Peter Walker Profile picture
Head of Insights @Cartainc | prev: @COVID19Tracking, @PublicRelay, @CovidWh Using data visualization to illuminate our world
Aviva Gabriel Profile picture David Robinson Profile picture 2 subscribed
Apr 23 12 tweets 4 min read
Big thread on startup fundraising on SAFEs and Convertible Notes - all data from before any priced rounds.

Pre-seed, "seed on SAFEs", etc.

Data from more than 100,000 SAFEs signed by Carta companies since 2020. Lets go Image Cash into pre-priced rounds stayed fairly steady in Q1, with the estimated final deal counts basically equal to Q4 2023 Image
Apr 19, 2021 6 tweets 2 min read
Optimistic COVID 🧵 - I think we may be rounding towards the mini summit of this final hump.

Cases by region here since Jan 1, 2021. Midwest seems to have turned for real, only real remaining worry is the South (but vaccinations continue to grow). 1/n Michigan, the recent hotspot - is that a top? Maybe a touch early to say for sure, but certainly the exponential growth is no longer there. 2/n
Mar 29, 2021 5 tweets 2 min read
Why @CDCDirector is leery of reopening across the country - brief 🧵 on current state of COVID metrics.

Nationally, cases are rising off of recent lows and that is matched by an uptick in hospital admissions (more of whom are younger, unvaccinated persons than before) 1/5 Clearly the case rise is happening in the Northeast and Midwest. 2/5
Dec 27, 2020 5 tweets 2 min read
Little reminder🧵on the holiday effect in COVID data.

Major holidays have had significant effects on reported COVID data this year. Below you can see how COVID cases have responded to holidays with sharp declines in almost every instance. 1/ So why is that? A myriad of reasons. Some state don't report anything on a holiday, as their staff is given a well-deserved day off. People may be less likely to get a test on a holiday. Some states report a fraction of their known cases...it's messy. 2/
Oct 4, 2020 4 tweets 3 min read
Final update 🧵 for the evening. Reminder that the initial tweet from the President was at 9:54pm PT on Thursday.

Since then, we've added 216 contacts to our tracker. 18 positive, 60 negative, 128 unknown.

Here's a breakdown of the positive cases thus far 1/n When viewed on an event basis, the SCOTUS Rose Garden is the most impactful. 47 contacts so far and there are many more unknowns.

We'd like to include all possible contacts - staff, essential workers, etc deserve the same level of concern here 2/n
Sep 13, 2020 8 tweets 4 min read
Rounding out another week of #COVID data - might be a good time to do a deeper dive into the numbers.

Optimistic 🧵 ahead...

First up - it was a good week. Holiday lag may have had some impact, but cases, hospitalizations, and deaths all fell from a week prior. 1/ Image In our previous major hotspots, cases per million people have seen precipitous falls. Testing, while not at peak levels, has not fallen as much as cases (especially outside of Florida) 2/ Image
Aug 7, 2020 4 tweets 2 min read
Short thread on (what feels like continual) HHS
#COVID hospitalization issues.

Changeover to HHS happened 3+ weeks ago. So the data should be stable now, right?

Not even close. 1/ A select few states to illustrate the point. All numbers are currently hospitalized ppl per 100k population:

8/5 8/6
AR: 54 ?? (not even included)
FL: 46 31
TX: 34 15
AL: 41 8
SC: 30 3

They're all like that (see @JasonSalemi for full list) 2/
Jul 28, 2020 4 tweets 2 min read
For those interested in how the HHS directive has changed our ability to get accurate figures for the currently hospitalized patients in the US: covidtracking.com/blog/whats-goi…

A short summary thread - please read the below for more detail! 1/ As you see in here, HHS is reporting a higher number of currently hospitalized patients than the states. So...why?

A few reasons:
- The switch to HHS was urgent and had to take place on legacy systems. Some hospitals may not be able to report to both state and HHS every day. 2/
Jun 14, 2020 4 tweets 3 min read
Brief thread re: reopening and the impact on new cases.

Here's the full USA. Weekly #COVID tests in bars, % of those tests that were positive in circles.

Testing up, % positive down...this is a good chart. 1/

@NateSilver538 @JamesSurowiecki Image Check in our original epicenter region, the Northeast. This includes: CT, MA, ME, NH, NJ, NY, PA, RI, and VT.

Wild that in late March, almost 40% of tests came back positive.

But this week only 1.9% did so.

Gradual reopening makes sense here 2/
@yayitsrob @DemFromCT Image
May 31, 2020 4 tweets 2 min read
Thread on Minnesota's #COVID metrics. If you are out protesting, please wear a mask / bring hand sanitizer / consider quarantining after

These demonstrations are essential for our democracy and almost certain to worsen the pandemic statewide. It's absolutely tragic. Cases 1/ Image Minnesota % positive 2/ Image
May 29, 2020 5 tweets 3 min read
The highlights from @COVID19Tracking today:

% positive (7-day avg) drops as testing hits a new high.

@JohnJHarwood @DemFromCT 1/ Image Test by week slightly exceeded last week (some real down days in testing from Memorial Day).

@AndyBiotech 2/ Image
May 16, 2020 4 tweets 4 min read
1/ #COVID hospitalization by state over time. Data by @COVID19Tracking

This reflects the # of ppl with COVID-19 who were in the hospital on that day, 4/15 - 5/15.

East first - Good trends! That NY number is still stunning, but dropping.

NH rising, however. @joshrobin Image 2/ #COVID hospitalization over time, South and Southwest. See first tweet in thread for data explanation.

Rising trend: AZ, KY, VA, & NM. Also a lack of data in many states. Is reporting a "currently hospitalized" number harder than I know?

@NateSilver538 @Noahpinion Image
May 14, 2020 9 tweets 5 min read
GA #COVID trends since reopen (dashed lines below).

The takeaway - things are going in the right direction.
% positive continues to drop as testing ramps up!

Let's hope other reopened states have data that follows this pattern. @NateSilver538 @GovKemp

covidcharts.tech/home Image Note - if you'd like to see your state in this format, give me a shout.
May 3, 2020 4 tweets 3 min read
% positive - what is it?

So testing is going up - this is a good thing! But we keep finding as many or more positive cases - is this bad?

Watch the % positive. 21% of tests were positive earlier this month. Now 12%. 1/
@CarlBialik @justin_hart @aginnt
covidcharts.tech/overlay Image But this drop is not universal across states. Iowa, for instance, is showing a real worrying trend. Tests are going up, but the % positive is rising alongside them. Not good.

covidcharts.tech/overlay 2/ Image
May 1, 2020 8 tweets 9 min read
May 1st - good moment to look back at the #COVID crisis data thru April. All data in the thread from @COVID19Tracking (follow them!)

Over 1 million positive cases found, 6 million tests.

@alexismadrigal @kissane @hackingdata @juledurg @mbsimon

covidcharts.tech/home 1/ Image New York has borne the burnt this in the US, accounting for almost a third of all deaths in April. But the overall growth month to month is stunning. @justin_hart @aginnt

covidcharts.tech/NYfocus 2/ Image
Apr 29, 2020 4 tweets 4 min read
Recap of today's @COVID19Tracking numbers: tests keep rising, but today marked a new high in deaths as well.

Tests: every state save for AZ has now tested at least 1% of its population. Sorry, Arizonans (ians?) @azcentral

see: covidcharts.tech/tests 1/ Image Positive cases: nationally, the % of positive tests is slowing declining (as we would hope would happen as tests increase). That figure is still rising in certain states like DE and PA, however.

see: covidcharts.tech/positives 2/ Image
Apr 25, 2020 5 tweets 6 min read
It's the weekend, so here's a new way to look at our #COVID19 data. Thanks to @AstorAaron for the idea.

As testing (hopefully) ramps up, the metric to track is:
% positive tests.

The US has been hovering around 20% in April.

Interactive Data: tabsoft.co/3eSmwqk 1/ Image In NY, the % of positive tests has begun to fall as more testing occurs. That's what we want! Obviously it's still really high, but the trend matters. @NateSilver538 has been on this for awhile. @nytimes @Noahpinion

Interactive: tabsoft.co/3eSmwqk 2/ Image
Apr 24, 2020 4 tweets 4 min read
The highlights from @COVID19Tracking's #COVID19 numbers on Apr. 24. Testing continued to grow nationwide, albeit not as quickly as we'd hope. Large increases in tests from yesterday in FL, IL, and TX.
@aginnt are those your tests? :) 1/

Interactive Map: tabsoft.co/2VAbOwT Image Hotspots - some definite positive news. Clear declines in positives cases in NY and MI, with a flatter trend in NJ. NY deaths fell to their lowest level since Apr. 2nd (...seems like a lifetime ago). Small victory, but something. 2/

Interactive: tabsoft.co/2VDSS0t Image
Apr 22, 2020 4 tweets 5 min read
Built a tool today for all wondering "How is the #COVID19 outbreak in my state?".

Data from @COVID19Tracking (and @kevin for RT work). Interactive data: tabsoft.co/3btKLJ7
@JeremyBWhite

Here's CA - Deaths are rising, positives leveled off. But 46th in tests/million? 1/ Image A better story in Michigan. Deaths have slowed, and positive cases have really turned. Now the focus should be ramping tests greatly. @GovWhitmer deserves some credit! Interactive: tabsoft.co/3btKLJ7 2/

@NateSilver538 @mikeyk Image
Apr 19, 2020 5 tweets 6 min read
Taking a look at the effective reproductive rate (Rt) of the virus in clusters of states. Data here: tinyurl.com/yde9fuwc. Indebted to @kevin and @mikeyk for the Rt work, @COVID19Tracking for data collection. Here's all on 4/18. @politicalmath @sarahfrier @emilychangtv 1/ Image First up - the West (CA, OR, WA). Trends down with a slightly worrying rise in WA. There have (obviously) been no change to shelter-in-place orders in these states - so why the uptick in Rt? Perhaps just a lag in positive case reporting? @justin_hart @aginnt @latimes 2/ Image
Apr 18, 2020 5 tweets 5 min read
Is #coronavirus replicating strongly in your state or dying out? Great work from @kevin producing a model of effective Rt (how many more people the virus will infect on average per positive case). Above 1 is bad, below 1 is good! I built on that with a map and over time view. 1/ Image As an example, here's Michigan's effective Rt over time. Peaks in early March around 3 and has been hovering around 1 for days now. Rt probably does lag case growth but the @COVID19Tracking data is the best we have 2/ Image