The Climate Change policy advocates never discuss odds. They have less descriptive "likely", "very likely" and "most likely" which provides ballparks and "fat tails", things that "could" happen. I "could" win the lottery. Earth "could" have an impact event in the next three months. A nuclear reactor core "could" burn through the Earth to China. Major alarmist Climate Scientists "could" be right.
Consider the upper range of climate change impact, 4.5C degrees per doubling of CO2. From an interglacial starting point, that is where most of the glacial ice sheets have melted and sea level in high like now, for the past 1 million years, there is no point were the Earth temperature made it to 4.5 C higher than it is today. The last time temperatures were 4.5C higher was ~ 30 million years ago and the last time there was a major impact event was ~2.7 million years ago. Alarmist Climate scientists have and continue to use 4.5 C as an upper boundary for a doubling while the current rate per doubling is on the order of 0.8 to 1.6 C per doubling. The odds of 4.5 C per doubling is likely on the order of an impact event. An asteroid or comet crashing into the Earth disrupting the geomagnetic field and causing a magnetic field reversal, major climate shift and new evolutionary paths.
"Sensitivity" is defined as the change in "surface" temperature per change in radiant "forcing". The estimated range of "sensitivity" is 1.5 to 4.5 C per ~3.7 Wm-2 with the "benchmark" response estimated at 3.3 Wm-2 per degree C. 3.7/3.3 yields 1.12 C per 3.7Wm-2 From the start, even the minimum impact is overestimated by ~0.38C degrees. The "benchmark" is based on observation from arguably a cooler period caused by greater than "normal" volcanic activity and/or other natural factors. A cooler object is easier to warm so the "benchmark" value is likely inflated, meaning the already inflated minimum "sensitivity" is even more inflated. Now 3C per doubling has "likely" the odds of an impact event or me winning the lottery.
A simple estimate of the new "benchmark" can be made using the "average" estimated Down Welling Longwave Radiation of 334 Wm-2 +/- 10 Wm-2. That produces a "benchmark" of 4.8Wm-2/C +/- ~0.4 or about 0.76 C per 3.7Wm-2 equivalent doubling. The curious would note that the "average" temperature of the oceans is 4C which has an approximate energy of 334.5 Wm-2 which also produces a "benchmark" of ~4.8Wm-2. Someone even more curious could look at the "environmental" lapse rate formula and notice that considering the specific heat of the atmosphere has a factor alpha ~0.19 which inverted is 5.25 Wm-2/C or should be if you work out all the units, though I have not done so.
You can even compare the ~0.8 C sensitivity to a 3.0 C sensitivity using reconstructed CO2 data to discover the potential impact of "guessing" that the 20th century was "normal" enough to use in determining the "benchmark".
Each of these "discoveries" reduces the odds that "sensitivity" to 3.7 Wm-2 of atmospheric forcing will be high. These "discoveries" also increase the possibility that alarmists Climate Scientists are whack jobs, cheese done slipped of their cracker, die hard "China Syndrome" DVD owners. .