This ended up being a lot longer than expected so we’re going to do a three parter. The premise of part 2: winning a world series is all about pitching. It takes a big jump in pitching production from the year before to win the championship and in the last ten years that jump has been unsustainable for reasons I will try to explain in part 3.
I recently reread Baseball Behind the Numbers by the Baseball Prospectus guys. Having just watched the Giants surprising postseason run I was eager to review the study where they unveil the factors that lead most directly to success in the postseason.
According to BP, the only four factors that have any kind of statistical significance (and it’s not that much statistically speaking) are: 1) strong starting pitching (and more specifically, power pitchers who can strike batters out and get swings and misses), 2) dominant relief pitching (especially from the closer), 3) defense that doesn’t make mistakes (doesn’t have to be great in terms of range but NO errors), 4) to a much lesser extent than the first three, power hitting (i.e. home runs).
Just based on that you can see why the Giants 2010 run makes a lot of sense. Their findings form the basis for the premise of this post and the importance of pitching:
Winning in the postseason, and thus the World Series, is, based on the stats, all about pitching. My observation is that over the last ten years the WS champions have seen a major addition or jump forward in their pitching talent, and then an inability to sustain that the next year. I am way more adept at reading statistical baseball analysis than performing it, so what follows is pretty elementary, but hopefully you will see my point.
Let’s look at the last 10 World Series Champions, how their pitching fared in the year before, the year of, and the year after their win, and their key staff additions (through trades, free agency, or a young talent emerging). I’ll be using K/9, ERA, ERA+ (note that an era+ score of 100 is league average), and WHIP data from BaseballReference.com.
2001 Arizona Diamondbacks
- 2000 83-79 (3rd in NL West): 7.6 K/9, 4.35 ERA, 110 ERA +, 1.34 WHIP
- 2001 92-70 (1st): 8.0, 3.87, 121 (11 pt increase), 1.24
- 2002 98-64 (1st): 8.1, 3.92, 117 (4 pt drop), 1.23
The Diamondbacks had 2 key additions in 2001: a full season of Curt Schilling and Miguel Bautista. They also had some great subtractions (no Omar Daal). They actually had a better regular season in their repeat year, but got swept out of the first round of the playoffs by the Cardinals. Thus begins the trend I hope to highlight here: it is very difficult to get the same production out of a pitching staff two post-seasons in a row. Especially staffs that are super top-heavy (It was Johnson and Schilling and then smoke and mirrors the rest of the way).
2002 Anaheim Angels
- 2001 75-87 (3rd in AL West): 5.9 K/9, 4.20 ERA, 108 ERA+, 1.38 WHIP
- 2002 (99-63 Wild Card): 6.2, 3.69, 120 (12 pt increase), 1.28
- 2003 77-5 (3rd): 6.2, 4.28, 103 (17 pt drop!), 1.35
A classic case of key additions (Kevin Appier which sent Scott Schoenwies to the bullpen) and the emergence of young talent (John Lackey, K-Rod, Scot Shields) providing a huge, but unsustainable, bump in production. The 2002 Angels actually saw a bigger salary increase from 2001 to 2002 than the infamous 1997 Marlins team. As a result, they got better in a lot of areas, not just pitching. In 2004 they saw a similar bump when they added Bartolo Colon and Kelvim Escobar and won the division before getting bounced from the playoffs in dramatic fashion by David Ortiz (I mean the Boston Red Sox).
2003 Florida Marlins
- 2002 79-83 (4th in NL East): 6.8 K/9, 4.36 ERA, 93 ERA+, 1.43 WHIP
- 2003 91-71 (Wild Card): 7.0, 4.04, 105 (12 pt increase), 1.35
- 2004 83-79 (3rd): 7.0, 4.10, 101 (4 pt drop), 133
Looking at this makes me mad because it reinforces what a flukey team this was. They were obviously helped by the additions of Donrelle Willis, Matt Redmond, and Uegeth Urbina as well as getting full/emergent seasons from Carl Pavano and Josh Beckett. But they were essentially the same (fairly average) staff all three years with a small spike in performance in 2003. One caveat (and some foreshadowing) injuries really hurt them in 2004 as none of their 5 main starters pitched a full season.
2004 Boston Red Sox
- 2003 95-67 (Wild Card): 7.0 K/9, 4.48 ERA, 104 ERA+, 1.36 WHIP
- 2004 98-64 (Wild Card): 7.0, 4.18, 117 (13 pt increase), 1.29
- 2005 95-67 (Wild Card): 6.0, 4.74, 96 (21 pt drop!), 1.39
The Red Sox were really good through this three-year window, yet they represent the best example so far of the cycle of major additions, big bump in production, followed by post-championship drop off. The 2004 team saw a huge jump because they added Curt Schilling and Keith Foulke. And the 2005 team suffered greatly the loss of Schilling (injury kept him out most of the season) and Pedro Martinez to the Mets. The 2005 additions of David Wells and Matt Clement weren’t enough to overcome the letdown. (The ERA numbers also show how hard it is to pitch in the Al East.)
2005 Chicago White Sox
- 2004 83-79 (2nd in AL Central): 6.4 K/9, 4.91 ERA, 97 ERA+, 1.42 WHIP
- 2005 99-63 (1st in AL Central): 6.3, 3.61, 125 (28 pt increase!), 1.25
- 2006 90-72 (3rd): 6.3, 4.61, 103 (22 pt drop!), 1.36
The White Sox trump the ’04 Red Sox in my pitching cycle theory. The addition of Orlando Hernandez, healthy/full seasons from Freddy Garcia and Jose Contreras, and the emergence of Jon Garland and Bobby Jenks help partially account for this crazy spike in ERA+ followed by a tremendous drop. The interesting thing about the White Sox is how the K rate stayed the same, but the other numbers changed so dramatically. This probably shows that in ’05 they had an exceptional year defensively and got a bit lucky (if I had more time I’d look into some other numbers like BABIP). The 2005 Chicago White Sox: the poster-children for my pitching theory.
2006 St. Louis Cardinals
- 2005 100-62 (1st in NL Central): 6.1 K/9, 3.49 ERA, 122 ERA+, 1.27 WHIP
- 2006 83-78 (1st): 6.1, 4.54, 98 (24 pt drop!), 1.38
- 2007 78-84 (3rd): 5.9, 4.65, 95 (3 pt drop), 1.41
If the ’05 White Sox are my well-behaved/front row students, then the ’06 Cardinals are my unruly/spit wad shooting/back row/trouble makers. Thanks for killing the drill guys. That said, keep in mind that the Cards had GREAT teams in 2004 and 2005 and that the ’06 team was basically the same core. Also, you do see a drop in the repeat year, mostly having to do with losing Chris Carpenter for the season (notice a big injury following a Championship has a lot to do with the depressed repeat season stats). However you slice it, this was a weird Championship team, period.
2007 Red Sox
- 2006 86-76 (3rd in AL East): 6.7 K/9, 4.83 ERA, 99 ERA+, 1.44 WHIP
- 2007 96-66 (1st): 7.2, 3.87, 123 (24 pt increase), 1.27
- 2008 95-67 (Wild Card): 7.4, 4.01, 116 (7 pt drop), 1.33
The Red Sox stats show that they had the best chance of anyone to repeat in the last 10 years. They went to 7 games against the Rays in the 2008 ALCS and from their statistical steadiness that should be no surprise. The big additions in ’07 were the Japanese imports (Matsuzaka and Okajima), plus Becket had the best year of his career as he fully adjusted to life in the AL East. Papelbon also benefited from his first full season in the closer role. The Sox did add a full season of John Lester in ’08 which helped, but the big problem with repeating was Becket broke down as the season wore on and pitched with half a shoulder in playoffs. Again, the injury problems to key starters in the year after a WS win is a major reason why teams have struggled to repeat.
2008 Philadelphia Phillies
- 2007 89-73 (1st in NL East): 6.5 K/9, 4.73 ERA, 97 ERA+, 1.45 WHIP
- 2008 92-70 (1st): 6.7, 3.88, 113 (16 pt increase), 1.36
- 2009 93-69 (1st): 7.1, 4.16, 101 (12 pt drop), 1.35
Record wise, the Phillies have gotten better each of the last 4 years (including 2010). And they seem to be the team that takes my theory “most seriously,” having added Cliff Lee in 2009 to try to repeat, Roy Halladay/Oswalt in 2010 to get back on top, and Cliff Lee again this offseason. Hats off for improving the most important part of the team four years running.
The 2008 championship team featured the key addition of Brad Lidge (who had the year of his life posting a 226 ERA+) which allowed them to put Brett Myers back in the rotation. 2008 also saw the emergence of Cole Hamels (ERA+ of 142). However, despite the addition of Lee in ’09, Lidge had a major setback that year posting an abysmal ERA+ of 59, and Hamels struggled with mental and physical issues all year (ERA+ of 97). Bullpen struggles have really hampered them the last two seasons.
2009 New York Yankees
- 2008 89-73 (3rd in AL East): 7.1 K/9, 4.28 ERA, 104 ERA+, 1.36 WHIP
- 2009 103-59 (1st): 7.8, 4.26, 108 (4 pt increase), 1.35
- 2010 95-67 (Wild Card): 7.2, 4.06, 106 (2 pt drop), 1.31
The Yankees are an interesting case. First, it’s clear that their success is more closely tied to offense than any team in the last 10 years. Despite some significant additions over the last three years, they have had essentially the same staff production each season. Each staff had one guy who had a great year (Mike Mussina in 2008…look it up, it was legit…and then CC Sabathia in ’09 and ‘10), one other guy who stepped up as a legitimate number 2 (Andy Pettitte in ’08, AJ Burnett in ’09, Phil Hughes in ‘10). The big difference in 2009 was Pettitte offered a solid third option, which they sorely lacked in 2008 and 2010. 2010 had a lot to do with injuries, per usual, with Pettitte/Javier Vasquez/Burnett all missing time or lacking effectiveness due to being hurt.
That’s it for part 2. Sorry for the length! Tomorrow we’ll wrap it up by looking at the 2010 Giants, drawing some conclusions, and deciding if this spells doom or not for 2011 Giants.
I’ve been writing about this for a number of years now, and in fact, I have a Giants business plan posted which incorporates these findings as part of the strategy for building a team for a couple of years now. I love the analysis you did on each World Series winner, very interesting take on how to look at things.
I have some quibbles, though, with your interpretation of BP’s findings: Baseball Prospectus did not say that home runs is a factor in winning and, much smaller quibble, the bullpen is not as important as you noted.
First, let me clarify what BP found: NO offensive measure is found to be a statistically significant factor in a team advancing deeper into the playoffs. Stolen Base Attempts is the only one that came close to being a factor, perhaps representing how speed on the team adds value to a team through more base hits, better base running, and better defense.
In fact, their correlation calculation found that runs scored had ZERO correlation with doing well in the playoffs.
HR do have some correlation, but BP did not point it out as a statistically significant factor. Here is what they wrote about offense: “Just as run production as a whole hasn’t had much relationship with playoff success, neither have any of the individual offensive metrics. The A’s postseason struggles have sometimes been attributed to their tendency to rely on walks and home runs, but there is no evidence that teams that play Smallball instead fare better in the postseason. Although stolen-base attempts have a slight (but statistically insignificant) positive relationship with [playoff success], sacrifice-hit attempts have a slight negative one. Speed Score, a composite of five different offensive statistics that provides evidence about a player’s wheels, has no relationship with [playoff success] at all. Nor do teams that hit well in the clutch in the regular season see that advantage carry forward into the playoffs.”
Regarding the bullpen, it is only the closer that they said was key to playoff success, as BP noted, “… the performance of non-closer relievers is of very little performance in the postseason.”
Yeah, agreed. I have to admit I simplified/dumbed down the BP analysis a little bit. Personally, I do think home runs in the post season are important even if they might not be statistically significant. Maybe I’m just being naive there. And I generalized the closer issue to the bullpen intentionally. Trying to make the broader point about pitching, and pitchers who strike batters out, to be more specific.
Re-reading, to your point, it is the staff’s overall K/9 that was important, and while that for the most part is weighted heavily by the starters, the relievers is a significant part of that too. So the bullpen has an affect there.
And I generally agree that the bullpen is important – Tom Tippett’s studies in the mid-2000’s showed how playoff teams had to have good bullpens – but for the purposes of BP’s study, the bullpen has a small but significant contribution to the metrics that they settled upon, via the K/9, and of course, the closer controls the WRXL.
What BP said was that they identified three factors which is tied to playoff success: pitching strikeout rate (which covers starting and relieving, but obviously dominated by starters); closer (as measured by BP’s WRXL metric); and fielding (as measured by BP’s FRAA metric).
With strikeout artists in Lincecum and Sanchez, plus steadily good Cain, and now Bumgarner, who looks to be steadily good at worse, or at best, making a top three, plus a strikeout bullpen led by Wilson, the Giants look to be pretty set at a high K/9 in 2011 again.
Wilson just looks to be dominant again (something Beane has still yet to learn, he still thinks he can easily find a closer to replace the last one he traded away).
And defense looks to be no worse than last season, perhaps better with Posey, a slimmer Sandoval, plus Torres and Ross full season in outfield, and DeRosa as super-utility guy.
That’s been Sabean’s formula for a while, he’s always looked for a good closer and good team defense, and now he has the strong pitching staff in terms of K/9. Wheeler and Runzler look to continue that for us going forward as well.
Do you see Runzler as a starter or reliever?
He’s probably gone now, but I’ve only viewed Runzler as a reliever. They were only starting him so that he could get to throw more pitches in game situations. Had he showed any progress, they could have brought him up as long relief plus some middle relief work, but unfortunately, his early success never returned, it was like he was now pitching to avoid failure, not achieve success.
Interesting. I need to ponder this a bit more before commenting further. Very in depth analysis
Pingback: Madison Bumgarner and the “Verducci” Effect « baseballmonk
Pingback: Why the Red Sox Will Not Win the World Series « baseballmonk
Pingback: The National League « baseballmonk
Pingback: 2011 AL West Predictions « baseballmonk
Pingback: Forecasting the Future (or something like that) « baseballmonk
Pingback: Contenders vs. Pretenders « baseballmonk
Pingback: Contenders vs. Pretenders, Pt. III « baseballmonk
Pingback: More Bad Offense! « baseballmonk
Pingback: Playoffs and Pitching « baseballmonk
Pingback: A Couple of Classics « baseballmonk
Pingback: Life With and Life WithOut Pablo #SFGiants #PabloSandoval #Pitching | baseballmonk