Greg Ferrar Data Analysis

This is a summary of COVID-19 trends today, by countries, US states, and California counties.

Summmary

	Daily Cases	Daily Growth
World	cold (49 DCPM)	cool (0.7% and warming)
United States	warm (150 DCPM)	hot (2.3% and warming)
California	cool (82 DCPM)	cool (-0% and cooling)

Graph of the Day

Europe is the new epicenter, and it’s much worse this time than it was in April, and it’s getting even worse, quickly, in a few European countries:

Countries of the World

Interactive source graphs, where you can select regions, zoom, etc., are here.

Countries Severity

It’s been a few weeks since I’ve written a blog, and things are really starting to change on the world scene. The graph above shows the twenty most severely-hit countries (“severity” is a metric that combines daily cases per million, and growth rate). All of these are severity hotspots, though on the bright side, that means that most of the rest of the countries of the world are not severity hotspots; most countries of the world are doing OK.

It is worth noting that the United States is not in this list of severity hotspots. The Americas in general have mostly recovered from a month or two ago, when they were the COVID-19 epicenter (especially Central/South America). We still see Argentina and Costa Rica on the list, but Brazil, for instance, is gone. These have been replaced by a new wave in Europe, far more severe than the first one.

Daily Cases: Europe is in Bad Shape

Looking at DCPM (above), Europe also leads the world; half of the worst-hit countries in DCPM, are in Europe. And remember how we thought Europe was having a bad wave back in April? It never even hit daily-cases hotspot territory then (see that little bump at the left? That was its April “wave”). But as of today, we have five European countries which are daily-cases hotspots. The Czech Republic, Belgium, and the Netherlands are the worst right now, over 300 DCPM—and the Czech Republic has exceeded 500 DCPM, a level seldom seen elsewhere. For comparison, we have occasionally seen DCPM each 1000, for instance in Qatar, so Europe hasn’t quite reached epic historical levels yet.

Counties Growth Rates

The graph above shows the weekly growth rates in daily new cases for the hardest-hit (in DCPM) countries. I generally consider anything above 20% to be a big problem, and the world has a lot of big problems right now. Switzerland is growing at an astounding 115% per week, and accelerating. Nine other countries (the top ten in the legend at the right above), are growing at above 20% weekly. There is no clear pattern as to the trend of these growth rates: Switzerland, Slovenia, and the Czech Republic are accelerating, but Jordan, the United Kingdom, Lebanon, and Moldova, to name a few, are decelerating. Europe, broadly, is accelerating, but there are exceptions. This is certainly not a good-news graph, but several of these hardest-hit countries are shrinking, and the general pattern is still that only about twenty of the world’s countries are really hard-hit right now, which mean that about 175 countries do not currently have a serious problem. The pandemic is severe, but only in a few red-hot spots.

States of the United States

Interactive source graphs, where you can select regions, zoom, etc., are here.

US States Severity

As noted in the Countries section, the United States is not one of the hard-hit countries right now. It used to be the worst, but now it’s gotten a lot better. But like the world, its recovery is very uneven. Most states that have seen a bad wave earlier in the year, have not seen another bad wave, but there are a whole flock of new hotspots arising in states that never had it very bad before, especially the central-north states: Montana, North and South Dakota (for some reason, probably a lack of volunteers, coronadatascraper.com does not collect data on North Dakota, but its trend is similar to South Dakota), Wisconsin, and a bunch of others. The east and west coasts remain relatively stable and are not experiencing a significant wave—but they haven’t fully recovered from their previous wave yet, either.

This is not a happy graph; almost all of the top-20 states in severity, are severity hotspots. This is much worse than it was a month ago. A few bright spots, though:

1. Nineteen severity hotspots means that 31+ states and territories are not currently severity hit.
2. Those who have been following me for a while will notice that almost none of these states are the ones that were severe in the March, May/Jun, or July waves. So the places we’re seeing strong surges, are places that haven’t seen strong surges in the past: we’re not seeing new waves in New York, Louisiana, Arizona, Florida, etc.
3. These are generally low-population states, so their surges do not contribute much to the total country numbers. This is why we have continued to see the United States as a whole look pretty good, in spite of these local spikes. This is small consolation to those who live in the central-north, though.

US State Growth Rates

The graph of growth rates for US states, for the top-twenty in DCPM, above, is pretty bad. I consider 20% weekly growth to be a problem, and seven states exceed that. Of the rest of the top twenty, most are above the zero line, indicating that they are still growing. The one saving grace is that none of them are accelerating upward: the cases are growing, but the growth rate itself has stabilized. This typically indicates that we’re about halfway to the peak. When those lines turn downward and cross the white horizontal zero line, we will have reached the peak of these regional waves. It is not clear from this graph when that will occur; they have been fairly steady for a couple weeks now, with no indication of the downward turn we would expect. But at least they aren’t heading even higher, as we see in the worst cases.

California and Colorado Growth Rates

In a nod to the fact that I live half in California and half in Colorado and tend to have more readership from there, the graph above shows the growth rates for both. California has had an extended period of shrinking daily cases (line below zero), though in recent days it has crept back almost to zero; still, this is much better than half the country, as California is at least stable, and arguably shrinking. Colorado, which is surrounded on three sides by hard-hit states, has partly avoided being pulled into the mess that most of the center-north states are in, but with a current growth rate at 20% and recent growth rate consistently above 0% and sometimes alarmingly above 40%, it is not in great shape, and it is entirely plausible that Colorado is just late to the party, and will soon catch the wave that has plagued its neighbors. Or maybe not: we’ll keep an eye on it. Either way, right now Colorado is a borderline growth hotspot.

California Counties

Interactive source graphs, where you can select regions, zoom, etc., are here.

Counties Severity

California is in relatively good shape, and unlike the world and the United States, it doesn’t even have regional hotspots. Even these twenty worst-hit counties (above) are at or below 200 severity (Sonoma and Kings are just barely above). Also, the trend lines are generally horizontal, so there is little evidence here that things are going to get worse in the short-term. This is still an epidemic—it is no time to relax—but relative to the hard-hit half of the country, California is golden.

Counties Growth

The growth graph by county, above, is similarly fairly good. These are the worst counties in DCMP, and of those most are shrinking. Only Stanislaus and Sonoma are really problematic, at 40% and 20% weekly growth and rising. The rest are below or fairly near zero growth rate, and roughly horizontal. Overall, California continues to shrink in daily cases, and though the shrinkage rate isn’t as fast as it was a few weeks ago, it’s still below zero, with no really problematic counties.

This is a summary of COVID-19 trends today, by countries, US states, and California counties.

Summary

	Daily Cases	Daily Growth
World	cold (37 DCPM)	cold (0% and warming)
United States	warm (115 DCPM)	cold (-1% and steady)
California*	warm (103 DCPM)	cold (-4% and cooling)

* California data is currently unreliable due to a computer system problem

Graph of the Day

The counties of California, graphed as a weekly growth rate of new cases. Almost all of them are shrinking in daily new cases, many of them very quickly.

Countries of the World

Interactive source graphs, where you can select regions, zoom, etc., are here.

Countries Severity

The countries of the world are looking mixed in severity, maybe a little worse than they have been, but not terrible. Severity hotspots are Bahrain (again), Israel, Argentina, Costa Rica, Hungary, Spain, and Kuwait. Only Bahrain and Israel are really severe, and even those are not as bad as other countries have been in the past. Also, those two countries are only recently severe, so there is some hope that those will turn out to be brief spikes or even data anomalies. So, there are no really bad country severity hotspots, but there are still 15+ which are warm or hot in severity:

Counties Growth Rates

Looking at the growth rate of the twenty hottest countries in DCPM, things look a little better. Though most of these are still growing in daily cases (they are above the horizontal white zero line), most of them are also flat or trending toward zero in their growth rate, so at least they’re not accelerating. Only three countries—Libya, Israel, and Bahrain—have the triple-whammy of (1) being in the top 20 in the world in DCPM, (2) having a positive growth rate, and (3) having an accelerating growth rate:

The United States

Looking at the United States, the news is very good. The two graphs below (daily cases, and growth rate) show that (1) the United States continues its strong decline from the daily-cases peak of late July, and (2) its growth (shrinkage) rate is fairly steady, and has been for many weeks, at -10% per week. The United States is barely even a daily-cases warmspot anymore.

I got a little worried about a week ago when it seemed like growth was leveling off and maybe even headed back toward zero, but it remains below zero and is now trending back toward the 10% mark. Given the large numbers of upward pressures on this metric—Labor Day, return to school, impending winter—it wouldn’t be surprising if we saw an upward bending. But so far, this has not occurred. The downward pressures (mitigating factors like physical distancing, lockdowns, school closures, and the now almost-universal mask-wearing) appear to be up to the task of resisting the upward pressures and keeping our metric dropping, so far.

India

I’ve brushed off India for months now saying, yeah, the trend doesn’t look great and the potential for explosive epidemic is there, but they have very low DCPM and are not yet worth worrying about. Well, India is now worth worrying about—their long slow rise has finally taken them within striking distance of 100 DCPM, i.e. daily cases warmspot territory, and due to their immense population, that’s a lot of cases. Not adjusting for population but just looking at raw numbers (as I seldom do), India is news, as it now has far more daily cases than the long-declining Brazil and United States:

More significantly for comparative purposes, India has risen to 65 DCPM (the dashed orange line is the United States, dotted pink is India). That’s not yet a daily-cases warmspot, but it’s closing in at a pace that might get it there in a few weeks:

In growth rate, India has also been consistently positive for a long time, and at the moment, is around 10%+ and recently accelerated from 5% to near 15% (for comparison, dashed-green is the United States, dotted-orange is India). Its grown rate is a little lower in recent weeks, but still positive. In summary: India is looking problematic, and there is currently no end in sight of its epidemic:

States of the United States

Interactive source graphs, where you can select regions, zoom, etc., are here.

US States Severity: Still A Lot Of Hard Hit States

The states of the United States, are not out of the woods yet. There are still five severity hotspots: South Dakota, Missouri, Guam, Oklahoma, and Kansas. South Dakota and Guam were particularly hard hit recently, but are quickly recovering. The rest of the top 20 are severity warmspots, but appear still to be trending upward:

States Daily Cases

Looking at the DCPM graph, the news is moderately good, mostly because the news was so bad in July that today looks good in comparison. The July peak is long behind us, and all states that were hard-hit then, have come a long way toward recovery. We do have a few new “fourth wave” states/territories, notably Guam and South Dakota, but those also seem to be headed toward recovery. Looking at this graph, it is easy to see why the United States as a whole has been declining in daily cases since that July peak. But still, all of these top-20 states, are daily-cases hotspots or daily-cases warmspots, so the United States is far from recovered. We’d really like to see almost all states below, say, 10 DCPM, before we can really relax and resume something like normal life, and right now there are still a lot of them over 100:

Here’s New York, by the way. This is what a successful recovery looks like. It’s worth noting that New York can’t seem to get below 35 DCPM either, maybe because it continues to have a influx from all the other infected states. But it is at least holding steady at a reasonably low daily-cases number, far below warmspot territory. We’ll see how the school openings, and coming winter, and Labor Day fallout, affect it:

States Growth: Trending Firmly Toward Recovery

The best news for the states is the graph below, showing that nearly all of the DCPM-hottest states, are declining in daily cases (below the zero line), or are decelerating toward zero. Nothing here is above 10% weekly growth, which isn’t too bad (I think of 20% as “bad”). There remain a few which stubbornly hold on to a positive-but-low growth rate, but there’s nothing here remotely like a triple-whammy (top 20 DCPM, high growth rate, and accelerating). Really, only four states look remotely bad in this graph: Missouri, Wisconsin, Oklahoma, and Utah, all with slight and steady growth:

The Metaphorical Immune System Hypothesis Lives!

Notice how when I talk about the hardest-hit states, I’m never talking about the states I talked about months ago? New York, New Jersey, Rhode Island, Michigan, Illinois, Louisiana, Arizona, Florida: these are nowhere to be seen in the latest top-20. This is evidence for my still-standing hypothesis, expanded at length in my Medium article, that once a states gets a strong wave, it does not get another one in the medium term. This is because it develops metaphorical “antibodies” at the social level, which fend off new waves. One of my favorite graphs, below, is the state trajectories graph, showing what happened to each state after it reached 100 DCPM.

Almost invariably, states’ peaks clump at the far left, following the pattern I’ve repeatedly pointed to in the past: rising for about 30 days, peaking, falling for about 40 days, and never rising again. Three apparent exceptions to this are Louisiana, Mississippi, and South Dakota; the latter two can be easily explained because they never really got their peak until they had hovered at 100 for a while, pushing them to the right of the graph; their trajectories really do follow the pattern. Louisiana is not so easily explained, but one way of looking at it is that it had a small first wave that was just New Orleans, then the rest of the state caught it much later and gave it its “real” wave.

In any case, the trend is clear that almost all states had their large wave, got over it, and never rose again. Now that all states and territories have had their wave, and no second waves have appeared, I again predict that the United States will not have a second large wave in the next three months. And with the likelihood of a vaccine in the next three months, I predict that the United States will never have another wave like the July wave. In the metaphor, every state has built metaphorical immunity to such waves, will keep it for months to come (until the metaphorical antibodies wear off, in the form of fatigue and growing popular non-cooperation, or premature official lifting of restrictions), and actual individual biological immunity in the form of a broadly-taken vaccine will prevent a future large wave.

California Counties

Interactive source graphs, where you can select regions, zoom, etc., are here.

California Counties Severity

California counties are looking better. Only four counties are hotspots: Marin, Kings, Butte, and Imperial. Marin is literally off the chart below, but after discussing this with the data source provider, I conclude that it didn’t really have a terrible spike and fast recovery, but that the data provider just switched from the county data to the Johns Hopkins data, which led to a sharp jump in reported cases (JHU says there are twice as many cases in Marin as Marin Country Health says). Though this discrepancy is odd, it’s not extreme, and it doesn’t change most of my analysis, which is mostly around growth rates. Marin’s growth rate does not seem to be exceptional. It will take a while for it to settle down to see where its DCPM really is, so we’ll return to Marin in a later episode. So the top 20 are all hotspots or warmspots, which isn’t good, but there are no really extreme cases, and most counties seem to be declining:

The entire history of all counties in DCPM shows a pretty good trend too: there are few daily-cases hotspots, and though there are still a lot of warmspots, the trend seems to be toward recovery, as shown by a higher snarl in the middle of the graph, than the snarl at the right:

Remember Los Angeles?

Los Angeles, once the dominant force in California COVID cases, has recovered sharply and is not even a warmspot anymore. This probably accounts for most of the recovery of the total California numbers, so we might now worry that California numbers will not decline as much, now that Los Angeles has given back most of the daily cases it had to give. But with the apparent downward trends of most counties, it seems likely that the state numbers will continue to drop even now—but maybe more slowly than recently:

County Growth Rates: The Best News Today

The really good news is in growth rates by county. The graph below shows that nearly all counties are shrinking in new cases (below the white line), many of them rapidly (20% per week or faster). Only Butte, Imperial, Kings, San Diego, and Humboldt are growing, and of those, only Humboldt is accelerating. This is what we like to see in a growth graph:

Definitions

DCPM	Daily cases per million
Daily-cases warmspot	>100 DCPM but <200 DCPM
Daily-cases hotspot	>200 DCPM
Growth warmspot	>1% daily (>7% weekly) growth in new cases
Growth hotspot	>5% daily (>35% weekly) growth in new cases
Severity	A metric combining DCPM with growth
Severity warmspot	>100 Severity but <200 Severity
Severity hotspot	>200 Severity

This is a summary of COVID-19 trends today, by countries, US states, and California counties.

Summary

	Daily Cases	Daily Growth
World	cold (37 DCPM)	cold (-0% and cooling)
United States	warm (166 DCPM)	cold (-2% and cooling)
California*	hot (235 DCPM)	cold (-1% and cooling)

* California data is currently unreliable due to a computer system problem

Graph of the Day

The top twenty hardest-hit states are all shrinking in daily new cases now; but we’re seeing strong growth in the Territories:

Countries of the World

Interactive source graphs, where you can select regions, zoom, etc., are here.

Countries Severity

Several countries continue to be severely hit, with severity above 100, for the worst twenty countries. Five countries are severity hotspots: Peru, Colombia, Panama, Bahrain, and Brazil. The rest in the graph below are severity warmspots. All other countries are not even warm. Most countries of these top twenty, however, appear to be rising slightly in severity (as we’ll see below, growth rates are dropping, so this is due to cases continuing to rise for those countries which have not yet peaked, especially in Central/South America):

The United States is going in a very nice direction. With 140 severity, it’s still a severity warmspot, but falling sharply:

Countries Growth

As we’ll see below, the world is actually lagging the United States in recovery now: states growth charts look even better than the chart below. But still, at the countries level, the news is pretty good. Of the hardest-hit (highest-DCPM) countries, most are now shrinking in daily growth, and almost all are seeing their growth rates declining. Growth hotspots are Peru, Colombia, Argentina, Moldova, Kazakstan, and Costa Rica. Note the Central/South America epicenter, still worst in the world. But note also the downward slope of growth rates in most of those countries, signaling a peak there in a few weeks:

The United States, once the worst in the world on almost all metrics, is doing much better, having peaked a week or two ago, now with a growth rate of -10% weekly; i.e., each week sees 10% fewer new cases than the week before:

The world as a whole has also now peaked and is declining in daily cases, and also declining in growth rate of daily cases, continuing this pleasant descent it has been on since early July:

States of the United States

Interactive source graphs, where you can select regions, zoom, etc., are here.

States Severity: Looking Better Except Hawaii and Territories

In severity, the states are looking much better than they were in July. Though the top 20 most severe states all remain severity hotspots and severity warmspots, the worst of the non-territories (territories tend to have more extreme numbers due to their small populations) is Hawaii, at 320 severity:

Hawaii is by far the worst of the states right now, clearly finally experiencing its first wave:

States Growth: All Hardest-Hit States Are Shrinking

The growth rate of states is looking very good now, with all of the hardest-hit (highest DCPM) states now shrinking in daily new cases. Puerto Rico and the Virgin Islands, like Hawaii, are far worst than the States, but do not affect the total US numbers much due to their small population. Nevertheless, these are not the places to visit right now:

Daily Cases Remain Very High In Most States!

But it’s not time to relax! Not yet. In terms of daily cases the top 20 states are still DCPM hotspots or warmspots. Though the daily cases graph is showing strong recovery (almost all states bending downward toward zero), the number are still very high across the country. Whatever it is we’re doing (masking, social distancing, staying home), is working, and if we keep doing these things, these graphs will probably continue to improve. If we stop doing these things because the trends are looking good, the situation may degenerate again. There is an end in sight, a couple months from now perhaps, when some of the difficult mitigating steps we’ve all been taking can be relaxed, but not yet:

California Counties

Interactive source graphs, where you can select regions, zoom, etc., are here.

Counties Severity: Reporting Glitch Causing Short-Term Upward Trend

The short-term severity trends for California counties is not looking good, with most of them rising. However, as we’ve discussed in past livestreams, this was expected, and does not necessarily indicate a rise in actual recent cases. California had a computer system problem which resulted in a backlog if 300,000 test results not being recorded; this has now been fixed, but all these results landed in the past few days, artificially driving up the total cases, and total tests, and even more, the growth rate, which is a major component of the Severity metric graphed below. It will take a few days, maybe a week, before this gets averaged into the medium-term trends of these graphs, so at this point, the California daily-cases, growth, and severity graph trends can’t be trusted to correlate at all with the underlying infection rate:

Because of this data problem, we don’t know what’s actually going on in California today, so I’ll stop here and wait for the data to settle down from this artificial spike.

Definitions

DCPM	Daily cases per million
Daily-cases warmspot	>100 DCPM but <200 DCPM
Daily-cases hotspot	>200 DCPM
Growth warmspot	>1% daily (>7% weekly) growth in new cases
Growth hotspot	>5% daily (>35% weekly) growth in new cases
Severity	A metric combining DCPM with growth
Severity warmspot	>100 Severity but <200 Severity
Severity hotspot	>200 Severity

This is a summary of COVID-19 trends today, by countries, US states, and California counties.

Summary

	Daily Cases	Daily Growth
World	cold (36 DCPM)	cold (0% and cooling)
United States	warm (180 DCMP)	cold (0% and cooling)
California	warm (190 DCPM)	cold (0% and cooling)

Graph of the Day

The states continue their recovery, with most of the hardest-hit states now showing negative growth of daily cases:

Countries of the World

Interactive source graphs, where you can select regions, zoom, etc., are here.

Countries Severity Hotspots: Not Many

Only a few countries remain as severity hotspots: Peru, Panama, Colombia, Brazil, and just barely, Kosovo (note the Central/South America cluster). None of these is over 300 severity, so not super hot. There are less than 20 severity warmspot countries in the world. These top-20 most severe countries are mixed in whether they’re getting better or worse; those which were most severe in July have universally been getting better, but there are a number of countries rising in severity, most notably Peru, Panama, and Colombia (Brazil is looking a little better than those in the short term):

Countries Growth: Looking Good

The “severity” metric includes the number of cases actually reported (daily), as well as the growth rate of cases, so it counts countries as severe even when the cases are dropping, if they are still high. For a pure look at growth, we have the countries growth chart, which picks the hottest countries in DCPM, but then graphs just their growth rates. This chart is looking very good these days, with the vast majority of hardest-hit countries showing a negative growth rate in daily cases (fewer cases every day), and those that are still growing, showing a strong trajectory toward zero growth:

States of the United States

Interactive source graphs, where you can select regions, zoom, etc., are here.

States Severity

There continue to be many, many severely-hit states in the United States. On the bright side, the last time I did this blog (four days ago), all twenty of the top-twenty were severity hotspots, and now six of them have dropped into severity warmspot territory. In general, these most-severe states are visibly getting better, as their severity lines trend downward, but at this rate, it’s still looking like a couple months before they all drop down into the teens, where we can really say their waves are over. Note that Hawaii, light blue, one of very few states which have not yet had a wave, is now having a wave, an has reached a very severe level of 300:

States Growth

The growth rate graph by state looks even better, with most hardest-hit states now shrinking in daily new cases, and the rest of the top twenty clearly trending downward to zero growth:

States Testing: The Are Problems Here

Returning to the Gang of Six, the center of the May wave of states (Alabama, Arkansas, Arizona, North Carolina, South Carolina, and Utah), here are their positive test ratios. A number like 8%, is around the cutoff to be considered good; and we’d really like to see 2% or better; anything above 10% is pretty bad. We can see below that these six states are still experiencing high test percentages, even though we’ve seen that their cases have been dropping for a while. This is part of a recent trend where, for various reasons, fewer tests are actually being performed, and tend to be done on those who are sickest, hence higher positive test rates:

California Counties

Interactive source graphs, where you can select regions, zoom, etc., are here.

Counties Severity

The top-20-most-severe counties are still pretty severe, with most of them as severity hotspots. Kern and Marin were extraordinarily high recently, but they quickly dove back down toward the pack, so those may have just been data anomalies. We know there have been testing backlogs and reporting infrastructure problems in California lately, so none of this recent information is particular reliable. But if we do choose to believe it at face value, most hard-hit counties are currently trending downward in severity:

Counties Growth: Heading In The Right Direction

Similarly to countries and states, the growth chart for counties of California is looking much better than a month ago, with about half of the hardest-hit counties shrinking, and most of the rest trending toward zero. The trajectory toward a recovery continues:

Definitions

DCPM	Daily cases per million
Daily-cases warmspot	>100 DCPM but <200 DCPM
Daily-cases hotspot	>200 DCPM
Growth warmspot	>1% daily (>7% weekly) growth in new cases
Growth hotspot	>5% daily (>35% weekly) growth in new cases
Severity	A metric combining DCPM with growth
Severity warmspot	>100 Severity but <200 Severity
Severity hotspot	>200 Severity

Note: I didn’t post a blog here on gregferrar.us last week because as I started to write it, I decided that Medium would be a better, uh, medium. Last week, the United States was clearly reaching another peak, so I took the opportunity to package that up for popular consumption (as I did on Medium in the first US peak in early April) together with my hypotheses based on my “metaphorical regional immune system” theory. That the US has peaked in the short term, is uncontroversial. My more-controversial hypothesis is that it will not peak again this high in the medium term. My even-more-controversial hypothesis is that it will never peak this high again. I wrote this all up in a much-longer-than-my-usual-blog Medium story, and it’s gotten a lot of attention. So here’s last week’s story, published on Medium. Now, back to our regularly scheduled blog.

---

This is a summary of COVID-19 trends today, by countries, US states, and California counties.

Summary

	Daily Cases	Daily Growth
World	cold (30 DCPM)	cool (0.6% and falling)
United States	warm (192 DCPM)	cold (-1.8% and falling)
California	cool (85 DCPM)	neutral (0% and falling)

Graph of the Day

Of the hardest-hit states in DCPM, half have peaked (daily growth rate below the thick white zero line), and are seeing fewer new case each day:

Countries of the World

Interactive source graphs, where you can select regions, zoom, etc., are here.

Country Severity: The United States Is Not Hot Anymore

There are no super-hot countries right now, but there are four severity hotspots: Brazil, Colombia, Peru, and Panama. Here we clearly see the Central/South America epicenter, and it is notable that the United States is not listed; the United States is no longer a severity hotspot, having dropped to 190 in severity. However, there is no clear trend downwards for the top-20-most-severe as a group (as there has recently been); these seem broadly to be holding stable, not getting worse, but not getting better either:

But the United States, still the largest source of cases, is getting better in severity:

Looking at the growth rate graph for these highest-DCPM countries, however, we can see that the trend is leaning toward the positive. Only seven of the top 20 are growing in DCPM (above the thicker white zero line), and most are trending downward in growth rate. So the other thirteen, have peaked in daily cases and are now declining. Only Kosovo and Peru show a slight upward trending growth rate, and even those are only around 20% weekly, which makes them just a growth warmspot. There are no growth hotspots.

States of the United States

Interactive source graphs, where you can select regions, zoom, etc., are here.

States Severity: Improving

In severity, the states of the United States look much worse than the countries of the world; there are still a lot of very severe severity hotspots; in fact, almost all of the 20 most severe states are severity hotspots. On the bright side, most of these are noticeable trending downward toward less severe. Only Hawaii, the bottom line, is showing a really bad trend right now, as it has recently and sharply turned upward and is nearly in hotspot territory (from the Medium article: Hawaii is one of just four states that have not yet had a strong wave, and might be expected to join together in a fourth wave of states soon):

States Growth: Half Of Hardest-Hits Are Shrinking

But, the growth rates of the states continues to point in a very nice direction. Unlike a month ago, when the highest-DCPM states were still accelerating upward, almost all of them are seeing decreasing growth rates now. In fact, half of all these worst-hit states have now crossed the zero line of growth, meaning they have turned the corner, peaked, and are now declining in daily new cases. Only Missouri has a really bad growth rate (40% weekly, hence the only growth hotspot), and that has been declining strongly for the past week:

California Doesn’t Even Register

It is perhaps worth noting that California did not appear in the above graphs because it’s not even a daily-cases warmspot; though it flirts with that 100 DCPM line, it hasn’t crossed it (in 14-day moving average). A DCPM of 90 still isn’t great, but because so many other states are very severely hit, California does not even make the cut to be in the “worst 20” states graphs above.

State Deaths: The “We’re Just Not Testing Enough” Narrative

There is a narrative going around, that reported cases are only declining because of insufficient testing. The following graph of deaths by state can provide some insight into the validity of this narrative. (All states are shown here, more than fit in the legend, so it is not generally possible to tell which line is which state; use the interactive version to explore this graph yourself).

This graph (above), does not, yet, directly prove or disprove the narrative. It supports it at first glance, by showing that deaths are rising, even while cases are falling; this might indicate that underlying infections are in fact rising, and are just not being reported due to testing lags (but deaths are reported without lag).

However, I would expect death reporting to lag case reporting by a couple weeks, since deaths happen weeks (sometimes months) after an infection happens, and even with delayed testing, deaths will probably still occur days after the case is confirmed by testing.

Consider the example of Arizona, the first of the current wave, hardest-hit in DCPM, and one that has now apparently peaked in deaths. Here are the graphs of Daily Cases, and Daily Deaths. Cases peaked around July 10th. Deaths peaked around July 25th, two weeks later. It does seem that the deaths graph is offset from the cases graph by about two weeks:

It’s too soon to say whether the other states will follow this pattern, but if they do, we’ll see a peak in deaths in most of the third-wave states (which is most of the states of the United States), in the coming week or two, since they peaked in cases about a week ago. If we see this pattern of peaking and dropping death graphs across the US, it will be evidence against the theory that we still have rising infections, but just undetected. So, it will take a couple weeks before we can see whether this narrative is true or not.

California Counties

Interactive source graphs, where you can select regions, zoom, etc., are here.

Counties Severity: Holding Steady But Not Getting Better

In severity, there have been three extraordinary counties in recent weeks, with severity in the thousands; recall that a Severity of 200 is a severity hotspot. These may be data outliers and “not real” in some sense, or at least not persistent, and they are not large counties, but I’ll show them briefly here before I discard them. My guess is that Kern (green) really did have a persistent spike, now recovering, while the other two (Marin and Stanislaus) were more data anomalies (from their triangular shape, indicating a single-day blip), due either to changes in reporting, or simply wrong numbers:

Discarding those, this is the counties severity graph. It’s pretty bad, with just about all of the top 20 counties as severity hotspots (recall that the state itself is not a hotspot, however). It’s also not clear that these counties are getting better in severity; they are perhaps horizontal, with a general trend for the worst ones to be trending a little better. None of them are accelerating strongly upward at this point:

Counties Growth: Also Holding Steady But Not Getting Better

Results from the growth rate graph are similar. A few counties are shrinking—Imperial especially, which was once by far the hottest county, continues to shrink steadily at 20% per week, and is no longer in the running. None of the counties are growth hotspots; only Merced is close. But the other hottest countries are broadly horizontal in growth, not accelerating in growth rate, but not getting better, and hence almost all still growing in daily new cases:

Bay Area Counties: Still Growing Slowly

In the San Francisco Bay Area, the counties are still generally growing, albeit slowly (mostly in the 0% to 20% range). Alameda (red) appears to be doing great, way down there at -40%, great enough that I wonder if the data source is accurate; if I had to guess, Alameda is really in the cluster with the others, and something has gone wrong with data reporting by the county, or collection by coronadatascraper.com. The others show a medium-term downward trend, but short-term they have leveled out, and while not accelerating in growth, they’re not obviously getting better either. So the Bay Area, like California, seems stable right now, but is not trending toward actual recovery:

Definitions

DCPM	Daily cases per million
Daily-cases warmspot	>100 DCPM but <200 DCPM
Daily-cases hotspot	>200 DCPM
Growth warmspot	>1% daily (>7% weekly) growth in new cases
Growth hotspot	>5% daily (>35% weekly) growth in new cases
Severity	A metric combining DCPM with growth
Severity warmspot	>100 Severity but <200 Severity
Severity hotspot	>200 Severity

This is a summary of COVID-19 trends today, by countries, US states, and California counties.

Summary

	Daily Cases	Daily Growth
World	cold (32 DCPM)	warm (1.2% and cooling)
United States	hot (202 DPM)	cool (0.5% and cooling)
California	hot (220 DCPM)	warm (2.2% and cooling)

Graph of the Day

Deaths are rising in the hardest-hit states, even those states that have started to show signs of recovery, as deaths lag a couple weeks behind cases:

Countries of the World

Interactive source graphs, where you can select regions, zoom, etc., are here.

The Most Severely Hit Countries Are Trending Toward Recovery

The severity graph shows that the worst-hit countries are Oman, Bahrain, Panama, South Africa, and Israel. Kyrgyzstan is probably also in this list, but with a severity of 2400, it is so hard hit that it wrecks the curve and is omitted from the graph below. Kyrgyzstan has suddenly leapt into a bad place recently, but it’s so recent and so sharp in its rise that I’m not yet sure it’s real, and not a data anomaly. We’ll keep an eye on Kyrgyzstan. But all the other hardest-hit countries are showing a slight and recent downward bend in their severity lines; though they are far from recovery, they are all at least trending that way (see United States, below).

Actually, the United States isn’t quite trending downward in severity (yellow line above), but it’s certainly flattening out there, and in weekly growth (below), it joins the others in a clear trend toward recovery: all of them are seeing growth rates dropping. At this rate, the growth rate of the United States will hit zero in another two or three weeks, and daily numbers will start to drop. But only if the trend continues—all these trends have shown themselves resistant to accurate modeling and prediction. Bahrain, however, appears already to have turned the corner, and daily cases are already dropping there:

States of the United States

Interactive source graphs, where you can select regions, zoom, etc., are here.

Severity By State

The states don’t look so good from a severity perspective: We have at least 20 severity hotspots. On the bright side, it was just a few weeks ago that we had four or five states at even higher levels of severity: Arizona, Florida, Louisiana, Idaho, and Georgia. As of today, all of these have recovered quite a bit, growth-wise; all are headed in the right direction toward lower growth of daily cases. But a number of states are competing to be the new most-severe: Alaska, Alabama, Nevada, Texas, Mississippi and others; and all of these are still getting worse.

States That Recover, Don’t Get A Second Wave

We continue to see a broad trend, not just among states but among regions at all levels: once a state recovers, from a place of high severity, to a place of low severity, it does not return to high severity. In other words, there are no second waves. The graph below shows that with just one exception (Louisiana), states that reach 100 DCPM typically peak about 30 days later, and have recovered back to 100 DCPM about 40 days later, and do not return to high levels. Louisiana never really recovered—not like, say, New York did—so we could justify its extraordinary performance now as a continuation of its first wave. Note especially that Floria and Arizona, which were some of the recent hardest-hit states, are following the typical trend closely, having peaked around 30 days and now appearing to be headed sharply downward in daily cases, toward a recovery around 70 days.

Along these same lines, here is the latest growth graph of the Gang of Six, which I’ve repeatedly focused on. The Gang of Six were bad news a couple months ago, and are mostly still growing in new cases; but their growth rates started trending downward in mid July, and have maintained this course steadily ever since. Arizona is actually shrinking, having turned the corner around July 11; four of the others are at less than 1% daily growth and dropping; and even Alabama, which has been bucking the trend for the past month, is finally flattening out at 3% growth. It is reasonable, having seen so many similar trajectories, to expect all of the currently-most-severe states, to behave similarly over the next couple months. And, there aren’t many states left to catch it now, so once the ones that are severe now recover, we may well rejoin Europe in a state of sustained low-severity.

Deaths Are Rising In The Latest-Wave States

Deaths might be expected to rise a couple weeks after cases rise, due to the progression of the disease: catch it, incubate it, test positive for it, got to the hospital for it, and die from it. (Fortunately, a lot of people recover at each stage). Indeed, deaths are now rising, and not yet falling, in the recent hard-hit states of Arizona, Florida, Louisiana, Texas, and Georgia. Most of these states are now starting to show signs of recovery in daily cases; in a couple weeks we can expect that these deaths graphs will also start to turn downward:

California Counties

Interactive source graphs, where you can select regions, zoom, etc., are here.

County Severity

Imperial and Kings, which once ruled the state in severity, have recovered back to the level of mere normal hotspots, but now we have Stanislaus and Kern. Stanislaus briefly had the distinction of being the most sever county ever, but is now looking much better—but it’s still worse that anyone but the extreme severity newcomer Kern.

Discarding the weirdly severe ones, we have this, showing that about half of the most severe counties are trending toward less-severe. The other half continue to get worse. This is, at least, better than a few weeks ago, when the trend was universally toward more severe:

The growth graphs, though, still shows that nearly all the highest-DCPM counties are growing in daily new cases (only Imperial is shrinking). So while California is looking somewhat better than a few weeks ago, and shows a trend toward recovery, it’s still in epidemic mode. The true turning-of-the-corner will come when the growth rates of half the counties are below zero:

Definitions

DCPM	Daily cases per million
Daily-cases warmspot	>100 DCPM but <200 DCPM
Daily-cases hotspot	>200 DCPM
Growth warmspot	>1% daily growth in new cases
Growth hotspot	>5% daily growth in new cases
Severity	A metric combining DCPM with growth
Severity warmspot	>100 Severity but <200 Severity
Severity hotspot	>200 Severity

This is a summary of COVID-19 trends today, by countries, US states, and California counties.

Summary

	Daily Cases	Daily Growth
World	cold (30 DCPM)	warm (1.4% and cooling)
United States	hot (220 DCPM)	warm (2.2% and cooling)
California	hot (220 DCPM)	warm (3.0% and slight warming)

Graph of the Day

The counties of California are finally starting to show some motion toward recovery; this graph shows that many of the hottest counties in daily-cases-per-million, are now trending downward in growth rate of daily cases. This is a leading indicator of peak-and-recovery of daily cases, so we should expect some of these counties to peak in the next few weeks:

Countries of the World

Interactive source graphs, where you can select regions, zoom, etc., are here.

Severity Hotspots

The severity hotspots are Oman (orange), Bahrain (pink), Panama (red), South Africa (yellow), Kyrgyzstan (dark blue), Israel (green), and the United States (light blue), top to bottom below. The new “severity” metric combines DCPM with growth, to show how bad it is in each region:

Worldwide Growth Rate is Dropping

Though cases are still rising every day worldwide (growth rate is above zero), the growth rate has finally peaked and is starting to drop. This is a leading indicator of recovery in daily cases; if the downward trajecctory of the growth graph continues, we will see total daily case ppeak worldwide in a few weeks. It doesn’t take too much squinting at that graph, in fact, to imagine that worldwide cases are now headed much more sharply downward, even than the trendline indicates:

States of the United States

Interactive source graphs, where you can select regions, zoom, etc., are here.

State Severity Hotspots

So many states are current severity hotspots or warmspots that it is hard to list them. A severity above 100 makes it a severity warmspot; here’s the whole tangle of states, provided to give a broad picture of how very many states we have in high severity (use the interactive graphs on a computer, to see which is which). Worst to best, we have Florida, Arizona, Louisiana, Georgia, South Carolina, Alabama, Texas, Idaho, Tennessee, Nevada, Mississippi, California, Arkansas, Virgin Islands, Utah, Utah, Oklahoma, and North Carolina.

The Worst States Are Starting to Look Better

However, the five worst-hit states (yellow dashed line is South Carolina) are starting to show visible recovery even in their DCPM graphs (only Louisiana shows no clear sign of a peak):

The Gang Of Six Is Recovering; Except Alabama

The rope chart of the Gang of Six, which were the worst-hit states a month or two ago, continues to show their recovery, and indeed Arizona is turning the corner toward daily shrinkage of new cases (crossing zero to negative growth), right now. (The upper orange line is South Carolina; the lower line is North Carolina). Alabama, however, continues to diverge strongly from the other five, showing no sign of recovery:

Deaths Are Not Dropping

I keep hearing a narrative that goes something like this, “Yes, cases are rising, but deaths are dropping!” The implication seems to be something like: so there isn’t as big a problem as we think. As far as I can determine, the premise is false; deaths are simply not dropping. Here are seven states I picked because they are forefront in my mind as recent or current hotspots. Top to bottom, they are Arizona, South Carolina, Florida, Alabama, Texas, California, and Georgia. In all seven cases, deaths are rising and have been for some time. If deaths trail cases by a couple weeks, as I might expect, these may start to drop in a couple weeks, since some of these states seem to be dropping in cases right now. But they are not dropping yet, and therefor this deaths graph confirms the severity of the current broad epidemic in the United States:

California Counties

Interactive source graphs, where you can select regions, zoom, etc., are here.

A Whole Mess Of Counties Have Severe Outbreaks

Here is the severity graph by county (use the interactive version from a computer to explore individual lines, and see which is which). Notice that most counties are above 100 severity, making them severity warmspots; and about half are severity hotspots. Stanislaus (dashed purple) is particularly extreme, and Imperial (yellow) was very high but is strongly recovering (but still very much a hotspot):

The Worst Counties are Turning Toward Recovery

At last, some good news for California. The graph below, which for a month has shown almost all California counties not just growing but accelerating, is finally showing many of them decelerating. All the downward-slanted lines indicate a trajectory toward recovery; when those lines hit zero, the counties will peak in new daily cases, and will start to see fewer and fewer cases each day.

Definitions

DCPM	Daily cases per million
Daily-cases warmspot	>100 DCPM but <200 DCPM
Daily-cases hotspot	>200 DCPM
Growth warmspot	>1% daily growth in new cases
Growth hotspot	>5% daily growth in new cases
Severity	A metric combining DCPM with growth
Severity warmspot	>100 Severity but <200 Severity
Severity hotspot	>200 Severity

This is summary of COVID-19 trends today, by countries, US states, and California counties.

Summary

	Daily Cases	Daily Growth
World	cold (29 DCPM)	warm (2.2% and rising)
United States	warm (193 DCPM)	hot (5.9% and rising)
California	warm (150 DCPM)	hot (4.4% and rising)

Graph of the Day

The Americas still have a problem, but it may be getting better. The graph below shows all the countries of the Americas; most of them are growing, but the general trend in growth rate is downward, suggesting that the Americas are heading toward a peak.

Countries of the World

Interactive source graphs, where you can select regions, zoom, etc., are here.

Country: Hotspots: Bahrain, Oman, Panama, and Brazil

Quite a few countries are daily-cases hotspots: Bahrain, Oman, Panama, and Brazil. Qatar, shown sharply recovering on the graph below, was the worse in the world for a long time, but is no longer even a daily-cases hotspot:

Brazil looked like it was headed for recovery a month or so ago, but faltered and has been rising slowly since. The other three are trending downward in growth rate, however. Qatar is shrinking rapidly in daily cases, and Oman is just turning the corner toward negative growth. None of the daily-cases hotspots are anywhere near being also growth hotspots:

There are a lot of daily-cases warmspots (see the interactive graphs to differentiate countries in this graph):

More than half of the warmspots are growing, but only two have the triple-whammy of being a daily-cases warmspot, a growth warmspot, and accelerating growth: Israel at 11% daily growth, and the United States at 6% (both yellow below; see the interactive charts to match up lines with countries):

The Americas Still Have a Problem

Here’s a graph of the growth rates of most of the countries of the Americas. I might have missed a few. This is a broad view; there are too many to list, and as always, you can look at the interactive graphs to differentiate countries (especially necessary this time since the legend is scrolled so it doesn’t even show most of them). Most are growing, and the worst are Paraguay (16%), Costa Rica (9%), Venezuela (8%), Ecuador (6%), Argentina (6%), and the United States (6%). All of these worst growth hotspots are accelerating in growth. Then there’s Cuba at the bottom, shrinking 9% a day; Chile, Haiti, and Canada are also doing well:

States of the United States

Interactive source graphs, where you can select regions, zoom, etc., are here.

The Big Picture Is Still Bad For The States

The graph below shows the top 25 states in DCPM (this is an aggregate view; use the interactive graphs to see which state is which). This is about as bad as it has ever looked: there are twelve daily-cases hotspots (Arizona, Florida, Louisiana, South Carolina, Georgia, Mississippi, Alabama, Arkansas, Texas, Tennessee, Utah, and Nevada), and eight additional daily-cases warmspots:

The rope chart for these top-25-states shows that they are all still growing, except Texas:

But notice the five states that are angled downward in growth, and the red line in the middle that is almost horizontal. These are the Gang of Six: Arizona, Arkansas, Alabama, North and South Carolina, and Utah. These were the big news six weeks ago, but they have all turned downward in growth rate, and are headed toward the zero line, which will mark the peak of their daily cases. The theme here, as has recently been my theme, is that states which are hard hit and accelerating, peak in daily cases in four to six weeks after they get hit, perhaps due to mitigating efforts by the residents and the governments of those hard-hit states. The Gang of Six is still strongly and unanimously indicating this trajectory; we can reasonably hope (maybe even expect) that all these other hard-hit states we see now, will follow a similar path in the next month or two.

Case Study: Florida

Much has been written recently about how terrible Florida is. And that’s true in many ways: Florida is showing a very troubling DCPM graph: it is a daily-cases hotspot with its trendline at 350 DCPM, and worse, almost all recent dots are above the trendline, indicating recent acceleration, and arguably an actual DCPM of about 500:

The growth graph of Florida, however, already shows signs pointing toward recovery. It is still technically a growth hotspot—its trendline is 5.8%—but all recent dots are below the trendline, most well below 5%, and couple even show significant shrinkage in daily cases (below 0%). We can see the conservative trendline bending downward already; it is likely that Florida will soon join the Gang of Six in a downward-angled trajectory of its growth rate trendline:

Similarly, the 14-day moving average has flattened out; given the trends we’re seeing in growth, it is also likely that this moving average will turn downward in just a few days, indicating an impending peak in DCPM for Florida. It’s early and this is speculative, but things are looking up, for this hard-hit state:

California Counties

Interactive source graphs, where you can select regions, zoom, etc., are here

Still No Good News For The Counties

The news by DCPM is the same as last time: almost all counties are growing, and a whole mess of them are hot or warm. If you want to pick out a particular line, use the interactive version:

County Trajectories: Up

The Counties trajectory chart, below, which shows the trajectory of each county from when it reached 100 DCPM, shows how outsized Imperial has been and remains; no other county has half its cases, adjusted by population, and though it’s trending downward, it’s not exactly plummeting; it’s still at 700 DCPM, a height seldom seen anywhere.

Notice that this chart above echoes the rope chart above it, by showing that there are very few counties getting better. The end of each thread marks today for that county. Kings is slightly descending, but not so much in the past week or two. All the other threads we see are poking upward: Tulare, Santa Barbara, Riverside, San Joaquin, San Bernardino, Stanislaus, Merced, Orange, and Los Angeles. And those are just the growth hotspots.

Below is another version of the Trajectories graph, but I clicked off the five longest-running warmspots. The snarl below, of the twenty remaining ones, shows how much has happened in the past month. For months and months, California sailed along in pretty good territory, with really only Imperial and Kings in trouble. But in the past month, there have been 20 new daily-cases warmspot counties. And as seen by the density at the left, there are a whole lot of counties that just entered warmspot territory in the past week or two:

Counties Growth: Still Growing

Almost all counties are still growing: see the rope chart below (as always go to the interactive version to see which line is which), with almost all lines ending above zero. This is about what California has looked like for a month. One of these weeks, we will surely see more of these lines turning back toward zero growth, but it’s not moving that way yet:

At 150 DCPM, California is not yet a daily-cases hotspot. And at 4.4% growth in daily cases, it is not yet a growth hotspot. But it is warm, on both counts. If things turn around in the next couple weeks, it may be able to avoid either or both labels; it is certainly not yet in the company of states like Arizona or Florida. But until it stops growing and accelerating, there’s no way to say when this will end for California.

Definitions

DCPM	Daily cases per million
Daily-cases warmspot	>100 DCPM but <200 DCPM
Daily-cases hotspot	>200 DCPM
Growth warmspot	>1% daily growth in new cases
Growth hotspot	>5% daily growth in new cases

This is summary of COVID-19 trends today, by countries, US states, and California counties.

Summary

	Daily Cases	Daily Growth
World	cold (22 DCPM)	warm (2.2% and rising)
United States	warm (168 DCPM)	hot (5.9% and rising)
California	warm (127 DCPM)	warm (3.3% and rising)

Graph of the Day

The Gang of Six (Alabama, Arkansas, Arizona, North and South Carolina, and Utah), trending toward recovery in growth rates.

Countries of the World

Interactive source graphs, where you can select regions, zoom, etc., are here.

India is Rising

I don’t usually focus on total cases, but India is worth at least a moment, because of the sheer number of cases. Though its DCPM numbers are still very low (17 DCPM), making it very far from a daily-cases warmspot, its total numbers are still very large due to its high population. India is solidly in third place worldwide in total cases, behind the United States and Brazil, and continues to grow in daily new cases. (See the interactive graphs to differentiate the smaller countries at the bottom of the graph).

Daily-cases Warmspots and Hotspots

About ten countries are either daily-cases hotspots or almost-hot and rising: Bahrain, Oman, Qatar, Panama, Brazil, Armenia, Kuwait, Chile, South Africa, and the United States. The epicenters, then, are the America, Middle East, and Africa; but only in the Americas is there a truly widespread outbreak. Africa has mostly been spared so far; the Middle East has mostly been spared so far; but each has a few major hotspots:

These same ten aren’t looking bad in growth, though. They’re roughly split around the zero line, with half of them growing in daily cases and half of them shrinking or steady. Most of them are seeing their growth rate dropping, indicating initial trajectory toward a recovery. Only two countries, United States and Kuwait, have all three red flags indicating full-fledged epidemic: (1) they are daily-cases warmspots, (2) they are growing in daily cases, and (3) they are accelerating.

The United States Still Isn’t Hot In Daily Cases But Is Headed There Fast

The United States, yellow line above, is arguably the worst in the world right now. Though it’s not yet a daily-cases hotspot, it is solidly a daily-cases warmspot (170 DCPM); and it is a growth hotspot at 5.9% growth in daily new cases; and its growth rate is still accelerating.

States of the United States

Interactive source graphs, where you can select regions, zoom, etc., are here.

The Gang of Six Have Turned Toward Recovery

The Gang of Six (Arizona, Arkansas, Alabama, North and South Carolina, and Utah), are no longer the big news; there are at least ten states doing worse than them now. But I’m putting them here at the top anyway because they provide more evidence for something important: even the worst states follow a predictable trajectory toward recovery. It’s been about three weeks since the inset graph for these six, which showed some very alarming numbers (Arizona was above 9% growth in daily new cases!; four of the six were at or above 5%! Four of them were accelerating rapidly!). But as we have seen in nearly all cases, it’s just a few weeks—about six, typically—from the time the state really starts to grow, until it peaks. Then it’s another six weeks or so until it is mostly recovered. The media alarms (dire headlines) start to go off perhaps two weeks into the rise, and it’s about four weeks from the time that everyone notices that, say, Arizona is in big trouble, until it peaks in daily cases. Now, Arizona hasn’t actually peaked yet, but it will peak when that blue line hits zero; and at this point we can clearly see that it’s headed in that direction. In one sense, Arizona and the other five have peaked already: in growth rate. Growth rate is clearly dropping for all except maybe Alabama, which is still slightly upward.

If we compare the trajectories-since-100-DCPM of the Gang of Six with a couple of fairly typical recovered states (New York and Michigan), it doesn’t immediately jump out that the gang of six are doing very well. All of them have been rising in new cases, and continue to rise in new cases. But at this point in the trajectories, the two other states (and many like them, not shown) were also rising. The rope chart above suggests that these six are starting to slow down, and it is entirely plausible that these six will peak and fall as the other recovered states have, and along roughly the same timelines. Arizona is exceptionally hard hit, and might take longer than the rest to recover; but my guess is, we’re going to see this same trajectory repeated for these six in the next couple months.

The Next Wave

As I mentioned above, the Gang of Six are not the problem. They are still getting a fair bit of press, and the trajectory graph above suggests some of why that it: they are all still rising. But their growth rates have dropped to less than 5%, and are falling, so at least they are visibly headed toward recovery in the “rope chart.” The same cannot be said of our new up-and-coming wave: Florida, Louisiana, Georgia, Texas, Nevada, Tennessee, Idaho, and Oklahoma. (California, by the way, sort of fits in the crack below this wave; it’s not quite bad enough to qualify for this wave). All these have the triple-whammy of being daily-cases warmspots, being growth hotspots, and having rapidly accelerating growth. You might think these were headed for a complete catastrophe—but remember that the Gang of Six were right here a month or so ago, and started to show signs of recovery in a few weeks, and are looking very different, and much better, now. I think it likely that these, too, will start to turn downward in a month or so.

A cautionary note. Let me take a moment to speculate (reiterate) about the reason hard-hit regions so far have recovered. It’s probably not just something that naturally happens to defeat the virus without any steps taken by the populace. I think it happens because the media and government notice that a region is in terrible trouble, and they start to inform an warn the populace, and to put lockdowns and restrictions in pace. The populace, with or without government mandates, starts to see the news reports about how bad it is right there where they live. They also start to see some of their friends sicken and die, and are driven by very-much-justified fear to take their own extreme precautions. And that is why the epidemic wanes: because of the behavioral changes of the populace. So please don’t read this as anything like permission to do nothing and wait for it to blow over in your hard-hit region: I believe it blows over because the residents take extreme measures for a few months.

The Next, Next Wave

I get the distinct sense that there is another wave now coming, not yet very strong but likely to follow this current one (which in turn followed the Gang of Six, which followed Michigan/Illinois/Louisiana, which followed the East Coast). Delaware, Iowa, Oklahoma, Kansas, New Mexico, Wisconsin, Washington, Ohio: these are all showing as growth hotspot or growth warmspots, even though they don’t yet have that many cases (California is somewhere in there too, though still somewhat tepid and unremarkable, without all that many cases or all that hot a growth rate).

History suggests that these states will get many cases soon, through the mechanism of being a growth hotspot. And then, like the other waves, it will pass in a couple months. And that might be it. Because at that point, there are precious few states left still to catch it (Kentucky? Hawaii?). Maybe where will see a few final, smaller waves through the remaining few states. But eventually, every state will have had it, and will have recovered, and the epidemic will be out of fuel. If my theory about the “metaphorical immune system” of regions is correct, few if any of states will get it again badly, for a long time; or if a vaccine is developed, maybe forever.

California Counties

Interactive source graphs, where you can select regions, zoom, etc., are here

The Counties Are a Mess

The graph below is very messy because the state of California is very messy. If you want to pick out a particular line, use the interactive version. It used to be that the counties were split nicely into certain leaders and a bunch of less-hit noise below. No longer: we have an even spread of the counties across the range of DCPM here, with no clear clumping except our usual off-the-charts Imperial. One consistent feature is that they are all still growing. There are so many daily-cases warmspots right now that it’s not even worth listing them; let’s just say half of the counties are warm. Even for daily-cases hotspots, there are a lot: Imperial, Tulare, Riverside, Kings, San Joaquin, Los Angeles, Merced, Fresno. As always, it is a big problem that Los Angeles is a daily-cases hotspot: that’s a about half of the state’s daily cases right there:

The rope chart for California is looking mostly has it has for weeks now: almost all counties are growth warmspots. Many of them are growth hotspots: Orange, Sutter, Yolo, San Diego, Madera, Fresno, Merced, Butte, and Napa are all growth hotspots with daily cases growth over 5%. Some of those are pretty big counties too, so this is not going in a great direction. There may be just the touch of a turn toward recovery in that we have at least four counties now trending downward in growth rates: Imperial and Marin (both actually shinking in daily cases, after having a hard time of it a while ago), San Benito, and Santa Barbara. If these counties do anything like the Gang of Six at the state level, we’ll see more and more of them bend downward in the next couple weeks; but California hasn’t followed the trends of other states very well so far, so that’s even more speculative than usual.

Definitions

DCPM	Daily cases per million
Daily-cases warmspot	>100 DCPM but <200 DCPM
Daily-cases hotspot	>200 DCPM
Growth warmspot	>1% daily growth in new cases
Growth hotspot	>5% daily growth in new cases

This is summary of COVID-19 trends today, by countries, US states, and California counties.

Summary

	Daily Cases	Daily Growth
World	cold (27 DCPM)	warm (2.2% and warming)
United States	warm (145 DCPM)	hot (5.9% and warming)
California	warm (115 DCPM)	warm (3.5% and warming)

Graph of the Day

This State Trajectories graph shows that a typical trajectory of COVID-19 in a state is that after it becomes a daily cases warmspot, it rises for about 30 days, peaks, and then falls below into cool territory in about another 30 days; and it does not return to warm again (no second waves, at least so far).

Countries of the World

Interactive source graphs, where you can select regions, zoom, etc., are here.

Country Hotspots

The only daily-cases hotspots are Barhain, Oman, Qatar, Panama, and Chile, Armenia, and Brazil. Brazil and Armenia have recently joined this list, and Oman has just passed Qatar, which is declining. Other than Brazil, then, there are no large countries in the world which are daily-cases hotspots—and Brazil is just barely a hotspot at 200 DCPM. Seven other countries are daily-cases warmspots, including the United States, solidly warm at 145 DCPM.

All of the daily-cases hotspot countries except Brazil are improving. Three of them (Armenia, Qatar, and Chile) are below zero in the following graph, indicating that they are dropping in new cases each day; these three are also accelerating downward. The other three are still rising, but their growth rate is falling. Brazil alone is rising in new cases, and also slightly accelerating. That recovery trend in Brazil, was short-lived, and never really took hold.

The United States Is Warm and Getting Warmer

The United States become a daily-cases warmspot around June 20, and is now at 145 DCPM. This is not yet a huge number of cases, after adjusting for population (it is not yet a daily-cases hotspot), but it continues to rise:

Indeed, the rope chart shows a very high 5.9% growth in daily cases, making the United States a growth hotspot. Furthermore, this growth rate continues to climb:

States of the United States

Interactive source graphs, where you can select regions, zoom, etc., are here.

The Gang of Six Is Doing Even Better

The Gang of Six (in order of the graph below, from the top, South Carolina, Arizona, Arkansas, Alabama, Utah, and North Carolina), are no longer the biggest news, but I’m going to emphasize them to point out how state trajectories typically go. Here is their growth graph, which is looking better and better; four of them have now peeled away from the main group and two are pointed headed downward (Utah and North Carolina). The other four are still at least slightly upward, but the graphs are visibly bending toward a peak. Once they peak, they will begin pointing downward and heading to zero; once they reach zero, they will start seeing fewer and fewer new cases each day. North Carolina, the bottom line, will take something like six weeks to peak at this slope; but as the slope has been pulling downward, it is likely that its actual peak will be sooner than that.

States Have Been Following A Predictable Trajectory

The following graph shows the trajectory of all states that have ever been daily-cases warmspots; all trajectories are overlaid starting on the day they first reached 100 DCPM. The pattern here is that once a state reaches 100 DCPM, it continues to rise for a few weeks until its peak, then it peaks and begins to fall, and drops back below 100 DCPM a couple months from when it first hit 100. After completing its first wave and dropping below 100 DCPM, it does not see a second wave. There are two exceptions to this “no second wave” rule: Louisiana (purple at the right) is hot again now after having recovered, and Delaware (red at the right) is barely warm now after having recovered, so there may be second waves coming in some other states as well. But so far, that has seldom happened. If second waves really are typically easily handled by the metaphorical immune system of the region, built up during the first wave, then once all states have seen a first wave, there will be no more fuel for the epidemic, and the United States will begin to decline in total cases again. If Louisiana and Delaware are bellwethers of what other states will do, however, then other states will begin seeing a second wave about 90 days after their first wave. Which one it is, will become clear in the next couple months.

The Gang of Six Has a Lot of Competition

The graph below shows the growth rates of all the growth hotspot states, among the top-25 warmest states in DCPM. None of these are member of the Gang of Six; and all of them are hotter in growth than even the worst of the Gang of Six. So as the Gang of Six begins to peak toward recovery, these are the new growth hotspots: 5% to 10% daily growth in new cases. From the top, they are Idaho (10%!), Florida (9%!), Nevada, Kansas, Delaware, Georgia, Guam, Texas, Tennessee, Louisiana, and Oklahoma. At a 5% daily growth rate, the number of daily new cases doubles every two weeks; at a 10% daily growth rate, the number of daily new cases doubles every week. These are alarming growth rates:

California Counties

Interactive source graphs, where you can select regions, zoom, etc., are here

Daily-Cases Hotspots: Imperial, Kings, Tulare, Riverside, San Joaquin, Los Angeles

Just like last time, these six counties are the only daily-cases hotspots in the state. There has been a little shuffling around the cluster at 200—Tulare and Riverside have just passed Kings, which drops to 4th place—but Imperial is still the big one in DCPM. Given the enormous size of Los Angeles, however, being a daily-cases hotspot is a very big deal in terms of the number of total cases:

There’s no news in the growth rates either; as it has been for many days, almost all California counties are growing in new cases, and accelerating. Imperial alone among the top-25 daily-cases warmest, is shrinking:

I do have a Counties version of the trajectories chart, below. But since California wasn’t even warm for a long time, there are no counties that have yet reached the “60 days since warm” mark that tends to be the recovery point for states. Kings, purple, is showing a trajectory similar to a typical state. Imperial looks like its curve, though broadly falling, will probably not make it back to 100 DCPM at the 60–70 day mark, so it appears to have an unusually long trajectory to recovery. And at the bottom, we can see Los Angeles, Tulare, and Santa Barbara, all pushing the daily-cases hotspot line, after a month or two in warm (the rightmost point of each line, is the point for today):

Definitions

DCPM	Daily cases per million
Daily-cases warmspot	>100 DCPM but <200 DCPM
Daily-cases hotspot	>200 DCPM
Growth warmspot	>1% daily growth in new cases
Growth hotspot	>5% daily growth in new cases