Surprisingly, yes. That guy is good.
As of this writing, the Mets still haven’t found that front-line starting pitcher that everybody’s looking for, but Omar’s plan to corner the market on backup catchers is coming along swimmingly. The dearth of Mets prospects still makes it unlikely that the team will be able outbid the Red Sox-Angels-Dodgers for Johan Santana, Dan Haren, or Erik Bedard; which leaves the Mets with a slowly dwindling supply of somewhat uninspiring free agents. We’ve shown that Livan Hernandez is no kind of solution in the rotation, but what about Carlos Silva? He’s still available, but is he any good?
Below, you can see the career peripheral rates (K/9, HR/9, etc., which are often better predictors of future success than ERA or win-loss records) for Carlos Silva, first as a reliever with Philadelphia in 2002 and 2003 and then as a starter for the Twins since 2004. Playing for the Twins in the AL Central, Silva wasn’t somebody that I knew a lot about. I vaguely remember him pitching against the Mets as a middle reliever, but those were dark days and I do my best to pretend they never happened. Finesse pitcher, I recalled. Not many strikeouts, but not many walks, either.
Wow, is that an understatement. My initial response when confronted by a starter who throws less than four strikeouts per nine innings is to run like hell. Last year Livan Hernandez threw 4.0 K/9—higher than Silva’s career 3.8 K/9—and I just got done telling you what a bum he is. So how the hell is Carlos Silva any good? The answer, of course lies in his even-more-absurdly-low walk rate.
These are Silva’s peripheral rates, expressed as a percentage better or worse than average for AL starters from 2004 through 2007 (we’re interested in Silva as a starter, so I only looked at those years). Silva strikes batters out at a rate 38% worse than the average AL starter, which is terrible, and ordinarily I would say that rate is unsustainable, but he also walks batters at a 45% better rate (meaning less). His strikeout-to-walk ratio is actually about 13% better than league average.
Silva had one really excellent year, 2005, in which he struck out nearly eight times as many as he walked. That’s elite pitching. And he did it, not by striking out more batters, but by walking even fewer, only 0.4 BB/9. He issued only nine walks that whole year—and two of them were intentional! That’s insane.
Strikeouts are an important indicator of pitching ability, but they’re by no means the only one. Tom Tippett wrote an article over at Diamond Mind Baseball back in 2003, as a refutation (or modification) of Voros McCracken’s then-controversial claim that pitchers had little-or-no control over whether or not a ball put into play would become a hit. Most of us are familiar with this idea now, either from Moneyball, DIPS (Defense-Independent Pitching Stats), BABIP (Batting Average on Balls In Play), DER (Defensive Efficiency Ratio) or, uh . . . this website, and accept it as true.
By looking at the career profiles of various pitchers to determine whether or not they had some influence on BABIP (it turns out they do), Tippett also showed how those pitchers had been successful. Some, like Pedro Martinez or Roger Clemens, succeeded with lots of strikeouts. Others found success by limiting their home runs (Tommy John) or walks allowed (Greg Maddux). Some even found success despite a high number of walks by limiting the base hits allowed on balls put into play (rough, tough Charlie Hough). The point is there’s always more than one way to skin a cat. As long as Silva can couple his few strikeouts with even fewer walks, he can be successful.
The danger for a pitcher who allows so many balls to be put into play, of course, is that bad luck will have more opportunities to affect him than a pitcher who strikes out a lot of batters. Last year, Silva faced 848 batters, and they managed to put the ball in play 82% of the time. By way of comparison, Oliver Perez faced 765 batters, but they only managed to put it in play 63% of the time. At those rates, if they each face 830 batters in 2008 (about 32 starts’ worth), Carlos will have 160 more opportunities to get bitten by a dying quail.
Of course, the flip side of that is that Carlos would also have 160 more opportunities to get lucky, too. In fact, one could make the argument that a good defensive team—like the Mets, with Carlos Beltran in center and Luis Castillo, Jose Reyes, and Gold Glover David Wright in the infield—is better off allowing batters to put the ball in play more often, because it gives the defense more opportunities to shine. Also, groundballs are more democratic.
The danger is that Silva’s walk rate wouldn’t have to climb much before it would match his strikeout rate and make him useless. It’s a pretty fine line that he’s got to walk, and without many strikeouts to fall back on, he doesn’t have much room for error. He’s still on the right side of thirty, though, so I have some faith that he’ll be able to maintain his present level of effectiveness, or something close to it. Combined with the fact that Silva would be coming to the slightly weaker and DH-less National League, I think he’d be a fine addition to the Mets rotation. Or at any rate, a better option than most, and one the Mets could acquire without shipping off any more young talent.
Special thanks, as always, to Baseball Reference, the Hardball Times, and David Pinto for making it possible for me to disseminate my unqualified opinions to the masses.