Mangled the tags there ... you get the idea ...

Steve, you'd have just cringed during the Mets-Astros game tonight, as Garner resolutely refused to bring in Lidge because it wasn't a save situation. Predictably, the Astros ended up losing 1-0 in 11 innings, without ever using their best reliever. I invoked you in the game chatter . . . . .

McKeon's one of my all time favorite managers. He uses platoons to good effect (rare these days), he's been at least open to a 4 man rotation IIRC, and the way he reined in the running Marlins of 03 was bold and effective. Larkin hated him in Cincy, but whadayagonnado?

I was pretty sure that before the labels of "closer by committee" or "bullpen by committee" appeared, and before the modern emphasis on the Save statistic, the notion of "using your best reliever as often as possible, in the most important situations" was around. And it was universally accepted. But hey, what do I know?

I was pretty sure that before the labels of "closer by committee" or "bullpen by committee" appeared, and before the modern emphasis on the Save statistic, the notion of "using your best reliever as often as possible, in the most important situations" was around. And it was universally accepted. But hey, what do I know?

You know the truth. Duh.

I'm hard at work compiling a comprehensive database of closer and LOOGy deployment since the 1960s, and will be writing it up over the next couple of weeks. I suspect it will be of interest.

the notion of "using your best reliever as often as possible, in the most important situations" was around. And it was universally accepted.

Those guys were the old-school firemen right?

These people think you are saying we can bring in just anyone to get the save

No these people, including Mike Emeigh, state that the modern utilization of closer maximizes LI over any previously developed deployment model. And that is a fact.

Now, I don't know if Emeigh says this, but others might say that the modern trend of increasing pitching staffs and creating defined roles increases the length of pitching careers. The latter is a fact.

Others might also say that using anybody as a closer is just about exactly what the Red Sox did a couple of years ago.

Others might also say that some pitchers have difficulty in the closer role, but perform admirably in many other roles. And you can look at a bunch of people including Timlin, Remlinger, Karsay, Hawkins for examples.

Others might also say that bullpen deployment is personnel dependent, but the best default model is closer, and its certainly better than ACE RELIEVER if you just look at LI.

Others might also say that diversification of innings promotes better staff effectiveness than concentration of innings and use the Atlanta Braves as an exemplar.

Others might also say that any tangential advantages with ACE RELIEVER is completely refuted bunk, but it won't keep certain people from positing it again.

Others might also say that "create a reliever" has failed many times and the failure by a closer is more manifest in actual W-L than any type of translation therom.

Others might also say that bringing in your best pitcher prior to the 8th would be pretty defeating. Their is a high possibility you will face the same LI in the 9th, so your other pitcher has to overcome that LI to begin with. If you use the other pitcher first and he fails, you still lose but you don't waste utilization of your best pitcher. If he succeeds, the same is accomplished with either model.

I say all those things. I'll leave it to your judgment if I'm openminded about sabermetrics.

I'm hard at work compiling a comprehensive database of closer and LOOGy deployment since the 1960s, and will be writing it up over the next couple of weeks. I suspect it will be of interest.

Hardly.

Others might also say that bringing in your best pitcher prior to the 8th would be pretty defeating. Their is a high possibility you will face the same LI in the 9th, so your other pitcher has to overcome that LI to begin with. If you use the other pitcher first and he fails, you still lose but you don't waste utilization of your best pitcher.

Using your best pitcher prior to the 8th inning would a lot of the times include him coming on with RUNNERS ON BASE, which makes the sitatuion an even higher leverage than say a 1 run lead going into the 9th.

Also, using your best reliever early is balanced by the possibility that your own team might score a bunchload of runs in the top of the 9th/bottom of the 8th, thus enabling you to use a mopup reliever for the new low-leverage situation.

Using your best pitcher prior to the 8th inning

I maent prior to the 9th.

Also, using your best reliever early is balanced by the possibility that your own team might score a bunchload of runs in the top of the 9th/bottom of the 8th, thus enabling you to use a mopup reliever for the new low-leverage situation.


First, this has been covered before. Its not "balanced", not even in the best stathead creative reading of words. Second if it did occur, you have now once again burned utilization for no reason, unless you think your closer caused the rally.

<i.>I maent prior to the 9th. </i>

Then you posted a rebuttal to a single point that is not even a rebuttal. Classic Wok. As such I'm not sure if I need to address your first point as you don't even really understand what I'm talking about.

I say all those things. I'll leave it to your judgment if I'm openminded about sabermetrics.

To the extent that I express an opinion, I see a lot of wisdom what you say, although I would (1) defer to the expertise of others (2) wonder whether "and that is a fact" is a little strong. I also see a lot of wisdom in the other point of view. I have no dog in this fight. I don't find pitching all that interesting, to be honest. Let other people think about it.

I was merely pointing out (1) That Hench appears open-minded about sabrmetrics (2) That those opposed to the "closer model" have done a poor job communicating their argument. Now, that argument may be correct or incorrect, but that's another issue entirely, and one - as I say - that I'm not particularly interested in discussing with you.

I think the two limited points I made are quite true, however. As for whether you are openminded - I think you can be quite dogmatic, but you're at the table making well-thought-out arguments, which is the main part. So you are openminded, yes.

Backlasher,

Using your best reliever in the most important situations means that you're using him in the situations where, if the other team scores, it most greatly decreases your team's chances of winning. If you want empirical proof (and measurement) of when those situations are, then you need look no further than Studes' work on Win Probability. In a later article, titled Closer, Studes gives a very detailed account of why the ninth inning, or more accurately, a Save situation, isn't always the best time to use your best reliever (aka your "Closer").

Although the two most important general situations are in the ninth inning (when your team is up by one run or the game is tied), there are other situations which are more important than the ninth inning where your team's chances of winning are more greatly harmed than allowing a run when there is a 2-run lead in the ninth.

I'm fairly certain that there are situations you will encounter in the eight or even seventh inning where using your best reliever is the best option, because that point in the game gives the other team the biggest boost to their chances of winning (should they score a run).

Sure, there is something to be said for having to use your closer "early" and potentially "wasting" him later on. But the fact of the matter is that you don't know what will happen in the future - but what you do know is that the situation you're currently facing is extremely important. Keeping a reserve is an old military axiom, and it does apply in many real-life situations, but sometimes you have to take things one day (or one inning) at a time.

Sure, there is something to be said for having to use your closer "early" and potentially "wasting" him later on. But the fact of the matter is that you don't know what will happen in the future - but what you do know is that the situation you're currently facing is extremely important. Keeping a reserve is an old military axiom, and it does apply in many real-life situations, but sometimes you have to take things one day (or one inning) at a time.

Inq,

So let me get this straight, in one instance you want to use probability and in the next instance you want to tell me you don't know what is going to happen.

I don't have to use Studes work, because I can use Tango's work to show an actual increase in LI over all other previously deployed models. You are leaving out availability in your implicit calculations.

This has all been done before. As recently as a month ago on this forum with much better composites than any blog article.

It's back to the old Diamond Mind thing for me: really, play a load of games and you tell me when the best time to bring your best reliever is.. You have ases loaded, one out in the 7th? Do you really send in anything but your relief ace?

I was playing with the 2002 Dodgers early in the year, and for an amusing part of the season Eric Gagne was leading the league in wins (up to about 6 wins). This is a fluke of course, but illustrates my point to an extent.

Baseball men would just dismiss it as a computer sim not reflecting reality, and they'd be half-right, but it drums home a lot of good common-sense too I think.

Then you posted a rebuttal to a single point that is not even a rebuttal. Classic Wok. As such I'm not sure if I need to address your first point as you don't even really understand what I'm talking about.

IT WAS A TYPO YOU SANCTIMONIOUS SELF RIGHTEOUS SOB

balanced by the possibility that your own team might score a bunchload of runs in the top of the 9th/bottom of the 8th

bunchload...I like that word.

Predictably, the Astros ended up losing 1-0 in 11 innings, without ever using their best reliever.

They used Dan Wheeler, right? That's their best reliever.

(Okay, he's not, but he is family, sort of.)

Did anyone associated with the Red Sox even use the term "closer-by-committee" that spring? I think Kevin's right that the plan was that one of the good relievers they had would emerge as the closer. They were hoping to create a closer rather than trading for an established one.

It's back to the old Diamond Mind thing for me:

If Diamond Mind modelled availability properly than it would be germaine. Although Tippett doesn't disclose, it would seem the only availability factor he uses is rolling pitch counts aggregated over four days.

IT WAS A TYPO YOU SANCTIMONIOUS SELF RIGHTEOUS SOB


Which one "before the 8th" or "before the 9th". You might want to go back and read to figure out why you are being non-responsive. You might want to read Tango and Emeigh's in thread works about contextual success rate. But that would be too hard for you when you can just trot out some non-probative talking point you heard somewhere else.

Which one "before the 8th" or "before the 9th". You might want to go back and read to figure out why you are being non-responsive. You might want to read Tango and Emeigh's in thread works about contextual success rate. But that would be too hard for you when you can just trot out some non-probative talking point you heard somewhere else.

You're either stupid or trying to push Chris's buttons. I'm sure it's the latter, but I just don't understand why. Are you a sadist? Does making others angry and upset give you pleasure?

You're either stupid or trying to push Chris's buttons.

How? No one has made a statement wrt LI potential in the 8th versus the 9th, yet this is what Wok is arguing after his amendment. Its not a typo its a reading comprehension error.

These are not the words of, say, Joe Morgan or Buster Olney. Hench may be wrong, but it stems from misunderstanding, not deceit.

I definitely wasn't charging deceit in regard to Hench (or anybody else for that matter). I also wasn't placing Hench in the Orthodox Stathead pigeonhole, either. I was inferring that many people refuse to accept (or can't remember a few decades back when it was operational) a different way of doing things. Instead of analyzing James' approach (which was around when James was a pup), they would rather lash out at sabermetrics, "number crunchers," and their like.

IOW, I can handle an article that is critical of the fireman approach, but I have little respect for writers who are busy doing more disparagement of the proponents than actual critiquing. In regard to Hench, it's a little surprising to see him on that side of the aisle.

I should also add that I would like to see the two schools of thought in regard to relief pitching duke it out on the field. Eventually, one of them would be left standing: if it's not the fireman model, then I'll be the first to say that I was wrong.

I should also add that I would like to see the two schools of thought in regard to relief pitching duke it out on the field. Eventually, one of them would be left standing: if it's not the fireman model, then I'll be the first to say that I was wrong.

I guess I don't agree with this...my problem with the closer model is not so much that it doesn't work, but rather that it's applied so slavishly at times. I don't see why the two schools (or more) can't co-exist. For some teams the closer model is great. I just wish teams would get a little more creative.

Which one "before the 8th" or "before the 9th".

I clearly meant "before the 9th" because I advocate using the ace in the 7th and 8th innings instead of the 9th. Had I remained with "before the 8th" in my original post, it woudl have meant that I advocated bringing the relief ace "before the 8th" (Which meant NOT the 8th, which is clearly an inning I want my ace to be pitching), and that's not what I was trying to say.

I actually do like the TLR model of bullpen usage. I mean, if you need say 8 outs, and you have 4 good pitchers in your pen and enough lefties, there's no reason to get as much platoon advantage as possible late in games where the leverage is high. The added advantage to this is that the pitchers can be specifically "briefed" on their assignments before hand (perhaps making them moer effective) and they'd be ready to go next day (since they probably only therw 15 pitches or so to 2 batters)

I see the relief ace model most effective in cases where you only have a relief ace, another pitcher that doesn't suck, and a bunch of krappy ones in which you want to avoid high lev situations as much as possible.

If I had a team that was out of the race early, I'd be tempted to put the relief pitchers on a rotation system; just like the starters. Interesting discussion.

I actually do like the TLR model of bullpen usage. I mean, if you need say 8 outs, and you have 4 good pitchers in your pen and enough lefties, there's no reason to get as much platoon advantage as possible late in games where the leverage is high.

That's fine, it's only that a few teams actually have four good pitchers in their bullpen. Most teams have one or two, only they behave as if they have four, giving high-leverage work late in games to mediocre pitchers who happen to throw with their left hand.

But if you've got the 'pen (as LaRussa did with those A's teams), then sure, it's a viable strategy.

And, of course, the other side of that argument is that the "closer" model has forced GMs to go out looking for, and spending money on, three or four quality relievers, when a field manager with some imagination can probably be almost as successful with one or two.

And that's the real problem here - most teams have limited financial resources, and the "closer" model forces them to spend those resources in places it probably isn't necessary to.

Most teams have one or two, only they behave as if they have four, giving high-leverage work late in games to mediocre pitchers who happen to throw with their left hand.

Yes.

And that's the real problem here - most teams have limited financial resources, and the "closer" model forces them to spend those resources in places it probably isn't necessary to.

Yes. And the investment of money leads to a commitment to the usage pattern, and the cycle feeds itself.

The issue with the Closer model isn't that it doesn't make sense; it makes a lot of sense -- under certain circumstances. My view is similar to that expressed by RP above: the problem is in the uncreative orthodoxy.

There are lots of ways bullpens can be deployed effectively, yet increasingly, a rigid orthodoxy around just one pattern pervades.

During the Mets/Braves game the other day the idiots on TBS (Chip Caray and his partner. Ron Gant was there too, IIRC) were talking about how Looper couldn't be a closer because he didn't have the closer makeup. Apparently they made that assertion because he blew the opening day game. I guess they never saw him pitch before this season.

They also said they saw countless pitchers who could pitch in the 8th but not the 9th, of course they didn't name anyone.

"I don't have to use Studes work, because I can use Tango's work to show an actual increase in LI over all other previously deployed models."

The closer model might be an extremely effective one. It might be more effective than anything that anyone has employed before, but that doesn't mean it can't be improved upon. Maybe you are better off not using your closer in a three-run game. Maybe it depends on who's coming to the plate. I seriously doubt that the model of using your best reliever in the ninth inning and only in the ninth innning with a one, two, or three run lead, regardless of all other circumstances, can't improved upon.

For some teams the closer model is great. I just wish teams would get a little more creative.

Good point, RP. If you only have one excellent relief pitcher, an argument can be made that the closer makes more sense than the fireman model.

If you only have one excellent relief pitcher, an argument can be made that the closer makes more sense than the fireman model.

Maybe. It's also a function of what kind of an innings workload that excellent reliever can be expected to handle; not all relievers are equally durable. The point, again, is that jamming all bullpens into one fixed usage pattern is the problem, almost regardless of what that pattern might be.

It's back to the old Diamond Mind thing for me: really, play a load of games and you tell me when the best time to bring your best reliever is.. You have bases loaded, one out in the 7th? Do you really send in anything but your relief ace?

Does Diamond Mind require you to warm pitchers up? I know Strat-O-Matic doesn't. If it doesn't require it, I think that is a major deviation from reality -- probably even more so than the psychological "players need to know their roles" stuff. Considering that you do in real life have to warm the guy up first... in this situation, are you going to start warming up the fireman in the beginning of the 7th? If you are, then what happens if you don't need him? Are you gonna sit him down when his team bats... and then start warming him up again in the 8th?

Point being, you can't automatically have the fireman ready for the game's highest-leverage situation, because 1) you need to identify that situation 2-4 batters in advance, which cannot be done perfectly, and 2) you can only warm him up a certain number of times, for a certain length of time.

Of course, that's not to say that "let the save rule manage your team" is the winning strategy. If A-Rod, Matsui and Sheffield are due up, maybe you do warm the fireman up at the beginning of the 7th, and if you end up having to sit him later because he threw too much in the bullpen, then oh well -- it's possible that you still made the percentage move. But, just to say that the probabilities you're crunching have to be looking a few batters ahead due to the need to warm up, and that that is not very easy.

I, for one, love the Angels model, at least up until this year:

1. Get three or four kickass relievers.
2. Make the old guy, who's the least kickass, the closer.
3. Use the other guys to get out of jams and get the game to the closer.

Now it's different, but I really liked the old way. But the job is made simpler when you actually have four kickass relievers.

During the Mets/Braves game the other day the idiots on TBS (Chip Caray and his partner. Ron Gant was there too, IIRC) were talking about how Looper couldn't be a closer because he didn't have the closer makeup. Apparently they made that assertion because he blew the opening day game. I guess they never saw him pitch before this season.

Well, to be fair (not that I desire to be fair to Don Sutton), people were saying that about Looper this time last year. Looper did a fine job last season, but that doesn't shut the door on the concerns of him being able to be a closer.

I know Strat-O-Matic doesn't.

In my house it always did! My brother and I had a rule that you couldn't bring a reliever into the game until you had indicated, at least 2 batters earlier, that you were warming him up. You couldn't warm up more than 2 relievers at a time.

Looper did a fine job last season, but that doesn't shut the door on the concerns of him being able to be a closer.

Looper has 75 saves in his career; he's hardly a neophyte 9th inning pitcher. I think a far more reasonable concern with Looper isn't that he doesn't have "closer makeup," it's that he really isn't all that great a pitcher.

Does Diamond Mind require you to warm pitchers up? I know Strat-O-Matic doesn't.

It's an optional rule in DMB, but only for human-managed teams. Computer-controlled teams don't have to warm up pitchers, regardless.

<i. I should also add that I would like to see the two schools of thought in regard to relief pitching duke it out on the field. Eventually, one of them would be left standing: if it's not the fireman model, then I'll be the first to say that I was wrong. </i>

That happened...closer won.

It might be more effective than anything that anyone has employed before, but that doesn't mean it can't be improved upon.

To my knowledge, no one here or anywhere, has ever even suggested that it can't be improved upon. Plenty of people have suggested that the ACE RELIEVER model is superior, and that is just wrong. So what is it you saberists call this type of rebuttal?

I clearly meant "before the 9th" because I advocate using the ace in the 7th and 8th innings instead of the 9th. Had I remained with "before the 8th" in my original post, it woudl have meant that I advocated bringing the relief ace "before the 8th" (Which meant NOT the 8th, which is clearly an inning I want my ace to be pitching), and that's not what I was trying to say.


So then go back and read my post. So what is it you saberists call this type of rebuttal?
Its just amazing that you don't see that word when someone is challenging the Primer orthodoxy.

The issue with the Closer model isn't that it doesn't make sense; it makes a lot of sense -- under certain circumstances. My view is similar to that expressed by RP above: the problem is in the uncreative orthodoxy.


BS, you have unambigously lobbied for and made futile and juvenile attempts to prove the ACE RELIEVER model is superior. Now you want to rewrite history. If nobody calls you on it, you will swear you never said X, and you didn't write that cute little pitch count estimator article.

The point, again, is that jamming all bullpens into one fixed usage pattern is the problem, almost regardless of what that pattern might be.

BS, see supra. More important, it is a good idea to start with a default workable situation and adapt based on the situation. Its never a good idea to risk health, performance and well being because some kook thinks DL trips are a measure of injury.

I see the relief ace model most effective in cases where you only have a relief ace, another pitcher that doesn't suck, and a bunch of krappy ones in which you want to avoid high lev situations as much as possible.


I see that roster as indicative of organizational failure.


And, of course, the other side of that argument is that the "closer" model has forced GMs to go out looking for, and spending money on, three or four quality relievers, when a field manager with some imagination can probably be almost as successful with one or two.


Like the Oakland A's trading away valauble assets for bullpen depth. And how does that square with a field manager who is just suppose to stand on the steps and project authority?

Looper has 75 saves in his career; he's hardly a neophyte 9th inning pitcher. I think a far more reasonable concern with Looper isn't that he doesn't have "closer makeup," it's that he really isn't all that great a pitcher.

I don't disagree with that. Perhaps I should have made it more clear, I was speaking in terms of media criticism, how one year as the regular closer from start to finish isn't going to stop the criticism that he might not be up for the job.

But the job is made simpler when you actually have four kickass relievers.

If you actually have four kickass relievers, you can use them pretty much any way you want to and still be successful. It's what to do if you only have one kickass reliever that represents the challenge...

Good points by a lot of people here.

One thing that tends to get forgotten in these discussions is RCL's point about warming up your ace reliever.

Sure, you'd like him to pitch with the bases loaded and one out in the 7th, but when do you start warming him up:

Groundout
Single
Double
Walk

Surely we can't start warming up Gagne with one out and a runner on first in the 7th. And if you start him after the double, you're likely to find that the other team has already scored, or the current pitcher got a double-play and you warmed him up for nothing.

Either way, you've put more pitches on his arm than necessary without really accomplishing much.

That's not to say that there is no value in being willing to use your ace reliever outside of a strict closer role. There is. I don't know that we need to go back to the fireman days, but structuring your bullpen to at least recognize the opposing lineup (bring your guy in to face the heart of the lineup in the 8th, for example) seems appropriate.

I also agree with LAWoBH about what might be close to an ideal bullpen structure. The Mariners in 2001 were a great example of this, too. Nelson and Rhodes were lights out all year and they got most of the really tough situations. Sasaki was certainly good, and clearly had no "closer hangups" and he finished off the game.

During the Mets/Braves game the other day the idiots on TBS (Chip Caray and his partner. Ron Gant was there too, IIRC) were talking about how Looper couldn't be a closer because he didn't have the closer makeup. Apparently they made that assertion because he blew the opening day game. I guess they never saw him pitch before this season.

Chip Caray specifically mentioned that Looper lost the closer role and Florida. I imagine that he also saw Looper being a baseball broadcater.

They also said they saw countless pitchers who could pitch in the 8th but not the 9th, of course they didn't name anyone.


Ok, I'll name some: Steve Karsay, Mike Remlinger etc. Of course these guys would have to blow 80 saves before you thought that you had enough of a sample size.

Who's the best reliever who (true or not) couldn't pitch in the ninth?

Like the Oakland A's trading away valauble assets for bullpen depth. And how does that square with a field manager who is just suppose to stand on the steps and project authority?

Bullpen depth is very important, particularly for a team like the A's with a lot of young starters, who can't be expected to go 7 innings every game. I think the point is that teams are a little too obsessed with LOOGYs, closers, and other relievers who average an inning per appearance at best.

The type of bullpen you love to see is one with guys like Duchscherer, Calero, and Cruz who can easily go 2 innings or more if necessary.

I think one conclusion to draw from all this is that bullpen needs of teams vary a lot more than we might expect. Not just from team to team, but also depending on situations for a single team.

Ideally, you'd love to have bullpen depth. And, if you have it, you should probably employ a closer model. But if a couple of your guys get hurt or prove to be ineffective, you should be willing to evolve your bullpen. Teams change their approach in the postseason; they should be willing to do it during the season, too. If the only reliable reliever you've got is Lidge, maybe you should use him more as an ace reliever than a closer. But, you should also be looking for good relievers to add to your roster. The two are not exclusive.

Looper has 75 saves in his career; he's hardly a neophyte 9th inning pitcher. I think a far more reasonable concern with Looper isn't that he doesn't have "closer makeup," it's that he really isn't all that great a pitcher.

I agree. He's a good pitcher, not a great one. It just seems stupid to criticize him for not having closer makeup when he's done a good job in that role. It's not as if his ERA jumped 2 runs.

If you actually have four kickass relievers, you can use them pretty much any way you want to and still be successful.

What teams have 4 kickass relievers?
The Angels last year (ie. K-Rod in high lev as "relief ace", Percival as "closer")...

Cardinals? (Izzy, King, Kline, Tavarez, Calero, Haren) all competent pitchers being used TLR style to maximize platoon advantage (I have no problems with this last year, since the personell was very good.) The relief ace model with this team would actually work too (bring out Izzy for the 7th and 8th, and still have competent pitcher(s) for the 9th, but King and to a certain extent Tavarez isn't designed to be throwing more than 1 inning.

The Pads had 3 kickass relievers right? Otsuka, Linebrink, and Hoffman right?

Who's the best reliever who (true or not) couldn't pitch in the ninth?


BTF Paging Mikael, BTF Paging Mikael, BHK man-love needed aisle 4

Chip Caray specifically mentioned that Looper lost the closer role and Florida.

He was moved out of that role because they got a better pitcher.

Do you think they would have talked about Looper's makeup if he had closed out the opening day game?

The problem with so many of these use your best RP's when the game is on the line arguements is they don't ever account for timming. When do you know the game is on the line? At the time it occurs. Say Derek Lowe is pitching in a 1 run game in the 7th. X gets on base, so you want your best RP ready so you start getting Gagne up. He gets the next better to fly out, and then he gets the next batter to GIDP... well you just got your best reliever up when you didn't need him. Now your likely to be forced to bring him in during the 8th, and he likely won't be able to finnish the game, unless your going to make the arguement that pitchers should be routinly going longer each game. Which may or may not be true, but I certainly haven't seen any evidence for it.

I'm perfectly willing to accept the notion that certain pitchers have a mental block about the ninth inning. Pitchers are human beings after all, not machines, and all the meaning and pressure that's been placed into the ninth inning has got to affect some of them.

The problem is, which ones? As Backlasher notes, what's an appropriate sample size for determining that closing games is beyond certain pitchers? Is closing games something that LaTroy Hawkins sucks at, or did he happen to hit a rough patch just when they were trying him out at the job?

It's tough to say, because pitchers who blow a couple of saves early on tend not to get second chances at the job.

As for the guys who can't seem to close. I don't doubt that there are some good (maybe even great relievers) who have a lot of trouble with the unique pressure of closing a game.

However, I say that with the following caveats:
1. Why is there a unique pressure to closing the game? I think most of it is due to the current obsession with closing. If saves didn't exist and the difference between Reliever A (no saves) and Reliever B (38 saves) with identical stats wasn't $5 million, you'd probably see less of this.
2. Even if some guys have trouble with the pressure, MOST don't. The list of relievers turned into closers is very very long. Most guys who are good relievers also turn out to be good closers. It's an exception for guys to bomb out who are otherwise very good.
3. Most of the "failures" are just guys who were pretty good candidates to start losing it anyways.
4. It's probably a good strategy to create closers. You'll have a failure every once in a while, but the money saved by not paying for a closer name is likely to help you more than a little extra security.

And finally, if you can get a top-10 reliever (Gagne, Rivera, Foulke, Hoffman), a guy who's demonstrated a clear ability over several seasons to dominate, you should probably do it. It's the Loopers and Mesas of the world that it's probably better to stay away from.

But if a couple of your guys get hurt or prove to be ineffective, you should be willing to evolve your bullpen.

Agreed.

Teams change their approach in the postseason; they should be willing to do it during the season, too.

How you change is very important. Rest patterns are much different in the postseason. Wins and losses have more value in the postseason than most wins and losses in the regular season.

If the only reliable reliever you've got is Lidge, maybe you should use him more as an ace reliever than a closer.

Unless you wear him out in June, then you are up sh1t creek. Its not just the closer, setup evolution has taken place beneath the closer. Anybody watch the Braves; anybody see Tom Gordon last year.

Of course Treder thinks you can solve all that by having the teenage kids throw 10% more.

unless your going to make the arguement that pitchers should be routinly going longer each game. Which may or may not be true, but I certainly haven't seen any evidence for it.

I think Treder has made many arguments for just this proposition.

Unless you wear him out in June, then you are up sh1t creek. Its not just the closer, setup evolution has taken place beneath the closer. Anybody watch the Braves; anybody see Tom Gordon last year.

And this is the part where the bullpen coach/pitching coach/training staff tells the manager that so and so is getting tired, and that they should try to get so and so more rest, not to mention having the pitching coach ensure that proper mechanics are in place.

I have a weird pet theory, and I don't know if anyone has ever looked at this. All of my evidence is anecdotal.

It seems to me that when relievers enter the game in a tough situation, and get out of it, they often struggle when they come out to pitch the next inning. This just happened to Brendan Donnelly last night, but I've witnessed this several times, and I'm sure you have, too.

If this is a real effect, that might hamper the James Ace Reliever Model. Also, it might be that some pitchers have a tendency to do this, in which case they're likely not your Ace Reliever, anyway, but who knows.

If there's any validity to this purely anecdotal theory, one might wonder if part of the cause is that pitchers are no longer used to long relief outings, and thus tire easily. Most closer can probably go two innings if they start the first one; but coming into a game with runners on base is a more stressful situation, so perhaps they get worn out quicker, so getting six outs is tougher in such a situation.

Has anyone ever seen a study on this? Or am I just crazy?

Unless you wear him out in June, then you are up sh1t creek. Its not just the closer, setup evolution has taken place beneath the closer. Anybody watch the Braves; anybody see Tom Gordon last year.

I don't disagree with this, necessarily. That's why you should also have the goal of bolstering your bullpen. Use Lidge an extra 10-15 innings now, win a couple games, keep yourself in contention, and work on fixing the problem before you burn him up.

But you could also help yourself out by not using him to save 3 run games.

He's got 4 appearances so far this year. One was just such a 3 run 9th inning save (to be fair, against the top of the Cards' order). And he didn't make an appearance in last night's 0-0 nailbiter.

What exactly is the "relief ace" model, and why does it wear pitchers out more quickly? Does it involve pitching more innings?

oooh, another battle between Backlasher and Ironbrain Chris Wok...this should be close (snicker)

I think Treder has made many arguments for just this proposition.

My proposition is that the modern bullpen usage pattern, which isn't just the Closer but is also multiple LOOGYs, and all relievers working ever-shorter stints, has costs as well as benefits, and the costs aren't nearly as regularly examined as the benefits.

Among the costs are:

- Fewer innings worked by the best relievers, regardless of LI, by definition requires more innings worked by inferior relievers

and

- Shorter stints by all relievers, in combination with ever-fewer complete games, by definition requires more roster spots devoted to relievers, which by definition removes roster spots from other positions

I think reliever use would be better optimized if pitchers were used in longer outings less frequently. So, instead of 75 games and 80 innings, maybe they pitch in 50 games but throw 100 innings.

I base this on my impression that pitchers often have worse stuff on their second consecutive day of working, and also on the generally understood physiology of the human body requiring time to repair itself after a physically taxing activity. Is there any validity in this idea?

What exactly is the "relief ace" model, and why does it wear pitchers out more quickly? Does it involve pitching more innings?

There isn't "a" relief ace model. But there are many varieties of different usages of relief aces throughout recent history, that provide lots of empirical data on alternatives to the Closer pattern.

In the NBJHBA, James presents a study of the prevailing models used throughout history (the Clint Brown, Elroy Face, Hoyt Wilhelm, Bruce Sutter, and Robb Nen patterns), and the results of a sim he ran testing various models. His conclusion is that, generally speaking, the most efficient use of the relief ace is in something along the lines of the Face or Wilhem patterns, in which the ace reliever typically gets 60-65 appearances in a season, and 90-110 innings; used in lots of Save situations but obviously not exclusively in 9th inning situations.

Here's a snippet of some of the data I'll be presenting in a THT article about Closers week after next.

Average innings/game of the reliever recording the most Saves on each team:

2004: 1.08
1999: 1.09
1994: 1.09
1989: 1.24
1984: 1.49

Steve -- do you address the fact that the saves were so much more spread out on most teams 20 years ago? i.e., Why were the save situations so spread out back then? What caused teams to start giving almost all of them to the "closer"?

think Treder has made many arguments for just this proposition.

Treder has posted a bunch of garbage then left the table, I haven't seen him advance a single argument.

To wit:

- Fewer innings worked by the best relievers, regardless of LI, by definition requires more innings worked by inferior relievers

and

- Shorter stints by all relievers, in combination with ever-fewer complete games, by definition requires more roster spots devoted to relievers, which by definition removes roster spots from other positions


Which concludes what? This is easily the thirtieth time he has posted this malarky.

In point 1, so what? The implication that Steve wants you to read into this is that an inning by Pedro Martinez is always better than an inning by Mike Timlin. You all know this isn't true.

Evidence has been presented that in closed environment, distribution of innings resulted in better aggregate bullpen performance than concentration of innings. Steve usually runs away at this point.

In point 2, so what? The implication that Steve wants you to read into this statement is that you could use that extra roster spot for Erubiel Durazo. That has never happened on no evidence or analytical construct suggests it could happen. The 25th man is the 25th man and in an aggregate analysis a run saved is a run earned. In a more germane analysis, if that 25th man can improve others efficiency, or prevent injury, he is more valuable. Pitchers do have influence on other pitchers because the availability equation for pitchers is different than for position players. Moreover, the injury potential is higher for pitchers. It pure Wok logic.

To wit:


And this is the part where the bullpen coach/pitching coach/training staff tells the manager that so and so is getting tired, and that they should try to get so and so more rest, not to mention having the pitching coach ensure that proper mechanics are in place.

Just please, stop and think for two minutes before you rush to post. You want to use a micro strategy to develop a macro plan. (and later you want to use your macro analysis to decide who to use in a micro context). That is so beyond dumb. You want someone to get hurt or face risk of injury before they get rest. You want to make game time decisions about how to pitch when you cannot even properly gauge the potential for incurring LI. That is just so juvenile, I can't even begin to comment on it.

Here's a snippet of some of the data I'll be presenting in a THT article about Closers week after next.

Average innings/game of the reliever recording the most Saves on each team:

2004: 1.08
1999: 1.09
1994: 1.09
1989: 1.24
1984: 1.49


Which as usual is reflective of nothing. I'm sure we will get a good old days argument at the end of the meaningless data. If you have data and an argument post it. I long ago grew weary of your post-and-run tactics, but I am really surprised BTF is letting you continually tease your articles for nothing. This isn't the Leno show.

"There isn't "a" relief ace model. But there are many varieties of different usages of relief aces throughout recent history, that provide lots of empirical data on alternatives to the Closer pattern."

Thanks. This is what I thought, which is why I don't understand what anyone is arguing about. Clearly there is *something* better than bringing in your best reliever in save situations.

If an omniscient (but not clairvoyant) being were the manager of a baseball team, taking into account all possible information (how different hitters are expected to perform against different pitchers, how different patterns affect injury and performance expectation, etc.); the result would alomst certainly be a situation where the best reliever is not *exclusively* used in the ninth inning or in save situations, a type of "relief ace" approach, I guess. The closer model isn't the worst, but it certainly isn't optimal. So at least theoretically, the relief ace is better.

I base this on my impression that pitchers often have worse stuff on their second consecutive day of working, and also on the generally understood physiology of the human body requiring time to repair itself after a physically taxing activity. Is there any validity in this idea?


ASMI would agree with your hypothesis wrt health, the problem is that this model would obstruct availability.

I have a weird pet theory, and I don't know if anyone has ever looked at this. All of my evidence is anecdotal.

It seems to me that when relievers enter the game in a tough situation, and get out of it, they often struggle when they come out to pitch the next inning. This just happened to Brendan Donnelly last night, but I've witnessed this several times, and I'm sure you have, too.

If this is a real effect, that might hamper the James Ace Reliever Model. Also, it might be that some pitchers have a tendency to do this, in which case they're likely not your Ace Reliever, anyway, but who knows.

If there's any validity to this purely anecdotal theory, one might wonder if part of the cause is that pitchers are no longer used to long relief outings, and thus tire easily. Most closer can probably go two innings if they start the first one; but coming into a game with runners on base is a more stressful situation, so perhaps they get worn out quicker, so getting six outs is tougher in such a situation.

Has anyone ever seen a study on this? Or am I just crazy?


You aren't crazy. Its the type of stress that causes injury or the level of resiliency. If you have to rare back, you likely increase the risk of an injury and reduce resiliency. I cannot say that I have seen a study on this effect.

Of course, you may have saberists throw rotten fruit at you if you propose it. I've seen Davis (and I'm waiting for Wok to say it this thread) say that modern pitchers go all out all the time. This of course is counter to the empiracle evidence that starters usually gain MPH on their fastball when converted to relief.

Nevermind, I do know what people are arguing about. For the record I have no idea idea how the "Wilhelm" pattern compares to the closer. Please don't lash at me.

Clearly there is *something* better than bringing in your best reliever in save situations.


No there is nothing clearly better. There are possible alternatives that could be better.

the result would alomst certainly be a situation where the best reliever is not *exclusively* used in the ninth inning or in save situations, a type of "relief ace" approach, I guess. The closer model isn't the worst, but it certainly isn't optimal. So at least theoretically, the relief ace is better.


No, the ACE RELIEVER model is theoretically and actually worse at least in regards to LI. It is indeterminate but analytically worse on all other elements.

You are letting Steve play both ends of a semantic game with you. ACE RELIEVER clearly and unambiguously refers to James abandoned hypothesis for which Steve carries the torch. Closer model does not specifically refer to the LaRussa model. LaRussa was the next evolutionary step, but the model has been improved.

Analytically, a reasonable hypothesis is to use any specific reliever in the appropriate situation for where his success measure is optimal and wrt your closer to use him in the ninth inning when your team is tied or up by three runs or less provided he has not pitched more than three consecutive days in a row. If you achieve a substantial success measure higher by using the closer in the 8th inning, where the LI is a certain level and the game has significant importance, then you can bring your closer in on rare occasions in the 8th inning provided he has not pitched the last two days or had a similar situation in the last four days.

That is not ACE RELIEVER, that is the plan that is a close reverse engineering of the most innovative and knowledgable pitching minds in today's game.

It is one that has evolved and preserved injury mitigation, and its not some kooky thing that was put together on the back of an envelope from a work of fiction, an imprecise simulator, and the delusional mind of a would be sabermetrician.

Analytically, a reasonable hypothesis is to use any specific reliever in the appropriate situation for where his success measure is optimal and wrt your closer to use him in the ninth inning when your team is tied or up by three runs or less provided he has not pitched more than three consecutive days in a row.

So if he's only pitched three consecutive days, you go ahead and bring your closer in on that fourth day?

It should be noted that, as a knuckleballer, Hoyt Wilhelm was almost unique in his ability to pitch multiple innings in relief on little or no rest.

He's likely an outlier, but guys like Goose Gossage and Bruce Sutter showed that it's not impossible for other types of relievers.

When we refer to the "Ace reliever" model, we're not talking about some wacky experiment that hasn't been tried. We're talking about a model that was used with much success by many teams in the past. I have yet to see convincing evidence as to why this wouldn't work today.

That is not ACE RELIEVER, that is the plan that is a close reverse engineering of the most innovative and knowledgable pitching minds in today's game.

Just curious - to whom are you referring here?

So if he's only pitched three consecutive days, you go ahead and bring your closer in on that fourth day?

Read it again, bootsy.

Just curious - to whom are you referring here?

Cox and Mazzone

I'm not sure why it matters that the innings are spread around the bullpen. The issue is, of course, leverage, since you're not going to ask Keith Foulke to pitch the 9th inning in an 8-1 game (um, too often). Unless you know of a way to find the next Wilhelm or Gossage without burning out a bunch of arms along the way, you're probably better off sticking with something close to what we have now, albeit with fewer outings for your ace (closer, whatever) with a 3-run lead and a few more when tied or down by a run (depending on recent usage).

As to the question of whether an 11 or 12 man pitching staff is overkill, I think the question is whether a team is better served by having those final 2-3 roster spots used by chasing the platoon advantage on the mound or at the bat. As a fan my preference is to spend it on hitters -- pinch hitting substitutions are a lot less time consuming than mid-inning pitching changes, and I do miss platooning. As a strategic matter, I'm not sure I've seen the question studied, though I may well have missed it.

Steve -- do you address the fact that the saves were so much more spread out on most teams 20 years ago?

Absolutely. Here is the average percentage of his team's total saves that were recorded by each team's top save-producer in a sampling of seasons:

2004 75.2%
1999 73.2%
1994 64.5%
1989 66.6%
1984 57.4%

Incidentally: total Saves per game for those seasons:

2004 25.3%
1999 25.1%
1994 24.3%
1989 25.4%
1984 23.6%

The historic peak of overall save production was in the 1990-1993 period, a very interesting fact that I'll be exploring in my article.

Why were the save situations so spread out back then?

Because teams were less rigid, more flexible, in the manner in which they used relievers, and the Save hadn't yet acquired its mystic reverence. :-)

What caused teams to start giving almost all of them to the "closer"?

That's a complicated issue to which there isn't a short answer. I'll explore some of the possibilities in the article.