Joe Posnanski with more data showing that closers simply don't matter

[Kc61]

Somebody needs to explain what the "anti-closer" group (for lack of a better term) is actually advocating.

If they are advocating not using one pitcher to pitch the ninth, not limiting him to one inning, in other words, being more flexible in use of late inning relievers, I can buy that although I think it is a minor quibble. Managers like set roles for relievers and while flexibility is great, it doesn't warrant all this debate.

If the group is advocating that lesser pitchers finish close games, that's a more substantive dispute. It's hard for me to believe that such an astute group of fans want lesser pitchers throwing the late innings of very close games.

If it's a question of salary, well, at any position you hope to get a very good player cheaply. It may be that young relievers with great stuff should be relied on as late inning pitchers -- guys with dynamite stuff certainly can be effective pitching one inning. However, experience is always important so there's a balancing in choosing late inning pitchers as with any position.

So I guess I don't know what is being advocated here.

Otherwise put -- Reds lead by one in a key pennant drive game, eighth inning, starter is gassed, do you want Marshall and Chapman getting ready? Marshall and Broxton? Arredondo and Simon? Parra and Ondrusek? Lecure and Hoover? It doesn't matter?

Do people really believe it doesn't matter?

[dougdirt]

I am advocating that the 9th inning isn't special. That just about any old "quality" reliever is going to save as many games as any "special" reliever will over the long haul if given the chance. Obviously bringing in Chad Reineke is going to make a difference. But any reliever who would be able to throw up a 3.75 or lower ERA in Cincinnati is going to convert a very high rate of saves.

I also am saying that saving your best reliever for the sole purpose of starting the 9th inning when you are leading by 3 or less is stupid.

[jojo]

Somebody needs to explain what the "anti-closer" group (for lack of a better term) is actually advocating.

If they are advocating not using one pitcher to pitch the ninth, not limiting him to one inning, in other words, being more flexible in use of late inning relievers, I can buy that although I think it is a minor quibble. Managers like set roles for relievers and while flexibility is great, it doesn't warrant all this debate.

If the group is advocating that lesser pitchers finish close games, that's a more substantive dispute. It's hard for me to believe that such an astute group of fans want lesser pitchers throwing the late innings of very close games.

If it's a question of salary, well, at any position you hope to get a very good player cheaply. It may be that young relievers with great stuff should be relied on as late inning pitchers -- guys with dynamite stuff certainly can be effective pitching one inning. However, experience is always important so there's a balancing in choosing late inning pitchers as with any position.

So I guess I don't know what is being advocated here.

Otherwise put -- Reds lead by one in a key pennant drive game, eighth inning, starter is gassed, do you want Marshall and Chapman getting ready? Marshall and Broxton? Arredondo and Simon? Parra and Ondrusek? Lecure and Hoover? It doesn't matter?

Do people really believe it doesn't matter?

Basically I think the argument is that the 9 th inning may not always be the best inning to use your best reliever.

[Kc61]

I am advocating that the 9th inning isn't special. That just about any old "quality" reliever is going to save as many games as any "special" reliever will over the long haul if given the chance. Obviously bringing in Chad Reineke is going to make a difference. But any reliever who would be able to throw up a 3.75 or lower ERA in Cincinnati is going to convert a very high rate of saves.

I also am saying that saving your best reliever for the sole purpose of starting the 9th inning when you are leading by 3 or less is stupid.

As to both points, I don't think as currently phrased they are very debatable. As I said in my last post, I don't really think there is a huge gulf here when one thinks about the issue.

On the first point, you say that most quality relievers can close. I agree. It doesn't have to be the guy with the best stuff. I look at David Weathers as an example. I thought he was extremely effective as a closer although he didn't have Chapman type stuff.

Of course, as a matter of team management, somebody has to decide which relievers should be trusted with the ninth inning. Doesn't have to be the best guy, but a reliable guy who they feel can thrive in that spot. It's just good team management not to go random -- to assign people (in general terms) with repeatable roles.

As to your second point, I agree as well. Of course it depends, if Rollins, Utley and Howard are coming up in the ninth, three runs up, I would use my closer.

But generally I would agree that there should be more flexibility in the ninth innings. I've always liked having a backup closer, for example, to avoid overuse. And adherence to the closer up three runs with the bottom of the order up, I agree that isn't necessary.

The main thing, on which I think there is (or should be agreement) is that we want a quality reliever finishing close games. Frankly, the rest of it is interesting debate but not so important IMO.

[RedsManRick]

Some data from 2012:

For qualified relievers from 2012 (n=136), averages by bands of gmLI (A pitcher’s average leverage index when he enters the game).

Leverage Index (LI) is merely an attempt to quantify this pressure so we can determine if a player has been used primarily in high-leverage or low-leverage situations. An average (or neutral) LI is 1. High leverage is 1.5 and above, and low leverage is below 1. 10% of all real game situations have a LI greater than 2, while 60% have a LI less than 1.

gmLI range	#	SV+BS	FIP	ERA	 gmLI
1.9 and Up	0	  --	  --	  --	  --
1.8 to 1.9	4	40.8	2.87	2.81	1.85
1.7 to 1.8	5	36.2	2.89	2.76	1.74
1.6 to 1.7	11	24.4	3.32	3.06	1.63
1.5 to 1.6	12	27.6	2.99	2.86	1.55
1.4 to 1.5	10	11.4	3.43	3.16	1.44
1.3 to 1.4	21	 9.3	3.27	2.91	1.34
1.2 to 1.3	10	 7.4	3.40	3.21	1.26
1.1 to 1.2	16	 3.4	3.73	3.30	1.14
1.0 to 1.1	16	 3.6	3.73	3.13	1.04
0.9 to 1.0	9	 2.2	3.91	3.53	0.96
0.8 to 0.9	9	 1.2	3.82	3.65	0.84
0.7 to 0.8	6	 1.0	4.48	4.45	0.76
0.6 to 0.7	4	 0.8	3.50	4.34	0.67
0.5 to 0.6	2	 0.0	4.71	3.15	0.55
0.4 to 0.5	0	 --	  --	  --	  --
0.3 to 0.4	1	 1.0	5.00	6.05	0.38
Under 0.3	0	 --	  --	  --	  --

From what I can see, save opportunities actually do tend to correlate fairly strongly with leverage index. While not all high leverage situations are save chances and not all save chances are high leverage, there's a reasonable degree of overlap.

Further, managers are, on average, pretty darn good at using their most effective relievers in those high leverage situations.

One potentially HUGE caveat: Leverage is calculated based on win likelihood. The problem with win likelihood is that it treats late game events as more influential than early game events, despite the fact that a run scored in the 9th inning counts the exact same towards the eventual won or loss as a run scored in the 7th.

The flip side of the argument, of course, is that if you slam the door in the 7th to protect a 1 run lead and then your offense goes out and scores 5 in the 8th, that door slamming ultimately wasn't as a big of a deal as we thought it was going to be.

Setting aside the data above, I think the biggest flaws in manager reliever use are:

• Not giving sufficient weight to situations in the middle innings when men are on base (saving the closer for a save chance that won't exist if you don't escape the current jam)
• Giving too much weight to low & middle leverage save chances, particularly when the closer has been worked hard and could use a break (any pitcher on a major league staff can protect a 3-run lead 90% of the time)
• Not giving sufficient weight to the spot in the lineup. Bringing in a Logan Ondrusek type to face the heart of the lineup in the 7th inning because Sean Marshall is the 8th inning guy is ridiculous, as is bringing in your closer to protect a 2 run lead in the 9th against the 6-7-8 hitters.
• Over-reacting to "bad luck" outcomes and pulling a generally effective reliever, especially for a less effective reliever (if I see one more blooper -> bad call walk -> nubber -> pulled reliever.....)
• Under-reacting to actual reliever ineffectiveness, as indicated by good real-time scouting (guy comes in but just doesn't have his stuff or command)
• Thinking that platoon splits should trump ability -- Sean Marshall is better against RHP than any Reds reliever not named Aroldis Chapman.
• Thinking that handedness implies platoon splits -- Arredondo, Simon and Ondrusek, all righties, have been significantly more effective against LH hitters than RH over their careers.

[RedsManRick]

management, somebody has to decide which relievers should be trusted with the ninth inning. Doesn't have to be the best guy, but a reliable guy who they feel can thrive in that spot. It's just good team management not to go random -- to assign people (in general terms) with repeatable roles.

I agree with this. But why do bullpen roles have to be defined by inning? Why not by leverage? Is it that much ask relievers who are available in a given day to "keep your head in the game and be mentally ready in the 7th thorough 9th" innings? Just how predictable do things have to be?

I know I'm not typically old school, but come on. Is it really that hard to:

1. Have your mop up guy(s), your elite/high leverage guy(s) and the rest and have everybody know whose who
2. Explain to the guys how you look at leverage. Once we get to the 7th:
- First cut is the score
- Second cut is the base/out situation
- Third cut is spot in the opponent's lineup
- Fourth cut is lateness in the game

Sure, you could layer in some nuance if you have a guy who has a giant platoon split or who owns certain types of hitters, but there's no way that players can't handle an approach that doesn't start with an arbitrary stat and works down from there.

Show me a manager who brings his closer in at the top of the 8th to face the 2-3-4 batters in a 2 run game and then uses his middle reliever to get 5-6-7 and get the "save" and I'll be happy.

[Kc61]

There are several problems with defining roles by leverage, to quote Rick's fine post.

1. You never know when the highest leverage situation arises. Sometimes it may appear clear, but using up Chapman in a sixth inning leverage situation may leave you stranded in the eighth and ninth.

2. If the same pitchers throw every leveraged situation, then they will die of overuse and perhaps over stress. There has to be an allocation to allow the workload and stress load be divided somewhat.

3. Using relievers by leverage could lead to constant warming up of top guys when you believe a high leverage situation is on the horizon. There's no surefire advance warning often that high leverage is about to happen and a manager would have to warm up his top guys whenever he thinks it MAY happen.

4. Some pitchers just like to know when they are likely to be used. Chapman probably doesn't want the irregularity of use by leverage. He wants to get ready in the eighth and pitch in the ninth.

I think the use of top relievers in the eighth and ninth is simply a managerial device to give pitchers a regular, repeatable role. It's just an allocation system.

And in my view, in the playoffs and other really key games, most managers DO change these roles. They do use starters as relievers, and do use their top relievers more in high leverage situations.

But they choose to go with more regular roles in allocating work over 162. I get that, even though in a particular game it may not work out.

[dougdirt]

There are several problems with defining roles by leverage, to quote Rick's fine post.

1. You never know when the highest leverage situation arises. Sometimes it may appear clear, but using up Chapman in a sixth inning leverage situation may leave you stranded in the eighth and ninth.

This is probably why a lot of managers do what they do. Of course, in doing so, they also worry about something that may never happen as opposed to know exactly the scenario presented to them right now that is happening. We see it all of the time (can't bat lefties back to back because a LOOGY might come in in the 8th inning) where managers worry more about things that might happen instead of things that absolutely are happening. I would rather know that I am using my best guy in a situation I know is very valuable instead of hold onto him later just in case another situation arises.

[scott91575]

My contention is two fold...

1) Your best bullpen arm should not be used exclusively in the 9th inning with a lead of 3 or less.
2) While still having value, a closer or any single bullpen pitcher for that matter does not have a ton of value. Bullpens are nice to have, and as a whole certainly adds up in value. Yet the bullpen should never be improved to the detriment of other areas of the team. In other words, a good pitcher who can start should be a starter and teams on a limited budget should never blow a ton of money on things like a closer (see Cordero).

[wheels]

All we need to look at is what Jack Mckeon did with Sullivan, Graves, and Williamson in 1999.

To me, that is the template for how to manage a modern bullpen. It was sheer perfection , and I don't remember people knashing teeth about who was used, and when.

[scott91575]

All we need to look at is what Jack Mckeon did with Sullivan, Graves, and Williamson in 1999.

To me, that is the template for how to manage a modern bullpen. It was sheer perfection , and I don't remember people knashing teeth about who was used, and when.

as I mentioned earlier, I think many managers now use a closer for morale reasons. Top end pitchers want those saves to pad their resume. There are still enough people in baseball that put a bunch of value in the closer spot. Rotating closers is very difficult to do and keep the bullpen guys happy (especially if there is a guy better than the rest). It also makes it tough to keep talent or sign free agents if you don't have a defined closer role. Until enough teams buck the trend, the best bullpen pitchers in the game will want to be closers in order to get those all important saves on their resume.

That is why the save stat should be abolished. At the very least combine saves and holds into a single stat.

[mth123]

All we need to look at is what Jack Mckeon did with Sullivan, Graves, and Williamson in 1999.

To me, that is the template for how to manage a modern bullpen. It was sheer perfection , and I don't remember people knashing teeth about who was used, and when.

Mostly agree, but it helps when you have a Scott Sullivan to lean on. This team has some good pitchers in the pen, but there isn't a Sullivan in the bunch. Graves and Williamson could alternate in the 9th because Sullivan's presence meant they didn't really need to ticket anyone for the 8th. Sullivan bridged the late inning gap by himself most of the time.

I like Marshall and Broxton, but those two guys alone aren't enough to handle the 8th and 9th, I think the quality drops off significantly after those two and the two next best guys are likely going to be in AAA (Hoover and Cingrani).

[wheels]

Mostly agree, but it helps when you have a Scott Sullivan to lean on. This team has some good pitchers in the pen, but there isn't a Sullivan in the bunch. Graves and Williamson could alternate in the 9th because Sullivan's presence meant they didn't really need to ticket anyone for the 8th. Sullivan bridged the late inning gap by himself most of the time.

I like Marshall and Broxton, but those two guys alone aren't enough to handle the 8th and 9th, I think the quality drops off significantly after those two and the two next best guys are likely going to be in AAA (Hoover and Cingrani).

We are definitely in agreement in regards to this year's club.

I think the kid with the Fu Manchu from Texas might come close, though.

[Griffey012]

This is probably why a lot of managers do what they do. Of course, in doing so, they also worry about something that may never happen as opposed to know exactly the scenario presented to them right now that is happening. We see it all of the time (can't bat lefties back to back because a LOOGY might come in in the 8th inning) where managers worry more about things that might happen instead of things that absolutely are happening. I would rather know that I am using my best guy in a situation I know is very valuable instead of hold onto him later just in case another situation arises.

While I agree in theory on using the best pitcher in the best situation. I just don't see it being plausible that often. If trouble occurs in the 8th inning, or you can easily predict the key portion of the other teams lineup is coming up in the 8th in a 1 or 2 run game then I believe you can and should get your closer in the game for that inning. But how often is it that easy to predict late in games?

I would like to hear your opinion on how to handle the fact that a lot of 7th and 8th inning leverage situations arise mid inning. You see a 3 or 2 run lead suddenly become a 1 run lead rapidly. Not rapid enough for a "closer" to go from sitting and waiting to warm, dependent on the "closer."

I would tend to believe the 9th would have the highest portion of the distribution followed closely by the 8th and so on. Could it be true that by saving the closer for the 9th the manager is maximizing the probability the closer pitches in the highest leverage situation?

A bigger issue I feel is the lack of using closers for more than 3 out saves. Many times we see a closer seem loose in the pen yet the manager refuses to bring him in for 4 or 5 outs.

[gonelong]

Also, the Boston Red Sox tried the Bill James approach to closing and it was an epic failure. That isn't to say that if another team tried the same approach, it wouldn't work. However with the short sample size of teams that have tried that, it doesn't have a great track record.

I am going totally from memory, so I reserve the right to be wrong here ...

My impression was that Bill James had advocated using your best reliever in high leverage situations in the 6/7/8/9 innings if the opportunity presented itself.

I seem to recall that Boston just didn't have a bonafide closer that year and went with closer by committee. They didn't have a "best reliever" to deploy in high leverage situations, they had a bunch of mediocrity.

When I think of using your top flight reliever in High leverage situations, I think it's closer to how the Reds used the Nasty Boys in 1990, than how Boston recently went about it.

GL