> One would expect that they'd also count "I will, "we will", etc. and show a ratio. But no. No - what you want to measure is the absolute number of occurrence

> One would expect that they'd also count "I will, "we will", etc. and show a ratio. But no. No - what you want to measure is the absolute number of occurrence, not the ratio, the reason being that "making negative predictions" is not the opposite of "making positive predictions". Similar to the fact that negative and positive affect aren't opposites but distinct factors (that share some correlation though!). 'Common sense' can easily lead you astray here. It's a good example for something that looks simple, but really isn't.