I think what he meant is that we don't have a good way of objectively dentifying what caused the Cubs to begin to play up to their "pytg. potential", if you know what I mean. Therefore we wind up attributing everything to "luck", a sort of catch-all phrase we fallback on when we can't explain the reason for or meaning behind certain phenomanon. I think the hope is that we can find a way to objectify that analysis in the future rather than saying "they just got lucky", which is demeaning in its own way. Luck is just a different way of saying variance. Players who hit .250 don't go 1 for 4 every day, and pitchers with a 4.50 ERA don't give up 3 ER over 6 IP every start. In the same way(or, "as a result" works better I guess), .500 teams don't alternate win-loss-win-loss all year. A different way to say it without using "luck", is that they "played to their potential". People say luck because the close games that make and break teams' records have a LOT of variance. I am not sure that is what everyone means when they use the term, but I'll play along. Assuming you are correct, do we have a way of quantifying or predicting this variance? What about for middle relievrs, which is the topic the hand? Also, would you be willing to concede that baseball in general is to a great extent influenced by this "variance", or do you think certain teams get "luckier" than others? And if it is the case that some get luckier, isn't it true that there is only so much you can do to build a team that wins the WS? I'll try to relate it to middle relievers. Variance comes because you can't predict when Marmol or Wuertz is going to have a bad outing. Obviously managers have some sway over this, and may be able to sniff out when a guy may have stayed out too late, or looked really good in his last side session. But even then, sometimes the guy makes his pitch and the hitter makes a play on it. All of a sudden you've lost that one run game. Over the course of a long season(kinda tying in what Sulley just said), these things tend to even out(although no one is saying it's guaranteed), hence the value of pythagorean. On a macro scale, it's a little easier to make distinctions like this. For example, Les Walrond is a bad pitcher, calling him up and letting him pitch in a bunch of close games is a bad idea. But over the course of less than 1/3rd of a season there's a near infinite amount of noise that can distort. That also ties in to how I feel about the playoffs. There may be a common denominator in successful playoff teams, but even isolating it and building upon it may not do much for your chances of playoff success given the short amount of games(at most it's half of the sample I was just talking about). That's why my objective would be to build the best team possible, with maybe small differences to an ideal roster to account for the slightly different dynamic of postseason team construction(i.e. having 2-3 studs and 1-2 lesser pitchers in the rotation instead of 1 stud and 4 decent-solid guys).