Do you agree that you were wrong when you stated I must prove Sampras has "a good understanding of past statistical trends, and the ability to take into account unknown variables"?
Not at all. Let me present your two options:
A.) You choose to believe Sampras' prediction is true, or likely to be true. To justify this belief, you must prove Sampras' credentials as a predictor. That he has in other words, "a good understanding of past statistical trends, and the ability to take into account unknown variables".
B.) You choose not to believe in Sampras' prediction. In other words Sampras' prediction ability is of no help to us, here. Selecting this option ends the debate.
The statement "There is more reason to believe Sampras than Roy Emerson" can be true if you select either A or
B. Hence, it settles nothing. Eg. it could be that we have more reason to believe than Sampras than Roy Emerson, while having no
reason to believe Sampras' prediction. Think about the lunatic asylum example.
What you need to do is pick one of the two options.