Reversal Statistics in CoLBeRT Paper suck

[aleifer]

I could either:

1. present the data differently, perhaps by showing all instances of reversals or non-reversals. This will obscure the point, but will be more accurate and straightforward

2. Take more data

3. Do nothing