They aren't a top 4 NL team by any stretch, but they could still very easily make the playoffs. Pythagoras says we are. Is there any evidence that pythagorean WL is a predictive stat? I'm not being argumentative here; I really want to know if there is any study that shows actual WL reverts to pyth WL instead of the other way around (or, possibly, no reversion in either direction). I've always looked at pyth record as nothing more than a summary of runs for/against, not a prediction of future winning pct - but maybe I'm wrong. It seems to be a fairly decent predictor, but certainly not without its flaws. In 2006, the Mets were 7 games better than "expected" and Atlanta and Colorado were 5 games worse than "expected". In the AL that year, Oakland was +7 and Cleveland was -12 (!). http://www.baseball-reference.com/leagues/AL_2006_standings.shtml