A closer look at the pitchers
Doing some browsing on the internet looking for something Pirate-related to talk about has finally brought me some inspiration today. While browsing the BBTF blog I found a link to this interesting post by the Baseball Crank. Using the new toys at Baseball Reference, BC compared a bunch of pitchers across generations to see how they pitched in their wins compared to their losses (just read the whole post, it's worth it). He found that as much as people bitch about wins and losses as a useful stat, things evened themselves out over time and in general even great pitchers pitch very well in their wins and poorly in their losses. This is incredibly intriguing to me, though I suppose maybe I shouldn't be that surprised. Still, career-wise, Shawn Chacon has a better ERA in games that he's won than Chris Carpenter, it's just that he pitches well enough to win on a much less consistent basis than Carp does. Anyways, I've decided to have some fun with these splits and a couple other interesting ones to me (OPS against by inning, specifically) on a much smaller scale, looking at the three pitchers that spent the whole year in the Pirates rotation (Duke, Maholm, and Snell). Specifically, I decided to look for what varied most drastically from wins to losses (it's unfortunate that it doesn't track no decisions, but we'll live for now). If at least part of what separates a great pitcher from an average (or an average one from a bad one) one is consistency, were the problems the Bucs' Big 3 had something they can control, or was it attributable to something like bad defense? If nothing else, it's just fun to look at stats like this.
Zach Duke
Key stats in 10 Ws: 72 and 1/3 IP, 2.74 ERA, 70 H, 3 HR, 45 K, 17 BB, .261/.315/.377 against
Key stats in 14 Ls: 88 and 2/3 IP, 6.50 ERA, 122 H, 12 HR, 48 K, 34 BB, .332/.391/.512 against
The two things that jump out, at least at me, are the walks (his BB/9 jumped from 2.12 to 3.45) and the home runs (HR/9 jumped from 0.37 to 1.22), two things that are entirely within his control (well, sometimes Pujols does what Pujols wants, but what I'm getting at is that you can't blame a fielder for a home run). Still, I'm not sure an extra walk and an extra homer every nine innings explains the huge disparity in numbers. Without being completely positive myself, it strikes me that holding the opponents to a SLG only about 110 points higher than their average is very good for a pitcher, whereas when they have a SLG about 200 points higher, things are not so good. Knowing that the Pirates were the second worse defense in the league on balls hit in the air (at least according to David Pinto's PMR rankings) perhaps at least some of Duke's struggles were due to poor defense (it would seem to me that balls hit in the air lead to extra base hits more often than balls hit on the ground). Still, knowing that his homer rate almost quadrupled in his losses, I don't know how much of that I buy as it would seem to be an indication he was just getting hit harder those nights. I suppose it's something to think about, if nothing else. If you scroll further down you'll see Duke's biggest problem in 2006, he got destroyed in the first inning, giving up at .951 OPS against. That's just plain bad.
Paul Maholm
Key stats in 8 W's: 52 and 2/3 IP, 2.73 ERA, 45 H, 3 HR, 29 K, 20 BB, .242/.333/.344 against
Key stats in 10 L's: 52 and 1/3 IP, 6.54 ERA, 70 H, 7 HR, 32 K, 28 BB, .324/.407/.500 against
Maholm is similar to Duke, though he really walks too many people all of the time (his BB/9 numbers jump from 3.42 to 4.81 in losses) for a guy that doesn't strike a ton of people out. His homers also more than doubled (0.51 per 9 to 1.20 per 9). We can make the same guesses about the defense that we do for Duke, and since the homer jump is smaller maybe it's an indication that the defense was in fact worse on nights Maholm lost, but there were some nights when Maholm simply got hit hard and it's hard to ignore that fact having watched a bunch of those games. Maholm also got killed in the first inning, giving up a whopping .995 OPS in the first and a .950 OPS in the third.
Ian Snell
Key stats in 14 W's: 90 and 2/3 IP, 2.28 ERA, 68 H, 8 HR, 87 K, 33 BB, .210/.282/.336 against
Key stats in 11 L's: 57 and 2/3 IP, 7.34 ERA, 79 H, 14 HR, 48 K, 21 BB, .333/.385/.574 against
Now here's some interesting stuff. Snell was better than Maholm or Duke in his wins and worse in his losses. His walks weren't the problem, they were 3.28/9 in both wins and losses. Instead, the problem for Snell in his losses seems to be a drop in strikeouts (8.64/9 in wins compared to 7.49/9 in losses) and a jump in homers (0.79/9 in wins to 2.18/9 in losses). A drop in strikeouts means more balls in play for the defense to screw up, but a huge jump in home run rate means Snell is getting shelled. Snell also had a massive OPS against in the first inning with opponents clocking in with a .990 rate. That mythical wall we talked about Snell hitting so many times is also evident in his splits, his OPS against was .905 in the 5th inning and .972 in the 6th inning.
So do we have a conclusion? It seems likely that some bad defense (probably bad flyball defense) cost the Pirates pitchers some bases during the season. Pinto's PMR rating only takes whether balls were recorded as outs into effect, but we can assume that if the Pirates were poor at turning fly balls and line drives into outs then they were also poor at keeping singles from turning into doubles and doubles from turning into triples, especially with the ridiculous set of outfield arms we trotted out this year. But the three pitchers were also maddeningly inconsistent. Maholm and Duke both walked more batters in their losses and all three of them gave up at least twice as many gopher balls in their losses than the did in their wins. Still, the most interesting split was one I didn't even intend to look at today; the first inning OPS against for all three pitchers was over .950. Since it dropped over the bulk of the game for all three of them, it would seem to me that other teams were simply incredibly prepared for our pitchers and they weren't ready for the opposition. They all adjusted well (save Snell, who got rocked the third time through the lineup pretty regularly), so I'm going to have to say that the advance scouting of Kent Tekulve and whatever advice Colborn was giving them from the tapes they watched was pretty useless this year, at least as far as helping our pitchers prepare for opposing batters. In fact, it might've been more hurtful than the defense was.
Zach Duke
Key stats in 10 Ws: 72 and 1/3 IP, 2.74 ERA, 70 H, 3 HR, 45 K, 17 BB, .261/.315/.377 against
Key stats in 14 Ls: 88 and 2/3 IP, 6.50 ERA, 122 H, 12 HR, 48 K, 34 BB, .332/.391/.512 against
The two things that jump out, at least at me, are the walks (his BB/9 jumped from 2.12 to 3.45) and the home runs (HR/9 jumped from 0.37 to 1.22), two things that are entirely within his control (well, sometimes Pujols does what Pujols wants, but what I'm getting at is that you can't blame a fielder for a home run). Still, I'm not sure an extra walk and an extra homer every nine innings explains the huge disparity in numbers. Without being completely positive myself, it strikes me that holding the opponents to a SLG only about 110 points higher than their average is very good for a pitcher, whereas when they have a SLG about 200 points higher, things are not so good. Knowing that the Pirates were the second worse defense in the league on balls hit in the air (at least according to David Pinto's PMR rankings) perhaps at least some of Duke's struggles were due to poor defense (it would seem to me that balls hit in the air lead to extra base hits more often than balls hit on the ground). Still, knowing that his homer rate almost quadrupled in his losses, I don't know how much of that I buy as it would seem to be an indication he was just getting hit harder those nights. I suppose it's something to think about, if nothing else. If you scroll further down you'll see Duke's biggest problem in 2006, he got destroyed in the first inning, giving up at .951 OPS against. That's just plain bad.
Paul Maholm
Key stats in 8 W's: 52 and 2/3 IP, 2.73 ERA, 45 H, 3 HR, 29 K, 20 BB, .242/.333/.344 against
Key stats in 10 L's: 52 and 1/3 IP, 6.54 ERA, 70 H, 7 HR, 32 K, 28 BB, .324/.407/.500 against
Maholm is similar to Duke, though he really walks too many people all of the time (his BB/9 numbers jump from 3.42 to 4.81 in losses) for a guy that doesn't strike a ton of people out. His homers also more than doubled (0.51 per 9 to 1.20 per 9). We can make the same guesses about the defense that we do for Duke, and since the homer jump is smaller maybe it's an indication that the defense was in fact worse on nights Maholm lost, but there were some nights when Maholm simply got hit hard and it's hard to ignore that fact having watched a bunch of those games. Maholm also got killed in the first inning, giving up a whopping .995 OPS in the first and a .950 OPS in the third.
Ian Snell
Key stats in 14 W's: 90 and 2/3 IP, 2.28 ERA, 68 H, 8 HR, 87 K, 33 BB, .210/.282/.336 against
Key stats in 11 L's: 57 and 2/3 IP, 7.34 ERA, 79 H, 14 HR, 48 K, 21 BB, .333/.385/.574 against
Now here's some interesting stuff. Snell was better than Maholm or Duke in his wins and worse in his losses. His walks weren't the problem, they were 3.28/9 in both wins and losses. Instead, the problem for Snell in his losses seems to be a drop in strikeouts (8.64/9 in wins compared to 7.49/9 in losses) and a jump in homers (0.79/9 in wins to 2.18/9 in losses). A drop in strikeouts means more balls in play for the defense to screw up, but a huge jump in home run rate means Snell is getting shelled. Snell also had a massive OPS against in the first inning with opponents clocking in with a .990 rate. That mythical wall we talked about Snell hitting so many times is also evident in his splits, his OPS against was .905 in the 5th inning and .972 in the 6th inning.
So do we have a conclusion? It seems likely that some bad defense (probably bad flyball defense) cost the Pirates pitchers some bases during the season. Pinto's PMR rating only takes whether balls were recorded as outs into effect, but we can assume that if the Pirates were poor at turning fly balls and line drives into outs then they were also poor at keeping singles from turning into doubles and doubles from turning into triples, especially with the ridiculous set of outfield arms we trotted out this year. But the three pitchers were also maddeningly inconsistent. Maholm and Duke both walked more batters in their losses and all three of them gave up at least twice as many gopher balls in their losses than the did in their wins. Still, the most interesting split was one I didn't even intend to look at today; the first inning OPS against for all three pitchers was over .950. Since it dropped over the bulk of the game for all three of them, it would seem to me that other teams were simply incredibly prepared for our pitchers and they weren't ready for the opposition. They all adjusted well (save Snell, who got rocked the third time through the lineup pretty regularly), so I'm going to have to say that the advance scouting of Kent Tekulve and whatever advice Colborn was giving them from the tapes they watched was pretty useless this year, at least as far as helping our pitchers prepare for opposing batters. In fact, it might've been more hurtful than the defense was.