The graph in question is from his blog post entitled 'Meltdown 2012 : Earth Has More Sea Ice Than It Did In 1980', and although it was posted on the April 1st I'm certain that it wasn't supposed to be as foolish as it seems.
The Cryosphere Today' site and I am assuming he has graphed the data correctly. Helpfully he rings this years data point and makes the correct claim; " Earth Has More Sea Ice Than It Did In 1980". So what?
Clearly this is an insinuation to some great truth about climate change in the Arctic regions, without saying anything meaningful about them. It appears that he has taken the current value and then hand picked the year furtherest back and at the start of the data set that was lower, i.e. 1980, while gleefully ignoring all other data points in between that might actually be used in determining a meaningful trend or credible conclusion. In fact, by only relying on two data points in a set of 33, he has ignored about 94% of the data to not say anything meaningful - that takes a concerted effort in my book.
Hand picking two data points to the exclusion of all others is how I would define 'Cherry Picking'. However when I mentioned the phase in a comment on Goddard's site he ironically replied;
"Cherry picking the current maximum? Are you as dense as you pretend to be?"Does the man realise what he says sometimes? Hand picking a maximum, and ignoring over 90% of all other maximums is cherry picking in my book. He repeats the exact same nonsense here, only lists more years from the 80s, and ignores the clear trends from the mid 90s to the present.
What can we really tell from Goddard's graph? Well there is more sea ice now than most years, and a lot more sea ice than many recent years. So although the sea ice area had recovered on that particular day compared to most years, the clear trend is that global sea ice is of decline when looking at a single day, April 1st, and most of that decreasing trend occurred from the mid 90s with more years than not being lower than this years Fools Day.