QED, Dept.
I see it, but I don't believe it

.

Dr. D zapped Igor in the armpit with a quick 20,000 to blow off a little frustration.  Then he went ahead and googled a very-carefully-thought-out phrase.  “Do pitching teams win more?” 

INSTANTLY he was buried in a mountain of old, crumpled Dixie cup articles proving, yes, of course pitching teams win more.  Just chew on this, for instance:  3% of World Series winners had below-average pitching staffs.  But 33% of them had below-average offenses.  And we got a lotsa mo where that came from.

....

A prelude:  Dr. D, and probably you, had ALWAYS argued that “the history of the World Series is written on the fabric of Cy Young starters.”  But he had also read “studies,” eleventy-nine times bare minimum, claiming that Greg Maddux and Roger Clemens don’t win you anything beyond what their WAR indicate.

So the shame runs deep, in DOV Labs Inc., that we’d been talked out of our love for rotations like Cliff Lee, Roy Halladay and Cole Hamels.  To say nothing of the Nationals’ last year.  Or the Braves of the 1990’s.   Poor Dr. D had given up even longing for rotations going Felix-Iwakuma-Greinke-Paxton.

....

EXHIBIT A (of A thru XZ)

Here is a 2011 Beyond the Box Score article that proves something that can’t possibly be true.  :- )  Mojo, do we file this in the Konspiracy Korner subdomain or the Sabermetrics area?

The Konspiracy being:  if you build around pitching and defense, you are (moderately) more likely to finish with a top-8 run differential.  (How in the world could that be true?  But it has been true.)

.

Pitching = Run Differential
Pitching = Run Differential

.

Not only that!  But given that you have a +100 run differential to start with, you (of course) win more games, the lower-scoring your games are.   If you score 900 and give up 800, you are scoring 1.12 runs for each run your opponent does.  But if you score 600 and give up 500, you’re hauling in a massive 1.20 runs per enemy plate-ster.  Compare a 112 OPS+ to a 120 OPS+.  It's like two standard deviations' worth.

Whoop!  Shoulda used ERA+ in that exampler.

.

EXHIBIT B

Hey, we're just gettin' loose.  That was just talking about what pitching does in the reg season.  Turns out this is way, WAY amplified in the postseason.

Here’s a 2011 Hardball Times article with a coy title:  Pitching (Almost) Always Wins Championships.  The exec sum would be,

  • Only 3 of 106 champions have done it with an ERA+ of less than 100
  • But!  33 of the 106 champions have sailed in with an OPS+ of <100
  • Since the Big Red Machine, only 5 of 33 champions have been hitting teams
  • The average champ (since 1900) had an OPS+ of 103 and an ERA+ of 114

For instance, the 2010 Giants had an offense of 95, a defense of 121, and were predicted by EVERYBODY to get wasted by the Phillies, if not the Braves.  What it turns out, though, is that the Giants had an absolutely model 1900-2010 World Series winner.  That is, they had Matt Cain, Tim Lincecum, Madison Bumgarner, and Jonathan Sanchez (3.05 ERA, 205 strikeouts).  They won the World Series, 4-1. 

The 2012 Giants won it in a clean sweep, 4-0.   Granted, their offense was 106 and pitching 96 this time, but EXHIBIT B is talking about the postseason.  The ’12 Giants had Bumgarner, they had Matt Cain whose postseason ERA is 2.10 lifetime, Lincecum whose life ERA is 2.40 in 13 games postseason, and Vogelsong was tough.

The 2014 Giants had, as even Dr. D recalls, the miracle stretch run from Madison Bumgarner with 53 IP and only 6 ER in October… including a shutout in the Wild Card game.  (You think a Wild Card slot is a dice roll, think again.  We got Felix.)

So, yeah.  It’s nice to be wrong.  Or right.  Or whichever.  If you like pitching, lean on back.  Loosen your belt buckle a notch.  It’s gonna be allll riiiggghhht.

Cheers,

Dr D

Comments

1

Some years ago you had a post that stimulated me to do some research.  While I can't remember the post, I remember the outcome of the research, or at least I think I do...  

My investigations surprised me, because I found that historically great hitting teams achieve an OPS+ of 115-120, but historically great run prevention teams have ERA+ values from 120-135.  For instance the historically great 1976 Reds had an OPS+ of 120, while the Phil Niekro led Braves of 1974 had an ERA+ of 124!?!

The higher maximum for run prevention also appears to be historically true.  The 1906 Cubs had an ERA+ of 151, while the 1927 Yankees Murder's Row had an OPS+ of 127. 

The great Cardinals teams of 1942-1944 had ERA+ of 135, 132, 134 in a league where the highest OPS+ in those three years was 107.

Between 1949 and 1964, the Yankees won 13 pennants.  The highest OPS+ they had during that time was 112.  Constrast this with the Braves of the 1990's, where their ERA+ from 1992 to 1999 exceeded 118 every year with a peak of 131.

++++++++++++++++++

While this is just a collection of anecdotal evidence, I am sure the thorough study would confirm that the extent to which run prevention can positively deviate from average is much larger than the range to which run scoring can positively deviate from average.

My explanation whould be that a well managed team with pitchers and fielders that can execute a game plan day-in and day-out suppress the randomness of baseball.  Think about it.  How you pitch a batter influences where the ball is hit and consequently where your fielders should be positioned.  That is a lot of potential control and also emphasizes the importance of controlling the strike zone.  Compare this level of control to the amount achieved by how you order your hitters.

Just my 2 cents.

 

2

The worst OPS+ in the AL last year was 90, by the Twins; the worst ERA+ was 85, by the Tigers.

In the NL the worst OPS+ and ERA+ were similar, but the best?  OPS+ high was 105, by LA with a $270M payroll; the ERA+ high was 135 (!!) by the Cardinals. 

I don't doubt your judgment that the dispersion for run prevention is broader.  Good stuff Dr. Kelly.

3

As to control ... agreed ... in baseball "the pitcher holds the ball."  In essence the pitcher is on offense, throwing a pass to a catcher who is like a WR, and the hitter is standing there like a cornerback with a stick.  Offense naturally dictates most of where the ball is going, whether you're talking NBA, NFL, NHL, whatever...

....

Baseball has that Pachinko machine / Pascal's triangle BABIP effect and maybe control of that is more important than we thought.  Not sure about this last point but it is provocative.

4

but I am struggling to see a better alternative.  If the pitcher and defense can execute a collective strategy, it could go beyond simply reducing BABIP.  People get mad when hitters won't go the other way against the shift, but it also means that the shift gets pull hitters out of their game.

Spectator's point about pitching depth (3-4 a game) versus hitting depth (need 9+ every night) is an intriguing alternative.  I buy this argument for the post season, not sure about the regular season. 

The relative importance of defense had been in consistent decline with the significant rise in strikeouts, but agressive defensive positioning based on balls in play data seems to have changed the trend.

5
  • Offenses must always use 9 hitters, 8 or 9 of whom must also play in the field -- and all 8 defensive positions must be covered.
  • Defense only needs to use one pitcher per game, and rarely more than 3 or 4, who may or may not have to hit a couple of times depending on the league/venue.
  • Even in the regular season, 5th starter can be skipped.  In the postseason, often the 4th and 5th starter can be skipped.  Teams with good starters can deploy their bullpens more strategically.
  • Offense can't skip the worst part of its lineup, ever.
  • Did you notice the Royals often had Alex Gordon hitting 6th or even 7th?  How many teams go that deep with tough outs? 
  • Hitters generate offense most effectively when they launch the ball hard in the air, but that also increases their susceptibility to strikeouts (except for the very most elite).  Good strikeout pitchers are very best at neutralizing exactly the most potent weapon of the offense.
6

That all makes perfect sense.  Now that I see it.  Especially your first point about the defensive spectrum - a good # of those hitters have to be fielders first.  ... of course, we could say that half the run-prevention unit has to be hitters first... still.

Fascinating to me, how the best-and-brightest SSI think tank'ers react to this light bulb.  Acknowledgment, open-mindedn reception, investigation, provocative attempts to explain.

Me personally, I felt pretty stupid having googled "Do pitching teams win more?"  :- )  Somehow I'd been talked out of this orientation.  For about 15 years.

7

The total staff impact over limiting BABIP.  Hmmmmmmmmm? 

Great staffs likely K more guys then run-of-the-mill staffs.  But you would assume they also give up less hard hit balls.  BABIP Allowed for those staffs should stand out.

And I'll add this, Doc:  I'm not sure which thread you mentioned it, and I have company coming and not enough time to search it up, but in the playoffs a 115 ERA+ staff becomes a 118 or 120 staff. Or better  Teams commonly enter the playoffs with 11 man staffs, and in some cases 10.  You've already eliminated your weakest arms. 

You can't quite eliminate your weakest positional bats (although rest is a smaller factor) but you can with your arms.

8

Without a doubt, Nolan Ryan - Jamie Moyer - Joaquin Benoit are going to give up less batted-ball velocity than Dave Burba ... or than Dr. D on Seniors Day...

Yep, I think we mentioned that in one of these three posts, that playoff teams roll their best pitchers at you for a disproportionate set of innings.

....

I'm annoyed with DiPoto over Iwakuma :- ) but the big picture is that --- > he not only gets this, but buys in harder than anybody around.  Very kewl.

9

Run prevention, scored on a percentage scale like ERA+, will not be linearly related to run scoring *proxy* data like OPS+ - the proxy won't pick up all of the ways in which runs are scored, and it will tend to be flatter.

Second cautionary note: When looking at the best teams at preventing runs, you will see a larger bias toward big percentages for pitching as well, and the reason has nothing to do with skill and everything to do with the way percentages work:

Take a league that averages allowing 650 runs.

One team adds 100 runs, which gets them 750 runs scored and 650 allowed.

One team subtracts 100 runs, which gets them 650 runs scored and 550 runs allowed.

The hitting team will have a pythag W% of .571 and an offense that is 15% better than normal, the pitching team will have a pythag W% of .583 and a defense that is 18% better than normal.

I think that is the main reason for the bias you folks have observed.

11

It should be noted that, in the post-season, the run scoring environment, for all time, has been about 3.75 R/G/Side, whereas in the regular season, the run scoring environment has been about 4.4 R/G/Side.

The lower the run scoring environment, the more important it is to have real skill in preventing runs, and I just showed why that is with my percentage comparison above.  The fewer runs you score, the more each run prevented improves your chance of winning.

Add comment

Filtered HTML

  • Web page addresses and e-mail addresses turn into links automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <blockquote> <code> <ul> <ol> <li> <dl> <dt> <dd><p><br>
  • Lines and paragraphs break automatically.

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.

shout_filter

  • Allowed HTML tags: <a> <em> <strong> <cite> <blockquote> <code> <ul> <ol> <li> <dl> <dt> <dd>
  • Lines and paragraphs break automatically.
  • Web page addresses and e-mail addresses turn into links automatically.