I'm not sure if this is a serious question, since the answer is so obvious. In the late innings, when you've already scored enough runs to win, it makes sense to put in the guy that will do the best job of preventing runs from scoring. In the early innings, obviously the light-hitting, all defense guy is a bigger liability than the guy that can hit but plays inferior defense. Of course it is a serious question. It makes no logical sense whatsoever. What does it matter that the poor defensive player lets in runs at the beginning or end of the game? But more to the point, how often does his replacement prevent a run from scoring in the 8th or 9th inning? Are you honestly saying you don't understand the logic behind why a manager might substitute a Doug Mientkiewitz for a Craig Wilson in the 9th inning with a 1-run lead? The premise is elementary: when you need your defense to hold the lead and secure the win, then you put your best defensive players in the game. It's a basic principle that applies to many sports. You can agree or disagree with the wisdom of such a strategy as it applies to baseball, but to fail to grasp the logic is pretty puzzling. It's pretty basic stuff. Ok the logic is faulty. Here's the deal, Why pay a guy and take up space on the 25 man roster when whatever minimal benefit he might provide is going to matter maybe once or twice a season, if that? Substituting Freddy Bynum for Matt Murton or Neifi Perez for Todd Walker is just foolish. Those two are 2/13 of the position players on the Cubs team (the kept 12 pitchers). If you take out Blanco they represent 1/6 of the entire team. I don't know if the figures could be generated but I would think that having Bynum and Perez on the bench cost the Cubs more games than if the Cubs had guys on the bench who could actually hit. The value of the bench lies in it's ability to create runs. If you have to rely on your bench for jacks-of-all trades-masters-of-none defensive replacements you are putting your team at a decided disatvantage. That's a different argument then. Sure, people on the bench who can hit are more valuable, and players should not be kept on the roster who are seen as only defensive replacements. If they are on a roster though and a team is up late, there is value in a defensive replacement-just not nearly as much usually as an offensive player on the bench. People thought you were arguing originally that a defensive replacemnt in-game is never a good move instead of the value between an offensively minded bench guy and a defensively minded bench guy. That is exactly what I'm saying. Baseball is a game of %. Having a guy on your bench simply becuase he is a good glove is never a good idea. the opportunity costs are just too high over the long haul.