Wednesday, August 1, 2012

Bullpen Depth: George Kontos

The Giants snapped a five game losing streak, beating the Mets 4-1 behind a strong seven inning performance by Tim Lincecum. Which means that the odds were very low that George Kontos would take the mound in relief. Here's why.

George Kontos was acquired at the start of the season from the Yankees for Chris Stewart, and he was called up on June 9th. Since then, he has posted some good numbers, which he has earned--when he's right--with a sharp slider.

Year W L ERA G IP ERA+ WHIP H/9 SO/9 SO/BB
2012 0 0 2.21 18 20.1 162 1.082 8.9 9.3 10.50
Provided by Baseball-Reference.com: View Original Table
Generated 8/1/2012.


Despite those numbers, especially the 9.3 SO/9 and 10.5 SO/BB rates, it doesn't seem like Kontos has been used in many high leverage situations. He has, for instance, neither decisions nor saves (nor save opportunities). Then, a comment by Grant Brisbee (I can't find it at the moment, but it was something like: "what the hell is Brad Penny doing out there when Kontos would be better for this high leverage situation?") motivated me to check out his game logs to see when he's brought out to pitch, and I discovered an odd pattern.

On the one hand, it might be just what you expect from Bruce Bochy: in the eighteen games that Kontos has appeared in, the Giants are 6-12. Bochy doesn't yet trust the rookie.

But then it gets slightly weirder. Verging on a variation of Jim-Tracy-Colorado-Rockies-four-man-rotation weird. When Kontos debuted for the Giants, on June 10th against the Rangers, it began an idiosyncratic pattern: he's pitched in every game that Lincecum has started and not made it into the seventh inning since June 10th (6 games), and only once when he has (July 20th).

Lincecum's other starts during that period: June 27th against the Dodgers (7IP with the win), July 14th versus the Astros (8IP and what should have been his win) and last night vs. the Mets (7IP and the win).

Otherwise, Kontos is typically relieving Barry Zito, for a total of six appearances, with four of those being blowouts (in either direction).

The only real exception to this rule was the last Dodgers debacle, when Kontos appeared in all three games. If we subtract these, then eighty percent of his appearances follow Lincecum or Zito (12 of 15), and even if we don't, then it's 72 percent of the time.

The conclusion: the numbers suggest that Kontos ought to be used in situations with higher leverage. However, the Giants did not acquire additional bullpen help, and Bochy seems set on using Casilla, Affeldt, Romo and Lopez (CARL) in these spots. But there are plenty of times where it also appears that Bochy calls for Kontos in situations that would be more appropriate for Brad Penny (and vice versa), whose numbers are a little more situationally appropriate for eating up innings  rather than holding tied games (5.02 ERA, 1.32 WHIP, 2.5 SO/BB rate). Using Kontos in the seventh or eight innings might also provide more flexibility with CARL in save-by-committee situations, especially if Romo (I doubt it) or Casilla (...) continue to struggle.

No comments: