Stephanie Doctor

Chakiera & Cheng:

This was a really fun project - I've always wanted to get more into football but never had a team to root for!

I agree with some of the previous comments - namely, that the color helped me distinguish the teams from one another and that flipping the axis to vertical would take away the "down = bad" bias. Then you could make the bars positive in both directions, so if the team on the left won one game that year and the team on the right won two, the latter team's bar would just be twice as long. At a glance you could see where the bars were longer over the ten-year period generally and if you wanted to narrow down to a specific year you could see the actual amount of games played and won by each team.

I also agree with Lance that it would be helpful to have a more general visualization to go along with the one you have. Head-to-head matchups are cool to look at but also make it hard to declare a winner. You might use overall win percentage out of the games with these four teams, or maybe just rank them 1-4 each year and show how that changes over time (ranking them would prevent them from overlapping so much in a line graph).

Great job!

Chana:

I thought this was a really cool, relevant, and important topic that can't be discussed enough (and isn't, as you mentioned). I also think you did a great job of creating a narrative that runs through the visualizations you created in Tableau, incorporating text with the graphics to tell a story. This is furthered through some of the choices you made running through all of the different visualizations, like the colors representing the regions.

The treemap in the first visualization looks nice and definitely gets the point across, but I wonder if it's the best choice for this data, since it's not exactly parts of a whole. Well, I guess the counts one is but as a percentage of population less so (though it's a really important point to make).

I like that you have a combination of maps and other types of plots (like the scatterplot). The scientist in me wants to make the scatterplot axes log transformed but I realize that's not exactly user friendly :(

I think the point made by the scatterplot - that some countries are prioritized over others - is really telling and could be incorporated into the map that follows if the ratio between accepted applications and applications made was mapped instead of the count... or perhaps the count normalized to the population of that country. A few thousand refugees from a tiny country would be more significant than a few thousand from the DRC. Though maybe that would have the opposite effect by overstating the significance of a few refugees from a small country like Belgium (refugees from Belgium?) as opposed to the few thousand from the DRC. I'm not sure... could be worth looking into.

Overall, great job! I really enjoyed learning from your project.