If you're talking about the quote from the IPCC on the last slide, I am VERY suspicious of anything coming out of a UN sponsored body. The UN is corrupt and it is highly possible that the report is tainted by UN bias. But that's a topic for another group. This quote is the same one that is used over and over. I'm sure that very few people repeating the quote have reviewed the data personally.
Hell if I know. I'm not a climatologist or a meteorologist. You tell me. I will tell you that if average temperature was just a random number then the odds of the top 4 in 100 being clustered together in the last 6 years would be about 1 in 6.7 million. But temperature is not a random number. It's not like rolling the dice. There are patterns and trends. Annual average temperatures are not independent of the years immediately before and after. Of this you can be certain. The trends can only be explained by a science that is still developing.
Let me rephrase your question and I will take a shot at answering it:
Q - What is the likelihood that extreme hot years or extreme cold years will be clustered in near proximity?
A - Pretty damn high. Records assembled for the last 100 years show that sudden drastic change on an annual basis is unlikely but it does occasionally happen. Therefore, if one year is a hot year, it is likely that the years immediately before and after will be hot as well. The exact mathematical odds cannot be calculated because the science to understand the factors that influence temperature are still developing.
http://apollo.lsc.vsc.edu/classes/met130/notes/
chapter18/recent_trend...
I haven't made any conclusions. Only posed questions and remained