Roll a die and ask "Will I get a 6?" A yes-or-no thinker guesses and is either right or wrong. A probabilistic thinker says "the chance is 61, so in 60 rolls I expect about 10 sixes, and 7 or 13 wouldn't be strange." That answer is more useful: it predicts what happens over many tries and how much the results can wobble, even though no single roll can be predicted.
This way of thinking changes plans. With a 30% chance of rain it will probably stay dry, but about 3 in 10 days with that forecast turn out wet, so you hold the picnic and pack a tarp. You are not predicting rain; you are preparing for a branch that has a real chance. How much you prepare should depend on both the chance and the cost: a fire in any one year is very unlikely, yet a smoke alarm is worth having because the harm would be so great.
Not every situation needs this. A vending machine that never fails, or the product 3×4, has one certain result, so plain reasoning is right. Probabilistic thinking also goes beyond computing a probability: finding P(6)=61 is arithmetic, while deciding what that number means over many tries, and what to do about it, is the thinking.
Think of it like this. A goalkeeper facing a penalty kick can't know where the ball will go. So she studies where this kicker usually shoots, leans toward the likely side, and still stays ready to dive the other way. Likely is not certain, and she plans for both.
Core idea. Probabilistic thinking is reasoning about uncertain outcomes in terms of likelihoods and ranges instead of yes/no certainties.