VisLies 2021 Gallery

Of course, what ultimately matters is not the year-to-year performance but the overall return on investment. To know how much an investment earns, you need to calculate the cumulative gains (or loss), which once again is best represented by a traditional x-y plot.

As we can see here, the asset allocation is not particularly special. The US equities were better performing without the volatility of investments like emerging markets. (Reminder: This is not financial advice. This is an example of data display, not claims about any particular investment.)

So, as can be seen, the asset allocation quilt makes a story that is easy to see with a proper x-y plot more complicated than it needs to be.

At Least it Looks Green

Just before the VisLies event Ben Simons encountered this unhelpful information about the Australian plan to reach zero emissions. For context, this visualization was used during this news report proceeding Scott Morrison’s (Australian prime minister) begrudging attendance of the Glasgow COP26 conference. What we are seeing is an overview of the plan to reduce Australia’s carbon emissions. The display is just a one-row table, but gosh it looks pretty.

Despite its simplicity, this graphic is difficult to interpret. If you add up the numbers, you get an emissions reduction of… up to 110%?

As ambitious of a goal that is, Australia is probably not planning to address more than all of its carbon emissions. Note that the first entry is “Reductions to date” (where somehow “today” is one year before the report was released). It’s likely that the other reductions are in proportion to the carbon emissions of today, so we can discount the initial “-20%” of carbon emission reductions from today.

But that raises an interesting question of how to interpret the remainder of the numbers. Are those percentages with respect to the carbon emissions of today or of the projected emissions at the time they are implemented. For example, for the -15% emissions labeled “Global tech trends,” is that 15% off the emissions of today, or 15% off what is left after the 40% from technology investments (that is, 15% of 60% of what we have today)? If the reductions are applied sequentially, the overall reduction can be as low as 61%.

Ben also notes that much of the plan relies on happenstance technology improvements that are innovated by others (“Global tech trends”) and future technology we assume will happen (“New tech”). It may be the case that the “Offsets” category is placed where it is based off technology improvements that we assume will happen.

In fairness, this graphic was created in response to a plan that was criticized for being vacuous on substance and was, in retrospect, never intended to be implemented.

Not as Cool as you Think

David Borland just got a new air conditioner. With it came a fancy new thermostat that records usage and comes with an app that allows you to review the energy consumption.

But David noticed something odd about the reports he was getting. In each bar chart is a dotted line that would seem to suggest some sort of trend or average. But in examples like this, that cannot be the case as the dotted line is clearly lower than all the bars. Such a value cannot, by definition, be an average.

Looking deeper into the reports, David found these details that seem to explain how this line came about. The numbers here correspond to the bar chart shown. The total value does seem to be the sum of the 3 bars. The “average” is the total divided by 7, the number of days in a week. That makes sense for data on a week that has already passed, but it is utter nonsense to consider days that have not yet happened for the average. The supposed average is just a useless item of non-information that is best ignored.

Less Energy, Smaller Home

Laura Tateosian recently received this summary as part of her energy usage report. The text is clear enough; it compares her energy use with that of an “efficient home.” But the text is accompanied with a couple of icons in the shapes of houses that represent the two numbers being provided, and these icons make little sense.

The icons are different sizes, and the smaller size does indeed represent the smaller number. But the relationship to the numbers end there. The smaller number is about 40% that of the larger number, so you would think that the sizes of the icons would be somehow proportional. The top of the smaller house is more than 40% than the top of the larger house. Likewise, the proportions of the height of the eaves is wrong. As far as we can tell, there appears to be no logical measurement of the icons to match the proportions of the numbers.

The icons are clearly meant to show the relative values of the numbers, but in no way do they actually do that.

Stately Elements

David Pugmire presented this Venn diagram comparing the abbreviations of known chemical elements with the postal codes for U.S. states. The diagram shows that of these two groups of things have 17 common items.

This commonality probably means… nothing. What does the U.S. state of North Dakota have to do with the element Neodymium? Just that they both happen to have the same abbreviation. Their names are not even that similar.

A simple internet search reveals a number of versions of this diagram. Likely, all of these diagrams are made in jest. After all, there is no real relationship between chemical elements and U.S. states.

Except that last statement is not true. Many elements are named after places, and two elements are specifically named after states: Californium and Tennessine are named after the U.S. states California and Tennessee, respectively. So, if we think about it, there are actually 2 elements that have a real overlap with states. These images are unfortunately hiding interesting features about the history and discovery of these elements behind a banal observation that 2 letter abbreviations sometimes overlap.

Ironically, neither of these element/state pairs are actually represented in the overlapped region of the Venn diagram. The Californium abbreviation, Cf, does not match the California postal code, CA. Likewise, the Tennessine abbreviation, Ts, does not match the Tennessee postal code, TN. So, the Venn diagram fails even in this simple respect.

Highlighting Sparsity

A problem with choropleth maps that has been discussed previously at VisLies is that the proportions of the visible elements often do not properly reflect the proportions of the data it is representing. The main problem is that data is marked on choropleth maps per their area whereas the actual data relate to some other property, such as population. Thus, certain areas, such as rural areas, get over represented.

Nathan Woods highlighted how imbalanced a distribution can be with this choropleth of the U.S. This image compares the population of Los Angeles County, the most populous county in the U.S., with the least populous counties. The map shows just how disparate the population across the U.S. is.

However, it should be noted that this map is an exaggeration, and to use it in pretty much any other context is probably a lie. Nathan pointed out some problematic elements of this map. A good portion of the western U.S. is marked red. But notice that some regions in the middle are conspicuously missing. The map is carving out all the bigger cities like Denver and Salt Lake City. This is clearly inflating the rural area while dividing up the geography in a way that makes it impractical for other comparisons.

Nathan also noted something else interesting about this population map. It looks suspiciously similar to the map of federally owned land shown here. The U.S. federal government owns land for a variety of reasons including parks, preserves, and defense. But one thing they all have in common is that federal land is closed to private ownership or occupation. So, it should be no surprise that counties with federal land have smaller populations. This is clearly reflected in the large amount of federal land in the western part of the country. But it can also be seen in several places on the eastern side including the north part of Minnesota, Wisconsin, and Michigan as well as near the border of Virginia and West Virginia.

This population map is best used as a cautionary tail to avoid making misleading observations. Unfortunately, this very map can be used to make those same misleading observations. For example, it has been used to argue for the availability of land for a growing population. However, just because an area is not densely populated does not mean it can sustain more people. As we just saw, much of the “available” land is really federally owned and cannot be used. Even if these restrictions are lifted, it is not the case that people can simply move into these regions. Much of these regions are mountains, desert, or otherwise impractical for sustaining large populations.

This population map is also used to demonstrate inequity in government representation (particularly with the U.S. legislature). This map is not appropriate for making any such claims. The outline of the regions are in no way representative of how any U.S. government body is selected, and, because of the way the map is draw, the implication is a gross exaggeration.