Right now Arctic sea ice (extent) is, well, behaving very differently than in years past. It's maximum (at least so far) was February 15th, several days ahead of the average. Since then it's been separating itself from the pack (bold purple line below):
NSIDC shows it's more than 2 standard deviation below the 1981-2010 average. According to my spreadsheet, it's 2.20 stdevs below the mean. Assuming the distribution of daily SIE is normal (Gaussian), that only happens 1.4% of the time. That's about half what you'd expect for the most anomalous year in a data record that's 37 years long (= 2.7%).
Neven has more here.