April 12, 2003

Troy Percival Slams the Bil...

Troy Percival Slams the Bill James Closer Theory: James, for the two of you out there who don't what he's up to yet might still care, is a revolutionary baseball analyst, whose research and hundreds of invented statistical methods have turned conventional wisdom on its ear, in three short decades. This year, he's crossing the line from analyst/observer to participant, joining the front office of the Boston Red Sox. Unfortunately for him (I think), his most noticeable effect is helping to convince the team that a bullpen by committee is better than assigning Proven Closers to perform in rigidly determined situations. So far, the experiment has been a disaster. In this article, Angels closer Troy Percival gets up all over James' fancy ideas. It will be ironic, to say the least, if after seeing his ideas travel from lunatic fringe to mainstream, James ends up being known mostly as the damned fool who thought defined roles in the bullpen were overrated.

Posted by at April 12, 2003 05:48 PM
Comments

I have to admit that I've never heard that bit of Jamesiana before. In any event, the problem with Percy's critique is that a team might run into the same problem that the Dodgers had in 1997: a closer who can't be relied on to hold the lead. The manager is reluctant to go with someone else, since it might be interpreted as a lack of confidence in your closer, but his team blows lead after lead in the ninth. In the case of the Red Sox, a bullpen-by-committee might make sense if they don't have a reliable closer, since it gives the manager greater flexibility to make situational decisions for each batter. Like most of James' theories, there is always a good deal of initial skepticism before it becomes a part of the conventional wisdom.

Posted by: Steve Smith at April 12, 2003 08:16 PM

1) The last out is always the hardest to get. Yes, too often it seems that games are lost in the 7th or 8th inning when you WISH you had a better pitcher in, but experienced fans have seen numerous games slip away with 2 outs in the ninth. As Yogi said, it ain't over until it's over, so you want to make sure that you have some reliable way of getting it over with you on the winning side as often as possible.

2) There is something to be said for intimidation factor. An established closer can be terribly intimidating. Being a closer isn't necessarily being the best guy out of the bullpen; part of the formula is the other team believing you're hard to beat. I think that's been part of Hoffman's success over the last few seasons, especially when he's closing out games at home.

That's my answer to this little dust up.

Posted by: Howard Owens at April 12, 2003 09:59 PM

Whatever are the pros and cons of the closer-by-committee theory, you can't judge its efficacy by the results achieved in Boston so far. It doesn't matter *how* you deploy a bullpen with a group ERA near 7.00 -- they're going to blow some games for you.

Posted by: RobertNAtl at April 13, 2003 12:11 AM

What makes the bullpen by commitee approach even more problematic, it seems to me, is that entire farm systems have been built around the one-inning Larussa closers. We know the routine - starters go 6 or 7 innings, setup guys pitch the 7th and 8th with a lefthanded relief specialist (Orosco et al.) being spotted and then the closer finishing up. Roles have been established for pitchers not only for their current team but for them throughout their careers - minor league to the bigs. And guys know each day they come to park, which situations they'll likely have to pitch in.

This new approach by the Red Sox throws that out the windows. Guys won't be able to prepare - physically and mentally - since they're not sure what role they'll be playing in that day. There has been a sort of "baseball Darwinism" at play here; it's been used because it works.

Posted by: SMGalbraith at April 13, 2003 03:07 PM

I'm on the Troy Percival side of this (being a Yanks fan, that's almost a default, I suppose, what with Mariano Rivera's skills and all).

But I think the best argument against "committee" closers comes near the bottom of the piece:

"It comes down to the people element. Coefficients don't play baseball. When pitchers know approximately when they'll pitch, they can prepare. When they can prepare, they can relax."

Of course, the argument could be made that you could eliminate that uncertainty by having a "closing rotation" of sorts. That would be pretty amusing.

Then again, I'm certainly not one to complain about the Sox experimenting if it means they miss the playoffs for another year. =)

Posted by: Nik Bonopartis at April 13, 2003 03:21 PM

SMGalbraith:

I don't know where you got your information about farm systems, but according to the good people at Baseball Prospectus (www.baseballprospectus.com), who's job it is to watch more minor league baseball than anyone else on the planet, bullpen prospects have a decidedly worse chance of sticking in the majors than starters. LOOGYs (Lefty One Out GuYs) are made, not born.

Nik:

Funny you should mention Mariano Rivera, since he's probably the biggest reason the Yankees won the WS in 1996 and came one out away from winning it in 2001. Those two years, not coincidentally, were when Torre used him for more than one inning at a time, and brought him in with runners on base if it was necessary. Also, I'm sure that you'll remember that in '96 he was brought in for the seventh and eighth innings; John Wetteland was used for the ninth, and I seem to remember lots of hand-wringing when he wasn't kept for the '97 campaign. I'm sure the Yankees regret it now...

Three other points:

- It is way too early to determine the success or failure of this experiment. Lets see how things have shaken out at the end of the year, though even then we won't really have enough information to go on.

- The concept of a "closer" was as far from a natural occurrence as there is in baseball. It started with Jerome Holtzman coining the term, and settled into its current usage with Tony LaRussa's usage of Dennis Eckersley in '88. Do you honestly think that relievers from 20-30 years ago were in *better* shape than pitchers today? If not, then why shouldn't today's relievers be able to chuck it for two innings at a time? This isn't Darwinism, this is fear. Fear of getting castigated like Bill James and Theo Epstein are right now.

- In case you missed it, Alan Embree went on the DL. I wouldn't be surprised if it turned out a few more of the current relievers were injured as well; Chad Fox has always been brittle, to name just one.

Posted by: Noel Steere at April 14, 2003 09:03 AM

I'd rather see a Proven Closer than bullpen by committee any day. It's certainly more fun from a drama and rooting standpoint.

But it's really a bitch when your closer blows up at a critical moment, ie the World Series.

Think Mitch Williams in the '93 Series (hello Joe Carter). Dennis Eckersley in '88 (Kirk Gibson). Mark Wohlers in '96 (Jim Leyritz). Trevor Hoffman in '98(Scott Brosius).

Each was invincible during the regular season and playoffs (though the wheels started to fall off Hoffman with a blown save opportunity in the LCS). Each got tagged at a critical moment, losing the aura of invincibility. Each of their teams therefore lost a key weapon -- and the series.

Posted by: Mark at April 14, 2003 09:21 AM

I think that the part the Red Six missed is that while you don't necessarily want your best reliever to pitch the ninth inning, you need a guy who can come in and dominate to get out of trouble. If I have a 3 run lead- I would rather have my best reliever come in with the bases loaded and 1 out in the 7th or 8th inning than with a 3 run lead and no one on in the ninth. But you still need to have that one guy who you can count on to get that crucial out whenever it might come in the game. The Bosox don't so much have a closer by committe as they have a bunch of average to below average relief pitchers.

Posted by: Damon at April 14, 2003 09:56 AM

Right you are Damon. Sometimes I think managers are more motivated by puffing the save stats for their main guy than they are with, you know, actually winning games. That said, I do think its important to rely on your baddest meanest reliever for the tight spots.

Posted by: Lloyd at April 14, 2003 10:27 AM

Noel:
Sorry, not so in this case.

I think you'll need to follow baseball more closely than to just do a quick Google search and scan the Baseball Prospectus. What's the old saying, "Beware the man of one book." Or in this case, one url.

What farm system DOESN'T follow the one inning closer style? I'm talking double A and up, not the rookie leagues or single A. Starters go 5-6 innings, setup guys come in for a couple, closers finish up. Admittedly, specific players/relievers are used in different capacities as players are called up or down or released or injured. One guy will close for awhile; if he's called up, one of the setup guys will move to close et cetera. Player movement necessitates shuffling of roles; but the roles are there.

SMG

Posted by: SMGalbraith at April 14, 2003 01:46 PM

Bill James spends 30 years developing theories based on over 100 years of baseball history and after 12 games, three of which were lost during "closer" situations, we are dismissing the man as a kook.

Wow, welcome to the blogosphere, the land of fisking and fact checking and intellectual rigor, where idiotarians are quickly outed and chased from the Town Square with torches and stones.

Of course Troy Percival, whose salary is inflated by the accumulation of saves, should be the final arbiter. Anybody who follows baseball can tell you how well versed baseball players are in the history of the sport.

If anyone thinks that the addition of Kevin Millar, David Ortiz, Todd Walker and the Other White Giambi won't make up for the loss of Urbina, they obviously didn't watch Tony Clark and Rey Sanchez try to hit major league pitching last season. The approach isn't based purely on effectiveness, but value. And unless you've got a Nenn, Percival, Rivera, you're probably gonna over pay.

FM

Contrary to what Matt may tell you, I am neither dissatisfied nor gay.

Posted by: Scott Ross at April 14, 2003 01:56 PM

Ross -- You *know* better than to say that I'm dismissing the Bosox Bullpen Revolution based on its disastrous performance so far. I'm *guessing* it won't work (though I'm hoping for James' sake that it does) ... but I'll reserve actual judgment for several years from now.

As for your alleged switch-hitting and relative satisfaction, I'll testify that you are the most satisfied Red Sox-supporting friend I have. Granted, *that* particular sample size is on the small side....

Posted by: Matt Welch at April 14, 2003 02:04 PM

SMG- the fact is that of the closers today, only a handful were brought up through the minors to be a closer. In the vast majority of cases the closers of today are just starters who couldn't cut it as starters. If the 49 players who got at least 30 saves ing the 90's, only 4 were a reliever for their entire career. Another 9 players made no starts above A ball. The other 36 (out of 49) players were all brought up through the system as starters. Look at a guy like Billy Wagner who made no relief appearances before he made the major leagues. There are very few minor league closers who end up making any dent at all in the major leagues.

Posted by: Damon at April 14, 2003 02:11 PM

Felipe Alou's stock may be on the rise again, after some grumblings from the commentariat. During his heyday as a manager, his single greatest accomplishment was managing a bullpen against the grain of management group think. He often had a reliever like his nephew Mel Rojas pitch the last 2 1/3 innings.

To me, the greatest reason to discount anything Todd Percival says is that the best reason for bullpen by committee is that you don't have to pay 5.5 million a year for a guy who fattens up his save totals on 3 leads in the ninth. You can pay 3 guys 1.4 million to get 12 saves each. I don't think it would be possible for an established closer to endorse such an idea.

Matt, who would you say was the most effective reliever on your 1988 vintage Dodgers?

Posted by: copans at April 14, 2003 02:19 PM

(Hint: he had one save.)

Posted by: copans at April 14, 2003 02:24 PM

It seems to me that if you have a Troy Percival in your bullpen, your going to use him in most save situations anyway, whether you use a standard bullpen or bullpen-by-committee. Very few teams have a stopper/closer who's that good, so simply assigning your hardest thrower that role, rather than splitting it up between 3-4 pitchers, isn't self-evident. It's kind of like a manager who follows a policy of placing the fastest runner on the team as his leadoff hitter. If it's Rickey Henderson, circa 1989, that's an obvious move; if it's Omar Moreno, circa 1985, you're screwed.

Posted by: Steve Smith at April 14, 2003 02:35 PM

Copans -- Dude, I'm an *Angels* fan (and that's "Troy" Percival to you, pally), and I grew up hating the Dodgers.... Still, I'll try to guess, without peeking ...

Uh, Doug Rau? No wait ... Was Alejandro Pena in the bullpen/on the team that year?

Posted by: Matt Welch at April 14, 2003 02:37 PM

Pena was in the bullpen, as was Jay Howell, Jesse Orozco, Dennis(?) Powell, and, in the post-season, Tim Leary and, on one memorable occasion, Orel Hershiser.

Posted by: Steve Smith at April 14, 2003 04:01 PM

In truth, the Dodgers had a deep bullpen with Howell and Pena (and Orosco??), but down the stretch, Brian Holton strung together a great streak. He gave up only one HR in 84 innings and had a 7-3 record with a 1.70 ERA and held his own in the playoffs and series (if memory serves: I'm a Giant fan so it is all painful, Angels or Dodgers.)

Posted by: copans at April 14, 2003 05:06 PM

Google to the stats page for the 88 Dodgers (http://www.baseball-reference.com/teams/LAD/1988.shtml) shows the following for the relief corps:

Howell: 21 saves, 5-3 w-l record, 2.08 era in 50 games

Pena: 12 saves, 6-7, 1.91 era in 60 games

Orosco: 9 saves, 3-2, 2.72 era in 55 games

Holton: 1 save, 7-3, 1.70 era in 45 games

Crews: 0 saves, 4-0, 3.14 era in 42 games

Others who picked up saves that year included starters Belcher (4), Hershiser (1) and Valenzuela (1).

Posted by: mark at April 14, 2003 06:03 PM

Thanks, Mark. That looks like a committee to me, and deep bullpens tend to do well in postseasons (at least on anecdotal evidence) and have done so since the Fingers/Knowles/Locker era of the A's. (Often the weakness of the Braves in last decade, though not last year.)

Posted by: copans at April 14, 2003 07:53 PM

I haven't read any specific comments by Bill James about the usage of the Red Sox bullpen. What I do remember is his analysis of the use of relievers over the decades, trying to figure out what the most productive use of one's best reliever is. He concluded that the most productive use of a star reliever is to put him in when the game was close (within a run or tied) in the late innings, regardless of the score (which was pretty much the standard use of the ace reliever from the late '50s to the late '70s, until the Cubs' manager decided to prevent Bruce Sutter from wearing out late in the season by only using him in "save" situations, meaning he would almost never come in when the game was tied or the Cubs were behind). This makes sense--any competent major league pitcher will be able to protect a three-run lead in the ninth inning over ninety percent of the time, making it a waste to use the team's best reliever in that spot, whereas if the game is within a run, the marginal gain involved in putting a reliever with a 1.50 ERA as opposed to a 3.20 ERA is going to be far more significant. Following the logic of this would tend to lead to the team's second or third best reliever coming in for situations where the closer coming in would be more or less automatic (two or three run lead in the ninth inning) under current doctrine. Of course, if your second and third relievers are putting up numbers like Greg Maddux has this season, this plan will be perceived as deeply stupid by your fans, particularly if the fans in question are feeling a bit abused by the baseball gods, as is justified given the history of the last eighty years in Boston.

Posted by: M. Scott Eiland at April 14, 2003 10:04 PM

"Think Mitch Williams in the '93 Series (hello Joe Carter). Dennis Eckersley in '88 (Kirk Gibson). Mark Wohlers in '96 (Jim Leyritz). Trevor Hoffman in '98(Scott Brosius)."

Don't forget Hyun-Young Kim in the 2001 World Series. Bob Brenly putting him out there in Game 7 to give up what looked like a game losing home run after he had exhausted himself the night before in a heartbreaking loss was one of the most unforgivable things I have ever seen a manager do--I would have fired his ass on the spot if I owned the Diamondbacks, miracle ninth inning win and World Championship be damned. That's the kind of thing that destroys careers, and there was no damned excuse for it.

Posted by: M. Scott Eiland at April 14, 2003 10:09 PM

Yeah, I was just about to post about that Kim fiasco. Man, that hurt to watch!

As far as the BoSox being abused by the gods, no way. The abused the fates and recieved their just reward.

I'm a White Sox fan, so I know how it feels. We are cursed together.

Anyway, I have to agree that if you don't have quality, then James' strategy makes sense.

From a financial point of view, I wonder how much profit (in games won) is made by spending more money in other areas? What is the most efficient allocation of salary by position -- all other things being equal? If you could determine the base algorithm and then adjust for the quality in the talent pool, you might have an interesting formula. Oh God, the baseball math rears its ugly head...

Posted by: Christopher Walker at April 15, 2003 07:05 AM

As a Red Sox diehard I can tell you the only curse on the Red Sox is the curse of not enough pitching. The Red Sox alleged closers in the last couple of World Series '86 Calvin Schiraldi, '75 Dick Drago, '67 John Wyatt. Not exactly HOFers. I think you can have different guys close and if you don't have a guy who's dominating there's no sense trotting out the same guy because he's your designated closer. The Yankees in the 70's had Gossage and Sparky Lyle and the '79 Pirates used Tekulve and Grant Jackson. I don't see the point of signing some hot dog like Ugie Urbina to big coin because you're not going to win with him closing. If I had Eckersley in the late 80's or Willie Hernandez in '84 or Rivera a couple of years ago I'd set him up but I wouldn't base my whole team around setting up Armando Benitez or Billy Koch.

Posted by: Jack Tanner at April 16, 2003 12:10 PM
Post a comment









Remember personal info?






looked like a game losing home run after he had exhausted himself the night before in a heartbreaking loss was one of the most unforgivable things I have ever seen a manager do--I would have fired his ass on the spot if I owned the Diamondbacks, miracle ninth inning win and World Championship be damned. That's the kind of thing that destroys careers, and there was no damned excuse for it.

Posted by: M. Scott Eiland at April 14, 2003 10:09 PM

Yeah, I was just about to post about that Kim fiasco. Man, that hurt to watch!

As far as the BoSox being abused by the gods, no way. The abused the fates and recieved their just reward.

I'm a White Sox fan, so I know how it feels. We are cursed together.

Anyway, I have to agree that if you don't have quality, then James' strategy makes sense.

From a financial point of view, I wonder how much profit (in games won) is made by spending more money in other areas? What is the most efficient allocation of salary by position -- all other things being equal? If you could determine the base algorithm and then adjust for the quality in the talent pool, you might have an interesting fofates and recieved their just reward.

I'm a White Sox fan, so I know how it feels. We are cursed together.

Anyway, I have to agree that if you don't have quality, then James' strategy makes sense.

From a financial point of view, I wonder how much profit (in games won) is made by spending more money in other areas? What is the most efficient allocation of salary by position -- all other things being equal? If you could determine the base algorithm and then adjust for the quality in the talent pool, you might have an interesting formula. Oh God, the baseball math rears its ugly head...

Posted by: Christopher Walker at April 15, 2003 07:05 AM

As a Red Sox diehard I can tell you the only curse on the Red Sox is the curse of not enough pitching. The Red Sox alleged closers in the last couple of World Series '86 Calvin Schiraldi, '75 Dick Drago, '67 John Wyatt. Not exactly HOFers. I think you can have different guys close and if you don't have a guy who's dominating there's no sense trotting out the same guy because he's your designated closer. The Yankees in the 70's had Gossage and Sparky Lyle and the '79 Pirates used Tekulve and Grant Jackson. I don't see the point of signing some hot dog like Ugie Urbina to big coin because you're not going to win with him closing. If I had Eckersley in the late 80's or Willie Hernandez in '84 or Rivera a couple of years ago I'd set him up but I wouldn't base my whole team around setting up Armando Benitez or Billy Koch.

Posted by: Jack Tanner at April 16, 2003 12:10 PM
Post a comment









Remember personal info?






/body>