Tuesday, November 19, 2013

Reasons MLB Run Scoring is So Low

People have differing reactions to the statement above.

First, some context...

Average Runs Scored per Game
1919: 3.88 R/G
1920: 4.36 R/G
Babe Ruth, in 1930, hit 49 home runs
with a .493 OBP (Getty Images)
1930: 5.55 R/G
1940: 4.68 R/G
1950: 4.85 R/G
1961: 4.53 R/G
1969: 4.07 R/G
1984: 4.26 R/G
1992: 4.12 R/G
1993: 4.60 R/G
1998: 4.79 R/G
2003: 4.73 R/G
2008: 4.65 R/G
2012: 4.32 R/G
2013: 4.17 R/G

What was your first reachtion to hearing that runs scored per game are lower than they've been in 20 years? I'm sure it wasn't that they raised the mound.

Some might jump to the conclusion that there are fewer steroids in baseball. That's probably a factor, but not more than half of the issue. My mind jumped to the idea that the strike zone has expanded noticeably the past four or five years.

I asked Toirtap what he thought about it:



I agree with Toirtap. Forcing your mind to boil things down to one factor is frustrating and closed-minded. Toirtap's initial claim that "ever-rising K rate is big" is true, but it's important to note that a larger strike zone helps K rates rise.

This post is designed to spark a discussion, rather than a lecture.

Below is a quick attempt to rank a few factors:
  1. Larger strike zone (fuels rising K-rate)
  2. Improved bullpens (fuels rising K-rate)
  3. Improved defensive strategy
  4. Less "juiced" baseballs
  5. Fewer bulked-up power hitters (steroids)
  6. Deterioration of plate discipline (fuels rising K-rate)
  7. Worse weather
  8. Yuniesky Betancourt's 409 plate appearances
  9. Alcides Escobar's 642 plate appearances
  10. Clayton Kershaw

Photo: Steve Mitchell, USA TODAY Sports

If I cared more about page views than the reader experience, I totally would have made this a slide show. Thanks for stopping by!

Sunday, November 17, 2013

3-Year Batting, Pitching and Fielding Stats

Congratulations to the Boston Red Sox on winning their third World Series in ten years. Narrative and praise can be bestowed upon any player, manager, coach, or executive, for their success the past decade. Many in the media have praised David Ortiz and Jonny Gomes for rallying the team with enthusiasm. Sure, encouraging players to focus and feel good through camaraderie exist and are a positive influence, but no-one knows how much influence. The acts of throwing, hitting, or fielding a baseball are affected by players' psyches at varying degrees. If we can just make up stories about the effectiveness of motivational speaking, I'll give a majority of the credit for their 3 most recent championships to Bill James. Having Bill James whispering sabermetric gold to front office executives should be the envy of every franchise.

Bill James, with a cup of Fenway dirt, after the 2013 World Series
This year's Red Sox gave many franchises hope for quick recoveries after catastrophic seasons, like the Bobby V Experience of 2012. Doesn't it feel like Valentine was hired way more than two years ago?


Before Michael Lewis wrote "Moneyball", Bill James pointed out that using three-year averages and trends of predictive statistics was more effective than primarily considering stats from the previous season.

Spend some time with Baseball-Reference or Fangraphs, and you're bound to fall into a rabbit hole of leader boards and player pages. Since you're on this blog, it's likely you'll enjoy Baseball-Reference Play Index.

For fun, we're going to share some of what we find tumbling along Fangraphs leaderbaords and the Baseball-Reference Play Index. If you see anything that strikes an emotion or nerve, let's talk about it in person, on Facebook, Twitter, the telephone, or in our forsaken comments section. Afterall, this blog is for you.

Batting

On the Play Index, we looked at runners in scoring position with 2 out, for their careers. No real surprises until I looked at a little known statistic called tOPS+. The Play Index Glossary defines tOPS+ as "OPS+ of this split relative to the player or team's overall OPS: 100*((split OBP/total OBP) + (split SLG/total SLG) - 1)"

In other words, it's like OPS+, an index above or below 100. If the number is above 100, that is how much the player's performance over-indexes compared to his overall performance. The same for team performance if looking at team statistical splits. Conversely, any number below 100 tells us the player was worse in that split than his overall performance. More examples: a 175 index is 75% better than average; a 73 index is 27% worse than average, and a 200 index is twice as good as average.

The player with the highest Career tOPS+ for batting with runners in scoring position and 2 out is Eric Davis's 133, meaning that Eric Davis was 33% better with 2 out and RISP than he was overall, throughout his career.

Eric Davis (Career BA/OBP/SLG)
Career: .269/.359/.482
RISP/2 out: .294/.435/.539

That's a pretty enormous difference in on-base percentage.

Other players with high tOPS+ with RISP/2 out include Jimmy Rollins and Joe Mauer.

I don't feel good about giving away too much content from paid sites. The J-Ro and Mauer numbers should be easy to check out on their player pages, anyway.

Getting back to the free stuff on Fangraphs, there are no real surprises in the leaderboards for qualified hitters the past 3 seasons. Dropping the minimum plate appearances to 300 tells us a few things.

Yasiel Puig has the 5th highest wRC+ among all batters with at least 300 plate appearances over the past three seasons.

2011-2013 MLB wRC+ (min 300 PA's)
Miguel Cabrera 178
Mike Trout 163
Joey Votto 162
Ryan Braun 161
Yasiel Puig 160

Too bad Ryan Braun lost his credibility.

Brandon Moss (144 wRC+) and Josh Donaldson (130 wRC+), of the Oakland A's, have the two highest wOBA+ among players who fall just beneath the 1,000 plate appearances threshold for "qualified" hitters the past three seasons.

In Wil Myers' 37 plate appearances this season, he had a 131 wRC+.

As a bonus to anyone who's made it this far into the post, let's take a quick peek at the baseball prospectus sortable stats. It should go without saying, but subscribing to Baseball Prospectus is a must.

The B-Pro stats will be just for the 2013 season...

Lowest 2013 double-play percentage (min 300 PA's)
Aaron Hicks 0% (0/60)
Lucas Duda 1.6% (1/61)
Jurickson Profar 1.9% (1/53)
Everth Cabrera 2.1% (1/48)
Adam Dunn 2.4% (2/84)

Highest 2013 double-play percentage (min 300 PA's)
Wilson Ramos 24% (12/50)
Jose Altuve 22.9% (24/105)
Matt Holliday 21.7% (31/143)
David Freese 21.7% (25/115)
Adam Lind 20.8% (20/96)
Placido Polanco 20.8% (15/72)

Best 2013 NET DP* (min 300 PA's)
Brandon Belt -9.70
Mike Trout -7.99
Chris Davis -7.97
Adam Dunn -7.73
Domonic Brown -7.63

*netDP is defined as "the number of additional double plays generated versus an average player with the same number of double play opportunities. Negative NET DP indicates that fewer double plays than added were produced.

Worst 2013 NET DP* (min 300 PA's)
Jose Altuve 12.51
Matt Holliday 12.05
Billy Butler 10.98
Martin Prado 10.47
David Freese 9.76

Let's talk for a minute about batters who know the strike zone. Batters can do damage to pitches outside of the zone, but for the most part, it is valuable to take pitches that are out of the zone. Something interesting we found is that 10 of the top 11 players with the lowest 2013 swing percentage of pitches outside of the strike zone are from the United States or Canada. The 11th, Alberto Callaspo, is from Venezuala. As baseball people have been saying for years, you don't walk off the island(s).

Looking at the statistic on Fangraphs, among active players for the past 3-years (2011-2013), 22 of the top 23 come from the US or Canada. Geovany Soto (19th), hails from Puerto Rico.

***

Pitching

2011-2013 Games Pitched
Joel Peralta 228
Matt Belisle 226
John Axford 224
Brad Ziegler 221
Tyler Clippard 218

2011-2013 Games Pitched (min 220 ip)
Matt Belisle 226
Tyler Clippard 218
Jim Johnson 214
Brian Duensing 160 (39 starts)
Jeff Samardzija 136 (61 starts)

Affix a red flag to Tyler Clippard, but he's still been really good. Over the past three seasons, across 911 plate appearances, opponents have a .173/.249/.309 BA/OBP/SLG slash line. His fastball velocity also has not dropped significantly. According to Pitch F/X, the average velocity of his four-seam fastball was 90.5 in 2009 trending to 92.2 in 2010, 92.6 in '11, 92.7 in '12, and 92.0 mph last season.

Part of the reason to remain bullish on Clippard is that he primarily succeeds with a devastating changeup. If Clippard were more reliant on a breaking ball, he would be more susceptible to a lack of effectiveness from overuse and/or injury.

Oh yeah, Tyler Clippard is also really good at strikeouts.

2011-2013 Strikeout % (min 220 ip)
Yu Darvish 30.1%
Tyler Clippard 28.7%
Stephen Strasburg 28.0%
Matt Harvey 27.9%
Max Scherzer 26.3%

Why the 220 innings pitched minimum? I don't know; it seems like a good place to only include relievers who've provided a constant presence the past 3 years. The original purpose of this blog was to look at the game from various perspectives, and time has often been the best variable to change.

It's cherry picking to pick different start/end points, or innings minimums, but we're doing it to see what story it tells, not to change it to a story we're searching for.

2011-2013 Ground Ball % (min 220 ip)
Jim Johnson 60.7%
Charlie Morton 59.6%
Trevor Cahill 57.8%
Alex Cobb 56.7%
Justin Masterson and Tim Hudson (tie) 56.1%

2011-2013 HR/9 (min 220 ip)
Charlie Morton 0.45
Matt Harvey 0.45
Jim Johnson 0.51
Clayton Kershaw 0.54
Adam Wainwright 0.61
Doug Fister 0.61
Justin Masterson 0.61
Tyson Ross 0.61
Tim Hudson 0.62
Johnny Cueto 0.62

Early stat of the decade nomination: Joe Blanton has a top 10 K/BB for the 3-year period between 2011-2013.

2011-2013 K/BB (min 220 ip)
Cliff Lee 6.54
Dan Haren 4.75
Adam Wainwright 4.63
Matt Harvey 4.58
Clayton Kershaw 4.20
Cole Hamels 4.19
Matt Belisle 4.02
Joe Blanton 4.01
Marco Estrada 4.01
Chris Sale 4.01

Strikeout-to-walk ratio (above) is one of the most predictive of all pitching stats. We never hear about the difference between total strikeouts and walks. Below is the leaderboard for strikeouts minus walks (K-BB):

2011-2013 K-BB (min 220 ip)
Cliff Lee 565
Clayton Kershaw 540
Justin Verlander 514
Felix Hernandez 492
Max Scherzer 473

A couple more variations:

2011-2013 K%-BB% (min 220 ip)
Matt Harvey 21.8%
Cliff Lee 21.3%
Stephen Strasburg 20.8%
Tyler Clippard 20.0%
Clayton Kershaw 19.9%
Yu Darvish 19.9%

2011-2013 (K-BB)/9 (min 220 ip)
Matt Harvey 7.74
Cliff Lee 7.63
Stephen Strasburg 7.52
Yu Darvish 7.38
Max Scherzer 7.13

We've been too quiet over here, so the remedial sabermetrics of these pitching stories can really help get the discussion started. Let's see which players have shown the biggest difference in speed between their fastball and changeup.

2011-2013 FBv-CHv* (min 220 ip)
Erik Bedard 13.2 mph (89.8-76.6)
Dallas Keuchel 12.3 mph (88.9-76.6)
Marco Estrada 12.2 mph (90.0-77.8)
Clay Buchholz 11.9 mph (92.2-80.3)
Tyler Clippard 11.8 mph (92.5-80.7)

*Average fastball velocity minus average change-up velocity

2011-2013 Opponents' OPS "with tie score" (min 220 ip)
Matt Harvey .516
Clayton Kershaw .535
Alexi Ogando .565
Jered Weaver .578
Tyler Clippard .581

Tyler Clippard shows up everywhere.

2011-2013 SIERA Leaders (min 220 ip)
Cliff Lee 2.87
Matt Harvey 2.89
Stephen Strasbug 2.97
Matt Belisle 2.98
Tyler Clippard 3.00

Matt Belisle has been one of the best relievers in baseball the past 3 years

***

Fielding

Defensive metrics are among the most volatile sabermetric stats. While few agree on the best defensive statistic, some old schoolers dismiss defensive stats altogether. It's fairly well known to use defensive data from at least three full seasons to get close to the true value and context of defensive performance. It'll be interesting to see how the release of Field F/X advances research enough to make smaller sample sizes statistically significant.

2011-13 Defensive Runs Saved (MLB Shortstops)
Andrelton Simmons 60
Brendan Ryan 51
Clint Barmes 39

Simmons is incredible. We advocate Andrelton Simmons enthusiastically. I'm wholeheartedly sold. He's the best defensive shortstop I've ever seen, and that includes late-80's Ozzie Smith and highlights of early-80's Ozzie Smith. Simmons has the best arm strength, accuracy, and balance that I've ever seen in a shortstop. Not the best combination of those three attributes, he's the best in all three.

My favorite part of the stat above is that Andrelton Simmons is on such a phenomenal rate of saving runs.

Andrelton Simmons MLB Games Played
2011: 0
2012: 49
2013: 157

Therefore, Simmons has only played one-and-a-third seasons. By comparison, Brendan Ryan has played 368 games the past three years (Barmes 375). Should Andrelton Simmons' defensive stats not matter until circa June 2015? The way Defensive Runs Saved is calculated makes a case for its validity with short-term score keeping, not necessarily forecasting. Plus, his numbers are aligned with results from the scouts' eye test.

On the flipside, there's some entertainment value to looking at the worst shortstops in Defensive Runs Saves the past three years:

2011-13 Lowest Defensive Runs Saved (MLB Shortstops)
Derek Jeter -38
Eduardo Nunez -38
Jose Reyes -33

Okay, Jeter was predictable, but who would have guessed Nunez and Reyes? The horrendous rate of Nunez's defense is most striking when looking at innings played.

2011-13 Lowest Defensive Runs Saved (MLB Shortstops)
Derek Jeter -38 (2,343 innings)
Eduardo Nunez -38 (1,110 innings)
Jose Reyes -33 (3,290 innings)

Looking at first base defense, according to Defensive Runs Saved the past three seasons, Adrian Gonzalez wins by a mile. Almost lapping the competition, Gonzalez has saved 39 runs since 2011. The next best is Anthony Rizzo at 22 and James Loney with 21 DRS. Rizzo has played 1,400 less innings than Gonzalez and almost 1,000 fewer innings than Loney. Let's calculate DRS as a rate stat, per one-thousand innings played, and call is "DRS/M" (M is the Roman numeral for 1,000).

2011-2013 DRS/M (First Base, min 1,000 inn.)
Adrian Gonzalez 10.03
Anthony Rizzo 8.85
Mark Teixeira 8.24
Nick Swisher 7.42
Albert Pujols 7.35
James Loney 4.96
Mike Napoli 5.80
Joey Votto 4.96
Adam LaRoche 4.71
Brandon Belt 4.67

You may be surprised that the player with the worst DRS/M is neither Adam Dunn, Chris Carter, Ryan Howard, Edwin Encarnacion, nor Paul Konerko. The Oakland A's Brandon Moss has logged 1,245 at first base the past three years, saving runs at a rate of -12.85 fewer runs per 1,000 innings. Adam Dunn was a close second:

2011-2013 lowest DRS/M (First Base, min 1,000 inn.)
Brandon Moss -12.85
Adam Dunn -12.46
Chris Carter -9.69
Edwin Encarnacion -8.15
Ryan Howard -7.85

Metrics for catching value are the furthest from being comprehensively accurate. There is still so much we do not know about the value of reading opposing batters' swings and calling a good game, pitch framing, nor how much value of controlling the opponents' running game should go to the pitcher or catcher.

Looking at defensive runs saved touts the value of Yadier Molina (24 DRS) and Russell Martin (17 DRS). Matt Wieters leads the rest of the pack with 9 DRS.

If you drop the minimum from "Qualified" to minimum 2,000 innings played, Salvador Perez appears with 21 DRS in just over 2,100 innings caught. Nick Hundley and Buster Posey also make a strong appearances with 8 and 7 DRS in almost 1,500 and 1,200 fewer innings, respectively, than Matt Wieters' 3,539 innings the past three years.

2011-13 Fangraphs C Def (min 2,000 innings played)
Matt Wieters 55.8 Def
Yadier Molina 52.2 Def
Russell Martin 39.8 Def
Miguel Montero 34.8 Def
Ryan Hanigan 31.6 Def

Ryan Hannigan!

Sincere question to sabermetricians out there, why does Fangraphs' overall value for catcher defense the past three years have Matt Wieters ahead of Yadier Molina? Is it all due to Wieters' extra 113 innings caught? A breakdown of their other stats below...

2011-13 Fangraphs C Def WAR
Matt Wieters: 3,539 innings caught, 55.8 Def WAR
13 rSB, -4 rGFP, 9 DRS, 137 CPP, 12.5 RPP, 33 FSR

Yadier Molina: 3,426 innings caught, 52.2 Def WAR
10 rSB, 7 rGFP, 24 DRS, 156 CPP, 12.8 RPP, 35 FSR

Perhaps a 3.6 difference in Def rating is minuscule. If anyone out there has some thoughts on the matter, please share.

The area of greatest advancement the past couple of years has been with the study of catcher framing. I'm having a tough time finding catcher framing data on Fangraphs, B-R, and Baseball Prospectus. It appears they don't publish value for this performance, unless it's hidden inside a composite stat.

Fortunately, we are able to look at some great details on the Catcher Framing Report of Matthew Carruth's impressive StatCorner. Catcher framing is murky art. Credit is given to catchers who have strikes called on pitches outside the zone. Conversely, they are penalized for pitches in the zone that are called balls. Much of that can be seen, as graceful catching and framing of the ball, so it looks like a strike. Other times, it could just be bad umpiring. The ideas is that with such a large sample size, of thousands and thousands of pitches, the good or bad luck associated with umpire quality will even out.

The StatCorner Catcher Framing Report doesn't provide 3-year averages or totals, but we can look back at prior years.

Here's a snapshot of Catcher Framing "Best and Worst"

2011 Catcher Framing: Best oStr%
Min 2000 innings caught
Jarrod Saltalamacchia 10.5%
Brian Schneider 10.1
Jonathan Lucroy 9.9%
Francisco Cervelli 9.9%
Chris Stewart 9.8
Wilson Ramos 9.8%

2011 Catcher Framing: Worst oStr%
Min 2000 innings caught
Ronny Paulino 4.4%
Ryan Doumit 4.7%
Michael McKenry 5.2%
Koyie Hill 5.2%
Salvador Perez 5.2%
Matt Treanor 5.2%

2011 Catcher Framing: Best zBall%
Min 2000 innings caught
Chris Snyder 10.3%
Jonathan Lucroy 10.4%
Ryan Hanigan 10.8%
Brian McCann 11.4%
Eli Whiteside 11.5%

2011 Catcher Framing: Worst zBall%
Min 2000 innings caught
Carlos Santana 20.7%
Lou Marson 20.6
Ryan Doumit 20.5%
Victor Martinez 20.4%
Matt Treanor 20.3%

2012 Catcher Framing: Best oStr%
Min 2000 innings caught
Jose Molina 9.9%
Chris Stewart 9.9%
David Ross 9.6%
Bobby Wilson 9.5%
Ryan Hanigan 9.4%

2012 Catcher Framing: Worst oStr%
Min 2000 innings caught
Ryan Doumit 4.2%
John Hester 4.6%
Michael McKenry 4.7%
Carlos Santana 5.1%
Drew Butera 5.3%

2012 Catcher Framing: Best zBall%
Min 2000 innings caught
Jonathan Lucroy 9.8%
Brian McCann 10.0%
Yasmani Grandal 10.1%
David Ross 10.2%
Jose Molina 10.4%

2012 Catcher Framing: Worst zBall%
Min 2000 innings caught
Gerald Laird 19.8%
Ryan Doumit 19.8%
John Hester 19.2%
Matt Treanor 18.6%
Mike Napoli 18.6%

2013 Catcher Framing: Best oStr%
Min 2000 innings caught
Yasmani Grandal 11.3%
Jose Molina 9.1%
David Ross 9.0%
Hank Conger 8.9%
Jonathan Lucroy 8.8%
Martin Maldonado 8.8%

2013 Catcher Framing: Worst oStr%
Min 2000 innings caught
Ryan Doumit 4.7%
Anthony Recker 4.7%
George Kottaras 4.8%
Gerald Laird 4.9%
John Buck 5.0%

2013 Catcher Framing: Best zBall%
Min 2000 innings caught
Hank Conger 9.5%
Martin Maldonado 10.1%
Jonathan Lucroy 10.2%
Evan Gattis 10.2%
Yan Gomes 10.3%

2013 Catcher Framing: Worst zBall%
Min 2000 innings caught
Ryan Doumit 20.2%
Kelly Shoppach 18.4%
Gerald Laird 17.6%
Henry Blanco 17.4%
Chris Iannetta 17.2%

Max Marchi opened eyes on evaluating catcher defense in 2011 at Hardball Times. Ben Lindbergh did a great job six months ago bringing pitch framing analysis to the masses on ESPN's Grantland. In case this prolonged experience on the reality tour hasn't satiated your appetite for reading, check it out.