El Niño is the main thing that springs to mind when thinking about predictive powers of (short term) climate modelling. They are getting reasonably good at that.
Also, your coin flip is not a predictive model any more than claiming that using the same numbers every week on the lottery is a predictive model. For one thing, it makes no attempt to model anything, and for another, it has no attempt to be predictive, it just attempts to be right nearly half the time (edge), through understanding of the likely odds, which is a different thing altogether.
Also, your coin flip is not a predictive model any more than claiming that using the same numbers every week on the lottery is a predictive model. For one thing, it makes no attempt to model anything, and for another, it has no attempt to be predictive, it just attempts to be right nearly half the time (edge), through understanding of the likely odds, which is a different thing altogether.