CMST: Visualizations!

[aseem.keyal]

Thanks to the work of Ophir, here are some wonderful visualizations of where people buzzed on each tossup. I've also included some tab separated data Ophir prepared, in case anyone else wants to do any visualizations (we encourage you to post the results here). Please do not post this link or any pictures publicly, for obvious reasons. Enjoy!

Note: the files are numbered round-tossup. Also Ophir was responsible for all the visualization, and I did all the rendering, so any image quality/display issues are my fault, not his.

Note 2: some sites are missing buzz points for various mods in some rounds, Ophir and I will try to recover those and update these later.

[ryanrosenberg]

I took a look at what the hardest categories were at CMST -- results here. Note that categories with high neg rates will be less-converted by this metric, since I only counted live buzzes.

[Mewto55555]

I love these. In particular, they provide a really good way to quickly tell how a question played, and draw conclusions from it. For example, just given the following categories/names/buzz distributions, you can tell that these two weren't amazing science tossups*, without even looking at the answerline.

*the first one was on a fake doubly-eponymous thing which dropped a different doubly-eponymous in power that can be learned from Wikipedia and not in classes
the second was an IJ math tossup that everyone bold enough to buzz early negged on because the clues and answerline selection were not too good