As many longtime readers know, I lived in Chicago for eight years. I probably had Chicago-style pizza fewer than eight times in my life. I grew up in the Philadelphia suburbs and for the last nine years I have lived in central Philadelphia, where pizza is very much a different thing. And in my life I have visited New York probably more times than I can count and obviously pizza is a culture of its own there.
The other day, xkcd posted about the various different regional styles of pizza.
Now I will not venture to throw my two cents in here on which pizza is the best. Rather, each has their own qualities that make them good fits for specific types of occasions.
What I will say is the mouseover state makes mention of Altoona-style pizza. As I am someone who visits the Altoona regional annually—and will be again next spring—I have indeed heard of the style. But, no, I have never tried it. Nor does it sound particularly appealing.
However, I had also heard of a Primanti Brothers sandwich of Pittsburgh fame, which includes French fries inside the actual sandwich. A few years ago when I visited Pittsburgh for the first time I did give one a go. But, and perhaps because I am not a fan of French fries, I found the sandwich disappointing.
So maybe some year I will give the Altoona-style pizza a similar go.
But right now, a slice of pizza, any other kind of pizza, sounds pretty good right about now.
One of the most debated questions one could ask at pub trivia: How many countries are there in the world?
To start, the question cannot be answered completely. What is a country? What is a state? What is a nation? Define recognition. Whose definition?
When I worked at Euromonitor International I had to edit a map for a Moroccan client because the maps and visualisations I designed showed Western Sahara as a disputed and occupied territory along with places like—surprise, surprise—the Gaza Strip and the West Bank, Kashmir, and in later years Abyei. The client was upset because Morocco occupies the majority of Western Sahara, particularly the coast and areas containing the area’s natural resources, and claims it to be an integral part of the kingdom. But the Polisario Front occupies the remainder of Western Sahara and claims the whole territory for its people. Some countries, particularly in Africa, recognise Western Sahara as independent—it belongs as a full member of the African Union. The rest of the world largely refuses to recognise Western Sahara as independent whilst also refusing to recognise Morocco’s occupation. Hence my maps.
You can see what I mean if you look closely at this un-Moroccanised map my team made. I have added some orange boxes to highlight Western Sahara along with the West Bank and Gaza, Abyei, and Kashmir. (Some geographies were simply not covered by the company and remained blank. Here the most notable example is Greenland.)
Do you count Western Sahara as a country?
At your pub trivia this week, will you count Palestine as a country?
This past weekend, the United Kingdom, Canada, Australia, and Portugal formally recognised Palestine statehood. They were followed this week by France, Belgium, Luxembourg, Andorra, San Marino, and Malta. (Maybe.) Notably, but not surprisingly, Israel and the United States refuse to follow suit.
My first draft of this post then continued with a history of the Israel–Palestine conflict from the most recent/today’s war to the formation of Israel to the Ottoman Empire and ancient history. But I wanted this post to focus on these graphics posted by the BBC. The first comes from earlier this week, prior to said announcements. The second from later.
I like this graphic. The designer chose to use circles as abstractions of countries rather than a more traditional choropleth map coloured by who recognises whom. In the context of an international conflict over geographic borders and statehood, removing the concept of borders shows a subtle attention to the nuances of the debate.
Critically, the yellow may be a touch light vís-a-vís the grey, but it clearly stands in stark contrast to the purple. I should note I saved this graphic from earlier this week before the official announcements of France et al. Today all that yellow is solidly purple.
As you can see in this second graphic, most of the yellows have become purples. Any of concerns over the lightness of the yellow faded.
Earlier I had noted how Belgium and Andorra were part of President Macron’s announcement, however this graphic shows them still awaiting to announce. I wonder if the discrepancy will be resolved when the representatives of the those countries speak at the United Nations. In essence, they have pledged to announce their recognition, but will do so formally in their address to happen later.
Once again, graphics need not be overly complicated or complex to clearly explain a point to the audience. You can easily see most of Africa, the Americas, Asia, and the Middle East recognises Palestine. Europe and Oceania remain the outliers at a regional level. But the use of a black stroke to outline the grey dot of the United States makes clear one point. Whilst the use of dots abstracted countries to discrete data points equal in theory, in practice not all dots are equal.
Credit for the piece goes to the BBC graphics department.
Last month Vox published an article about the trend in America wherein people are drinking less alcohol. They cited a Gallup poll conducted since 1939 and which reported only 54% of Americans reported partaking in America’s national tipple—except for that brief dalliance with Prohibition—making this the least-drinking society since, well, at least 1939.
Vox charted the data in the following graphic.
Overall, the graphic is good. Here the use of individual dots makes clear the years wherein Gallup conducted the survey. In some of the earlier periods it was not an annual question. The line weight is just thick enough to be distinct and the axis lines are lighter and in a light grey to create contrast.
The line uses the colour black. You really do not need colour in a line chart—or a bar chart or anything really charting only a single variable—unless it stems from your branding. Admittedly, Vox is now membership-supported and I am not a member so I only can read a few articles and am unsure if their chart brand standards now use black as their primary colour.
I can quibble with the axis label, “Drink alcoholic beverages”, because that should or could be included in the graphic header or sub-header. But putting in the graphic space is fine. And I like it next to the line instead of in a legend above the chart.
But the thing that irks me is the use of data labels at specific years. You could argue for the inclusion of the label at the current year, the 54%. You could argue for the maximum value, the 71%, but I would not.
Labels distract the reader’s eyes from the line itself and the line is the story. Calling out the 54% and maybe the 71%, fine, but the random 67%? A second random 67%? Those are unnecessary distractions that take away from the chart’s communicative value.
I have mentioned it before and I will mention it again, the use of excessive data labels plagues data visualisation these days. I need to write it up in a longer piece and someday I will. Here is a crude mockup of the graphic without the data labels.
The line’s pattern is easier to spot—fewer distractions for the eyes. The pattern is clear that since the late 1970s roughly 65% of Americans drank alcohol, with the occasional dip, including one in the late 1980s and early 1990s and a briefer one in the mid-90s. I wonder if the 1989–90 dip relates to the recession. The sudden dip in the mid-90s confounds me. But the point is these things are easier to spot without labels the sparkling distraction of the labelling.
Again, overall, the graphic is good. And these days the state of information design and data visualisation is…not great, Bob. So I do not want to critique this graphic too heavily. But a tweak or two would make it even better.
Another week has ended. Another weekend begins. I saw this a little while ago and flagged it for myself to share on a just for fun Friday. And since I posted a few things this week I figured I would share this as well.
It is remarkable it took the similarity of the two coastlines to “discover” continental drift, but that is, very broadly, the truth of the matter.
Last month the Philadelphia Inquirer published an article examining the geographic distribution of Teslas and Cybertrucks and whether or not your car is liberal or conservative. The interactive graphics focused more on a sortable table, which allowed you to find your vehicle type.
The sortable list offers users option by brand and body type—not model. But whilst the newspaper claims to include 11 million vehicles in Pennsylvania, its classifications are lacking. For example, I drive a Ford Focus hatchback, but the list includes only 1400 hatchbacks, of the Toyota four-door variety in the entire Commonwealth.
More interesting to me was the geographic map of Tesla electric vehicles and Cybertrucks. The screenshot below is of the Tesla.
The choropleth maps use shades of green to indicate the share of vehicles in the particular zip code. Teslas represent relatively small shares of the market across Pennsylvania, but their limited presence is concentrated in the more affluent communities surrounding the big cities of Philadelphia and Pittsburgh. Cybertrucks are even less popular in Pennsylvania and barely register statewide.
Overall the graphics work, but I wish the data was a bit more robust. I would be curious how political my vehicle choice is. Of course choice is relative here. I purchased my car back in 2010 and I cared little for the brand. I just wanted something small to drive and park in Chicago and that could lug cargo. Of course 2010 was a less polarised time when the outrages of the day centred upon tan suits.
Credit for the piece goes to Lizzie Mulvey and Dylan Purcell.
Conflict—a brutal civil war—continues unabated in Sudan. In the country’s west opposition forces have laid siege to the city of el-Fasher for over a year now. And a recent BBC News article provided readers recent satellite imagery showing the devastation within the city and, most interestingly, one of the most ancient of mankind’s tactics in siege warfare: encircling the town or fortress with a wall. No one gets in. And no one gets out.
The screenshot from the article above is of a village to the west of the city and it reveals the brutality of siege warfare. Paramilitary opposition forces constructed their siege wall through the middle of the village, which the image reveals barely exists any longer.
I love images like this because sometimes complex graphics are not necessary to tell the story. The scale of obliteration is obvious to the audience. And it only requires a moment’s extra thought to realise each building was someone’s home or business.
This second screenshot is of an annotated map from the Humanitarian Research Lab at Yale University. Coloured lines indicate the siege wall’s size and progress. Unlike the obvious nature of the first set of side-by-side images, here the story requires annotations.
Colours indicate when the paramilitaries built the segment of wall. The village in the satellite photos can be seen on the left of the map with the green dashes appropriately cutting through the text label. There are only four colours relevant to the wall and I perhaps would have explored using a light-to-dark gradient instead of four distinct hues as a gradient implies a progression. The wall began in the north, then went to the west and worked clockwise, i.e. cyan to green to yellow to red. A gradient probably would have been clearer in that regard.
I think the drawback with that particular graphic is the legend is not in the graphic itself. Instead, the BBC had to explain it via the article’s text. The lack of an integrated graphic hurts in cases just like this when, say, a screenshot of the image is used without the accompanying text. Here is an example of how that could work.
Overall I like the graphics. And by like I mean the design of the graphics. I do not like the ongoing civil war. The paramilitary forces have made clear their intention to starve and bombard the city’s population to death. Alas, I doubt we will see anything change anytime soon.
Credit for the original pieces go to the Humanitarian Research Lab at Yale University and the BBC graphics team.
Last weekend, Major League Baseball Commissioner Rob Manfred created a mild furore when he discussed the sport’s looming expansion and how it would likely prompt a geographic realignment. I am old enough I still recall baseball’s two leagues—the American and National—organised into only two divisions—East and West. In the early 1990s, baseball expanded and created a new Central Division. Afterwards, teams changed divisions, teams changed leagues, the sport expanded again to add new teams. As of today, MLB comprises 30 teams, 15 each in each league divided five teams per division. The only potential downside is whereas for decades the leagues only played amongst themselves, since the late 1990s, the odd numbers of teams have required “interleague” play on a daily basis.
But baseball wants to expand to 32 teams in a process yet to begin. Expansion requires the interested parties to pay an enormous fee, making other owners significant amounts of sweet, sweet cash. The question on everyone’s mind is where will the two expansion teams locate? And once they start play, how will the sport organise its teams?
The Athletic posted an article about this very issue earlier this week and it included two nice and simple maps showing one potential geographic realignment. This image is of author Stephen J. Nesbitt’s idea for the American League. (I care about the American League because as my regular readers by now know, I am an unapologetic fan of the Red Sox.)
As maps go, I think it works well. Personally, I prefer lighter backgrounds to darker, but a grey map is a grey map all the same. I think the colours work well and visually group the proposed divisions clearly. The white box makes clear the new division. (The map for the National League also works well.)
Beyond the information design or data visualisation level, I generally like the realignment proposed here as it offers a decent compromise between those wanting a basketball-style apportionment into two conferences, one eastern and one western, and those traditionalists like myself who still value the distinction between the American and National Leagues.
The biggest issue is where will the teams be? The author assumes one team on the East Coast or at least east of the Mississippi, and one West Coast or west of the Rocky Mountains. But baseball returning to Canada and Montreal would be welcome as would an expansion team in Texas, say Austin or San Antonio. Then of course you have the issue of Oakland. And the longshots have always been expansion into Mexico.
Personally, I have long been in favour of an even larger expansion to 36 teams. For sake of argument: Portland, Salt Lake City, San Antonio, Charlotte, Nashville, and Montreal have always made sense to me. Of course six new teams in one quick go would be impractical and thus it would be a long-term goal. But three 6-team divisions would mean 18 teams per league and you could still eliminate or reduce interleague play to special occasion weekends.
A good friend of mine sent me this graphic earlier this week. The World Wars fascinate me—to be fair, most history does, and yes, that even includes the obligatory guy thinking of the Roman Empire—and I can see on my bookshelf as I type this post up my books on naval warships from World War I and a history of the Great War itself. But the graphic comes from the Onion and satirises the Great War with a few choice bits relevant over 100 years later.
World War I is the what happens when you leave the default settings on and fail to make changes to your account after you set it up. Because someone declared war on someone else, another someone had to declare war on that person, and that person had to be declared war upon by another person, and so on and so forth until millions were dead.
Credit for the piece goes to the Onion’s graphics staff.
Last week two of the largest American freight railroads agreed to a merger with Union Pacific purchasing Norfolk Southern. Railroads have long played an important part in the history of the United States, from the Second Industrial Revolution to settlement and development of the West, through to the time zones in which we live and the laws on monopolies and corporations by which we live. Even in my own family history, my 3×great-grandfather was crushed between two rail cars and instantly killed whilst working for the Reading Railroad.
This merger, however, will create a truly transcontinental railroad spanning from the East Coast to the West Coast. Fortunately the New York Timescovered the news and included a graphic to showcase the new network—if approved by the Administration, which…really?…do we have any doubt?—with the Union Pacific in red and Norfolk Southern in orange.
The graphic is great. It need not be overly complex. With other railroads shown in light grey for context, the graphic highlights both the expansiveness and pervasiveness of the two networks. In a normally operating world, we probably would see significant concerns from the government about monopolistic concentrations of freight—we are talking about 40% of freight traffic here being controlled by one company. But this ain’t exactly Kansas. Except, it kinda is. Regardless, I doubt this merger gets held up.
A minor note, I appreciate that this map includes the state-level administrative units for both Mexico and Canada—Cuba too, but that is less important for this article. The only concern would be, I would probably have used a thicker white stroke for the international borders, because how many readers would know on the above map where the United States ends and “not the United States” begins…
In my last post, I commented upon a graphic from the Philadelphia Inquirer where a min/max axis line would have been helpful. This post is a quick follow-up of sorts, because a week ago I flagged something similar for me to perhaps mention on Coffee Spoons. So here I shall mention away.
We have another graphic from the Inquirer in an article about the Philadelphia region’s oppressive humidity this summer. The chart presents its information straightforwardly—bars representing the percentage of hours wherein the dew point sat above 70ºF. Muggy. Muggy as hell. Because I guarantee you the heat in Hell is a humid one. None of the dry dessert heat.
Overall, the graphic works well. It is interactive so you can mouse over the bar and read the precise data point. I love that far more than the increasingly prevalent let’s-label-every-data-point-on-the-chart-and-distract-the-eyeball-from-the-actual-pattern-of-what-is-going-on approach.
This summer has been the third muggiest in Philadelphia in the last three-quarters of a century. The designer highlighted 2025 at the end of the series—not necessary, but I can live with it. But what then stands out are the two muggiest years—two very tall bars. But note that there is no axis line above them. No upper bound. Nothing to help inform the user what percentage point they approach.
I do not always use a maximum or minimum axis line, but usually the outlier has to be extreme, and in that case I will add extra lines around that point to give the user the vital context of scale. Otherwise, the outlier should be just a wee bit above or below the line. I thought I would find a relevant example in my work quickly, but it took nearly 20 minutes of reviewing old work to find one such example.
Here you can see in Figure 6 the pink line barely and briefly rises above the 80% maximum. The reader can see the value just pokes above 80% but was otherwise below it during the entire span of time. And that works great.
Again, this is a small critique of the mugginess chart, but I feel an axis line significantly helps the reader see just how muggy those summers were. Spoiler: nearly 54% of the time was “oppressive”.
To play devil’s advocate, perhaps if the article were not about how this is third muggiest summer, the designer could have skipped adding an axis line at 60% or so. But, because such the author placed such emphasis on the third-most bit, the graphic really would benefit from the context of how the 45% thus far for 2025 compares to the top-two summers.
