The "super-informative ping pong graphic" was mentioned on Junk Charts. At the same level of VC deals (roughly 1000 deals), the number of ping-pong tables sold ranged from 150 to 230, about half of the horizontal range of the original chart.Here is what the same chart looks like after taking out the 2016 Q1 data point:
At the same level of ping-pong-table sales (roughly 150 tables), the number of VC deals ranged from 920 to 1020, about one-third of the vertical range shown in the original chart.
I assumed that WSJ was a being a bit tongue-in-cheek. That said, you could very well break the deal/sales into months rather than quarters and test correlation. I'd say that the biggest problem here is that the time frame is too short. Junk Charts doesn't prove a strong point with their graph by reducing the time frame. As far as seasonality, location, etc. It doesn't really matter when you are testing for correlation. Everyone knows that a reduction in pirates is highly correlated to global warming.
I agree that it is a design choice, but it obfuscates the scale and indirectly the significance of the numbers used so I generally find it a bad choice. Here's the same graph with axes to zero: R^2 is 0,44 with n=7 but drops to 0,10 when you leave out the Q1 2016 outlier.
The question is: Is it an outlier? If it's the beginning of a change, then it might not be. As I said above, the range needs to be greater. That's why I think Junk Charts doesn't make an effective point by dropping it. IMO there's nothing wrong with a silly hypothesis, only the poor test of one. 0.44 says it all.when you leave out the Q1 2016 outlier.
I will never forget me social media buzz around Paul the Octopus during the World Cup. People will believe the darndest things. PS. this might just be the best sentence on Wikipedia ever:The President of Iran, Mahmoud Ahmadinejad, criticised Paul, accusing him of being a symbol of Western decadence and decay. [29]