Rands covers this, more generally, when he talks about "changes in routine". What you're looking for isn't the absolute value of, e.g., HN participation, it's

Rands covers this, more generally, when he talks about "changes in routine". What you're looking for isn't the absolute value of, e.g., HN participation, it's the first or second derivative.