Its pretty obvious what they did. They didn't adjust the signal for random noise and then cherry picked a certain segment. Thats why you see so many data points however when you go into what BEST actually does when looking for trends they do a bounded convolution sum. BEST does a good job of showing the sigmas in their data and the evolution of the graphs from the raw data with all the noise to one year averages, three year averages etc.
They then use their own statistically justified normalization. What GWPF most likely did is take raw data for individual data points, showed the noise and then tried to call it a day. Nothing to see here its getting cold in january global warming is a lie.
The point is that the GWPF who made the graph is not BEST and when asked what the context and where the graph came from you said it was BEST.
When someone wants to find that graph they go here:
http://www.sourcewatch.org/index.php...icy_Foundation
and not
http://berkeleyearth.org/
That you see no problem with trying to mislead people like that may not be surprising but it demonstrates quite clearly what kind of person that you are: a deceptive one.