Plate tectonics and volcanic activity have also influenced climate in the past, e.g. the closing of the Panamanian isthmus, or the formation of the Deccan Traps.
Interestingly, the original paper proposing AGW (in 1896) was actually intended to explain Ice Ages:
"_All_ the studies are measuring the same tendency : the temperature is rising"
There could be lots of reasons for that. Anyway, the temperatures were not always rising, clear temperatures rise was observed in 1930-40s and in 1980-90s. Cooling in 1960-1970s. And yes, prediction has to be precise. If a model predicts rise of 3K in 100years, and you measure 1.3K - your model is wrong. It's even more wrong if you don't take into account any of the natural cycles, even if prediction is accidentally correct.