I wonder if this will receive attention every evening on SportsCenter as this approach did with the Red Sox a few years back. My guess is no given that, well, it's the Twins. And nobody at ESPN notices MN until about September.

What killed the Red Sox that year was that they blew saves on like two of the first three games or something like that.

Even by the end of the year when the Red Sox had pretty much fixed their bullpen with Kim and Williamson the narrative of "oh noes! the bullpen iz a dizastur" was still being pushed.

I'd like to see some sane, in-depth coverage of the move. I always find the whole closer-by-committee thing interesting even though the msm unaccountably corkscrews into hysterics over the idea of it.

What killed the Red Sox that year was that they blew saves on like two of the first three games or something like that.

There was also considerable backlash against the Bill James era beginning in Boston. Boston's big mistake was trying closer-by-committee with a crappy bullpen. I don't think that will be an issue with the Twins, as they have several valuable relievers, and this move should allow Gardenhire to leverage them properly.

Closers are so overrated....until of coure you don't have one.

There was also considerable backlash against the Bill James era beginning in Boston. Boston's big mistake was trying closer-by-committee with a crappy bullpen. I don't think that will be an issue with the Twins, as they have several valuable relievers, and this move should allow Gardenhire to leverage them properly.

They do? I'm not particularly comfortable with any of the Twins relievers. I suppose I trust Mayor McCheese (Guerrier) and Mijares the most, and Crain always has electric stuff with horrible luck, but none of those guys has me thinking they are particularly "valuable." Don't get me wrong, I think closer-by-committee is the right choice here because it allows Gardy to treat the 9th like any other inning and use his best matchups, but the bullpen was a pretty frightening spot last year.

I remember some analyst saying Boston's manager misinterpreted bullpen-by-committee to mean "everybody pitches."
I'd like to see some manager, faced with the "problem" of not having an established closer, develop a reliever in the role of "smokejumper" - a guy who comes in, no matter the inning, to put out the fire in a critical situation.

They do? I'm not particularly comfortable with any of the Twins relievers. I suppose I trust Mayor McCheese (Guerrier) and Mijares the most, and Crain always has electric stuff with horrible luck, but none of those guys has me thinking they are particularly "valuable." Don't get me wrong, I think closer-by-committee is the right choice here because it allows Gardy to treat the 9th like any other inning and use his best matchups, but the bullpen was a pretty frightening spot last year.

Rauch performed well after his arrival last year, and Neshek is back now. I think the bullpen depth is certainly better than it was a year ago.

GS4RB,

The one down side to that usage is the psychological impact of having a job where most of the hits you give up lead to giving up a lead. From a leverage basis it obvious would be the optimal way to use your best relief pitcher, but I just don't know enough about human psychology to know whether it would actually work in practice. My guess is that most baseball folk don't know enough about human psychology to judge it as well, so a complete rejection of the idea probably isn't warranted either.

Voros,
That's a good point. But I think a top-notch reliever, paired with a forward-thinking manager who deploys the smokejumper at the right moments, might lead to the opposite: The other team deflating because not only is the lights-out pitcher coming in, he's coming in at a moment where if they don't score they lose.

That's a good point. But I think a top-notch reliever, paired with a forward-thinking manager who deploys the smokejumper at the right moments, might lead to the opposite: The other team deflating because not only is the lights-out pitcher coming in, he's coming in at a moment where if they don't score they lose.

How do you predict those moments - men on base, game on the line? There is a chance the "smoke-jumper" warms up more than he pitches.

How do you predict those moments - men on base, game on the line? There is a chance the "smoke-jumper" warms up more than he pitches.

Good question - I imagine someone can run a cost-benefit analysis of having the smokejumper warm up but not get into a game vs. having the 3rd- or 4th-best reliever enter into the game's highest-leverage situation.

I'd like to see some manager, faced with the "problem" of not having an established closer, develop a reliever in the role of "smokejumper"

You would have liked baseball in the 1970s. Gossage, Marshall and Fingers were leveraged like this. They called it a "fireman".

I guess we'll get to see just how good Gardenhire is at managing a bullpen. I don't see a committee as a long-term solution. Either someone earns a closing role, they acquire a closer or they struggle all season.

You would have liked baseball in the 1970s. Gossage, Marshall and Fingers were leveraged like this. They called it a "fireman".

My baseball consciousness arrived in the mid-80s, when the 9th-inning-and-only-the-9th-inning closer mindset was coming into the fore.

Gossage complained regularly that Billy Martin "had me throw a f*cking season just warming up".

Gossage complained regularly that Billy Martin "had me throw a f*cking season just warming up".

It seems like that'd be inconvenient, but didn't Gossage understand why it was happening?

Then again, if Gossage didn't, then Jon Papelbon certainly wouldn't.

But you don't know when the highest leverage situations will be until the game is over. Even forget about the warming-up issue. It's a tie game, 2 on, 1 out in the 7th. Do you bring in your "smoke-jumper"?

Say you bring in your "smoke-jumper", and he gets you out of the jam. Now who's going to pitch the 8th and the 9th? Your 3rd/4th best relievers? Who now get into even higher leverage situations, which you can't bail them out of, because you already shot your bullets? Or do you keep your "smoke-jumper" in there for the 8th too, in which case (1) he looks a lot like an old-fashioned fireman (2) he won't be able to "smoke-jump" tomorrow (3) over 162 games you are asking for an injury (4) you still need to worry about the 9th inning.

Or... do you say you'll never bring in your smoke-jumper in a situation where he can't finish the game, i.e. 8th inning or later? Well frankly that's how the dependable closers are used these days.

The advantage of the closer model is that it takes a lot of the thinking away - everyone knows what is going to happen. You do see closers come in before the ninth, and you do see them in tie games. Managers aren't stupid. But you don't see them in tie games before the ninth, and you don't see them in with leads before the 8th, because leverage just grows sequentially by inning - unless your offense is scoring, in which case you can afford to give up that run anyway.

Then again, if Gossage didn't, then Jon Papelbon certainly wouldn't.

Papelbon strikes me as the kind of man who doesn't understand why the sun rises in the east.

I wonder if this will receive attention every evening on SportsCenter as this approach did with the Red Sox a few years back. My guess is no given that, well, it's the Twins. And nobody at ESPN notices MN until about September.

I'm guessing a team that turns to a "closer by committee" as a result of their superstar closer going down with an injury in March won't receive quite as much attention as a team actively planning a closer-by-committee from the beginning, and announcing to the world that established closers are overrated.

Just looking at the "leverage" numbers in the splits on BRef, it sure looks like Rivera faced a greater percentage of hisa hitters in high leverage situations than Gossage did in 1978-1980. Gossage of course pitched a hell of a lot more innings, even though he missed some of 1979 with an injury.

Rivera seems to get over half of his PA s in high leverage spots, while Gossage was generally in the 40% range during that span. It wasn't so much that Gossage was saved for high leverage situations, it was that Gossage was used enough to cover both Rivera's role and about half the role of the set up man as well.

To a certain extent, having it be the 9th inning adds a considerable amount of leverage in and of itself. So even the oft lamented "coming in to start the 9th with a three run lead" type save isn't a total waste. I mean those types of saves can be and have been blown, and doing so greatly reduces your chances of winning.

Now I don't know the full ins and outs of how Sean is measuring leverage in those numbers, so there may be some ways to dispute how well it's capturing what we truly want to measure when it comes to leverage.

EDIT: I think me and AlouGoodbye are trying to get at the same thing in different ways: modern closer use may not be optimal, but I don't think it's as far from what is optimally doable as people sometimes think. If I have a criticism of the modern closer, it would be that sometimes I think many of the best ones would have made very good to excellent starting pitchers if given the chance.

[20]

Gossage was used in high leverage situations as much as Rivera. Plate appearances alone doesn't capture that fact, however.

Baseball-reference shows that during 1978-1980, when Gossage entered the game, 64% of the first PA of the pitcher's appearance had a leverage greater than 1.5, versus 60% for Rivera during 1997-2009. (During these periods, Gossage entered the game before the 9th inning about 60% of the time versus about 20% for Rivera.)

EDIT: Gossage entered 56% of games with runners on base versus 21% for Rivera. Gossage entered 69% of games with the lead versus 82% for Rivera. These are not criticisms of Rivera, but just comparisons of how these two relievers were used.

I think me and AlouGoodbye are trying to get at the same thing in different ways: modern closer use may not be optimal, but I don't think it's as far from what is optimally doable as people sometimes think.

Didn't James conclude this in the NHBBA or somewhere else?

Also, haven't others concluded that the only real issue with the current closer mentality is the use of your closer with a 3-run lead? Thought we just saw a link to an article saying this within the last week or two.

[22] Based on a rough look at Baseball-reference Play Index, it looks like 20% of Rivera's games were of the 9th inning, 3-run lead variety. For Gossage during 1978-1980, it looks like 9%.

I would bet that Gossage was much more likely to face the opponents' best hitters than Rivera too. One, he'd be pulled in the 8th if they were coming up and he had more multi-inning appearances.

If there's one thing that annoys me about modern bullpen usage, it's when a team has its 2nd best reliever pitching to the heart of the order in the 8th then the closer gets the #7-9 batters in the 9th. Leverage ain't just about score and inning y'know. (Of course, on some teams, the "2nd best reliever" is actually the best reliever.)

Boston's big mistake was trying closer-by-committee with a crappy bullpen. I don't think that will be an issue with the Twins, as they have several valuable relievers

I'm so sick of this ####. NOBODY in here thought Boston had a crappy bullpen when the season began. It may have turned out that way, but Mendoza, Embree, and Timlin were all coming off very good seasons. You really think Guerrier and Rauch project to be much better pitchers than those guys did?

Gardenhire mentioned right-handers Matt Guerrier, Jon Rauch and Jesse Crain and left-hander Jose Mijares as possible members of the committee.

Can we call them the Gang of Four?

To hell with poverty we’ll get saves with cheap arms
To hell with poverty the closer will arrive we’ll beat the Sox again

Best Dressed Chicken, I predicted that the 2003 Red Sox bullpen would be crappy, that "bullpen by committee" would fail, and that "bullpen by committee," not crappy pitchers, would get the blame. This is exactly what happened. And I am no great prognosticator. I'm proud of that one, but I don't think I was off on an island in the spring in saying the Sox bullpen just didn't have the horses.

Jose Mijares don't run on time
He believes it's no coincidence
He thinks fatness will drag him down
Home! It's no castle!
He misses the Metrodome!

I suppose I trust Mayor McCheese (Guerrier) and Mijares the most, and Crain always has electric stuff with horrible luck

"Mayor McCheese"? He doesn't have a weirdly-shaped head that would suggest such a nickname.

BURN THE WITCH!

Guerrier ---> Gruyère, a type of swiss cheese ---> Mayor McCheese. At least, that's how I guess that nickname was developed.

Didn't James conclude this in the NHBBA or somewhere else?

James' suggested use aims not just for higher leverage, but more use from the elite relievers.

two innings a game when the game is tied
two innings a game when you have a one run lead, and
one inning at a time in other games when the game is close
at the end and the relief ace hasn't been used for a game or two.

"Never" before the 8th. (Yeah you can probably come up with a scenario where it would make sense to use your ace sooner)

In his sims this produced a workload of around 69 games, 113 IP.

I believe Brock Hanke would argue that this is a tad high. And BL has argued that it makes sense to be risk adverse.

The obvious place to cut back is 2 inning appearances with the bottom of the order up in the 8th.

Guerrier ---> Gruyère, a type of swiss cheese ---> Mayor McCheese.

That's 100% accurate. It beats "Matty" as a nickname, for sure.

#25 Timlin was fine in 2002. But why would you ignore 2001 or 2000? Overall Timlin pitched a little better than I'd expected in 2003.

Mendoza was a huge part of the disaster season.

Embree had been very good in 2002. It was also his second good season for a guy who'd been around for a while. Expecting him to repeat a good season was pretty much wishful thinking.

Notwithstanding Guerrier's poor 2008 I'd have him comfortably ahead of any of the 3 you named. I'd have Rauch below Mendoza and roughly equal with Timlin.

I'm so sick of this ####. NOBODY in here thought Boston had a crappy bullpen when the season began. It may have turned out that way, but Mendoza, Embree, and Timlin were all coming off very good seasons. You really think Guerrier and Rauch project to be much better pitchers than those guys did?

Just because nobody expected them to pitch poorly doesn't mean they didn't wind up pitching poorly. The problem wasn't Timlin or Embree though it was that beyond those two guys (and that's giving Embree - 4.25 ERA - benefit of the doubt) the Sox had nothing. Mendoza was awful, Bobby Howry, Chad Fox, Steve Woodard and Jason Shiell were all on the Opening Day roster with dreadful results. In retrospect only Howry has gone on to pitch well since.

Like any other bullpen plan, you need a few guys to pitch well to make it work and the Sox simply did not have guys pitching well enough to make it work.

There was also considerable backlash against the Bill James era beginning in Boston.

Wasn't part of that backlash being the team hired a guy who thought pitchers didn't have any effect on balls in play? :-)

I'd like to see some manager, faced with the "problem" of not having an established closer, develop a reliever in the role of "smokejumper" - a guy who comes in, no matter the inning, to put out the fire in a critical situation.

You do that in the 7th. You are successful. When the 8th inning comes around and the score is still 5-4 do you take your best pitcher out? How about the 9th? If you leave him in then you've got the Gossage workload. If you take him out then you may not have used your best pitcher optimally. As high leverage as a bases loaded situation with a one run lead is in the 7th, it is even more high leverage in the 9th.

By limiting appearances and innings teams have been able to get higher average leverage, better performance out of the elite relievers, and they are more likely to stick around instead of burning their arms out. I did an article on this for the THT annual. My mistake in the article was the example I used - Joe Nathan. Sorry Twins fans for the jinx. You guys probably already hate my ilk for Homerun Kennedy.

for all the talk on how the 2003 Sox pen was a disaster, proved you need a closer, etc etc: the team won 95 games using a bunch of different guys to close games, the bullpen was unbelievable in the playoffs and bailed out the offense (not the other way around), 3 different guys were used as 'closer' in the playoffs, and the team came within a few outs of winning the pennant.

So even if you believe their strategy was a mistake, it was pretty easily overcome.

retroactively and cumulatively, that team's weaknesses were Derek Lowe and John Burkett, not the relief crew.

I think Closer by Committee is a horrible mistake. The should go with a College of Closers. Or a Closer Panel. Preferably a Blue Ribbon Panel on Closing.

Or they could just hire Kyra Sedgwick.

You do that in the 7th. You are successful. When the 8th inning comes around and the score is still 5-4 do you take your best pitcher out? How about the 9th? If you leave him in then you've got the Gossage workload. If you take him out then you may not have used your best pitcher optimally. As high leverage as a bases loaded situation with a one run lead is in the 7th, it is even more high leverage in the 9th.

Yes, but you have no idea whether the game will be as close in the 9th. You have no guarantee of having used him "optimally" by waiting. I haven't looked at the numbers, but I would rather use my ace reliever in the 6th inning, bases loaded, tie game, other team's best hitter at the plate than the 9th inning with a 3-run lead. But nobody would ever do this.

You have no guarantee of having used him "optimally" by waiting.

no guarantee either way. If the guy up before the other team's best hitter hits a 3-run homer and your ace reliever is all warmed up and ready . . .

And you have no idea that there will be a 3 run lead in the 9th. That is something that could be looked at though: If the 6th inning ends in a tie, how often will there still be a tie in the 9? A one run lead? a 12 run lead?, etc.

And with such a distribution you could calculate the probability that the 6th inning will be your highest leverage situation.

I much prefer a bullpen by committee.

I would rather use my ace reliever in the 6th inning, bases loaded, tie game, other team's best hitter at the plate than the 9th inning with a 3-run lead.

Again, this used to be the norm. The thing that no one would ever do, however, was to take that guy out of the game unless he coughed up the lead. The effect of this usage was that relievers rarely had effective seasons back-to-back and generally had short careers. We know that we got one Goose Gossage as a result of this approach. We have no way of knowing how many Trevor Hoffman's we missed out on.

You do that in the 7th. You are successful. When the 8th inning comes around and the score is still 5-4 do you take your best pitcher out? How about the 9th? If you leave him in then you've got the Gossage workload. If you take him out then you may not have used your best pitcher optimally. As high leverage as a bases loaded situation with a one run lead is in the 7th, it is even more high leverage in the 9th.

Right. It doesn't make sense to use your best reliever in the 7th inning of every 1-run game, but there may be games where it makes sense -- depending on the quality (and handedness) of the hitters up that inning and the base-out situation. I don't think anyone is saying you should do that, they are just arguing against inflexibility with reliever usage.

The thing that no one would ever do, however, was to take that guy out of the game unless he coughed up the lead.

But there's no reason you couldn't.

But there's no reason you couldn't.

Sure there is. It's exactly the same reason that present-day usage patterns have so rapidly evolved into the set-in-twelve-feet-of-reinforced-concrete roles that we all love to rail against. The second guess. How on earth do we really expect any manager to stand in front of a 21st century post-game press conference and talk about how he had to have his best guy in there to win the game in seventh after he lost the game with his second or third best in the eighth or ninth?

And you have no idea that there will be a 3 run lead in the 9th. That is something that could be looked at though: If the 6th inning ends in a tie, how often will there still be a tie in the 9? A one run lead? a 12 run lead?, etc.

And with such a distribution you could calculate the probability that the 6th inning will be your highest leverage situation.

Has such a model not been done? I figured that was what WPA is based on. In any case, a tie game in the 6th inning is approximately 50-50. But it's unlikely that a game that was tied in the 6th inning will be tied in the 9th inning, regardless of what reliever you use. So while the 9th inning in a tie game is a higher "leverage" situation than the 6th in the sense that there's a smaller window to recover from the what transpires, it's not higher leverage in the sense that taken alone it has a higher impact on the outcome, of course.

Sure there is. It's exactly the same reason that present-day usage patterns have so rapidly evolved into the set-in-twelve-feet-of-reinforced-concrete roles that we all love to rail against. The second guess. How on earth do we really expect any manager to stand in front of a 21st century post-game press conference and talk about how he had to have his best guy in there to win the game in seventh after he lost the game with his second or third best in the eighth or ninth?

You say there would have been no lead to lose in the 9th if we don't get out of the 7th.

What are we talking about, maybe 3 or 4 games a year where the "fireman" is needed in the 7th? Generally, he be brought in in no earlier than the 8th and could finish. Of those 3 or 4, maybe for 2 he's gets the outs easily, or is well rested enough that he can get the 2+ IP save.

So, we're talking maybe once or twice where the "fireman" has to be pulled in the 9th. The manager should have the guts to face one extra nasty press conference a year.

I think, in general if you move to fewer, longer outings, a RP should be able to pitch more innings due to less warming up, greater rest between outings. Instead of 65 IP in 65 Gs, Rivera could give the Yankees 80 IP in 50 Gs.

What are we talking about, maybe 3 or 4 games a year where the "fireman" is needed in the 7th? Generally, he be brought in in no earlier than the 8th and could finish. Of those 3 or 4, maybe for 2 he's gets the outs easily, or is well rested enough that he can get the 2+ IP save.

In my opinion, there are far more than 3 or 4 games a year where the fireman should be used in the 7th or earlier. Generally speaking, if it's a tie or one-run game, there are men in scoring position, the heart of the order is up, I'm going with my fireman over any other reliever.

How on earth do we really expect any manager to stand in front of a 21st century post-game press conference and talk about how he had to have his best guy in there to win the game in seventh after he lost the game with his second or third best in the eighth or ninth?

Absolutely right. It would be much easier if he could stand in front of them and say "the numbers show we had a 50% chance of winning based on the circumstances in the 7th, but had we given up a run there it would have dropped to 33%."

That's 100% accurate. It beats "Matty" as a nickname, for sure.

What's wrong with Shaggy?


If Gardy is as brilliant with managing his bullpen as many think he is, then this is the correct approach. Let him look at his relievers performance and usage and the game situation and use them as best he can divine. If instead he just knows what boxes to put relievers in and uses them thoughtlessly when they are in their boxes, then this could be more interesting.

Has such a model not been done? I figured that was what WPA is based on. In any case, a tie game in the 6th inning is approximately 50-50. But it's unlikely that a game that was tied in the 6th inning will be tied in the 9th inning, regardless of what reliever you use.

Not that I know of. On average, the 9th inning should be the same leverage as other innings, as it includes 13-1 blowouts as well as tied games. But if we know that the game is tied through 6, that should make it much more likely that it will be tied or within a run in the 9th inning. How much more, I don't know.

Too bad such discussions that would inspire research usually happen when I'm at work, away from my retrosheet database. And by the time I get home, I will have little time outside of playing with a 2 year old until her bedtime and then watch 24. Maybe in another year I'll have my little one sufficiently skilled in SQL that we can look at retrosheet together.

Too bad such discussions that would inspire research usually happen when I'm at work, away from my retrosheet database. And by the time I get home, I will have little time outside of playing with a 2 year old until her bedtime and then watch 24. Maybe in another year I'll have my little one sufficiently skilled in SQL that we can look at retrosheet together.

Or you could just copy all the data to your work computer. Why waste valuable personal time doing this research? Waste your employer's time.

Or a Closer Panel. Preferably a Blue Ribbon Panel on Closing.

A death panel?

It beats "Matty" as a nickname, for sure.

Some might disagree.

What are we talking about, maybe 3 or 4 games a year where the "fireman" is needed in the 7th?

If you're right about this (and NYCTigersfan is wrong) then it just isn't worth the trouble. The times you don't needlessly warm up the ace and the certainty of assigned roles easily make up for the closer missing 3 or 4 high leverage innings.

I think, in general if you move to fewer, longer outings, a RP should be able to pitch more innings due to less warming up, greater rest between outings. Instead of 65 IP in 65 Gs, Rivera could give the Yankees 80 IP in 50 Gs.

This is the obviously correct usage pattern if you have two or three more guys who are all about equally good and also nearly as good as your ace. Unfortunately, that's a very big if for most teams in most seasons.

If you're right about this (and NYCTigersfan is wrong) then it just isn't worth the trouble. The times you don't needlessly warm up the ace and the certainty of assigned roles easily make up for the closer missing 3 or 4 high leverage innings.

Well, I think the 7th inning opportunities are few, say 5 to be generous, but there are probably 20+ games where it makes sense to use the "fireman" in the 8th. hat's what makes it worth while.

This is the obviously correct usage pattern if you have two or three more guys who are all about equally good and also nearly as good as your ace. Unfortunately, that's a very big if for most teams in most seasons.

I don't think so. You just need somebody who can pretty reliably pick up those 10-15 easy-saves. Againg using Rivera, last season he had 9 1 IP, 3-run lead saves. Any competent RP can handle those games. You only need to transfer a handful of other outings (2-run leads against the 7,8,9 hitters etc.) to get him down to 50 G.