This story at the UK Daily Mail has all the details in plain English, with graphs. (H/T Ace of Spades via ECM)
First, they got rid of the Medieval Warming Period by cherry-picking data:
Some tree-ring data eliminates the medieval warmth altogether, while others reflect it. In September 1999, Jones’s IPCC colleague Michael Mann of Penn State University in America – who is now also the subject of an official investigation –was working with Jones on the hockey stick. As they debated which data to use, they discussed a long tree-ring analysis carried out by Keith Briffa.
Briffa knew exactly why they wanted it, writing in an email on September 22: ‘I know there is pressure to present a nice tidy story as regards “apparent unprecedented warming in a thousand years or more”.’ But his conscience was troubled. ‘In reality the situation is not quite so simple – I believe that the recent warmth was probably matched about 1,000 years ago.’
[…]Finally, Briffa changed the way he computed his data and submitted a revised version. This brought his work into line for earlier centuries, and ‘cooled’ them significantly.
Then, they hid the decline in temperature after 1960:
According to his tree rings, the period since 1960 had not seen a steep rise in temperature, as actual temperature readings showed – but a large and steady decline, so calling into question the accuracy of the earlier data derived from tree rings.
This is the context in which, seven weeks later, Jones presented his ‘trick’ – as simple as it was deceptive.
All he had to do was cut off Briffa’s inconvenient data at the point where the decline started, in 1961, and replace it with actual temperature readings, which showed an increase.
On the hockey stick graph, his line is abruptly terminated – but the end of the line is obscured by the other lines.
Ace is calling this a must-read. I agree. Drop everything you are doing and go read it. If you have a blog, blog about it. Submit the Daily Mail post to Stumble Upon, Digg and Reddit.