In Bayesian Analysis, to what degree(s) does:

• “Probability” mean prediction(s) of a future event(s), and

• “Likelihood” means an assessment of the cause(s) of a
  past event(s)?

Assuming the percentages below are known to be accurate from prior knowledge; for example:

*IF a stack of wood in a fireplace is lit on fire, THEN the probability of observable smoke exiting the chimney is 84%. (Allows for the smoke-producing properties of various types of wood.)

IF smoke is observed exiting the chimney, THEN the likelihood smoke-producing wood was lit afire in the fireplace is 99%.

Do the above examples accurately portray the difference between “Probability as Prediction” vs. “Likelihood as Explanation”?

Your thoughts?