Picking winners straight is a valuable skill, but it is far less valuable in betting because all fights aren't even money odds. Because we as bettors are more interested in line value since that means profit in the long run, the only real way to measure performance for picking all fights like Stumberg is doing is to take the line and adjust it to ROI for each fight. So getting a -500 fight would only return 0.2u but picking a +500 right would get you 5.0u. If he ends up +ev after all those calculations, then he could call himself a good predictor. It's kinda like being an investor and saying you realized x% of gains. It's really the % over the S&P that matters in terms of performance, since that is the baseline.
But betting is even more complicated than that, since the other big skill in betting is seeing line movement before it happens. Assuming closing lines are more accurate in the long run than openers and mid-week lines, a good way to gauge whether you are really calling these odds correctly, or whether you're just getting lucky, is how often you beat the closing lines, and by how much. It's theoretically possible to only play closing lines and still win, but I don't think I've seen anyone successful doing that. It certainly isn't possible to be a winner over the long run taking worse lines than the closing line.
But betting is even more complicated than that, since the other big skill in betting is seeing line movement before it happens. Assuming closing lines are more accurate in the long run than openers and mid-week lines, a good way to gauge whether you are really calling these odds correctly, or whether you're just getting lucky, is how often you beat the closing lines, and by how much. It's theoretically possible to only play closing lines and still win, but I don't think I've seen anyone successful doing that. It certainly isn't possible to be a winner over the long run taking worse lines than the closing line.