|Average yearly temperature versus the average temperature for the first four months of each year. The red dot is 2016 and the error bars represent 2 standard deviations around that point.|
To get that percentage, I regressed the yearly average temperature against the temperature for the first four months. As you can see, it's a pretty good fit (R2 = 0.9495). I then found the standard deviation of the residuals and calculated the z-score between the predicted yearly average for 2016 (0.996ºC) and the actual average for 2015 (0.8275ºC). So congrats, 2016. May you reign long as the new starting point for deniers' claim of "No warming since _____!"