The failed theory of James sabermetrics on how “anyone could close in 2003” led to a porous bullpen and the need to sign Keith Foulke in the offseason.

Really? People are still propagating this misconception? Given that OP used quotes around "anyone could close in 2003" it would be swell of him to provide a link evidencing that quote.

it's too easy of a straw man for him to give it up

Silva's a joke. Doesn't even pretend to care about being logical.

Is Silva (a) sincere in his desire to know more about sabermetrics but incapable of understanding the subject or (b) intentionally provocative for the purposes of getting attention?

The failed theory of James sabermetrics on how “anyone could close in 2003” led to a porous bullpen and the need to sign Keith Foulke in the offseason.

Which is why the 2004 season ended up with the same Save total as 2003...36 Saves.

Never mind.

The failed theory of James sabermetrics on how “anyone could close in 2003” led to a porous bullpen and the need to sign Keith Foulke in the offseason.

Which is why the 2004 season ended up with the same Save total as 2003...36 Saves.

Because saves are the be all end all statistic to evaluate bullpen performance?

Because saves are the be all end all statistic to evaluate bullpen performance?

I think the point is you're supposed to be evaluating their cojones, not their statistics.

It was obviously the porous bullpen which blew it for the Red Sox in the 8th inning of Game 7 of the 2003 ALCS.

Do we still use that chain of wooden egg-shaped stones, or has science provided us a more efficient method of cojone-measuring?

Do we still use that chain of wooden egg-shaped stones, or has science provided us a more efficient method of cojone-measuring?

For awhile it was the cojone questec system, but they upgraded it to cojones-FX.

While the the overall cojone measuring has become much more standardized with the new system, a lot of umpires still claim it is worthless.

A few things

1. The piece was not "anti stats" if you read the entire thing I praise Toronto for bringing in Tango, as well as Tango's position that he is just a piece of the puzzle. I think Tom has found a way to market his ideas so that layman baseball people can incorporate it into their infrastructure. If you want to focus on the James paragraph so be it. He understands this stuff is more on the financial side than on the field.

2. Where is the venom about factual integrity? Cribb's statement that James was the reason behind the Sox 04' title is absurd. That team was started under the Duquette regime. I am hesistant to give Epstein 100% credit because he inherited the group.

3. We give Epstein credit for 07, yet Lowell and Beckett wouldn't have been in Boston if Theo wasn't off in his gorilla suit- am I wrong in saying that?


James has interesting work, but I am not about to give him credit for that 2004 title, or very much may I add. I think he probably has as much impact in Boston (if you could measure it) as Tango did in Seattle. And if you read Tango's quote he was a "piece" of all that.

You guys are focused on the wrong paragraph - this wasn't a slam

Go to the game log for 2003

Chad Fox and Brandon Lyon were closing early in the season. Not until they put Kim in as the closer did the bullpen stabilize. Even Fossum was thrown in there.

I remember thinking how ridiculous the whole "anyone can close" theory was.

No Keith Foulke in 04 = no title. Don't need James to tell me that

http://www.baseball-reference.com/teams/BOS/2003-schedule-scores.shtml

Is there any subject not covered on Wikipedia? :)

Measurement in the living adult is done in two basic ways:

• comparing the testicle with ellipsoids of known sizes (orchidometer).

• measuring the length, depth and width with a ruler, a pair of calipers or ultrasound imaging.

The volume is then calculated using the formula for the volume of an ellipsoid: 4/3 Pi × (length/2) × (width/2) × (depth/2).

I also found this uncited etymology amusing. Do they cover this in law school?

The etymology of the word is based on Roman law. The Latin word "testis", witness, was used in the firmly established legal principle "Testis unus, testis nullus" (one witness [equals] no witness), meaning that testimony by any one person in court was to be disregarded unless corroborated by the testimony of at least another. This led to the common practice of producing two witnesses, bribed to testify the same way in cases of lawsuits with ulterior motives. Since such "witnesses" always came in pairs, the meaning was accordingly extended, often in the diminutive (testiculus, testiculi).

Mike, the factual integrity issue is:

the failed theory of James sabermetrics on how “anyone could close in 2003”

Go ahead and find anything close to this (or the phrase "closer by committee") in anything that James has written.

Chad Fox and Brandon Lyon were closing early in the season. Not until they put Kim in as the closer did the bullpen stabilize. Even Fossum was thrown in there.

I remember thinking how ridiculous the whole "anyone can close" theory was.

The problem with this logic is that it assumes that Fox, Lyon and Fossum failed because they lacked some sort of "guts" gene. Look at the pitchers the Sox employed in the early part of the season as relievers, almost to a man they were pretty poor (or at least poor that season).

Mendoza, Howry, Fox, Todd Jones, these guys were all counted on and pitched very poorly. Lyon actually had pitched quite well as a closer saving his first 9 games then struggled a bit in June when he was likely already injured. It should be noted that Lyon (9 for 12, 75%) and Foulke (32 for 39, 82%) were pretty similar in their success rates as closers for the Sox.

The problem with the '03 Sox bullpen wasn't the lack of a closer, it was the lack of good pitchers.

I wrote this in my 2003 season preview:

Another big story was the Boston Red Sox hiring 28-year-old Theo Epstein as their general manager, the youngest in history. Epstein then did two interesting things. He hired Bill James, the father of modern, sophisticated "stathead" statistical analysis of baseball, as an advisor, and he let closer Ugueth Urbina and his 40 saves go, announcing that the Sox would take a "closer by committee" approach.

The idea that dominant closers are way, way overrated is bedrock gospel in the world of sabermetrics, as Jamesian analysis is called. If the save statistic isn't the most meaningless in all of sports, it's certainly in the running. A setup guy protects a two-run lead by coming in with the bases loaded in the eighth and retiring the side, but the save goes to the charismatic, dominant closer who waltzes in, heavy metal blaring on the P.A. system, to start the ninth. And the closer gets the big money. Why? Lots of saves!

So the Red Sox are making a good move by getting away from that idea, but here's the thing: I think the Boston bullpen is going to extend the illogical reign of the overpaid closer.

James has shown that the best time to use your best reliever is in the late innings with the score tied -- not in the ninth with a lead, which is current baseball thinking. No one argues that a dominant relief pitcher isn't a good thing to have. The question is when to use him if you have him. The Sox don't have that guy. They have a collection of pretty good relievers, guys like Alan Embree and Mike Timlin and Ramiro Mendoza. So even if they stay injury free, they'll just be, well, pretty good. People all over baseball, who are mostly suspicious of James and the wonky statheads who have grown up in his shadow, will point to Boston's performance and say, "See? You need a closer."

OK, I picked the Phillies to win the East that year. Missed by 15 games. But I think I nailed this one. The 2003 Boston bullpen did serious damage to the prospects for proper bullpen usage by being lousy, even though they weren't lousy because they were "closer by committee," they were lousy because they were lousy. I was too kind to Mendoza et al.

I don't remember now if the quoted phrase "closer by committee" was uttered by Epstein at some point, though from the way I used it I suspect so. It was certainly the vernacular for what the Sox were doing that offseason.

The problem with the '03 Sox bullpen wasn't the lack of a closer, it was the lack of good pitchers.

Exactly what I was going to say. The difference with Foulke wasn't some intangible thing, he was just better than all those other guys.
A relief ace is important; a "closer", in and of itself, is not.

One, James has probably never said anyone can close. You're confusing Bill James with Chef Gusteau from Ratatouille.

Two, for all of the guts (or performance under pressure, because, frankly, I'm not sure what guts actually look like -- as far as I'm concerned a player can have guts under pressure and still fail if he's not all that good to begin with) that Keith Foulke showed in 2004, Foulke was NOT famous for having those same guts in previous playoff incarnations on other teams. If the team's need was a a closer who'd be "gutsy" in the playoffs, there was no obvious way to know that Foulke was that person before 2004. His previous playoff performances (if I remember correctly) were hardly a testament to his gutsiness.

And in fact, for as mediocre as the pen was in 2003 during the regular season, it was actually quite good in the playoffs. Foulke was ridiculously good in the 2004 playoffs, but if one can justify acquisitions solely through performances in the playoffs (which I think is a ridiculous thing to do but it seems to follow from your line of argument), then there was no obvious need for Foulke since the 2003 playoff pen pitched well. (But that's silly -- you don't say, "Hey, we're good enough already." You get as many good pitchers as you can. Foulke was good, therefore it made sense to try to get him).

Three, aside from people trying to sell a book (see the baseballprospectus people, or, better, don't), to my knowledge, there are plenty of sabermetric people who acknowledge Duquette's contributions to the 2004 team. (I myself would give him about half the credit for the 2004 championship, though I've always tended to be biased towards Duquette). Of course, Duquette was famously pro-sabermetric at one point, so I'm not sure why giving him credit for 2004 would make sabermetricians totally re-think their position on the efficacy of sabermetrics.

First, you didn’t need advanced statistics to trade for Curt Schilling

Actually "you" did....or it was at least a great part of the presentation.

Did tango leave the M's? Or is he working for Toronto and Seattle?

Did tango leave the M's? Or is he working for Toronto and Seattle?

My understanding is that he is providing consulting services for both organizations.

I never said James said "anyone can close", but that is the basis of his theory? The quotes were used to accentuate the irony - I do believe I am allowed to use it in such situations.

Anyone can pitch in the ninth inning?

He basis the ninth on runs ahead, not factoring in his data is predicated on closers. Ask the Mets if anyone can pitch the ninth inning in 2008?

James' point was never "anyone can pitch in the ninth," Mike. The idea is that you should try to use your best pitcher in the highest-leverage situation, and that that is not always in the ninth.

It's really not a particularly difficult concept. In fact, it's put into play every single day in baseball, without a whiff of controversy. Pinch-hitters. No team keeps its best pinch-hitter in reserve for the ninth inning if a big, high-leverage pinch-hitting situation comes up in the seventh or eighth. It's the same logic, and yet no one suggests that using your best pinch-hitter in the seventh inning means "anyone can pinch-hit in the ninth" or "you don't need a best pinch-hitter."

I never said James said "anyone can close", but that is the basis of his theory?

Not really. The basis of his theory is that the massive, massive majority of pitchers can hold a three run lead for one inning, and that the gap between the best and the worst in those situations in terms of percentages is fairly slim. As a result, if you've got a great reliever or two in the bullpen, it makes a lot more sense to deploy him ealier in the game, when you're facing a smaller lead (or deficit) than to leave him sitting until than 3 run lead turns up in the ninth.

It's basically a proposal to return to the firemen of old, like Gossage.

The funny thing is, for all this Bll James-Red Sox-reliever talk....The Sox rarely follow this high-leverage business, and have often lost in the 10th or later inning on the road with Papelbon sitting on the bench picking his nose.

And no, I don't agree with it.

[24], [25]

Didn't The Book make the same points about leverage and relievers with 3-run leads in the ninth inning?

Didn't The Book make the same points about leverage and relievers with 3-run leads in the ninth inning?

Possibly. I haven't read The Book, although it is on the list of things to do.

Didn't The Book make the same points about leverage and relievers with 3-run leads in the ninth inning?

Yes. A lot of people and publications have. It's pretty basic stuff.

Mike, you got caught faking a quote and now you're being intentionally obtuse.

In fact, here's a particular odious piece of nonsense...lost in the 11th, with Mike Timlin, TIMLIN on the mound.

TIMLIN

In fact, here's a particular odious piece of nonsense...lost in the 11th, with Mike Timlin, TIMLIN on the mound.

Those bother me less on the road than at home. On the road in that situation even if the Sox get through the 11th they still would need to get through the 12th so it's reasonably defensible. If you don't want to ask multiple innings out of Papelbon you are committed to saving him for the save situation. There's no excuse for it in a home game.

In fact, here's a particular odious piece of nonsense...lost in the 11th, with Mike Timlin, TIMLIN on the mound.

TIMLIN

And, Papelbon hadn't pitched the two previous days, so should have been available for >1 IP.

It did seem in 2009, the Red Sox came around more on using Paps in high leverage situations on the road. But change in Boston is glacial sometimes.

I also found this uncited etymology amusing. Do they cover this in law school?

Is this the point whee I bring up the etymologay of "orchids"? You'll never look at those flowers the same way again...

Am I to understand from the excerpt that somebody pays Silva for this?

Dumber than ten dogs.

I'm not sure what guts actually look like

"Guts" wears a hockey mask and favors a Gibson Flying V bass.

Let's see:

I did find laughable the author, Robert Cribb, used Bill James with the 2004 Red Sox as an example of how sabermetrics works. First, you didn’t need advanced statistics to trade for Curt Schilling, nor did it develop the farm system (that was Dan Duquette) which allowed them to make such a trade. Key players such as Johnny Damon, Manny Ramirez, Derek Lowe, and Jason Varitek were acquired during the Duquette regime. The failed theory of James sabermetrics on how “anyone could close in 2003” led to a porous bullpen and the need to sign Keith Foulke in the offseason.

Okay--that's excruciatingly bad writing. A course in Grammar 101 could only help. With that it mind it's hardly clear Silva's claiming James actually said the quoted passage. Indeed, the utterly confused syntax throughout Silva's article supports his claim that it wasn't a direct quote.

I never said James said "anyone can close", but that is the basis of his theory? The quotes were used to accentuate the irony - I do believe I am allowed to use it in such situations.

No. No, you're not. Witness the confusion you've caused. You also owe it to people, in this case James, to not risk making them sound like they said something they didn't say. That's a common authorial courtesy you would do well to learn.

Mike, I know I've pointed out to you before what James has in fact said on the subject. You didn't acknowledge it last time. This time however you're not operating from ignorance.

From the comments section of the blog:

Spike
March 21st, 2010 at 2:51 pm

The failed theory of James sabermetrics on how “anyone could close in 2003”

I never said James said “anyone can close”

Your particular brand of chutzpah is beyond imagination.

"James sabermetrics" is murky, at best, but doesn't refer only to what James himself specifically asserted. The lack of an apostrophe (or a form such as "Jamesian sabermetrics") is atrocious, but not damning. "Arthurian legend" doesn't mean legends propounded by Arthur, nor does "The failed theory of Nixonian geopolitics including 'the domino effect'..." mean that Nixon himself actually used the term "the domino effect". It's bad writing, sure, but it's not an attribution. Look, it's not at all clear which is worse, Silva's thinking or his writing, but in this case I'm giving him a Jamesian pass.

Like many of us his age, Bill did most of his great work in the 80's. In fact as far as baseball analysis is concerned, and props out to Pete Palmer too, some might say Bill did most of 'the' great work in the 80's. The fact that it's hard to reproduce that sea-change now is largely because of how successful what he started has been. Most teams utilize statistical analysts. The field has gone so far it's impossible to be in the forefront as he once was. How about a moratorium on hacking at Bill. He's entitled to emeritus status where we don't expect the same production, and we show some deference.

If it's so hard to pitch in the 9th then it only makes sense that it's easier to hit. And that clutch-hitting only exists in innings 1-8 (and extra innings I suppose). So next time you see a walk-off celebration you should yawn and say "Big deal, it was the easiest part of the game".

[34] From what I can recall wrt reliever usage, Tito tends to play by the book during the regular season , then opts for heavy leveraging during the playoffs (w/ the 2004 ALCS, & the wanton ab/use of Foulke, being Exhibit A -- I think he also did this w/ Papelbon in the 2007 ALCS).

There's always, always, always been a disconnect between what James argued for regarding bullpen usage and what the Sox tried to do in 2003. Remember that the manager tasked with trying to implement the new bullpen strategy was so mentally lacking that he was lucky to get his shoes on the correct feet each morning. If nothing else, 2003 showed Theo that all the advanced baseball ideas in the world won't do you a damn bit of good if you have a grinning idiot managing games.

Also, it's once again worth noting that the '03 Sox bullpen got much better results once Kim was acquired; Grinnin' Gump absolutely rode him into the ground from July onwards, because he was the only decent guy in the pen for months on end. Once the playoffs rolled around Kim's arm was pretty much shot, and the Sox went with Embree, Timlin, and Williamson as their main pen guys, all of whom were fantastic in the playoffs. In fact, Grady ran Williamson into the ground just like he did Kim (had him pitch in all 5 games of the ALDS, throwing 2 innings in Game 4).

Not that idiot Grady noticed when it mattered most.

One thing Tito has done on a couple of occassions is use Papelbon in the 8th and when the Sox opened up the lead in the bottom of the inning yanked him for the 9th. It's only happened 2 or 3 times but I like that he's done it at all.

Sorry, Mike, your use of "anyone can close" re James was a mistake, and you should acknowledge that.

No disgrace in misunderstanding a concept. It happens to anyone. Now just recognize it and move on.

The idea is that you should try to use your best pitcher in the highest-leverage situation, and that that is not always in the ninth.

However, it is generally true that (a) the highest leverage situations DO occur in the ninth and (b) by reserving a specific pitcher for the ninth inning you are very likely to be using that pitcher in the highest leverage situation available in that particular game.

The problem with the argument that you should always use your best reliever in a high-leverage situation, even if that situation occurs before the ninth, is that you can't know in the seventh or eighth innings how the game is going to evolve going forward from that point. The current usage model is designed to have the best reliever in the game at the point where it is most likely to be the highest-leverage situation that the team will face - and when you see that the pitcher with the highest leverage and the most appearances in high-leverage situations on most teams is almost always the closer, it seems to be doing the job.

-- MWE

#48 If you look at James' proposed model the big change would be trying to get your ace into more games when you are tied or down by just one.

Tied in the 8th with the heart of the order due up is a very high weight situation. You won't get a higher leverage situation by holding him back.

Tied in the 8th with the heart of the order due up is a very high weight situation. You won't get a higher leverage situation by holding him back.

Agree. What I really want to see someday is up 5-3, 2 or 3 men on in the 7th, none out, closer comes in.

Even in the playoffs, teams use their 3rd or 4th best pitcher.

Agree. What I really want to see someday is up 5-3, 2 or 3 men on in the 7th, none out, closer comes in.

Even in the playoffs, teams use their 3rd or 4th best pitcher.

True, to an extent -- you often see managers Lefty/Righty these situations, to the extent they have the arms for the matchups. It's true that your LOOGY and ROOGY are often no better than your 3rd or 4th best relievers, but the point of them is that they can get a key out or two (and you wouldn't want them in for longer).

The problem with the argument that you should always use your best reliever in a high-leverage situation, even if that situation occurs before the ninth, is that you can't know in the seventh or eighth innings how the game is going to evolve going forward from that point.

I defer to better minds than mine, but this throws up my whole problem with leverage, theoretically.

1) Why should you use your best reliever in a high-leverage situation? If you had used your best relievers earlier on, then you are less likely to have a high-leverage situation later in the 9th (your team establishes an 'unassailable' lead) or that's probably why you have a high leverage situation (reasonable chance to win) at the end of the game. Considering all games start leverage-equal (minus home-field adv.), all leverage must be 'lost' at some point, so why not make sure your team ends up on the right side of that?

Imagine a one-game playoff, or better yet, Game 7 of the World Series, with everyone fully reseted, that the whole season comes down to 9 innings. The important thing is not when you use your best pitchers, it's making sure you use them, if we accept that high-leverage performance is entirely based on pitcher ability (which I know is a fairly controversial thing to accept).

This problem, I suppose, can be answered by making reference to future nights a reliever could pitch, or pitchers having a limited number of innings in their arm per season, but that seems a fairly unsatisfactory conclusion, considering it would involve a lot of assumptions about usage past and present that wouldn't apply in theoretical examples.

2) Leverage can only be adequately judged after the event with the knowledge of what comes later. The only reasonably-leveraged inning that can be predicted with any degree of certainty is top-of-the-first, and I'm not sure we should send the ace reliever out for that one.

Indeed, run differential, more than innings differential, is the crucial determinant of leverage; if the Brewers lead the Twins 10-0, it doesn't matter if it's the first, fifth or ninth, we pretty much know the result (and I think there's a fairly obvious inherent bias in discussions of WPA to discuss games where high-leverage situations can be found, rather than when the result has been 'essentially' decided by inning 4).

This in turn makes me wonder about whether false-consciousness of leverage could be caused by the structure of baseball as a game. Look at cricket, where one team (in the 50-over variant) amasses a score, and then the other 'chases' it - they tend to use their ace bowlers at the start and at the death, reasoning that the first few overs dictate the pace of the game and the last may decide the result. I wonder if baseball teams had 27 outs to score runs, prior to the opposition being given the same chance, would we see changes in pitcher usage? For the first team, all innings would essentially have comparable leverage. Then do you just go 'best arm', '2nd best arm' etc. (until each one is tired)? The second team would face increasing leverage until the last inning, but you should go in order from worst-to-best arms, if you were to use a leverage-based model.


As I say, there's loads of people here that know way more about leverage than my own poor mind can handle, but can someone solve those problems for me?

1. We excerpted the "three-run lead" part of the relief chapter in SI when The Book first came out. There's some good stuff in there. I'd say it's worth five minutes of your time.

2. James was mostly right, but partly wrong, regarding leverage, especially the tie-game. Really, it just comes down to the 3-run lead in the ninth inning. It sometimes comes down to letting your ace pitch two innings with a 1-run or 2-run lead in the 8th. It often comes down to your ace reliever pitching in blowout games, as he's been sitting around waiting for a "save" situation that never came, and so, he's going to pitch in games with no game impact just to keep him fresh.

3. Rob Wood did a fantastic piece, that should be re-read annually.