barrage of misunderstandings and new questions 

But it should be obvious that we have multiple outbreaks of flu. That's why we refer to it as "seasonal flu", because every year we get a new pandemic.

[SoulMonster]

Well, I can't help myself, I just like Excel. So I added ten more dates to make the declining trend of new covid-19 cases in China even more obvious to @action:

 

Again, it goes a bit up and down, but the overall trend is clear. This mean that the little increase seen at the end, for March 5, can't be interpreted as the infection increasing again, but can just be attributed to radom fluctuations in the number of new cases reported every given day. We simply cannot conclude, like you do, action, that it is on the rise again.

Edited March 6, 2020 by SoulMonster

[action]

20 minutes ago, SoulMonster said:

What I have been saying is that the epidemic has been slowing down in China. And it has. That's a fact. Then in my latest post on this issue I stated that it could have be increasing again, but that it is too early to say. We need to wait a few days to be able to be certain.

This has all to do with statistical uncertainty and I don't think I should spend a lot of time trying to teach you statistics. You will just have to take my word for it.

But let's look at the numbers of new cases in China over the last 10 days:

New cases reported in China:

Feb 25: 518

Feb 26: 412

Feb 27: 439

Feb 28: 331

Feb 29: 435

Mar 01: 579

Mar 02: 206

Mar 03: 130

Mar 04: 120

Mar 05: 143

Can you see that the trend is that the number of new cases is going down? It can be a bit difficult to see because it isn't a steady decrease, it goes a bit up and down, but generally speaking the trend is downward. Do you see this?

I have plotted it as a graph for you:

Perhaps it is easier to see now? This is a simple graph but let me know if I need to explain it.

Here is the graph again, but now I have added a trend line that shows the trend much better:

The trend line is the dotted line crossing the bars. Can you see it is sloping downwards? That means that the trend is that less people are getting infected now in china than at the beginning of these selected dates. There is a small increase now at the end of the graph, but if you see around the dates Feb 29- Mar 1, we had a significant increase there, too. But that's how real case data often behave, they don't confirm perfectly to trends. In fact, a trend is what you get when you cancel out all such local deviations. 

 

Can you think of any reasons why the data isn't completely linear (straight line?), and here a more challenging question for you, can you understand why, as the numbers gets smaller, that coincidences will tend to have a larger effect on the data?

 

 

you say, it's too early to say if numbers are rising again. But I believe, this only is true if you want to prove "trends". Yes, to prove "trends", the dotted line so to speak, it is too early. But it is not too early to see the absolute numbers. It is a perfectly valid statement to say the numbers are rising on thursdag, compared to wednesday.

that is what is reported: numbers are rising again, and this is a fact. As for this being a trend? indeed: too early to say.

[action]

4 minutes ago, SoulMonster said:

Well, I can't help myself, I just like Excel. So I added ten more dates to make the declining trend of new covid-19 cases in China even more obvious to @action:

 

Again, it goes a bit up and down, but the overall trend is clear. This mean that the little increase seen at the end, for March 5, can't be interpreted as the infection increasing again, but can just be attributed to radom fluctuations in the number of new cases reported every given day. We simply cannot conclude, like you do, action, that it is on the rise again.

it's too early to conclude about trends.

[SoulMonster]

2 minutes ago, action said:

you say, it's too early to say if numbers are rising again. But I believe, this only is true if you want to prove "trends". Yes, to prove "trends", the dotted line so to speak, it is too early. But it is not too early to see the absolute numbers. It is a perfectly valid statement to say the numbers are rising on thursdag, compared to wednesday.

that is what is reported: numbers are rising again, and this is a fact. As for this being a trend? indeed: too early to say.

But you said I "wrong to claim that the infections in china are slowing down" . And I am obviously not. I am right in saying the epidemic in China is slowing down. It is evident for anyone with eyes in the heads. The infection seems to be petering out. And the media is reporting on this, too: https://apnews.com/494f9ed0137e976c0b38cec58bc1af53 

What you did, on the other hand, was to insinuate that it wasn't slowing down, but was in fact developing into a new secondary outbreak, because this is what you wrote: "Unlike a common flu epidemy, where you see a spike and then a decline, here we had a spike, a decline and a rise again." And there is simply nothing in the data that suggests you are correct in this.

What we are observing is the expected ups and downs in an overall declining trend. Exactly what we would expect from statistical theory and empirical data from other epidemics. 

What you are doing, is your normal hysterical attempts at making this is more than it is.

[SoulMonster]

15 minutes ago, action said:

it's too early to conclude about trends.

No, it simply isn't  There is more than enough data here to model trends with confidence intervals, and the trend is clear: The number of infected people in China is declining.

You don't accept this because you know very little about statistics and apparently even less about epidemics. Moreover, it is hard to convince you with facts and theory, because you have no respect for either, preferring to believe your own "eyes and ears" and what you amusingly refer to as your "common sense." And lastly, you are now emotionally invested in your own unfounded beliefs and just won't let them go despite the overwhelming contrary evidence.

What is too early, on the other hand, is to say that the recent small increase in new cases mean that we have a secondary outbreak happening. There is simply not enough data for making that conclusion. Based on the data, it is much more likely just another random, local spike in the trend. But as I have said before, we will only know for certain in some days. 

Edited March 6, 2020 by SoulMonster

[action]

9 minutes ago, SoulMonster said:

But you said I "wrong to claim that the infections in china are slowing down" . And I am obviously not. I am right in saying the epidemic in China is slowing down. It is evident for anyone with eyes in the heads. The infection seems to be petering out. And the media is reporting on this, too: https://apnews.com/494f9ed0137e976c0b38cec58bc1af53 

What you did, on the other hand, was to insinuate that it wasn't slowing down, but was in fact developing into a new secondary outbreak, because this is what you wrote: "Unlike a common flu epidemy, where you see a spike and then a decline, here we had a spike, a decline and a rise again." And there is simply nothing in the data that suggests you are correct in this.

What we are observing is the expected ups and downs in an overall declining trend. Exactly what we would expect from statistical theory and empirical data from other epidemics. 

What you are doing, is your normal hysterical attempts at making this is more than it is.

the absolute numbers are what they are. a notable rise on thursday. everyone can make their own conclusions about that.

the rest, is just a tiring discussion about semantics. Which, I can say with confidence, you are notorious for sometimes.

[SoulMonster]

1 minute ago, action said:

the absolute numbers are what they are. a notable rise on thursday. everyone can make their own conclusions about that.

Yes, of course everybody can make their own conclusions. 

I am concluding that this is just another temporary up in a dataset that contains many such temporary ups. It confirms with theory on epidemic curves, it confirms with statistical theory when the numbers are low, and it confirms with the dataset from earlier where there has been other such ups. It is simply unlikely that this would be the initial stages of another outbreak. The data don't support this, our understanding of epidemics don't support this, and there is no support from this from the statistics. I cannot be 100 % sure it isn't another outbreak, but I am pretty close to it.

But nothing of this has any effect on your conclusion, of course, because you are not limited by rationality, you easily reject statistical theory, reject common epidemical trends, reject logic. It's almost like a superpower, just really shitty.

[SoulMonster]

7 minutes ago, action said:

the rest, is just a tiring discussion about semantics. Which, I can say with confidence, you are notorious for sometimes.

Ah, you take a left out of the Diesel's book by conflating an argument based on statistical theory and knowledge on epidemic curves, while pointing out that you have a problem with facts and logic, to it being about the meaning of words. Nice!

[action]

2 minutes ago, SoulMonster said:

Yes, of course everybody can make their own conclusions. 

I am concluding that this is just another temporary up in a dataset that contains many such temporary ups. It confirms with theory on epidemic curves, it confirms with statistical theory when the numbers are low, and it confirms with the dataset from earlier where there has been other such ups. It is simply unlikely that this would be the initial stages of another outbreak. The data don't support this, our understanding of epidemics don't support this, and there is no support from this from the statistics. I cannot be 100 % sure it isn't another outbreak, but I am pretty close to it.

But nothing of this has any effect on your conclusion, of course, because you are not limited by rationality, you easily reject statistical theory, reject common epidemical trends, reject logic. It's almost like a superpower, just really shitty.

by the way, have you considered the fact, that the numbers from say Italy have far more inherent value than the numbers provided by the chinese government? You know, the government that jails journalists for telling the truth, for jailing doctors who want to warn people, the government which even lazy Axl found the time to release an album about?

statistics are a tool, that stands or falls with the validity of the numbers. A large part of the numbers are provided by frickin China, for fucks sake. You don't account for this, at all!

[SoulMonster]

3 minutes ago, action said:

by the way, have you considered the fact, that the numbers from say Italy have far more inherent value than the numbers provided by the chinese government? You know, the government that jails journalists for telling the truth, for jailing doctors who want to warn people, the government which even lazy Axl found the time to release an album about?

statistics are a tool, that stands or falls with the validity of the numbers. A large part of the numbers are provided by frickin China, for fucks sake. You don't account for this, at all!

I am just responding to your posts and it was you who recently tried to argue that the infection was on the rise in China again.

[action]

15 minutes ago, SoulMonster said:

I am just responding to your posts and it was you who recently tried to argue that the infection was on the rise in China again.

I just want to show, that when china does report an increase in infections, they certainly won't lie about that because that would be against their proven track record. Why would they lie about an increase, when in the past they have done everything in their powers to minimalise the epidemy? 

[SoulMonster]

2 minutes ago, action said:

I just want to show, that when china does report an increase in infections, they certainly won't lie about that because that would be against their proven track record. Why would they lie about an increase, when in the past they have done everything in their powers to minimalise the epidemy? 

Not sure what you are getting at, but yes, I don't see any reason why they would lie about an increase. That's almost as ridiculous as lying about the fatality rate for children, which is more in your territory of thinking.

[action]

Just now, SoulMonster said:

Not sure what you are getting at, but yes, I don't see any reason why they would lie about an increase. That's almost as ridiculous as lying about the fatality rate for children, which is more in your territory of thinking.

To summarise my view on statistics, i'll say the following.

you can not look past the fact that china has done everything in their powers to minimalise the epidemy. This has included jailing people who wanted to tell the truth and false communication, at least in the beginning phase of the epidemy. It is therefore hard, to determine when they stopped this policy, and decided to communicate 100 % accurate information, and this includes numbers of infections and deaths, as a first and most important parameter.

You might call me paranoid and hysterical, but I'm sure everyone is intelligent enough to look at china's track record, and take everything they say with a pinch of salt.

To find the truth, to appreciate these graphs for their real value, you have to create a "virtual" graph, that takes into account the questionable integrity of the chinese government. Not an easy task to do. I don't pretend to make any scientifically valid statement when making the exercise. I'm in nature very cautious, I wouldn't say hysterical but certainly worried and confused, so I'd come to different results than someone like you who seems more relaxed in all of this. All of this, shouldnt even be necessary had the chinese government be integer from the first day, but that is really not the case.

[Dazey]

1 hour ago, SoulMonster said:

What we are observing is the expected ups and downs in an overall trend.

*cough* Climate Change! *cough*  

[SoulMonster]

7 minutes ago, action said:

To summarise my view on statistics, i'll say the following.

you can not look past the fact that china has done everything in their powers to minimalise the epidemy. This has included jailing people who wanted to tell the truth and false communication, at least in the beginning phase of the epidemy. It is therefore hard, to determine when they stopped this policy, and decided to communicate 100 % accurate information, and this includes numbers of infections and deaths, as a first and most important parameter.

You might call me paranoid and hysterical, but I'm sure everyone is intelligent enough to look at china's track record, and take everything they say with a pinch of salt.

I don't blame anyone for taking official statements from China with a pinch of salt. But you can't have it both ways, you can't trust their numbers when they align with your belief that the epidemic in on the rise again, and disregard their numbers when they show a downward trend. 

As for trusting the numbers. I am sure WHO have people on the ground helping to ascertain the numbers for their situation reports. I know social media is full of "reports" from the area claiming people are dropping like flies. But that's social media. I trust WHO over random dudes on TikTok.

And besides, what we are seeing, looking at the epidemic curve, is really not that surprising. It is more plausible that the epidemic is dwindling down than that it is picking up for a second outbreak.

 

10 minutes ago, action said:

To find the truth, to appreciate these graphs for their real value, you have to create a "virtual" graph, that takes into account the questionable integrity of the chinese government. Not an easy task to do.

[Dazey]

6 minutes ago, SoulMonster said:

I trust WHO over random dudes on TikTok.

Plus it would be a win for humanity as a whole if everybody on TikTok caught Coronavirus. 

[action]

6 minutes ago, SoulMonster said:

 

but that's something for my book. In it, I've already dedicated a whole chapter on this disease. This virtual graph is something I need to work on a little more. so when I'm dead, people who should stumble upon my diary, can know the truth. 

[Dazey]

1 minute ago, action said:

When I'm dead, people who should stumble upon my diary, can know the truth. 

Well if your predictions are anything to go by I think you'd better get writing fast.  

[action]

Just now, Dazey said:

Well if your predictions are anything to go by I think you'd better get writing fast.  

luckily, my right hand is well trained

[SoulMonster]

3 minutes ago, action said:

luckily, my right hand is well trained

Yeah, I think you should go more in the direction of comedy.