The miraculous is always "unlikely", but reported miracles are not always "improbable". Yes, both statements are true; it's just difficult to reconcile the Math with the Language on this one. To explore how we talk about such things, let's use Jesus' Resurrection as example 'A'.
If we frame the discussion in terms of actual occurrence, then the miraculous is always "improbable". That's mathematically verifiable because human history absolutely records countless moments during which a dead body did NOT miraculously rise from the grave. The data set on this one, in fact, is ridiculously huge. The improbability of human resurrection occurring within days after undergoing circumstances such as Jesus' execution... well, it's a very big number approaching infinity.
Note (1): The probability of a dead body returning to life is virtually and effectively nil.
By comparison, the probability of a dead body staying dead is also mathematically verifiable, based on a very large and well documented set of data. That is, we have a nearly unlimited record of dead bodies that stayed dead after dying. Again, if we frame this discussion in terms of comparable known events which have actually happened, there's no possible way to equivocate.
Note (2): The probability of a dead body remaining dead is/was virtually infinite - as close to certainty as it comes.
However, since measuring probability requires such large data sets the game changes when we begin to consider reports of unverified and unprecedented events. Although we still have a very large data set of dead bodies remaining dead, we do NOT have a very large data set of people attesting to resurrections that subsequently turn out to be false or imagined. Likewise, we do NOT have a very large data set of people attesting to resurrections that subsequently turn out to be verified as actually having occurred. Therefore, if we give any consideration at all to testimony in such a case, then all talk of "probability" should be ended.
Note (3): Unique testimony of miraculous events may seem credible or incredible, and historians may assert one or the other on whatever grounds seem justifiable, but there is no mathematical basis for measuring the "probability" of whether such testimony may be true or untrue.
To sum up - if we frame the discussion in terms of what typically happens to dead bodies, then the resurrection of anyone seems unlikely, and this does include Jesus. However, if we also take into account the various testimonies which purport Jesus did miraculously resurrect, then we are left with no measurable likelihood, no "odds", and certainly no "probability" that these people were or were not telling the truth. The data set of such claims is unsurprisingly small.
Miracles, by definition, are "unlikely". However, a reported miracle under unique circumstances is neither "improbable" nor "probable". It's unverifiable. Trust the report or do not. There is no "verify". ;-)
To conclude - here's my personal takeaway.
Let's quit arguing over factuality and discuss hypothetical scenarios. If Jesus did rise, then what does that mean? If Jesus did NOT rise, then what does that mean? Those are much more interesting conversations to have, in my personal experience. They are often more fruitful as well...