Do you know what a save is? Sure, you say, a pitcher gets one for finishing a close game for his team, which is more or less correct, but what is a "close game"? What does it mean to "finish" a game? What if you enter with a 10-run lead, give up 9, and then get the win?
In order to be awarded a save, a pitcher has to: 1. Finish the game. 2. Be credited with at least 1/3 of an inning pitched. 3. Not be the winning pitcher. 4. Either enter the game with a lead not exceeding three runs and pitch at least an inning, or enter the game with the potential tying run on base, at bat or on deck, OR enter the game and pitch at least three innings to the game's conclusion, provided the pitcher's team wins, but, again, the pitcher cannot be the winning pitcher, which is a different yet still somehow entirely meaningless concept.
Let's recap. Joe Assballs enters the game with two outs in the ninth and his team up one with the pitcher up to bat. He throws one pitch, which is grounded weakly to first. Save! Assballs for the Cy Young! Tommy Peendick enters the game in the eighth inning with his team up 4, strikes out the heart of the lineup over the next two innings and finishes the game. Fuck you, Peendick! That's not a save, because arbitrariness!
Why am I all het up about saves, now? Consider this, from Braves manager Fredi Gonzalez after last night's game:
"We were thinking, and we had it set up. We double-switched and put our best defense out there. We had it set up to bring him in for four outs. I think six outs was something that we weren't even talking about in the dugout. But I think with two outs we were planning to do that. We set up the eighth inning to be able to do that."
That's a pretty good plan - not the best plan, in my estimation, which would have been to trust your best reliever to throw two friggin' innings, but I digress - and it went to hell once Carpenter gave up a double to notorious not-playing-the-game-the-right-way-er Yasiel Puig. Now, what's a manager to do in this situation? Bring in your utterly dominant closer, who strikes out like four guys per inning and can throw 143 miles per hour and poses like some weird huge bird while getting the signal from his catcher? That seems smart - or, it would seem smart, if you weren't some moron for whom it is unquestionable knowledge that a closer pitches only the ninth inning, or maybe also the last out of the eighth inning if it's really important. I don't know if Gonzalez considered deviating from his well-laid plan, but he didn't, and then Carpenter threw the laziest, hangingest breaking ball to Uribe, who hit the absolute crap out of it. The Dodgers took the lead off of Uribe's homer, BROUGHT IN THEIR CLOSER FOR THE SAVE!!1!!, and the Braves' best reliever did not throw a pitch in an elimination game in which they never led by more than one run.
Managers outsmart themselves and come up with "clever" plans like the one above because saves exist. The save, much like the "win," has survived the statistical revolution by having a name that sounds good - seriously, if wins were called anything other than "wins," and idiot baseball commentators couldn't use their favorite false equivalence "this pitcher has the most wins and the point of the game is to win and that's why he's my Cy Young pick," would we still think about them at all, given how confusing and arbitrary they are by definition? - and by being defined using precisely zero mathematics. This is why old-timey baseball farts love these stats so much: It doesn't matter how you have to contort your grammar to write out the definition, as long as there aren't any fucking decimals, because this is baseball not mathball you goddamn nerd! But the save - well, it's a save! He saved the game! Why wouldn't I want my best reliever to get saves? I'm only gonna use my best reliever in the most common save situation, because if those situations weren't the most important, why would they be called saves, you dork?
Saves are ingrained in the culture of baseball, and this occasionally leads managers use someone other than their best reliever in the highest-leverage situations, and this sometimes causes teams to lose. It's time for the save to die.