How many times must it be said?
Fargo is a R-A-T-I-N-G tool, not a prediction tool.
Well it's sort of a prediction tool in the sense that the accuracy of the ratings lies in how close a player performs to the expected performance that his rating suggests he is at.
We just had a cool round robin which Mike Page introduced. 8 players and each set was five games. Not a race to five, five games. So each player played 35 games and the fargo rating had an expected win amount for each player.
Payouts were based on who outperformed their expectation the most.
I was expected to win 21 games and I won 25 so I ended up winning because I had the best performance over my expectation.
A local shortstop was expected to win 25 and he won 26.
So that type of tournament clearly uses Fargo as a prediction tool.
But you're right, Fargo can't ever be considered to be a perfect crystal ball. All it can be is an analysis of past data that can say what is more likely to happen based on thousands of similar past matchups. To me, and as I understand it, the entire foundation of Fargo depends entirely on the predictive accuracy OVER many many games. So any given set can go entirely against prediction but if players perform over many sets close to the predictive mean then it can be said to be accurate.
It's not much different than when people predict that player A will lose to Player B because of their subjectively assigned ratings and player A steps up and plays over his head and defies the predictions.
The above mentioned shortstop was predicted to lose a $5000 seven ahead one pocket set two weeks ago against another shortstop. Instead he closed out the set in about six hours.