Specific examples of lying crap on this graph:
a) They invented the green straight line out of thin air. Notice that it is a terrible match to the data (the line is too high) for the entire 18th century, for example; there is no reason whatsoever to invent such a line and call it a "0.29 degree per century trend"; indeed, picking different time-ranges will give you whatever trendline you want. I suspect they picked this one specifically because (by overshooting the entire 19th century) it happens to makes the post-1980 warming look smallish.
b) They drew the CO2 *emissions* and the temperature on the same graph. No one ever said that temperature follows year-by-year emissions---it follows *accumulated* emissions, i.e. actual atmospheric CO2 content. Why choose the emissions graph? To suggest that "because the temperature doesn't do THIS, scientists are wrong", presumably.
c) Why put that dashed line way on top? To suggest "look, sheeple, temperatures would be up HERE if scientists were right"? No, if scientists are right temperatures will be up there in 100 years. What's wrong, denialists? Afraid to extend the x-axis to 2100 and show the actual prediction? Of course they are---the actual prediction is terrifying, and there's no fuzzy straight line game that makes it look non-terrifying.
Classic case of "how to lie with statistics".