The Kimbrel contract carries risk. As I wrote last summer, Kimbrel is part of a class of pitchers who burn hot and fast, mega-strikeout relievers who throw very hard. He's probably the best of the class -- which includes pitchers such as Rob Dibble and Carlos Marmol -- but I would not have looked to make a long-term commitment to him, and in fact I would have been trying to trade him off of 2012's historic season, and continued doing so after his 2013. The vast majority of relievers don't have staying power, and to whatever extent Kimbrel might be the exception, he's highly unlikely to keep pitching at his current level. The Braves are paying for three years of performance that is not repeatable, burning much of the surplus value they earned off Kimbrel from 2011 through 2014.
By signing Kimbrel to a long-term deal, however, the Braves have created an opportunity for themselves. They have largely divorced Kimbrel's compensation from his statistics over the next four years. (Kimbrel does have a maximum of $4.5 million in contract incentives tied to award voting and games finished.) He doesn't have to worry about save totals and the impact of those numbers on his arbitration cases, or his negotiations for a multi-year deal. Now that Kimbrel is getting paid no matter how many saves he accumulates, the Braves can use him accordingly without taking money out of his pocket.
For example, Kimbrel is perhaps the greatest strikeout pitcher in baseball history. However, that skill is largely wasted given that Kimbrel comes in to start an inning in the vast majority of his appearances. Kimbrel inherited just three baserunners all year long, and just one in scoring position, that when the Braves held a four-run lead. The greatest strikeout pitcher in baseball history was never once used when a strikeout would do the most good relative to other outcomes. David Carpenter, a perfectly nice pitcher who wasn't signed to a $42 million contact, inherited more runners in his first two appearances -- five -- than Kimbrel did all year. Carpenter inherited more runners in scoring position in his first two appearances -- three -- than Kimbrel did all year. It's an unbelievable waste of resources to have Craig Kimbrel and yet never use him with the tying run on third and one out.
Then there's the matter of the opposition. Kimbrel faced the #7 spot in the order more than he did any other spot. Coming in second? The #8 spot, then #6, then #9. In all, Kimbrel faced the bottom four spots in the order 134 times, while facing the top four spots 100 times. Those numbers for Carpenter, to continue a theme, are 109 and 113, respectively. David Carpenter, journeyman, faced the top of the lineup more than did Craig Kimbrel, superstar. This isn't an accident; the way lineups fall in the average game, the eighth inning will generally see the top and middle of the order more often, and the ninth will generally see the bottom and perhaps the #1 and #2 slot.
This is entirely elective. The following chart shows the lineup spot Kimbrel faced most often when he first came into the game.
1 xxxx (4)
2 xxxxxxxx (8)
3 xxxxxxxxx (9)
4 xxxx (4)
5 xxxxxxx (7)
6 xxxxxxxxxxxxx (13)
7 xxxxxxxx (8)
8 xxxxxxx (7)
9 xxxxxxxx (8)
By limiting the variables of reliever usage to "save situation?" and "inning," you set up all kinds of ridiculous scenarios in which lesser pitchers are given harder jobs against better hitters. On July 8 in Miami, Fredi Gonzalez used Anthony Varvaro to pitch the bottom of the tenth of a 1-1 game, bringing him in to face the Marlins' 3-4-5 hitters -- with Giancarlo Stanton the first in line. An inning later, he used Kimbrel to face the Marlins' 6-7-8 hitters, then brought in Carpenter in the 12th to face 9-1-2. Craig Kimbrel's mom couldn't have given him an easier job. A few months later, back in Miami, it was Luis Ayala facing the Marlins' 4-5-6 hitters in the eighth, and Kimbrel facing 7-8-9 in the ninth. Look at the chart above, and go through the Braves' boxscores, and you find this happening a lot -- inferior pitchers facing the meat of the order, and Kimbrel mopping up -- for the save -- against the bottom.
That September outing involved protecting a three-run lead with three outs to go, something that happened quite a lot. Kimbrel made 68 appearances last season; in 18 of them, he was brought in to start the ninth or an extra inning with a three-run lead, the easiest possible scenario still qualifying as a save situation. Kimbrel entered in a tied game just ten times, and with the Braves trailing by one run once. Again, contrast that with Carpenter, who entered a tied game 11 times and, as mentioned, frequently came in with runners on base. Varvaro, Avilan, Jordan Walden…all of these pitchers, none as good as Kimbrel, none as able to get strikeouts and prevent runs, were more often tasked with situations more difficult than "three-run lead, three outs to go, scrubs coming up." Stick all the former closers you want in front of all the cameras in all the studios in the world, and you'll never make "the last three outs" enough of a factor to counter the difference in the situations save specialists and non-save specialists face.
