A comment from Anthony’s blog
Paul K2 says:
Thank goodness there are at least some people here who can spot the mathematical mistake in this post from Real-Science dot blog, so I don’t have to address it. The conclusions the post draws are nonsensical. Maybe Real-Science should actually be called Fake-Science.
Enough stupidity please, from the left.
The graph shows how many all time daily temperature records were set during each year. It is calculated by parsing all the yearly records at a particular station for a particular day of the year. The year which has the highest temperature wins. All years have an equal chance of setting the record. Every station has exactly 366 daily high temperature records associated with it.
What Paul K-9 is trying to suggest is that the graph plots successive temperature records. Obviously that is not the case, because if it were – the first year would set a record at every station on every day. There would be an asymptote coming down the Y-axis on the left side.
There is a small bias to the left because the record is set by the first occurrence of the record high. In a warming world, this wouldn’t have much effect, because later temperatures would be warmer on the same date.
The graph below has the opposite bias. In this case, the most recent date of a tie wins. Even though the bias favors recent temperatures, the 1930s is still by far the hottest.
If there were thousands and thousands of daily records being set (as claimed by Masters et al) the spike would be on the right side of the graph – not on the left side.
My advice to alarmists. Try using your brain once in a while.