Sunday, October 12, 2008

The 2008 Starting Rotation: An Analysis of the Record of Opposing Batters


Prior to the 2008 season, I studied the batting line of the projected three frontline starters for the Twins (Hernandez, Bonser and Baker).  This was a technique suggested by Bill James when trying to look beyond the standard record and ERA statistics.  The idea is that if your starting pitcher's batting line was comparable to Houston's Michael Bourne (.229/.288/.300) you have a pitcher of ace quality.  Conversely, if your pitcher's batting line looks strikingly similar to Albert Pojuls (.357/.462/.653), that hurler might already be booked for a flight back to one of the team's minor league affiliates (if not back to his hometown). 
Livan Hernandez
Comparable: Brian McCann
The findings showed that in 2007, Livan Hernandez made every batter hit like Victor Martinez - the 2007 version of Martinez, that is - who batted .301/.374/.505.  At the time I wrote that employment would be short-lived if Hernandez's results in 2008 emulated those of 2007.  In 24 starts in 2008 Hernandez's opponents batted .342/.375/.520, a lofty step backwards.  Hernandez saw the 2007 Victor Martinez and raised a 2008 Brian McCann, who hit .301/.373/.523 this past season.  The Twins continued to sign his paycheck well into June mostly because of a 6-2 record through 13 starts despite a .339/.369/.506 batting line against in those outings.  It would take a serious slide in victories over the course of the next 10 games (4-6) for the organization to consider cutting Hernandez loose.  Some teams, like the New York Mets, excrete desperation for starting pitching.  When Hernandez was designated for assigment by the Twins, Tony Bernazard, the Mets' vice president for player development, told the NY Times “We will look into anyone who is available.”  It would be the lowly Rockies that claimed him on the waivers.  His stint in Colorado would turn really ugly as National League opponents hit .345/.399/.571 off of him.  While most organizations would be intelligent enough to keep their distance from this sort of volatile there will be one or two ballclubs that will be willing to give Hernandez a spring training invitation because he finished the season with a winning record (13-11) and was able to work 180 excruciating innings.    
Boof Bonser
Comparable: James Loney
Boof Bonser is a curious case study.  The hopes of the organization going into 2008 was the Bonser would shed excess body weight return to the pitcher that appeared at the end of the 2006 season moving, but after making 12 starts and sporting a 2-6 record the Twins banished him to the bullpen.  His 5.97 ERA and record was a poor reflection of his performance as he was actually pitching better than either statistic suggested.  In those 12 starts opponents were batting just .267/.316/.418 (comparable to Miguel Tejada's season).  Once in the bullpen though, Bonser collapsed.  He faced 236 batters in relief, opponents hit .307/.353/.474 (Jose Reyes), while he couldn't keep runners from scoring as indicated by his 57.9% left-on-base percentage.  This was the worst strand-rate by any American League pitcher to log over 100 innings. The underlying problem that manifested itself in 2007 continued into 2008, which was that Bonser could no longer retire left-handed batters.  In 2007, Bonser turned all lefties into Kevin Youkilis. The group of wrong-handers collectively hit .349/.407/.563 off of Boof in 411 plate appearances.  This past season, lefties hit .315/.378/.489 in 243 plate appearances, a comp resembling Magglio Ordonez. This was an oddity considering that in 198 plate appearances in 2006, Bonser had limited left-handed opponents to .251/.305/.454 split.  What had happened over the course of the two seasons?  In 2008, Bonser used mostly fastball (63%), curves (19%) and sliders (11%) when facing lefties.  He would rarely use a changeup (6%).  The results of 2007 were the same: fastballs (58%), curves (23%), sliders (10%) and the occasional changeup (8%).  Yet in 2006, it was his changeup, not his slider that was his third most used pitch.  He used the changeup 12% of the time against lefties and was much more effective.  Because he is out of options, the Twins have tentatively labeled him the long relief guy in 2009, though judging from his splits, he would be better suited as a ROOGY. In order to return to the success that he experienced in 2006, Bonser should implement his changeup as his third pitch when facing lefties.     
Scott Baker
Comparable: Carlos Gomez

In 2008, Baker emerged and grabbed the title of staff ace, leading the rotation with the lowest batting line against of .247/.294/.381.  Coming into the season, Scott Baker also had the best record of opponents batting among the three starters in 2007 and I noted that Baker had the stuff to pitch his way into Brandon Inge territory, who in 2007 had a depressed .236/.312/.376 batting line.  Going sheerly off of OPS, Baker actually did slightly better than turning batters into Inge (675 OPS versus 678 OPS), but unfortunately for Twins fans, his closest comparable in 2008 was none other than Carlos Gomez (who had an OPS of 656). has Boof Bonser listed as the closest comparable at age 25 for Baker.  Now that he is entering his 27-year-old season in 2009, the similarities between him and Bonser will deviate significantly.  Baker will assume his rightful position as a number one or number two starter in the 2009 rotation, giving the Twins an excellent right-left combo with Francisco Liriano, and, if he can remain healthy, should accumulate 200 innings for the first time in his career.   

Nick Blackburn
Nick Blackburn is the successful version of Boof Bonser.  They both had similar groundball rates (44.9% gb% versus 40.7%) and line drive rates (20.9% ld% versus 20.4% ld%) and opponents hit Blackburn just as well as they did Bonser.  Yet one of them was asked to start the one-game playoff against Chicago and the other was picking splinters out of his bulbous ass in the bullpen.  What separates Blackburn from Bonser?  The simple answer is that Blackburn can pitch from the stretch and Bonser can't.  With runners on base, opponents were hitting just .274/.307/.390 off of Blackburn as opposed to .306/.348/.481 when he was pitching from the wind-up.  Bonser, meanwhile, implodes when runners reach.  Opponents were batting .322/.349/.467 when Bonser was throwing from the stretch - which has led to Bonser's 57.9% LOB% versus Blackburn's 70.7%.  Blackburn's numbers leave me to believe that he is a number four/five starter, certainly a valuable one who gained postseason experience in his matchup with the Chicago White Sox and had thrown nearly 200 innings in his first full major league season, but if Blackburn is traded during the offseason to obtain a shortstop or third baseman, I wouldn't be surprised either.   
Kevin Slowey
Comparable: Kevin Kouzmanoff
Kevin Slowey's 2008 season reaffirmed that his minor league pedigree would follow him to the majors as he posted a 123/24 K/BB ratio, barely losing to Scott Baker for the team's lowest OBP allowed.  Twins fans might balk at seeing Kouzmanoff as Slowey's comparable considering that Kouzmanoff has been one of the rumored third basemen that the team has been interested in, but he almost has the identical K/BB (139/23) and hit just one home run fewer than Slowey allowed (23 versus 22).  Though they are very similar, Kouzmanoff's sophomore season was a huge regression while Slowey's sophomore year was a huge progression in development.  As a flyball pitcher, Slowey benefited from the speedy outfield duo in Carlos Gomez and Denard Span and also took a big step forward in preventing the ball from leaving the park as his home runs to flyball ratio decreased from 16.5% in 2007 to 11.5% in 2008.  Slowey's improvements suggest that in 2009 he will be a frontline starter with Baker and Liriano giving the Twins three solid starters to enter any big series.  
Glen Perkins
Comparable: Raul Ibanez/Melvin Mora
Yes, seeing Ibanez as the most comparable probably isn't a good thing but it does fit as Perkins surrendered two of his 27 career home runs to Ibanez.  Perkins was lauded because of his 12-4 record but led the staff in runs allowed per nine innings (4.83) and home runs allowed (25).  His success is exaggerated because Perkins received 6.23 runs per start from his offense - a full run higher than Slowey's 5.20 per game, who finished the season 12-11.  His high opponent average against is because batters are able to make solid contact, as indicated by his 22.6% line drive rate.  Perkins fell apart in September going 0-1 with a 7.45 ERA while opponents bashed him for seven home runs and hit .384/.421/.733 off of him in five starts.  His minor league stats suggest that he is probably better than what he showed in 2008 but not that much better.  The Twins should consider selling high now for infield help.  Both Seattle and Milwaukee will be looking for pitching help this offseason and have Adrian Beltre and JJ Hardy, respectively.  Then again, with the bind that the Twins were in for starting pitching after trading away Matt Garza, the organization might shy away from trading away pitchers with major league experience.   
Francisco Liriano
Francisco Liriano's 2008 season is hard to label.  His first three starts in May were disastrous in a foolhardy move by the front office.  In those three starts, opponents hit .366/.509/.415 skewing the final opponent batting record, unfairly making him seem comparable to Jeremy Hermida or Mark Teahan.  When Liriano returned from Rochester the second time, opponents hit just .258/.296/.360 - once again turning opposing team's lineups in Carlos Gomez.  Had Liriano had a full season with this performance, his 656 OPS against would have overtaken Baker's 675 as the team's best.  With an entire offseason to prepare for the next year instead of rehabbing, Liriano should be expected to be the Twins' number one starter in 2009.