No time for commentary today, but here are the pretty pictures.
Friday, July 31, 2020
Thursday, July 30, 2020
COVID-19 July 30, 2020
The number of new cases has now clearly started to trend downward. That's driven largely by the data coming from Florida, California, and Texas which were responsible for the bulk of the second wave. However, there are still many states that are going up, up, up. These smaller states are getting washed out.
Deaths continue to increase. If there's any good news, it is that in two weeks or so we might expect those numbers to level off and then start to decrease since there is (or should be) a direct link between cases and deaths.
Tuesday, July 28, 2020
COVID-19 July 28, 2020
I wasn't able to post yesterday, so this entry covers two days. Nothing much has changed. The number of new cases each day remains flat and deaths are increasing.
I've had a number of questions about the time of the flattening curve vis-a-vis the change in national policy with the CDC. I see no evidence that there is anything funny going on and it appears to be coincidence. That doesn't mean numbers aren't being fudged in some way by the feds. It just means there's no obvious evidence.
I obtain my numbers from an aggregation of data reported by the state health departments and not from the federal government, at least not directly from the feds. I don't actually know the exact path that the data takes in getting to each of the state health departments. My suspicion is that in most cases individual hospitals and testing centers report either to their county or directly to the state. Then, counties report to the state. It could be that hospitals and counties report to the federal government and then the feds distribute that data back to the states. That would a rather convoluted path, but since we're talking about the government anything is possible even though it's unlikely. When I randomly dig into the state health department web sites, it looks to me like they are getting their data from the bottoms up and not from the top down.
While conspiracy theories about federal interference are bantered about I think it is more important to recognize that it is far more likely that numbers are being cooked at the state level. There have been credible reports of issues in Florida, for example, and these date back for months. Still, extraordinary claims require extraordinary evidence. I do my best to avoid conspiratorial thinking in favor of the more usual culprits: stupidity and incompetence. In other words, Hanlon's Razor. Never attribute to malice that which is adequately explained by stupidity.
When I first started tracking numbers I reviewed a good number of data sources. That includes Johns Hopkins, Worldometer, and other popular places people seem to go for data. I've posted on this previously and discussed why I went with the source I did. I intentionally avoided the federal government as a source of data, because it was clear from the outset that they were not up to the job because of failed leadership that begins at the very top.
I do believe the number of daily infections has leveled off on average. I don't know that it will stay that way. The current numbers are dominated by just a few states with large populations. Other, less populous states are still rising (e.g., my state of Colorado), but it takes a whole lot of those to overcome the inertia of the big states. On the local and state level there are a lot of places in a whole lot of trouble regardless of what the national averages may say. See Simpson's Paradox, which I've also previously discussed.
I've had a number of questions about the time of the flattening curve vis-a-vis the change in national policy with the CDC. I see no evidence that there is anything funny going on and it appears to be coincidence. That doesn't mean numbers aren't being fudged in some way by the feds. It just means there's no obvious evidence.
I obtain my numbers from an aggregation of data reported by the state health departments and not from the federal government, at least not directly from the feds. I don't actually know the exact path that the data takes in getting to each of the state health departments. My suspicion is that in most cases individual hospitals and testing centers report either to their county or directly to the state. Then, counties report to the state. It could be that hospitals and counties report to the federal government and then the feds distribute that data back to the states. That would a rather convoluted path, but since we're talking about the government anything is possible even though it's unlikely. When I randomly dig into the state health department web sites, it looks to me like they are getting their data from the bottoms up and not from the top down.
While conspiracy theories about federal interference are bantered about I think it is more important to recognize that it is far more likely that numbers are being cooked at the state level. There have been credible reports of issues in Florida, for example, and these date back for months. Still, extraordinary claims require extraordinary evidence. I do my best to avoid conspiratorial thinking in favor of the more usual culprits: stupidity and incompetence. In other words, Hanlon's Razor. Never attribute to malice that which is adequately explained by stupidity.
When I first started tracking numbers I reviewed a good number of data sources. That includes Johns Hopkins, Worldometer, and other popular places people seem to go for data. I've posted on this previously and discussed why I went with the source I did. I intentionally avoided the federal government as a source of data, because it was clear from the outset that they were not up to the job because of failed leadership that begins at the very top.
I do believe the number of daily infections has leveled off on average. I don't know that it will stay that way. The current numbers are dominated by just a few states with large populations. Other, less populous states are still rising (e.g., my state of Colorado), but it takes a whole lot of those to overcome the inertia of the big states. On the local and state level there are a lot of places in a whole lot of trouble regardless of what the national averages may say. See Simpson's Paradox, which I've also previously discussed.
Sunday, July 26, 2020
COVID-19 July 26, 2020
A short post today because it's the weekend and I've got work, yard work, and BASEBALL to watch. There's nothing really new anyway. New infection rates continue to hold steady. Deaths continue to increase. Your state may vary.
Saturday, July 25, 2020
COVID-19 July 25, 2020 : COLORADO SPECIAL
It's been a while since I did a "Colorado Special". I figured this might be a good time, especially since decisions need to be made about sending kids back to school.
The 7-day average daily infection rate is higher than ever in Colorado, and the trend indicates a continued increase. Some caution needs to be exercised with this statement. First, testing is much higher now that it was early on. All other things being equal, the number of identified infections should go up even if the actual infection rate were the same. At the same time, Colorado started reporting only confirmed cases (i.e., positive tests) rather than confirmed plus presumptive on April 25. This is why there is a big discontinuity in the daily infection data at that date. See the big downward drop? I've circled it in red. Even so, it didn't take more than a couple of weeks for new infections to go above and beyond the numbers prior to the change. I'll also add that I personally know some people in the state that have been waiting for almost three weeks to get test results. Many that were infected weeks ago won't show up in the stats until the tests come back.
Around May 5th, the daily infections began to drop fairly steadily. We were in the height of the lock-down and it worked. Then, around June 14 the numbers began to climb. This coincides, roughly, with the loosening of restrictions. Shocking, I know. It's almost as if...as if science works. Who knew? Colorado numbers are going up even though the 7-day average for the country has stabilized in the last week. Be careful with averages. Even in Colorado, the trends in counties and cities can be extremely different.
I'll make this point explicit: Cases are increasing even though we haven't yet sent thousands and thousands of students and staff back into enclosed buildings for 7 hours per day. You do the math. If anyone is wondering, I'm keeping my high-schooler at home; he will be attending via an on-line school option. I'm lucky that my school has provided this option. We do have the option to send the kids back to school on a hybrid system that includes both in-class (50% of the time) and on-line (50% of the time) instruction. I'm fortunate that I'm in a position to do the on-line option. Not everyone can. Some will be forced to send their kids into the grand petri dish experiment because they have to work or for other reasons. Who will be held responsible if this all goes horribly wrong?
I'll make this point explicit: Cases are increasing even though we haven't yet sent thousands and thousands of students and staff back into enclosed buildings for 7 hours per day. You do the math. If anyone is wondering, I'm keeping my high-schooler at home; he will be attending via an on-line school option. I'm lucky that my school has provided this option. We do have the option to send the kids back to school on a hybrid system that includes both in-class (50% of the time) and on-line (50% of the time) instruction. I'm fortunate that I'm in a position to do the on-line option. Not everyone can. Some will be forced to send their kids into the grand petri dish experiment because they have to work or for other reasons. Who will be held responsible if this all goes horribly wrong?
The 7-day average death rate bottomed out around July 4. It was darn near zero. Since that time, the death rates have been climbing. This is expected given the increase in infections and the lag between an infection and death. As cases increase and hospitals begin to get filled again, deaths should also increase. If there's any good news it's that medical professionals have learned a lot about how to best treat the infections with the tools at hand. No doubt that many who would have died in March and April will survive because of improved treatment options.
COVID-19 July 25, 2020
Infection rates continue to hold steady at about 65,000 per day. I've added a "zoom" plot that shows the infection data since about June 9. This makes it easier to see the most recent data and trends. It's clear by eyeball that things are more or less steady.
Deaths, on the other hand, continue to rise. Not much more to say about that.
Deaths, on the other hand, continue to rise. Not much more to say about that.
Friday, July 24, 2020
COVID-19 July 24, 2020
Despite the news headlines, the national 7-day average infection rate has slowed and nearly leveled off. That's an average across the country. That doesn't mean there aren't states where cases are continuing to grow rapidly. There are! But, on average things have stabilized. It also doesn't mean it will stay this way. I doubt it will. I suspect the new and upcoming states will start to take over and we will see the curve tick up again.
The death rates are continuing to grow, as expected. It is interesting to compare the current growth rate of deaths to that of the initial growth back in March. It's quite clear that the deaths are growing more slowly now. Why? There are lots of possibilities. Better treatments? Certainly. Different demographics of the afflicted? Seems likely. Mutation or dominance of a less aggressive virus? No strong evidence for this, but still possible. Change in how deaths are counted or attributed to COVID-19? To some degree, but it probably isn't a huge factor. Other possibilities? Almost certainly.
Subscribe to:
Posts (Atom)