Seeds are earned rather than rewarded. In that way they are more ranking than prediction. But rankings contain a degree of subjectivity (whether the Flo rankers want to admit it or not), so it can be hard to differentiate between a ranking and a prediction sometimes. An example, is quality wins. If two undefeated wrestlers do not wrestle the exact same schedule then some subjectivity comes into play to decide who is more deserving of the higher seed based on the quality of their wins. At the margin this is purely subjective and there is an element of a positive feedback loop. If wrestler A beats the #10 ranked wrestler and wrestler B beats the #11 ranked wrestler it is tempting to say wrestler A had a higher quality win, but who is to say that the original ranks were 100% accurate?