Sunday, April 28, 2013

What Happens to Ground Balls Hit Off Tigers Pitchers?

As often happens in a Rick Porcello start, a discussion about Tigers infield defense broke out on Twitter yesterday afternoon.  The Tigers defense was not terrible behind Porcello Saturday, but there were a couple of seeing-eye singles, which may or may not have been converted into outs by better infieldersThe theory is that, because Porcello does not strike out many batters out and induces a lot of ground balls, he would benefit more from a team with a better infield defense than the Tigers.

I decided to look into this issue further using the retrosheet databases.  The 2013 data are not yet available, but I was able to look at what happened on ground balls from 2010-2012.  Table 1 below shows that there were 5,913 ground balls off the Tigers staff during that period.  Of those, 1,523 resulted in hits yielding a batting average of .258.  That was the highest batting average in the majors which probably doesn't surprise anyone that has watched the Tigers defenders stagger around the infield in recent years.

The MLB batting average on ground balls was .241, so the Tigers were .017 higher than average.  There are about 2,000 ground balls per year, so that is 34 more hits per year on ground balls more than average.  The average failure to convert a batted ball into an out is 0.75 runs, so that comes out to about 25 runs per season, which is significant.

The lowest batting average on ground balls was .221 achieved by the Blue Jays.  In comparison to Toronto, the Tigers allowed 74 extra hits on ground balls per year which translates into about 55 runs.  That illustrates the value of a superior infield defense.

Table 1: Batting Average on Ground Balls Versus Pitchers, 2010-2012

Team
GB
H
BAA
TOR
6,104
1,349
.221
WAS
5,975
1,330
.223
ANA
5,934
1,334
.225
OAK
5,894
1,346
.228
TBA
5,629
1,299
.231
BOS
5,892
1,364
.232
LAN
5,640
1,306
.232
SEA
5,908
1,372
.232
SDN
5,938
1,384
.233
ATL
6,155
1,440
.234
TEX
5,619
1,317
.234
SLN
6,602
1,565
.237
SFN
5,717
1,356
.237
PHI
5,905
1,413
.239
COL
6,182
1,483
.240
CLE
6,525
1,566
.240
CIN
5,977
1,437
.240
NYN
6,010
1,464
.244
ARI
5,857
1,433
.245
PIT
6,275
1,541
.246
MIL
5,698
1,411
.248
BAL
6,112
1,515
.248
CHN
5,714
1,422
.249
HOU
6,060
1,540
.254
NYA
5,861
1,490
.254
MIA
5,976
1,527
.256
KCA
5,978
1,529
.256
CHA
5,910
1,518
.257
MIN
6,416
1,649
.257
DET
5,913
1,523
.258
 The information used here was obtained free of charge from and is copyrighted by Retrosheet.

I next looked at individual Tigers pitchers to see who might have been affected most by the porous infield defense.  Porcello's BAA on 962 ground balls was .262.  That translates into about seven hits per year, so the direct effect is not huge - an additional 0.25 to his ERA.  However, these data do not account for the effect of extra pitches thrown in order to get more outs.  Given that Porcello typically tires after about 75 pitches, this could be a problem.  It's also possible that a lack of confidence in his in infield could negatively affect pitch selection.

Interestingly, the pitcher that seems to have been affected the most by extra hits on ground balls was Max Scherzer who had a .291 BAA on ground balls.  Luckily, Scherzer does not rely nearly as much on ground balls as Porcello, but it still amounts to an estimated 8 more runs or 0.40 in ERA per year compared to the average pitcher.  This may explain why his ERA was higher than his FIP the last couple of years.

On the other hand, Doug Fister (in games with the Tigers only) and Justin Verlander each had a .212 BAA on ground balls which was considerably lower than average.  This may have been random luck or it could be that they induced weaker contact on grounders than other pitchers.  Unfortunately, data such as speed of batted balls is not yet available to the public.

Most of the Tigers pitchers did have higher BAAs than average though.  Joaquin Benoit, Phil Coke, Anibal Sanchez and Drew Smyly were all between .250-.265.  So, it does appear that in general that Tigers were adversely affected by infield defense.  I'll get back to this after the season when 2013 data are available.   

Friday, April 26, 2013

Anibal Sanchez Breaks Tigers K Record, Blanks Braves 10-0

It didn't look too promising when the struggling Tigers went into the weekend series versus the sizzling  Braves, without ace right hander.Justin Verlander.  We were reminded tonight though that Verlander is not the only Tigers starter capable of pitching a dominant game.  Anibal Sanchez set a franchise record by striking out 17 Atlanta batters in eight innings.  That broke a record set by Mickey Lolich who twice retired 16 batters on strikes in 1969.   
 
The final punch out for Sanchez came on a swinging strike on a curve to Braves second baseman Dan Uggla to end the eighth.  Sanchez left the field to a standing ovation from an enthusiastic Comerica Park crowd that seemed to understand the significance of the 17th strikeout.  He finished strong striking out the side around a single in the eighth and some fans may wanted to see him pitch the ninth, but with a pitch count of 121 there was no way that was going to happen.  

Sanchez is no stranger to remarkable performances having pitched a no hitter for the Marlins in 2006.  He also had three one hitters as well as a 14-strikeout game during his time in Miami.  For that matter, any of the top four pitchers on this Tigers staff is capable of doing something special on a given night.

The Tigers now have two pitchers on their staff that have struck out more batters in a game than the great Verlander. The other is Max Scherzer who had 15 in a game last year.  The most Verlander ever had was 14, a feat he has accomplished twice.  Not only that, but fellow blogger Matt Wallace reminded me that their other big starter, Doug Fister, holds the American League record for consecutive strikeouts with nine.  The Tigers fearsome foursome has combined for a 2.23 ERA with 127/31 K/BB ratio in 117 innings so far this year.  That is some outstanding pitching.

One more note about Sanchez's game tonight was that it was only the second 17-strikeout game in the majors since 2007.  The other was by Blue Jays right hander Brandon Morrow who struck out 17 Rays in a game on 2010.

What can the Tigers due for an encore?  Well, it won't be 17 K with Rick  Porcello on the mound tomorrow.  It usually takes him about a month to get that many.  Let's hope though that he keeps the momentum going with a solid game.

Tuesday, April 23, 2013

The New Tigers Closer is...Jose Valverde

The Tigers game with the Royals has been postponed tonight, but they were busy off the field today.  First, the team announced that rookie reliever Bruce Rondon has been recalled from Triple-A Toledo to replace Octavio Dotel who was placed on the 15-day disabled list with an inflamed elbow.  Later, they signed former closer Jose Valverde to a major league contract and optioned Brayan Villarreal to Toledo.

Jon Morosi just tweeted that Valverde is the new closer.  Not after pitching a couple games, but right away as in tomorrow night:
Jon Morosi@jonmorosi 3m
Meet the Tigers’ new closer, same as the Tigers’ old closer – Jose Valverde.
The Tigers are bringing back Valverde after only three innings with High-A Lakeland. On the surface, this looks a bit like a desperation move.  His fastball had lost a lot of zip in the last couple of seasons and he became increasingly reluctant to throw his splitter.  It resulted in a mediocre regular season in 2012 with a 3.78 ERA and 6.3 strikeouts per nine innings (down from a career rate of around 10).  And nobody needs to be reminded of his embarrassing performance in the playoffs.

Reports are that Valverde is throwing harder this spring and that he has regained confidence in his splitter.  I am not very optimistic about this move, but I won't write him off until we get a chance to see him pitch.

According to the Detroit Free Press. manager Jim Leyland said earlier today that Rondon would not close:
“I’d like to break him in a little bit easier, maybe the sixth inning or seven inning to start with,” Leyland said. “He’s not going to be the closer to start with, as we speak. That doesn’t mean he would never close a game, that just means right now I want to break him in, get his feet wet in the big leagues a little bit and see what he looks like.”
While general manager Dave Dombrowski raved about Rondon all winter, Leyland has spoken more conservatively about the right-handed flame thrower's talents.  Thus, it is not surprising that he won't be thrust into the closer role. Rondon had been pitching well for the MudHens though allowing no runs and posting a 9/2 K/BB ratio in 7 2/3 innings

Sunday, April 21, 2013

Tigers Pitcher Similarity Scores

The concept of similarity scores for both batters and pitchers has been around for a while.  Bill James first introduced them in 1986 in his annual abstract and Baseball Prospectus uses them as part of their PECOTA system.  The idea is that players are compared according to their career statistics and each pair of players gets a score which tells how similar they are.  For pitchers, the comparison includes metrics such as innings pitched, ERA, strikeouts and walks and, in the case of PECOTA, characteristics such as height and weight.

These similarity scores are useful in seeing how closely your favorite players compare to players from different eras.  For example, the Bill James similarity scores at Baseball-Reference suggest that Tigers ace Justin Verlander compares most similarly to Phillies left hander Cliff Lee and Angels righty Jered Weaver.  They also indicate that he was similar to old time pitchers such as Sal Maglie and Gary Nolan.

These similarity scores are fun to look at but they don't say much about what kind of pitches a pitcher throws.  If you were to compare a prospect to a major league pitcher to give you an idea of what kind of pitcher he might be in the future, it does not make a lot of sense to use career statistics.  It would be more useful to compare his pitching arsenal.

To that end, Stephen Loftus of Beyond the Box Score recently published a method for comparing pitchers based on their "stuff" rather than their performances.  It's a work in progress but it's some of the most interesting and potentially valuable baseball research I've seen in a while.  His article is very math heavy, but very worth reading even if you need to skip over some of the math.

Very simply, what Mr. Loftus did was compare pitchers using the following Pitch f/x data:
  • Pitch types
  • Pitch velocity
  • Pitch break (horizontal and vertical)
  • Release point (horizontal and vertical)
  • Pitch location
For each pair of pitchers with over 1,000 pitches thrown in 2012, he calculated a similarity score between 0 and 1 with the most similar pitchers close to 1.  These scores are broken down into categories such as 0.8-1.0 extremely similar and 0.6-0.8 reasonable similar.  The results for all pairs of pitchers can be found at a link to an excel spreadsheet at the bottom of his article

It turns out that the two most similar pitchers in baseball in 2012 were Anibal Sanchez of the Marlins and Tigers and Anthony Bass of the Padres.  A closer look at the data for Sanchez and Bass at Brooks Baseball tells us why they had such a high score. As Loftus points out, they shared three pitches - fastball, slider and change-up.  Sanchez also threw a curve, but not that often.  What makes them so close though is that they threw these pitches with remarkably similar velocity, break and release point.

I took Loftus' spreadsheet and determined the top three comparisons for all pitchers and included them in a Google spreadsheet.  The results for the Tigers are shown in Table 1 below.  The fact that Verlander was most similar to thirty-year old Orioles right hander Jason Hammell may seem more amusing than useful given their results, but Hammel was pretty effective last year.  They might not have been so similar in past years.

Table 1: Top Similarity Scores for Tigers Pitchers

Pitcher
Comp 1
Sim Score 1
Comp 2
Sim Score 2
Comp 3
Sim Score 3
Anibal Sanchez
ANTHONY BASS
0.97
MAT LATOS
0.96
YOVANI GALLARDO
0.91
Doug Fister
P J  WALTERS
0.87
JASON VARGAS
0.81
IAN KENNEDY
0.76
Drew Smyly
JAMES MCDONALD
0.79
CHRISTIAN FRIEDRICH
0.75
BUD NORRIS
0.74
Joaquin Benoit
FRANKLIN MORALES
0.81
BRANDON MORROW
0.80
HOMER BAILEY
0.80
Jose Valverde
MANNY PARRA
0.83
BRAD LINCOLN
0.72
LANCE LYNN
0.71
Justin Verlander
JASON HAMMEL
0.86
PHIL HUMBER
0.82
JEREMY GUTHRIE
0.82
Max Scherzer
HECTOR NOESI
0.77
JOE KELLY
0.77
HENDERSON ALVAREZ
0.75
Rick Porcello
CHRIS VOLSTAD
0.88
HENDERSON ALVAREZ
0.85
LUIS MENDOZA
0.83
Data Source: Beyond The Boxscore

This research is just a start and will be come more useful when more years are available and especially when we are able to compare minor leaguers to major leaguers.  Other data such as pitch sequencing could also be added to give better comparisons.  I look forward to see what Loftus and others can do with this concept going forward.   

Twitter

Blog Archive

Subscribe

My Sabermetrics Book

My Sabermetrics Book
One of Baseball America's top ten books of 2010

Other Sabermetrics Books

Stat Counter