Cherry Picking The Data

I've been messing around with graphs today (yes, it's nearly time for my end-of-year Power Usage update), and got to thinking about the famous "cherry-picking" that the climate sceptics like to do to illustrate that Global Warming stopped in 1998.

If you don't know this argument, it is a simple and seductive one: 1998 was the warmest year on record, so a graph of global temperatures starting in 1998 will show a declining trend - thereby "proving" that Global Warming stopped in 1998.

Of course this completely ignores every other temperature before 1998, and also ignores the fact that 1998 might not actually have been the warmest year on record after all. Skeptical Science has a nice round-up of the records to indicate why the whole premise is a false one.

In the last year, Conservative peer Lord Lawson has emerged as one of the leading sceptics in Britain, so I thought I would pick on the Tories to illustrate my point while I was playing with my graphs.

Here are the results of a YouGov/Daily Telegraph Political Trends Poll from May 2005 to December 2007. The trendline clearly shows that voting intentions for the Tories over that period are rising:


And here, for your viewing pleasure, is my "cherry-picked" poll. This is conclusive proof that the Conservatives have been shedding voters since 2006:


Incidentally, 2010 is predicted to be even warmer than 1998. I wonder how the sceptics will frame the argument then?