Suggestion for improvements

[kklepper]

Data is great, but it is hard to make sense of it. Some ideas to get a grip of what they mean...

Long lists are misleading and of little use. You want to compare only a small number of countries, for example

• corona switzerland sweden

You might allow wildcards for countries you don't know the name of, e.g.

• corona %orea%

It is inconvenient and impractical to write country names. Use ISO codes instead or better alternatively

• corona be ch de se uk us

To give you an idea of how these numbers can be used to provide insight, here are some examples created with SQL. All results shown are based on https://data.europa.eu/euodp/de/data/dataset/covid-19-coronavirus-data from 2020-04-17.

A very important ratio is missing: cases and deaths per million or 10,000 inhabitants

• Here you see at first glance that countries with fewer inhabitants may be much more afflicted (deathsPer1mCapita) although the number of cases in other states (like US) is much higher as this result is ordered by that ratio.

It would be instructive to normalize this data with respect to a single country to get a feeling which country does better, e.g. with respect to de (having ratio 1.00)

• You see here that Belgium has 9 times as many deaths per 1 million inhabitants than Germany. That number has more meaning than the ratio alone.

Next the ratio of cases and deaths is interesting. Well, we know that death numbers are not reliable (Corona may be interpreted as pneumonia) and case numbers are not reliable either (which people are tested, only those with severe symptoms or all infected?). But you get an idea nevertheless.

• With the death-cases-ratio you see which country scores better (higher numbers are better here). At that time, Turkey did better than Germany and UK worse than Spain. Switzerland isn't much better than the US, and India follows directly behind.

Also, you see that for countries with a high population (especially India, China) the absolute numbers aren't helpful at all.

Next it would be interesting to see these ratios develop over time. This would give a clue whether the measures of different countries are working.

[kklepper]

Some more instructive results, same query,

• all countries except ship incident,
• different order,
• data from https://data.europa.eu/euodp/de/data/dataset/covid-19-coronavirus-data as of 2020-05-01 23:20,
• limited to the first 20 results each: