Now, let me show you a few possible alternatives for the graphs that we discussed so far. The first one is an alternative for this kind of plot graph. And this graph is called the slope chart. How does the slope chart work? Well, you can imagine starting from a scatter plot, which is two orthogonal axes. In the slope chart, we have two parallel axes. Every item in the dataset can be represented by a line connecting the two values that are in the two axes. So, let's have very different kind of representation, not necessarily as effective as a scatter plot. What the slope chart is useful for is to see when the values are going in one direction or in the other direction. So, like, for instance, in these examples that we have here, we have information coming from the major baseball leagues. On the left hand side axes, we have the proportion between wins and losses of each team. So, every item that you see seen the role is a team. And on the other axis on the right hand side, you have the amount that has been spent by the team to hire the players. So, here what you see is that the slopes that go up mean teams that are not performing particularly well but spend a lot of money. And slopes that goes down is exactly the opposite. So, that's a possible alternative to the scatter plot. So let's look at possible alternative designs for the line chart. So, here I put three different designs that are all based on the idea of using different symbols or different graphical strategies to represent the values. So, in the one that you see on the right, in place of lines, we have bars like in a bar chart, which is not as effective as the line chart to represent the actual trend, but it's useful in case you want to read the values, every single value accurately. The next one is the area chart, which is basically a version of the line chart where the area below the line is completely filled up. And the last one is one where we use just the simple dots. But the problem with this one is that we can't really trace the line and we can't really see the trend the same way we can see it in the standard line graph. So, once again, the goal of this exercise and even for me showing you these alternative designs is just to start reasoning about the fact that for exactly the same information, we can have very different visual representations. Now, for the matrix representation here, I'm showing you two possible alternatives. The first one is what is called a stack bar chart, which we're going to discuss in more depth in a moment. So, here the idea is that one way to arrange two categories is to use the rows for one of the categories. And now, every bar that represents each of the categories is split in a number of smaller bars, colored bars, where every color represents the values of the other category. So, in this case, it's pretty hard to see the distribution of the values across the two categories. As we will see in a moment, the stacked bar chart is good for other purposes. And the last graph, the last alternative that I want to show you, the last alternative to the matrix representation, is a graph where the two categories are nested. So we use a bar graph, where we first have the first category. So, we have rows for each categories. But within each category nested, we have the other categories. So this is somewhat useful and interesting, but one problem here is that it doesn't scale visually as easily as the matrix representation. So that's one problem of this nested design. And finally, for the symbol map, one possible alternative is to get rid of the spatial metaphor. So, get rid of the map completely and just create a list of locations, which in this case is a list of zip codes. And for every zip code, we represent the value with a bar. So there will be a classic bar chart, like the one that we have seen so far, and we are treating locations as categories. Every single zip code is a category and every bar is associated to a zip code and as a length that represents the quantity that is associated to the zip code.
