Linear Saves 2009

Rivera Impresses, Rodney Surprises, Lidge Astounds

by Keith Glab,
November 25, 2009

In all honesty, our 2009 Linear Saves data isn't as exciting as our 2008 data.  There was no perfect save season, no new single-season saves leader, and no one passed Lee Smith in career Linear Saves.  Still, there are some interesting nuggets, including Mariano Rivera passing another milestone, Brad Lidge staking his claim as the least consistent reliever of all time, and someone with an ERA of 4.40 nearly leading baseball in Linear Saves.

If you are new to Linear Saves, you may want to follow this link, read about the concept, and peruse the historical data.  In a nutshell, Linear Saves compares a closer's saves and blown saves to that of his peers in a single number, expressed in the number of saves above or below what the league's average closer would convert given the same number of opportunities. 

LSV= SV-(Opps*LgSV%)

Mariano Rivera 1.76 44 2 4.9
Fernando Rodney 4.40 37 1 4.7
Huston Street 3.06 35 2 3.5
Jonathan Papelbon 1.85 38 3 3.1
Joe Nathan 2.10 47 5 2.7
Francisco Cordero 2.16 39 4 24
David Aardsma 2.52 38 4 2.3
Mike MacDougal 4.31 20 1 2.1
Trevor Hoffman 1.83 37 4 2.1
Joakim Soria 2.21 30 3 1.9
Ryan Franklin 1.92 38 5 1.4
Brian Fuentes 3.93 48 7 1.2
Heath Bell 2.71 42 6 1.2

Why do we need Linear Saves?  The fact that Randy Velarde had more hits than Nomar Garciaparra had in 1999 didn't make Velarde a better hitter, yet there are still people out there who judge a closer's effectiveness based on his save total alone.  Sole examination of a rate stat like save percentage is flawed as well.  Dwight Smith wasn't a more valuable hitter in 1993 than Craig Biggio was simply because he had a higher batting average.  Finally, since Linear Saves rates a closer against his peers, making comparisons across eras is possible.  Carl Yastrzemski  has a lower career OPS than does Jim Bottomley, but an examination of their OPS+ shows Yaz to be the superior hitter.  

It is important to note that for the purposes of Linear Saves, only pitchers with at least 10 saves in a season are counted when compiling the league save percentage.  This is a somewhat arbitrary number, but its purpose is to filter out the middle relievers who technically enter save situations routinely but are rarely allowed to pitch into the ninth inning and therefore have artificially low save percentages. 

The arbitrary figure of 10 came into play this year, however.  The Linear Saves trailers each year tend to be peppered with relievers who were only closers for part of the year and therefore have lower save percentages than they would had they been full-time closers.  This year, three of the five worst Linear Saves totals came from such part-timers that had exactly 10 saves: Jim Johnson (-3.7), Ryan Madson (-3.7), and Mike Gonzalez (-4.5).

Why does this matter?  Because if we had set the minimum number of saves to for a reliever to be considered a closer at 11 instead of 10, those three relievers would be taken out of our database and the league save percentage would jump from 85.1% to 86.1%.  That would cost Linear Saves leader Mariano Rivera half a linear save, saves leader Brian Fuentes 0.6 Linear Saves, and so forth.  Bump up our closer cutoff point to 12 saves, and we eliminate Jason Frasor (-0.9) and LaTroy Hawkins (-1.8) from the database, making our closer save percentage 86.4%.

Certainly, bumping up that cutoff would improve the league save percentage every year, but this year, the numbers fell in such a way as to make the effect substantial. It doesn't make sense to go year-by-year and selectively change the cutoff point for each season, but at the same time, we can't ignore that Mariano Rivera's 4.9 Linear Saves are not quite as impressive as they first appear.  The league save percentage was not nearly four percent worse than it was in 2008 because it was so much more difficult to save a game this year than it was in 2009.  That difference was largely a product of random variation and a fluke of the numbers.

This is not to diminish Rivera's season, however.  The league's save percentage would have needed to have been over 87.5% for Rivera not to have led the majors in Linear Saves.  By doing so, he has become the fifth player since 1969 to lead MLB in Linear Saves in multiple seasons, joining Trevor Hoffman, Eric Gagne, Rod Beck, and Dennis Eckersley.  He also joins Hoffman as the only player with over 30 (or even 25) career Linear Saves, as he now has 32.3 (or 29.8 if you decide to count his five saves in nine "chances" between 1995 and 1996 against him).  Regardless, he is closing the gap on Hoffman, who increased his total to 35.8 (or 34.1) this season.

45-3 Member YR ERA SV BS LSV
John Smoltz '02 3.25 55 4 4.7
Antonio Alfonseca '00 4.23 45 4 4.4
Rod Beck '98 3.02 51 7 2.5
Lee Smith '93 3.88 46 7 1.7
Rob Nen '01 3.01 45 7 0.9
Kazuhiro Sasaki '01 3.24 45 7 0.9
Joe Borowski '07 5.07 45 8 -0.2

Who would have led baseball in Linear Saves had the closer save percentage been over 87.5%?  Fernando Rodney, the 4.40 ERA Wonder, of course.  Despite allowing 41 walks, eight homers, and 37 earned runs in 75.2 innings, Rodney blew only one save opportunity all year, missing the illustrious 45-3 club by eight saves.  That 4.40 ERA would have been the highest ever from a Linear Saves leader and only the fifth season in which a Linear Saves leader posted an ERA over 3.00.  The others were Dan Quisenberry (3.09 in 1980), Ron Davis (3.34 in 1983), Antonio Alfonseca (4.24 in 2000) and John Smoltz (3.25 in 2002).        

While Rodney just missed the distinction of having the worst season ever for a Linear Saves Leader, Brad Lidge may have had the worst season ever for a closer.  His -5.2 Linear Saves are not the lowest total ever by a long shot.  Ambiorix Burgos (-7.3 in 2006), Francisco Cordero (-5.8 in 2006 and -5.9 in 2003), Ugueth Urbina (-5.7 in 2005), Chris Reitsma (-6.0 in 2005), and Bob Wells (-6.6 in 2000),  have all been worse in this decade alone. But when you couple his awful Linear Saves total with his 0-8 record and 7.21 ERA, is there any doubt that Lidge 2009 is the worst ever season for a closer?

Last season, Lidge became the first closer to ever go from the worst in Linear Saves to the best in consecutive seasons.  This year, he becomes the first to go from best to worst in back-to-back years.  Only three other relievers have both led the majors and trailed the majors in Linear Saves, and those three all have extenuating circumstances for their trail years.  John Hiller's 1976 season came before managers used the save rule to govern their managerial decisions, and Hiller had a dozen wins and a 2.38 ERA in 121 innings to go with his -5.9 Linear Saves.  You can't fully explain away 13 blown saves in 26 chances, but he averaged over two innings per relief appearance that year and was clearly being used in tougher situations than was the average closer in 1976.  In an interesting bit of trivia, Hiller's only start that year was a shutout of the Milwaukee Brewers in a meaningless October 1st game.

But I could write about John Hiller all day.   Francisco Rodriguez trailed the majors in Linear Saves with -4.4 in 2004 and would lead the majors with 4.0 in 2006.  Similar to Rivera in 1996, K-Rod was being used primarily as a setup man and had a 1.82 ERA, so it's tough to berate him.  Antonio Alfonseca doesn't really have an alibi for his poor season in 2002, but by practically every measure other than saves and Linear Saves, that was actually a better season for him than 2000 was.  2002 saw Alfonseca with a better ERA, WHIP, strikeout rate, and home run rate than in 2000, but unfortunately, he played on a much worse team.  J.J. Putz, Bruce Sutter, and Rollie Fingers each narrowly missed joining Lidge, Hiller, K-Rod, and Sixer in the leader/trailer club.

Lidge and Goose Gossage are the only relievers to ever finish last in the majors in Linear Saves twice, however.  Gossage trailed with -4.6 LSV in 1983 and with -5.2 in 1988.  His 1983 season was excellent in every other right as Gossage went 13-5 with a 2.27 ERA, but he suffered from Hiller syndrome a bit in terms of usage.  The 1988 season was indeed terrible, as the aging reliever was traded to the Cubs and floundered.  The Cubs can take some solace in the fact that Keith Moreland was nearly as awful for the Padres that year.  While Lidge's 2009 season was unadulterated trash, he was actually doing quite well in 2007 until he was given the closer's role (2.34 ERA/0 SV 1st half, 4.45 ERA/19 SV 2nd half).

I am certainly curious to see how Lidge will perform in 2010, as his ride over the past three years is quite unprecedented.  I am fairly confident that he will fail to live up to the $11.5 million dollars that he is owed in each of the next two seasons, though.  Mariano Rivera is signed to a relative bargain at $15 million next year, while Fernando Rodney is a free agent.  Will Rodney's agent cite Linear Saves in an effort to get his client a richer contract this winter?  If not, the man deserves to be fired. 

Disagree with something? Got something to add? Wanna bring up something totally new? Keith resides in Chicago, Illinois and can be reached at