I’ve been keeping an eye on the COVID race for a few months now. The players I watch are all standout players: FL, TX, PA, and GA. Those are the places that have people that I know, so I watch their progress.
I remember when FL was the star, I seem to recall it was in the top 5 for a while. But PA put forth a massive effort and shot right up the charts. TX was a slow starter, but it’s been doing pretty well lately. GA has always been mediocre, which I suppose is a good thing, honestly. But FL is recently finding its mojo and is climbing in rank again. Go, FL! Obviously, no one is going to take the crown from NY, and NJ is probably always going to be second to NY (in everything), so there’s only so far you can go.
For three of my players, I watch their personal progress dashboards. Two of them, FL and TX, use the same software, so it kind of gives some equal comparison of the numbers. But in both cases, they use graphs that are misleading. Well, they aren’t if you understand the data, but for casual observers and those that don’t want to put the minimal effort into understanding, the response could be either, "this isn’t so bad", or "this is great" when the reality is neither of those sentiments.
Take a graph from FLs dashboard:
Wow, that’s impressive. Deaths are falling, and dramatically at that! This is all behind us, let’s go party!
But there is a small disclaimer below the chart, for those that care to read: "Death data often has significant delays in reporting…". That means that those low numbers in the near term are low because there’s no data yet. Those numbers will rise as time goes on, but that’s just fine, because there will be newer, lower numbers to report as time goes on as well.
Here’s a graph from TX’s dashboard.
This is just dumb design: plotting two values, one that will constantly increase, and one that will remain relatively constant on the same scale. This will have two effects. First, the number of deaths per day (in blue) looks like a really small value. And comparing 20 to 1,698 does make 20 seem very small. But as the total number grows, and it will, every day, the scale is going to eventually have to be adjusted, and the daily value is going to be insignificant.
TX does the same charting with the number of cases, with the same effects.
Maybe its incompetency that made these charts, but in the current political climate, and judging that these two states have pushed very hard to justify their reopening plans, it might not be a stretch to think this is just propaganda. The numbers don’t lie, they’re not telling you anything false. It’s just being presented in a way that looks most favorable.
Comments are closed.