Baseball Toaster Mike's Baseball Rants
Help
This is my site with my opinions, but I hope that, like Irish Spring, you like it, too.
Frozen Toast
Search
Google Search
Web
Toaster
Mike's Baseball Rants
Archives

2009
01 

2008
10  09  07 
06  05  04  03 

2007
12  11  10  09  08  07 
06  05  04  03  02  01 

2006
12  11  10  09  08  07 
06  05  04  03  02  01 

2005
12  11  10  09  08  07 
06  05  04  03  02  01 

2004
12  11  10  09  08  07 
06  05  04  03  02  01 

2003
12  11  10  09  08  07 
06  05  04  03  02  01 

2002
12  11  10  09  08  07 
Links to MBBR
Setting 'Em Up And Knocking
2003-02-24 09:03
by Mike Carminati

Setting 'Em Up And Knocking 'Em Down: A Brief History of Middle Relieving

So 1987 became a transition year for the closer. The "modern" closer established by Bruce Sutter was replaced by the "post-modern" closer, one that was even more geared to producing higher save numbers: shorter outings, more strikeouts, better control, fewer wins and losses, etc. The transition was spurred on by an offensive barrage that overwhelmed the suddenly outmoded 100-inning closer approach. The new approach has become calcified in the last 14 years, but more on that in the Nineties section.

What I am interested in finding out here is how closers were able to be used in the "post-modern" way. I believe that the rest of the bullpen, which had been secondary closers and extra starting pitchers in the past was now a unit with more individualized and more clearly defined roles. I feel that the bulk of the involvement in relief wins and losses was transferred from the closer to the middle relievers. Closers now just saved games. I also feel that this transfer of responsibility may have obviated the need for a closer, but more on that later.

For now we will look at the development of the non-closing relievers and explore their impact on team success. First, we must define our terms, which gets a bit difficult with relievers and especially closers since its sort of the negation of certain things that define what a pitcher does. That is, he doesn't start games or at least does not very often. With closers, you can go by save thresholds but they vary over time and across teams.

So what constitutes a closer? One would say that relief appearances (i.e., games pitched minus games started) and saves. So thresholds like, say, 20 saves won't work, but what about percentage of team totals? A closer is someone who records, say, 75% of a team's saves.

That seems reasonable, right? Well, maybe not. Check out this table of percentage of saves by team leaders. For each year the percentage of major league clubs that have save leaders that reach the prescribed levels is listed:

Year	100%	90%	75%	50%	25%	10%	0 saves
1871	22%	22%	22%	22%	22%	22%	78%
1872	18%	18%	18%	18%	18%	18%	82%
1873	11%	11%	11%	22%	22%	22%	78%
1874	25%	25%	25%	25%	25%	25%	75%
1875	8%	8%	8%	8%	15%	15%	85%
1876	13%	13%	13%	38%	38%	38%	63%
1877	17%	17%	17%	33%	33%	33%	67%
1878	17%	17%	17%	17%	17%	17%	83%
1879	0%	0%	0%	13%	13%	13%	88%
1880	25%	25%	25%	63%	63%	63%	38%
1881	0%	0%	0%	13%	13%	13%	88%
1882	14%	14%	14%	14%	14%	14%	86%
1883	31%	31%	31%	50%	50%	50%	50%
1884	24%	24%	24%	30%	30%	30%	70%
1885	38%	38%	44%	50%	50%	50%	50%
1886	19%	19%	19%	31%	38%	38%	63%
1887	63%	63%	63%	69%	75%	75%	25%
1888	25%	25%	25%	38%	38%	38%	63%
1889	44%	44%	44%	81%	94%	94%	6%
1890	28%	28%	28%	64%	68%	68%	32%
1891	18%	18%	18%	47%	76%	82%	18%
1892	42%	42%	42%	83%	83%	92%	8%
1893	17%	17%	17%	58%	92%	92%	8%
1894	25%	25%	33%	42%	75%	75%	25%
1895	33%	33%	42%	67%	100%	100%	0%
1896	42%	42%	42%	83%	100%	100%	0%
1897	33%	33%	33%	67%	75%	75%	25%
1898	25%	25%	25%	50%	50%	50%	50%
1899	17%	17%	25%	58%	67%	75%	25%
1900	50%	50%	50%	75%	75%	75%	25%
1901	19%	19%	25%	75%	81%	81%	19%
1902	38%	38%	44%	75%	94%	94%	6%
1903	31%	31%	31%	81%	100%	100%	0%
1904	31%	31%	38%	69%	81%	81%	19%
1905	25%	25%	25%	44%	75%	75%	25%
1906	6%	6%	19%	50%	94%	94%	6%
1907	6%	6%	13%	50%	100%	100%	0%
1908	6%	6%	6%	38%	88%	100%	0%
1909	0%	0%	13%	44%	94%	100%	0%
1910	0%	0%	0%	38%	81%	100%	0%
1911	6%	6%	13%	31%	100%	100%	0%
1912	6%	6%	6%	38%	94%	100%	0%
1913	0%	0%	0%	44%	88%	100%	0%
1914	0%	0%	0%	13%	79%	100%	0%
1915	0%	4%	4%	29%	96%	100%	0%
1916	0%	0%	0%	13%	94%	100%	0%
1917	0%	0%	0%	0%	94%	100%	0%
1918	6%	6%	6%	25%	94%	94%	6%
1919	6%	6%	13%	50%	94%	100%	0%
1920	0%	0%	6%	31%	94%	100%	0%
1921	0%	0%	0%	13%	100%	100%	0%
1922	0%	0%	0%	13%	100%	100%	0%
1923	0%	0%	0%	31%	100%	100%	0%
1924	0%	0%	0%	50%	94%	100%	0%
1925	0%	0%	0%	25%	94%	100%	0%
1926	0%	0%	6%	19%	94%	100%	0%
1927	0%	0%	0%	25%	100%	100%	0%
1928	0%	0%	0%	19%	75%	100%	0%
1929	0%	0%	0%	25%	94%	100%	0%
1930	0%	0%	0%	6%	94%	100%	0%
1931	6%	6%	13%	25%	88%	100%	0%
1932	0%	0%	0%	38%	100%	100%	0%
1933	0%	0%	0%	19%	100%	100%	0%
1934	0%	0%	0%	19%	88%	100%	0%
1935	6%	6%	6%	19%	94%	100%	0%
1936	0%	0%	0%	19%	81%	100%	0%
1937	0%	0%	6%	19%	100%	100%	0%
1938	0%	0%	6%	25%	94%	100%	0%
1939	0%	0%	6%	31%	94%	100%	0%
1940	0%	0%	0%	25%	88%	100%	0%
1941	0%	0%	0%	25%	100%	100%	0%
1942	6%	6%	25%	44%	100%	100%	0%
1943	0%	0%	0%	31%	94%	100%	0%
1944	0%	0%	6%	50%	94%	100%	0%
1945	0%	0%	0%	31%	81%	100%	0%
1946	0%	0%	0%	6%	88%	100%	0%
1947	0%	0%	25%	44%	94%	100%	0%
1948	0%	0%	6%	31%	94%	100%	0%
1949	0%	0%	6%	19%	81%	100%	0%
1950	0%	0%	13%	25%	88%	100%	0%
1951	0%	0%	0%	25%	94%	100%	0%
1952	0%	0%	0%	44%	94%	100%	0%
1953	0%	0%	6%	38%	94%	100%	0%
1954	0%	0%	0%	38%	94%	100%	0%
1955	0%	0%	13%	38%	100%	100%	0%
1956	0%	0%	6%	31%	94%	100%	0%
1957	0%	0%	0%	31%	100%	100%	0%
1958	0%	0%	0%	19%	100%	100%	0%
1959	0%	0%	6%	31%	88%	100%	0%
1960	0%	0%	6%	25%	100%	100%	0%
1961	0%	0%	0%	44%	94%	100%	0%
1962	0%	0%	0%	20%	90%	100%	0%
1963	0%	0%	10%	60%	100%	100%	0%
1964	0%	0%	15%	50%	100%	100%	0%
1965	0%	0%	10%	55%	95%	100%	0%
1966	0%	0%	0%	50%	100%	100%	0%
1967	0%	0%	0%	40%	100%	100%	0%
1968	0%	0%	10%	20%	90%	100%	0%
1969	0%	0%	4%	50%	100%	100%	0%
1970	0%	0%	0%	58%	100%	100%	0%
1971	0%	8%	8%	42%	100%	100%	0%
1972	0%	0%	8%	42%	96%	100%	0%
1973	0%	0%	8%	38%	92%	100%	0%
1974	4%	8%	21%	67%	100%	100%	0%
1975	0%	0%	4%	42%	96%	100%	0%
1976	0%	0%	13%	50%	100%	100%	0%
1977	0%	0%	12%	62%	100%	100%	0%
1978	0%	0%	12%	62%	100%	100%	0%
1979	0%	4%	8%	58%	96%	100%	0%
1980	0%	0%	12%	58%	100%	100%	0%
1981	0%	0%	19%	54%	100%	100%	0%
1982	0%	0%	12%	58%	100%	100%	0%
1983	0%	4%	15%	50%	100%	100%	0%
1984	0%	0%	19%	62%	96%	100%	0%
1985	0%	4%	27%	62%	100%	100%	0%
1986	0%	0%	12%	58%	100%	100%	0%
1987	0%	0%	15%	42%	100%	100%	0%
1988	0%	4%	27%	85%	100%	100%	0%
1989	0%	0%	31%	88%	100%	100%	0%
1990	0%	4%	23%	69%	100%	100%	0%
1991	0%	8%	35%	65%	100%	100%	0%
1992	0%	4%	35%	69%	100%	100%	0%
1993	0%	21%	43%	86%	100%	100%	0%
1994	0%	4%	32%	79%	96%	100%	0%
1995	0%	11%	54%	82%	100%	100%	0%
1996	4%	7%	54%	86%	100%	100%	0%
1997	0%	11%	46%	75%	100%	100%	0%
1998	0%	17%	53%	87%	100%	100%	0%
1999	0%	30%	57%	83%	100%	100%	0%
2000	0%	17%	53%	83%	100%	100%	0%
2001	0%	20%	57%	90%	100%	100%	0%
2002	3%	40%	77%	90%	100%	100%	0%

I included all years because the cumulative effect is to present the history of relief pitching in microcosm. At first very few clubs employ relievers at all. By the middle of the first decade in the 1900s, John McGraw has changed that approach forever (only one team has zero saves after 1906). The notion of a saver in varying degrees fell in and out of favor throughout the next fifty to sixty years. It wasn't until 1995 that over half the teams had closers who registered at least three-quarters of the team saves. It wasn't until the mid-Sixties that the closer recorded at least half of the saves on a regular basis. Also, the momentum is continuing until today. Note that 2002 was the first year since 1900 in which at least 40% of the teams had a closer who recorded at 90% of all team saves. So percentages don't appear to be the solution either.

I think the only definition that holds up throughout baseball history is that the closer is the man or men with the most saves on the team. I will define setup men as everyone else. It's somewhat arbitrary: to regard Jesse Orosco in 1986 as a setup man because he had one fewer save -22 to 21-than Roger McDowell seems a bit unfair. But it's the best I can do. Historically, there are only three teams with more than one 20-save pitcher. They are the 1986 Mets (above), the 1991 Blue Jays (Henke, 32, and Ward, 23), and the '92 Reds (Charlton, 26, and Dibble, 25). It's surprising that no one else has tried to duplicate their success given that the first two teams won a World Series and the third won 90 games. However, the ability to have dual savers may be a result of a highly talented organization, not the reverse. Also, the last two rounds of expansion may have put a crimp in this strategy.

The reason I say men is that a team could have co-leaders in saves. This happened last year in Comiskey when Keith Foullke and Antonio Osuna led the White Sox with 11 saves each (rookie Damaso Marte also had 10 of his own). Here are the occurrences of two pitchers leading their teams with 10 or more saves:

Year	Team	Team Saves	Saves
1989	KCA	38	18
1985	NYN	37	17
1987	OAK	40	16
1961	CIN	40	16
1999	BOS	50	15
1992	NYN	34	15
1975	PHI	30	14
1973	CIN	43	14
1982	BOS	33	14
1987	LAN	32	11
1963	SLN	32	11
2002	CHA	35	11
1981	CHA	23	10
1984	TOR	33	10

One more comment regarding multiple closers, the concept of a bullpen- or closer-by-committee further complicates this. How can you call a pitcher the closer if he just happens to have 12 saves when there are three other pitchers with 7, 8, and 10, for example? My defense is that somebody has to be the closer. Well, maybe not though. What if all of the relievers are simply designated short relievers under the middle reliever category if the team leader does not reach a certain threshold? Well, then we're back to our mess of setting thresholds that work over eras. I think that assigning the team save leader as the closer will work because it will produce average numbers that are not that much different, perhaps, from the average reliever when the by-committee approach is in vogue.

Okay, we now have a closer defined. On to middle relievers. So what's the difference between middle relievers and swingman and part-time starters filling in? Middle relievers are a negation of a negation: they are not closers, i.e., save leaders, who in turn are not starters. So what defines a middle reliever? First, he makes a good number of relief appearances. I would think 20 appearances would be a good minimum threshold. He may start (this is especially true historically) but not often. Let's limit him to under 15 starts.

A starting pitcher then will be a pitcher who starts at least 15 games. The rest of the pitchers will be lumped into "other", anyone who did not start 15 games or relieve in 20 games. This includes late season call-ups, early season injuries, cup-of-coffee guys, little used men without a defined role, and position players finishing up a shellacking. Note that by our definition some early ace relievers like Iron Joe McGinnity will appear in both the starter and the closer categories. I'm going to allow this because they really did have a dual role. That's why the called him Iron Joe. I want to capture how the roles have changed over time, so I will allow them to by defined as they were used.

Also, I want to incorporate how the roles were used differently by different teams. Therefore, a team leader with a dozen saves and one with fifty will both be considered closers. Take a look at the varying use of closers in the table below containing the major-league leader, the minimum team leader, and the maximum and minimum totals for all team in a given year from 1960 to today:

Year	MLB Ldr 	Min	Max Team	Min Team
		Tm Ldr	Total	Total
1960	26	4	42	14
1961	29	5	40	13
1962	28	4	47	10
1963	27	4	43	12
1964	29	5	45	15
1965	31	3	53	14
1966	32	7	51	22
1967	28	6	46	19
1968	25	6	40	16
1969	31	6	44	17
1970	35	8	60	20
1971	31	4	48	12
1972	37	4	60	13
1973	38	6	46	19
1974	24	3	29	12
1975	26	5	50	16
1976	26	6	46	15
1977	35	8	47	20
1978	37	9	55	20
1979	37	6	52	11
1980	33	6	50	13
1981	28	3	35	10
1982	36	7	51	22
1983	45	7	49	23
1984	45	8	51	21
1985	41	9	53	24
1986	46	10	58	24
1987	40	8	51	16
1988	45	13	64	25
1989	44	15	57	26
1990	57	8	68	29
1991	47	12	60	33
1992	51	13	58	29
1993	53	10	61	22
1994	33	5	46	20
1995	46	12	50	22
1996	44	8	52	22
1997	45	14	59	29
1998	53	9	59	24
1999	45	12	55	29
2000	45	13	53	27
2001	50	11	57	26
2002	55	11	57	23


(By the way, the 1981 team with only 3 saves for its leader was the Billy Ball A's, who completed 60 of 109 games (Jeff Jones and Dave Beard led the team). The next highest team had just 33 complete games. In 1994, Cleveland was led by Paul Shuey and mid-season acquisition Jeff Russell with 5 saves each-Jose Mesa, Steve Farr, and Jerry DiPoto, who all closed at some point in their careers, were also on the team.)

Now that we have our categories, let's compare them historically and, specifically, in the "modern" and the "post-modern" eras. Below is a comparison of statistics for closers, middle relievers, and starters over time. Please note that no one qualifies for the middle reliever role until 1910. That year both Lew Richie (with the Braves and Cubs) and Lou Schettler (with the Phillies in his only major-league year) qualified.

The first man to qualify multiple times-if we take that to mean the first with career role as middle reliever-was the seminal reliever, the Giants' Doc Crandall (1912-'13). The first man to be used solely as a middle reliever (i.e., no starts but 20 relief appearances for a someone who was not the team leader in saves) was Hooks Wiltse, also of McGraw's Giants, in 1914. The first used solely as a middle reliever and who did not "record" any saves was Jesse Winters, of course of the Giants, in 1920. McGraw as we discussed earlier evolved the bullpen into having men of specialized roles throughout his tenure at the Giants' helm.

I could not fit all three roles (closer, middle reliever, and starter) in one line so I have broken down the comparisons to closer versus middle reliever and then starter separate. For each role the yearly average for the typical stats are listed. There are also a few atypical ones that need explanation:

Age: based on player's age as of mid-season, i.e., July 1.

ERA+: the ERA adjusted to the league average, where 100 is average, over 100 is good, and under 100 is subpar. Basically, the Baseball-Reference.com statistic.

/Tm indicates the number of players for each role per team in the majors.
%Tm Sv is used only for closer. It indicates the percentage of average team saves is represented by the average closer save total.

(Also note that the criteria for the strike years-1981, 1994, and 1995-were readjusted to take into account the lower number of games played.)

 (By the way, the 1981 team with only 3 saves for its leader was the Billy Ball A's, who completed 60 of 109 games (Jeff Jones and Dave Beard led the team). The next highest team had just 33 complete games.  In 1994, Cleveland was led by Paul Shuey and mid-season acquisition Jeff Russell with 5 saves each-Jose Mesa, Steve Farr, and Jerry DiPoto, who all closed at some point in their careers, were also on the team.)

Now that we have our categories, let's compare them historically and, specifically, in the "modern" and the  "post-modern" eras.  Below is a comparison of statistics for closers, middle relievers, and starters over time. Please note that no one qualifies for the middle reliever role until 1910.  That year both Lew Richie (with the Braves and Cubs) and Lou Schettler (with the Phillies in his only major-league year) qualified.  

The first man to qualify multiple times-if we take that to mean the first with career role as middle reliever-was the seminal reliever, the Giants' Doc Crandall (1912-'13).  The first man to be used solely as a middle reliever (i.e., no starts but 20 relief appearances for a someone who was not the team leader in saves) was Hooks Wiltse, also of McGraw's Giants, in 1914.  The first used solely as a middle reliever and who did not "record" any saves was Jesse Winters, of course of the Giants, in 1920.  McGraw as we discussed earlier evolved the bullpen into having men of specialized roles throughout his tenure at the Giants' helm.

I could not fit all three roles (closer, middle reliever, and starter) in one line so I have broken down the comparisons to closer versus middle reliever and then starter separate.  For each role the yearly average for the typical stats are listed.  There are also a few atypical ones that need explanation:
Age: based on player's age as of mid-season, i.e., July 1. ERA+: the ERA adjusted to the league average, where 100 is average, over 100 is good, and under 100 is subpar. Basically, the Baseball-Reference.com statistic. /Tm indicates the number of players for each role per team in the majors. %Tm Sv is used only for closer. It indicates the percentage of average team saves is represented by the average closer save total.
(Also note that the criteria for the strike years-1981, 1994, and 1995-were readjusted to take into account the lower number of games played.)

Comment status: comments have been closed. Baseball Toaster is now out of business.