June 23, 2006
BASEBALL: The Elements of A Great Pitching Team
Recall my comparison of the 2006 Tigers to the all-time great pitching teams (as measured by ERA compared to the league average). The Tigers were, at that writing, 39.9% better than the league, good for sixth best all-time; in coming back down to earth, they have settled at this writing to 31% better (3.52 ERA compared to 4.61 for the AL as a whole), which would put them just off the top 25, although that's still a plenty impressive place to be. You will also recall that I found that the Tigers' pitching success likely owes a lot to their defense.
Which got me wondering: if you look at the other top pitching teams - this time I expanded the list to the top 27 teams other than the Tigers, to include everyone since 1893 who ended 30% or better compared to the league - what most characterizes their great pitching? I decided to list out where each of these teams ranked vs. the league in the four main components of pitching: the three that the pitcher controls (preventing HR and BB and striking guys out), and the one that is mainly the responsibility of the defense (Defensive Efficiency, i.e., the percentage of balls in play turned to outs). As usual, I looked at Baseball-Reference.com for the answers. Let's see how these teams did it - "Lg" is how many teams were in the team's league at the time, and the rest are ranks from best to worst:
|1964 White Sox||10||2||1||8||1|
|1905 White Sox||8||1||2||5||1|
|1967 White Sox||10||1||2||7||1|
*Well, the obvious one that jumps out at me is how much more crucial the defense has been to making a great pitching team than any other element. I was stunned, frankly, that the results were this dramatic. Of the 28 teams, five finished in the bottom half of the league in strikeouts (including two 1960s White Sox teams), four in walks, two in homers (both pre-1920 teams), but not a one in DER, with the 1919 Cubs and 1912 Giants the only ones as low as fourth.
*More studies are still needed on which of these four elements are most subject to park effects.
*Granting that they had a lot of help from their park, the 1966 Dodgers were unsurprisingly the one team that swept the three pitching-only categories, unsurprising because - last I checked, in 2001 - they held the all-time (since the 4-ball/3-strike era began in 1889) record for best team K/BB ratio.
*Looking at the top 4 teams gives you more sympathy for the voters who put Tinker, Evers and Chance in the Hall of Fame. Ditto the voters who gave Marty Marion the 1944 MVP and put Luis Aparicio in Cooperstown.
*The 1939 Yankees' defense is also discussed here in the context of a broader look at the centrality of defense to the great Yankee teams of the 40s through 60s.
*Bear in mind as well that defense was more crucial before homers, walks and strikeouts became as commonn as they are today.
Amazing. I'm shocked too.
One thing that strikes me: yet *another* bit of ammo for the argument that the '39 Yanks are the greatest team ever. We knew about the picthing & the offense. Now this?
Seems that every year that sabermetrics advances, we come closer & closer to the conclusion that the '39 Yanks (as a one-year team and the culmination of the 4 year run), and not the '27 Yanks or '75 Reds or '70 O's, are the greatest team of all time.
The most important cog to the Braves success of the late 90s was Andruw Jones' brilliance in CF as much as it was Leo Mazzone's pitching and Bobby Cox's managing. As we suffer through the quagmire that is the '06 season, we're seeing our mediocre pitchers and abysmal relievers throw fatties that result in rocket shots that not even the great Andruw can catch up to.
Which supports my argument that unless Big Papi hits around .400 with 70 homers and 170 RBI, it's laughable to include him (or any DH) in any MVP conversation with the likes of A-Rod, Tejada or Vlad. But, then again, that contention is pushed by the likes of ESPN so it's not surprising.
DER isn't a perfect indicator of fielding. Batted ball types and zones enter into the picture a lot. The best correlation I've found between DER and the best fielding stats, such as UZR, is .50. Good, but still not enough to say that a specific team was a great fielding team.
IMO, the more proper conclusion is that great pitching teams either have great fielders, or pitchers that give up more hittable balls, or just plain luck.
Also, IMO, the best thing to do is calculate DER over a couple of years, which would at least take luck out of the equation (hopefully).
Here's a link to some analysis I performed on this:
Plus, here's my analysis of great fielding teams, with the DER caveat (and noting your previous research!):
Also overlooking a hard-to-quantify part of defense, which is defensive preparedness (for lack of a better term).
That is, teams are better defensively because their pitchers are throwing strikes and they are mentally prepared. A pitcher who labors and throws lots of balls puts his defense on it's heels. Somebody like Maddux or Buerhle, who gets ahead early and has lots of 2 and 3-pitch ABs keeps their defense fresh and free of mental fatigue. Believe me, it happens.
[XP'd on Baseball Musings]
Interesting that rate of conversion of batted balls to outs is the most important factor. Though, as noted by the Crank, most of these teams played in an era where BB, K & HR were much rarer than they are today. HR might become the most important factor if the study focused on a more recent time period (e.g. past decade).
It's fairly clear from Crank's study that low ERA and high DER are correlated, but there's no indication that low ERA is caused by high DER. It could be the reverse: that the best pitchers tend to induce the easiest-to-field balls (low ERA causes high DER). Causation is a tricky beast.
Anyway, cool study!
I see this as pretty unremarkable.
How could this not be true? The only way is if some good pitching teams got some inordinate number of their outs by K. This strikes me as pretty unremarkable. DER measures the out rate (minus BB, K and HR all of which come in far smaller numbers than hits). Great pitching teams surely get outs at a very high rate. This pretty much says nothing about how they get them (quality defense, quality scouting, easier to field batted balls).
In thinking about this in a very unscientific way, I am trying to reconcile good pitching on bad teams, and good pitching on good teams. I think the basic assumption is that good teams both his and field well, and bad teams may do one or the other, but never both.
Tom Seaver and Steve Carlton come to mind as good pitchers on bad teams. Wilbur Wood for that matter too, but Seaver and Carlton both won zillions of games when the teams around them lost with everyone else (especially Carlton). Gaylord Perry also pitched great for some stinkers, but he lost more too.
So it stands to reason that the Mets and Phils both hit poorly, which we know. I don't really recall how they fielded, I must have put the Mets then out of my memory and who can blame me. Seaver and Carlton were both great strikeout pitchers who walked few, thereby giving fewer chances for the fielders to screw up. I can see Maddux losing big time for those teams--he made you hit what he wanted you to, but the fielders had to do their jobs as well. Also, like Palmer, they gave up some home runs, but not generally 3 run homers. They limited the damage.
It doesn't answer the basic question of course. You put Seaver on the 1939 Yankees and he wins 32 games; but you put him on the 1962 Mets and he probably wins 17--OK, 14 (I didn't put them totally out of my memory I guess, they were really awful).
It does seem that the K and BB rates may be the most important categories for what a pitcher can do to affect a game: limit just what kind of damage any fielder can do. So I would want my pitchers to throw lots of strikes, and work quickly. Sounds kind of trite I guess. Cliches are cliches for a reason. And why Nolan Ryan ws not as good as he could have been: too many walks.