This must be a record for posting the most new HoM threads in a week. :-)

Can we start the Danny Kolb thread now? I'm ready to enshrine him!

What's going on with Keith Foulke?

I know! He's killing me! I can't relax with a 3-run lead in the 9th anymore!

Who is the all-times saves leader as of 1950?

Why Grandma Murphy! That's who!

Eligible in 1953. Save the date.

I'm not gonna say this is THE way to rank relievers, but by one method (my so-called Reputation Monitor, which is similar in spirit to the HoF Monitor) this is now I see them.

1. Wilhelm
2. Eck
3. Quiz
4. Sutter
5. Fingers
6. Gossage
7. Fingers
8. Smith
9. Lyle
10. Kinder

Active relievers not included. And I know this is not THE way to rank relievers because I have given the matter a little study not reflected in this particular method, and it became clear to me that Gossage is perhaps as high as the #2 reliever. But Wilhelm is a clear #1 and a clear HoMer. Everybody else depends on your view of the significance of their IP--too few? or does leverage make up for that? I'm not sure there is AN answer, rather that may be a matter of taste. But Wilhelm alone on this list transcends that problem.

Oops, try that again.

1. Wilhelm
2. Eck
3. Quiz
4. Sutter
5. Gossage
6. Fingers
7. Smith
8. Lyle
9. Kinder
10. Hiller

Kinder is not a controversial choice at all, IMHO. For the record, here is the Peak WS (3 yr. + 5 yr) for the top 15 relievers by Reputation Monitor (RM) (which includes the above 10 plus Myers, Marshall, Perranoski, Radatz and Face).

Quiz 182
Sutter 166
Eck 161
Gossage 159
Hiller 157
Kinder 152
Marshall 151
Wilhelm 148
Radatz 141
Smith 136
Face 126
Lyle 119
Perranoski 118
Fingers 117
Myers 106

Best ERA+

Wilhelm and Quiz 146
Sutter 136
Hiller 134
Smith 132

But for peak ERA+ Gossage just kills everybody, well, except Quiz.

On HoF Monitor + HoF Stds + BI + GI

Eck 338
Wilhelm 232
Fingers 212
Smith 209
Sutter 192
Gossage 187
Lyle 155
big drop off....

I am not sure how you can have a top 10 all-time without Rivera. Unless, of course, you decide not to rank him because he is still playing, which is defensible.

The question with Rivera, I guess, is ho wmuch weight should we give his playoff record?

Well, Rivera's postseason case thru what, 2000, was pretty compelling.
Since then, he's given back some significant territory. Still a definite plus, but I think ordinary baseball fans, especially Yankees backers, tend to remember his role in winning titles while downplaying his role in losing them.
Or one could downplay his influence in both cases, I suppose.

I was thinking about RPs during my brief lunch break at the Henry B. Gonzalez Convention Center....

There are essentially five questions to ask about Relief Pitchers:
1) Was he an effective pitcher?
2) What was his specific role in relief? (aka usage pattern)
3) How successfully did he execute that role?
4) How important was that role to actually winning games?
5) Does his career include any other important contributions (esp. including starting pitching), and does that contribution add, detract, or do nothing to his case?

To measure these things...

1) Effectiveness: We have the usual assortment of ERA+, DERA, WS, WARP, etc.... These esentially measure run prevention ability (though WS also includes some opportunity/role stuff). But things like RSI and adj defense don't necessarily tell us as much because the sample isn't as large.

2) Role: Realistically, we need game-by-game analysis to understand a pitcher's role, but we can proxy it pretty well by seeing how many decisions, saves, and GF he had per appearance. If his role was bullpen leader, he'll have lots of them, if not, he won't. Also, part of describing role would be how often he came on to start an inning versus come on to put out fires.

3) Success in role: Ideally we'd have bequeathed/inherited runner information for all RP. We have most of it going back to 1972ish thanks to BP (with ARP and other similar measures) and retrosheet, and for many seasons before that retorsheet has at least partial records. So we can piece together some of what we don't already know.

4) Importance of role: Tango Tiger computed leverage indexes for numerous (all?) RPs since 1972ish, and they help a lot in showing how important each appearance a reliever made toward contributing to a win, especially during peak/prime years. In this group I might also include performance in October games or stretch-drive games since their importance is so magnified.

5) Here's the usual assortment of stuff we use to evaluate starting pitching and pitcher batting as needed (which is not often in later cases and occasional in earlier ones for both starting and batting). The key issue here is appropriately weighting starting versus relieving.

I put up a WS-based list of best RPs previously, and it's not ideal, but it offers a very generalized starting point for discussion. I believe that ultimately we will have to blend all these ideas together to get a strong sense of all things relieverly, just as Sunnyday has begun to do in post #7.

5) Does his career include any other important contributions

Terry "Fat Tub of Goo" Forster and his magic bat!

OK, I'm kidding. It's a little hard to turn 31 lifetime hits (with only 6 extra bases and 2 walks) into "important." But he's still fun to talk about.

Well, Rivera's postseason case thru what, 2000, was pretty compelling.
Since then, he's given back some significant territory. Still a definite plus, but I think ordinary baseball fans, especially Yankees backers, tend to remember his role in winning titles while downplaying his role in losing them.
Or one could downplay his influence in both cases, I suppose.

1.42 ERA in the 2001 WS
1.29 ERA in the 2004 ALCS

He's still pitched fairly well even when the Yankees have lost.

The current group of Yankee veterans has amassed a very large amount of playoff numbers. Rivera has 108.7 playoff innings with 8-1 W-L, 32 Sv, 0.75 ERA. That's a full season of HOM-level relief numbers.

108 IP is a lot more than 1 season's worth of closing! How many games?

But c'mon, he won't even be eligible when this project ends! About Chet Brewer...

sunnyday2
my so-called Reputation Monitor, which is similar in spirit to the HoF Monitor

Do you mean that the purpose is to predict or retrodict reputation among baseball writers?

Some other recent relievers for the consideration set: Tekulve, Henke, Reardon. Older guys: Marberry, Murphy, McMahon. 'McDaniel' also begins with m. (Jim) Brewer begins with b.

THIS ARTICLE IS NOT MINE
by Dr. Chaleeko, moved from "1951 Ballot Discussion"

I've taken a look at relief some of the more dominant relief pitchers since the Fireman model of the McCarthy Yankees started to take hold.

Here's a top-ten list to chew on in rough rank order that is highly subject to change:

1 Wilhelm
2 Gossage
3 McDaniel
4 Sutter
5 Marshall
6 Quisenberry
7 Ellis Kinder
8 Lee Smith
9 Stu Miller
10 Rivera

Disclosure statement: This does NOT account for any leverage indexing, nor for any BP IHR stats and is based purely on a WS adjustment method. In addition, it does indeed count and account for RPs' starting contributions.

Kinder is obviously the potentially controversial name on this list, and he benefits from some SP credit, but he was also a very highly effective reliever, at least according to WS. IIRC, he comes up in the early 1960s.

I haven't identified a great way to rank modern closers yet because they tend to score very low in WS. Thus I am probably short-changing Rivera, Hoffman, and a few others.

Hey, for kicks, here's a second ten:

11 Hiller
12 Face
13 Doug Jones
14 Montgomery
15 McGraw
16 Garber
17 Peranoski
18 Franco
19 Fingers
20 Tekulve

Hoffman could be on this list, as could Wetteland or Eck or Henke.

Dr. Chaleeko, from the article imported just above.
1 Wilhelm
2 Gossage
3 McDaniel

Lindy McDaniel #3 and Don McMahon not in the Top 23?

:
Marberry, Murphy, McMahon, McDaniel, Miller, Marshall, McGraw, Moffat*, Myers, Montgomery

* Moffat as in Billie Jean? Yes, we use the decimal system here.

I was in the same building with Don McMahon when he died.

(Of course, the building was Dodger Stadium.)

As with all positions, I always believe rank vs. your peers is VERY important, so here's my list of careers by their primary decade:

1910's
1. Doc Crandall
2. ??

1920's
1. Firpo Marberry
2. ??

1930's
1. Johnny Murphy
2. Clint Brown

1940's:
1. Al Brazle
2. Al Benton
3. Ted Wilks

1950's:
1. Bobby Shantz
2. Ellis Kinder
3. Don Mossi

1960's:
1. Hoyt Wilhelm
2. Stu Miller
3. Ron Perranoski
4. Dick Radatz

1970's:
1. Rollie Fingers
2. Kent Tekelve
3. John Hiller
4. Sparky Lyle
5. Mike Marshall

1980's:
1. Dennis Eckersley
2. Rich Gossage
3. Tom Henke
4. Bruce Sutter
5. Dan Quisenberry

1990's:
1. Lee Smith
2. John Franco
3. Mariano Rivera (still moving up)
4. Trevor Hoffman
5. John Wettland

Maybe it is because I just turned 24 yesterday and so I missed the peak of these guys but to me Lee Smitha nd John franco were never that Dominate and this should't be put ahead of Rivera. Of course I am a peak voter who probably wasn't as cognizant during the peaks of those two players. I have no stats as usual.

Bump!

Here's that reliever thread!

From Sinnis encyclopedia, Top RSAA trying to get all the relevant "reliever" guys:

CAREER RSAA

GAMES STARTED <= 365
GAMES > 585

RSAA                           RSAA      GS        G       IP       SV       ERA    
1    Hoyt Wilhelm                282       52     1070   2253        227     2.52   
2    John Smoltz                 239      361      602   2699.2      154     3.27   
3    Dennis Eckersley            208      361     1071   3285.2      390     3.50   
4    Mariano Rivera              175       10      586    728.1      336     2.43   
5    Goose Gossage               160       37     1002   1809.1      310     3.01   
6    John Franco                 154        0     1088   1230.2      424     2.84   
T7   Dan Quisenberry             148        0      674   1043.1      244     2.76   
T7   Kent Tekulve                148        0     1050   1436.1      184     2.85   
9    Lee Smith                   143        6     1022   1289.1      478     3.03   
T10  Charlie Root                134      341      632   3197         40     3.59   
T10  John Candelaria             134      356      600   2525.2       29     3.33   
12   Bob Stanley                 132       85      637   1707        132     3.64   
T13  Steve Reed                  128        0      803    837.2       18     3.51   
T13  Tom Henke                   128        0      642    789.2      311     2.67   
T13  Murry Dickson               128      338      625   3053.1       23     3.66   
T16  Bruce Sutter                123        0      661   1042.1      300     2.83   
T16  Wilbur Wood                 123      297      651   2682         57     3.24   
18   Sparky Lyle                 122        0      899   1390.1      238     2.88   
19   John Wetteland              119       17      618    765        330     2.93   
20   Doug Jones                  117        4      846   1128.1      303     3.30   
21   Mike Jackson                115        7     1005   1188.1      142     3.42   
22   Roberto Hernandez           113        3      825    891.2      320     3.39   
T23  Jesse Orosco                111        4     1252   1295        144     3.16   
T23  Tom Gordon                  111      203      671   1896.2      114     3.99   
T23  Paul Quantrill              111       64      791   1186.2       21     3.74   
26   Jeff Montgomery             110        1      700    868.2      304     3.27   
T27  Armando Benitez             108        0      628    654        244     2.85   
T27  Trevor Hoffman              108        0      696    764.2      393     2.74   
29   Rollie Fingers              103       37      944   1701.1      341     2.90   
30   Gene Garber                 102        9      931   1510        218     3.34   
31   Mike Timlin                  97        4      812    955.1      117     3.61   
32   Jeff Nelson                  95        0      743    745.1       32     3.38   
33   Clay Carroll                 93       28      731   1353        143     2.94   
T34  Stu Miller                   89       93      704   1694        154     3.24   
T34  Tom Burgmeier                89        3      745   1258.2      102     3.23   
T36  Mike Marshall                88       24      723   1386.2      188     3.14   
T36  Gary Lavelle                 88        3      745   1085        136     2.93   
38   Robb Nen                     87        4      643    715        314     2.98   
T39  Rick Aguilera                85       89      732   1291.1      318     3.57   
T39  Tug McGraw                   85       39      824   1514.2      180     3.14   
T39  Dave Veres                   85        0      605    694         95     3.44  

That is actually not a bad ranking, though Smoltz is obviously buoyed by his many years as a top starter. Eckersly as well I guess, but he has more value as a closer than Smoltz.

Question, how is RSAA calculated for relivers. For instance, Rivera has 175 RSAA in only 728 innings. Is there a leverage factor here?

Also, I think that Rivera, (if we ever get to him) Wilhelm, and Gossage are definites. After that things will get interesting.

Excuse me if I am missing the point, but my understanding is that the "Leverage Index" is intended to show how important an inning is compared to an "average inning" -- not how important a reliever inning is compared to a starter inning.

It's one thing to say that "Mariano Rivera pitching in the ninth has a leverage inning of 2.0." That is a completely different thing -- if I understand correctly -- than saying that "An inning of Mariano Rivera is twice as important as an inning by Clark Griffith or Eppa Rixey."

My point is that starters pitch ninth innings too. Rixey pitched 290 Complete Games (leaving aside his 97 Games Finished as a reliever), which should work out to some Leverage as well.

It's one thing to say "Double a reliever's innings to account for leverage," but it matters if the starter you are comparing him to has a LI of 1.0 or 1.2 or 1.4.

Since starters have many early innings with LIs less than 1, they average out to 1 or less than 1 typically. At least, II am Ring C.

Yes, "average inning" and "starter inning" will be almost the same thing, with average inning being by definition a 1.0 leverage, and with the average stater innings typically slightly less than 1.0 leverage in "modern" years. Guys like Griffith and Rixey will possibly be slightly over 1.0 I would guess...

It's guys like Rixey and Griffith, though, whom I'm most interest in comparing closers to, as they are the high-ballot competition.

Rivera, note, has only pitched 800 innings, but you can't really "double that to 1600" for leverage, as he was the Yankees closer for the first 200 or so innings.

But let's take someone with a longer career -- Goose Gossage. 681 "Games Finished" over about 1575 innings (I'm assuming no credit for his ill-fated 1976 "Let's make him a starter!" foray.)

Compare that to Clark Griffith (415 CG+GF) or Red Ruffing (406 CG+GF) or Eppa Rixey (387 CG+GF), and it looks like while Gossage pitched the 9th inning a lot, his "marginal ninth innings" are much less impressive.

In an era in which starting pitchers pitched almost all of their innings, the average leverage index of their innings will necessarily be just about 1.0. Starting pitchers who also picked up leveraged relief innings would be the group of "starters plus" who end up with leverages a bit above 1.0. But these innings are such a small percentage of their overall workload that they wouldn't affect their leverage index much.

So the leverage index of modern relievers remains a valid element in comparisons between modern relievers and pre-modern starters.

This might make things look bad for modern starters, who may well have leveraged indices of less than 1.0. They gain back ground against pitchers in earlier eras, however, because they take a higher share of defensive value in every inning that they pitch than did the earlier pitchers. This advantage accrues to modern relievers as well, of course.

But let's take someone with a longer career -- Goose Gossage. 681 "Games Finished" over about 1575 innings (I'm assuming no credit for his ill-fated 1976 "Let's make him a starter!" foray.)

Tangotiger has already calculated a career leverage index for Gossage (you can follow the link to it posted in the 1960 ballot results thread). It includes his starter season, and that lowers his career leverage index to 1.63 or thereabouts, quite a bit down from Sutter's career index of 1.90. Tangotiger also has Rivera's seasonal leverage indices calculated, and you can see how they rise and fall with his usage patterns, and how they are damped down by his play for great teams, which create fewer high-leverage situations for their pitchers.

bump

That's a pretty reasonable list for a guy like me that is thinking about 6 or 7 full-time relievers is reasonable - I could see cutting it off right there, not counting Smoltz and Eck as true relievers.

Franco is kind of the Beckley of the group, so I could see dropping him (I could see being more peak oriented on the relief front). But this is RSAA, not RSAR, so really, Franco was probably just better than I realized.

I'm very surprised at how bad Rollie Fingers does on this metric. But he had a very long career for a reliever, so using something a little lower than average as the baseline would help him.

Does anyone know what a replacement level ace/closer's ERA+ is? I'd imagine it's much closer to average than the ERA+ of a starter, for example.

Wilhelm, Goose, Rivera and Quisenberry are the no-brainers to me. Eck too I suppose.

I wonder if closer isn't an area where I might change my mind on peak vs. career. We have, in the past 20 years, seen many relievers have one or two massive seasons and fade hard. Whether that's small sample size (my vote) or just that the demands of the closer role tend to wear an arm after a couple years, it just feels like high peak/short career is a typical closer path.

With that in mind, perhaps the career achievements of the Francos and Lee Smiths are more impressive than at first glance? Or maybe not...those good, not outstanding closers weren't provoding a ton of value over replacement (which is of course high for the position).

Right now, my no-brainer list is Wilhelm, Goose, and Rivera. Probably Eck, too, given his starting career as supporting value. I'm disinclined to give the benefit of the doubt to non-nobrainer RPs, but we're still many elections from this being an issue, so I could change my mind or be persuaded quite easily still.

Quiz looks more like Goose to me (for peak) than like any other reliever. I think he is top tier.

Considering only modern, full-career relievers, my preliminary analysis has Rivera, Sutter, and Quisenberry a cut above the rest (maybe Billy Wagner is on track to join them). I have no idea how many such relievers belong in the HoM, but 3 is the minimum that I'm comfortable with.

Many relievers seem to have a similar case to Sutter. He might still make my ballot, I don't know yet, but at any rate, I've concluded from recent events that he's a better Hall of Famer than HoMer.

Kinder is obviously the potentially controversial name on this list, and he benefits from some SP credit, but he was also a very highly effective reliever, at least according to WS.

Kinder is first in adjusted pitching runs (relief only) until Lindy McDaniel 1975, who pitched almost three times as many career innings. Among old-timers "who made notable contributions" (Palmer 1993), only Grove and Kinder achieved excellent rates of runs saved per nine innings in relief.

'HallofMerit' egroup subscribers know that I uploaded some Palmer data last week (relief.csv). I will now copy here some of what I mistakenly added to the ballot thread at that time.

A version of this article appeared in the 2006 HOF ballot thread.

While I contemplated voting, I compiled some relief pitcher data from TB3 and BBE2005, which I have now posted in the HallofMerit egroup Files.

Relief Pitcher career data, Pete Palmer 1993 & 2005

<b>Description</b>
Relief Pitcher career data, Pete Palmer 1993 & 2005.
explanatory labels in the first three rows

Scope for data thru 1992 (a subset of Palmer's "all those who made notable contributions"): mlb debut < 1940, >= 800 relief innings, or >= 50 adjusted relief runs.

Scope for data thru 2004 (see BBE2005 page 1652): career Top 50 relief innings, career top 50 adjusted pitching runs, career Top 49 "RNK" rating. Plus three non-qualifiers with 200 career saves and a few other non-qualifiers.

For everyone in the selection who worked fewer than 1500 innings as a starting pitcher, field one gives career Win Shares.

<hr>
The file includes 173 relief pitcher records: 151 from the TB1993 "Relief Pitching Register" (about half of the total, evidently including non-qualifiers Aker, Assenmacher, Drago, Farrell, Regan); 19 more from one of the three BBE2005 leader boards; and Beck, Mesa, Urbina with 200 saves.

Twenty of the leaders in adjusted relief runs (Top 50 thru 2004) retired before 1992, so that both published editions of Palmer relief pitching measures cover their full careers. For the 20, this table shows the changes in adjusted runs between editions.

<order>
R1992 R2004 ratio finale
245 252 1.03 1972 Wilhelm Hoyt
147 147 1.00 1990 Quisenberry Dan
142 139 0.98 1989 Tekulve Kent
121 115 0.95 1989 Stanley Bob
120 129 1.08 1985 Fingers Rollie
118 121 1.03 1982 Lyle Sparky
117 110 0.94 1988 Sutter Bruce
115 110 0.96 1980 Hiller John
113 113 1.00 1984 McGraw Tug
108 105 0.97 1975 McDaniel Lindy
103 103 1.00 1981 Marshall Mike
99 93 0.94 1957 Kinder Ellis
97 98 1.01 1978 Carroll Clay
96 97 1.01 1987 Lavelle Gary
93 90 0.97 1988 Garber Gene
87 85 0.98 1984 Burgmeier Tom
85 86 1.01 1974 McMahon Don
84 84 1.00 1989 Hernandez Willie
84 88 1.05 1973 Perranoski Ron
83 85 1.02 1968 Miller Stu
</order>

Nine more players active in 1992 would qualify for the Top 50 thru 2004 based on their careers thru 1992. One of them, career saves leader Jeff Reardon, barely played after 1992 and is now credited with fewer runs (was 88, now 84), by some unknown combination of negative scores in 1993-1994 and revision in the measure.

--
BTW, Murry Dickson 1939-1959 is the last of the "debut < 1940" qualifiers for inclusion in that file. I have written to Pete Palmer about a probable error in his record (see his thread). He was better than Grandma Murphy.

<div class="codeblock">
<u>'92 '04 ratio</u> finale name
245 252 1.03 1972 Wilhelm Hoyt
147 147 1.00 1990 Quisenberry Dan
142 139 0.98 1989 Tekulve Kent
121 115 0.95 1989 Stanley Bob
120 129 1.08 1985 Fingers Rollie
118 121 1.03 1982 Lyle Sparky
117 110 0.94 1988 Sutter Bruce
115 110 0.96 1980 Hiller John
113 113 1.00 1984 McGraw Tug
108 105 0.97 1975 McDaniel Lindy
103 103 1.00 1981 Marshall Mike
99 93 0.94 1957 Kinder Ellis
97 98 1.01 1978 Carroll Clay
96 97 1.01 1987 Lavelle Gary
93 90 0.97 1988 Garber Gene
87 85 0.98 1984 Burgmeier Tom
85 86 1.01 1974 McMahon Don
84 84 1.00 1989 Hernandez Willie
84 88 1.05 1973 Perranoski Ron
83 85 1.02 1968 Miller Stu
</div>

The median closer in 2005 had an ERA+ of 163 (Joe Nathan).

The seemingly impressive 151 ERA+ put up by Danys Baez of the D-Rays only ranks 20th amongst closers.

I don't think voters give much credit to players who were the 20th best third baseman in a season (in 2005, about Alex Gonzalez, also of the D-Rays).

So, remember that when you see Quisenberry's 153 ERA+ in 1986, you are essentially looking at "average", not "great"; his contemporaries that year were Eichhorn (246), Clear (198), Worrell (175), Righetti (168), Garber (157), Tekulve (153), Harris (152), Orosco (152), Plesac (147), Aase (139), Moore (138), most with as many or more innings.

Closers are also very replaceable. How many teams go more than a year or two without a closer whose ERA+ is above 125?

(1) The median closer in Quisenberry's time had a much lower ERA+ (around 120) than the median closer in 2005.

(2) Quiz was in decline in 1986. He exceeded 154 ERA+ in 1981, 1982, 1983, and 1985.

(1) The median closer in Quisenberry's time had a much lower ERA+ (around 120) than the median closer in 2005.

(2) Quiz was in decline in 1986. He exceeded 154 ERA+ in 1981, 1982, 1983, and 1985.

Alex Gonzalez had a WARP1 of 1.5 (WARP3 2.3) in just 383 PA in 2005. There's definitely some value there. Average players have value, and he was below average (20th).

From what you said there PhillyBooster, if the list is all-inclusive of the top 11 contemporaries, was that Quisenberry was tied for 6th in baseball. That's pretty good. He was also throwing a lot more innings per season. He threw 128-139 IP every non-strike year from 1980-85, so saying he didn't throw a lot of innings in 1986 doesn't really paint a fair picture of him. You act as if that's a typical year for him in his prime.

In his prime he was good for 135 IP at a 170 ERA+ and he was basically that pitcher for 6 years. He's the Hughie Jennings or Sandy Koufax of relievers - only he was great for 6 years instead of 4 or 5. He led his league in Games 3 times, Saves 5 times and GF 4 times. He was as work-horse as you get in his prime, in an era where relievers were worked much harder than today. I'm not saying that's as valuable as Jennings or Koufax, but a peak voter should love him if he gives any credence to relief pitching at all.

I think Quisenberry and not Gossage, Eckersley, Rivera or Sutter had the most valuable extended prime of any reliever, aside from Wilhelm, who is the Babe Ruth of relief pitchers (off the charts great).

Rivera and Eck didn't pitch enough and Sutter wasn't nearly as good. Gossage is very close, but he was only worked as hard as Quisenberry 3 times in his career (1975, 1977-78). In the early 80s he was as effective as Quis when out there, but only pitched 70% of the innings. Whether you prefer Gossage or Quisenberry is directly proportional to how much you value career or peak.

I think Quisenberry and not Gossage, Eckersley, Rivera or Sutter had the most valuable extended prime of any reliever, aside from Wilhelm.

That 6 year prime you mention is fantastic -- 700 innings at a 170 ERA+. But how is it better than Rivera's 10 year prime of 740 innings at a 201 ERA+? Or better than Gossage's 1977-1983 which had the same amount of innings with the following ERA+'s: 465, 246, 180, 180, 173, 173, 156?

I think you are right that Quis is the poor man's Hughie Jennings, but I think Michael Bass is right in post 30 when he says that even peak voters should consider career value for closers much more closely since there are many bright/short lived closing stars but few who can maintain it for a decade or more. Maybe that (the long career combined with a great peak) is what makes a closer Hall Worthy.

I've been a peak/prime voter from day one, and I am looking at relievers the same way. Yes there are lots who have that great 2-3 year peak but no real extended peak or prime. What sets the Quiz' and Gossages apart is that their peak is in that 6-7 year range, or 10 years in Goose' case.

You've still got a fair number of relievers with that 6-7 year peak/prime but now you can start to separate on the basis of how effective they really were, and that is where Goose and Quiz beat Sutter.

I think career is exactly the wrong way to evaluate relievers (Franco et al) because some of them, the more recent ones, pitch so few innings that while they've compiled some interesting career totals, the impact on any one pennant race is so small. Unless indeed they are one of these ERA+ 200-300-400 guys.

Marc,

How does someone like Henke do in your system? Fewer than 800 career innings but 7 seasons with an era+ over 180. Are those 420 innings enough? The more basic point being that if you use peak to evaluate the modern reliever you are evaluating fewer than 500 innings. Is that too few? Mickey Welch pitched more than that in his rookie season. I'm not saying Jesse Orosco is a HoM, I'm just saying that even a shift in focus away from peak to extended prime might be a better way of evaluating modern relievers.

So far Henke is not really on my radar and I can't yet say why. I studied relievers some years ago and need to take another look--first, to gather up some of the more recent ones and second, to revisit the "second tier."

My system bonuses relief pitchers just as it does catchers. Take away the bonus and no reliever would be in my PHoM or PHoF. Take away the bonus and I'm not sure if any are even candidates. But I need to revisit. That includes the "career" candidates because my whole system doesn't give a lot of credit for mere above-averageness.

In his prime he was good for 135 IP at a 170 ERA+ and he was basically that pitcher for 6 years. He's the Hughie Jennings or Sandy Koufax of relievers - only he was great for 6 years instead of 4 or 5.

Let's not oversell his credentials here. In fact, there is only one year in which Quisenberry pitched 135 innings with a 170 ERA+ (1983). 1981 doesn't pro-rate to it either.

Jennings and Koufax had a run when they were the BEST for a short period. I think the better comp for Quis's run is Smokey Joe Wood, minus 400 innings and a 110 OPS+. In some years, Q was the best, I guess, but was he that dominant?

How does Q's 1980 (33 SV, 128 IP, 131 ERA+) compare to Doug Corbett (23 SV, 136 IP, 220 ERA+), or for that matter anyone in the Yankee's bullpen that year? From the looks of it, Gossage, Davis, and May were all as valuable or more valuable that year.

How does Q's 1981 (18 SV, 62 IP, 209 ERA+) compare to Gossage (20 SV, 47 IP, 465 ERA+) or Sammy Stewart (4 SV, 112 IP, 156 ERA+)?

How does Q's 1982 (35 SV, 137 IP, 159 ERA+) compare to Bob Stanley (14 SV, 168 IP, 140 ERA+) or Dan SPillner (21 SV, 134 IP, 164 ERA+). (Also, Goose Gossage and Bill Caudill)

In some of the years in Q's 6-year-run, he was arguable the best, or about equal with the best. In other's he wasn't. Really, 1983 was the only year when he was Koufax/Jennings dominant in the AL, and even that year Lee Smith was comparable with the Cubs.

From Greg Spira, posted on SABR-L:

if you use peak to evaluate the modern reliever you are evaluating fewer than 500 innings. Is that too few? Mickey Welch pitched more than that in his rookie season. I'm not saying Jesse Orosco is a HoM, I'm just saying that even a shift in focus away from peak to extended prime might be a better way of evaluating modern relievers.

There is another issue regarding Jesse Orosco. He pitches with the platoon advantage much more often than any other LHP with his raw credentials and the platoon advantage is purchased at some cost.

Chris Cobb yesterday mentioned platooning in a genuine HOM context, in the Richie Ashburn thread.

I agree about Sutter getting in, if for no other reason than it might open the door for other relievers to get in.

Of course, the greatest reliever of all-time discussion cannot be complete without mentioning Wilfredo Arguelles and Charles States.

Are you reading, Duke!?! Come back to us!!! :)

hot topics

Everybody forgets this thread, it seems. :-)

See #33-35 and #45.
Back in January, I told Greg Spira that I was pleased to see Kent Tekulve again > 200. He replied in a day or two that Pete Palmer suspects a glitch, to be corrected in the 2007 edition if so. Apparently, he doesn't endorse minor rating changes for retired players between editions.

Yes, plum fergot this thread was open.

Rating relievers. I plan to spend more time on this than most players, but at least I'll only have to work a few of them.

My method:

1. Assess them against the SPs of their day, using the same basic tools (IP, ERA+, DERA, etc.) but adjusted for differences between SP and RP.

2. quantity: Adjust reliever’s IP by fireman leverage.

This will be tricky, unless someone has done the research to get a good number for each reliever. However, even if we find that Quisenberry’s career leverage was 1.8, for example, I would argue that this is NOT the right number to apply to all of Quiz’s efforts; because the extra leverage is only his value above the next guy the team would use instead, who is probably an average or better pitcher.
My proposed IP adjustment would that we find the 'best' number for whatever studies exist, either for the typical usage of the day, or for the individual if we have better data for that player. But I will only use the 'leverage' for runs svaed above average, not above 'replacement'; if the Yankees did not have Mariano, they woudl find someone else to close games who was not a poor pitcher.

3. quality: Adjust reliever’s DERA by typical fireman ERA advantage gained from
a. inherited runners charged to SPs
b. possibility of easier pitching when only once through an order, not a full game
....(a) is pretty straightforward; there are lots of studies out there that show a reliever's ERA advanatge in the modern game is about .20 (off top of my head), altho it might be different in the 1970s.
....(b) is more nebulous. It would be good if we had studies that indicated how much better pitchers were at relieving than they were at starting, for those who did both. If we don't have that data, I may not penalize bullpen guys any further.

As an example if we find Gossage saved 180 runs above average from 1975 to 1985, pitching 1200 innings, I might dock the RSAA by 30 (to 150), but up his leveraged innings to 2000, which would increase his 'effective' RSAA to 250. To compare his against a replacement pitcher, I would only use 1200 innings, not 2000.

There was an article many years ago in SABR's The Baseball Research Journal that quantified an ERA advantage for relief pitchers. I don't feel like digging it up now, but I think it showed that impact of entering games in mid-inning made relievers ERA's .30 better than starters. Since that article, closers have largely stopped entering games in mid-inning, so this ERA advantage is probably less.

This phenomenon is something I think needs to be considered in the study of relievers.

As I've noted before, a big advantage I see with today's one-inning closers is that they often simply can't have a whopper of a bad game.

3-2 lead, top 9, Lidge allows two singles, a walk, and a ball in the gap - and he walks off the mound a 4-3 loser.
Yes, 2 ER, 0.0 IP is bad.
BUT if it wasn't the 9th and it wasn't this situation: the ball in the gap scores a 3rd ER, and now that guy is on 2nd or 3rd with no out and is likely to score as well.

That's not just a one-time thing - the closer usually walks into the game with a 2 ER or 3 ER maximum, regardless of how dreadful he is. And with the responsibility for only one inning, he doesn't have to pace himself like other pitchers.

I agree wholeheartedly that we will need to compare relievers to other relievers in their own eras, and see who is best from the group. The ERA+ expectations, I assume, will rise with each era.

With relievers, among the easily available data, I tend to look at opponents' BA and OBA more than with starters where runs are the thing.

I think for Wilhelm and subsequent relievers I'm going to add 50% to the innings and subtract 10 points from the ERA+. That gives Hoyt 3300IP and a 136 ERA+, which is a slam dunk. I'll probably do a 50% correction (+25% to innings minus 5 ERA+) for Eckersley, which makes him 4100/111, somewhere below Rixey around the bottom of the ballot. Gossage becomes 2700/116, probably not, Smith 1900/122 (borderline) Franco 1900/127 (probably) Rivera (so far) 1200/187 -- needs to pitch till 45 to get the career length needed. This is quick and dirty, and rejects most career relievers, which I think is right. If Rivera retires soon, he becomes a sort of ultra-Koufax, which I'd regard as borderline.

Add in postseason, even without weighting them, and Rivera (adjusted) becomes 1400/ about 190. Gets tough to keep him out. We will need to add postseason for modern pitchers, especially relievers, as it becomes a significant part of the total (was about 8% even for Caruthers, IIRC, but the numbers get much bigger when you have multiple postseason series, as post 1969 (surprisngly, Gossage postseason only 3% of total; Fingers would presumably be more.)

Also, add 1/3 of saves to wins, makes Hoyt 219-122, Gossage 227-107 and Smith 227-92, Franco 231-87. The 700 WPct for Smith and Franco are cuckoo, but this pushes relievers in general slightly towards the ballot. Rivera becomes 180-35, still not quite enough career but superlative quality.

Leverage, IHR, BQR, and DERA are all BP stats, and getable.

The main problem with relief innings viz ERA is that relievers enter the game throwing at maximum effort, unlike starters who have to pace themselves a little more. This should show up as an ERA advantage for relievers.

Doc, yes it does, about 0.30 according to the study DanG cited.

This of course assumes entering mid-inning and so does not apply to the modern closer. There are constraints (a ceiling) that works to the closer's advantage as well, but I'm not aware it's ever been quantified.

Comparing starters and relievers has no more in common than a deadball catcher versus a 1990s cornerman. I don't see how it won't require a lot of the dreaded subjectivity.

0.3 would mean a 4.2 ERA in a 4.5 environment or a 107 ERA+, correct?

yes, DL.

I think Doc's point about "relievers enter the game throwing at maximum effort, unlike starters who have to pace themselves a little more" is different. We might find that guys who have been both RP and SP (Smoltz/Eck/others) allow a lower ERA (and WHIP, etc) when relieving simply because of the 'full steam' approach they can use when relieving. The trick is IDing and eliminating any selection effects; that maybe this applies to only certain types of pitchers.

I think Doc's point about "relievers enter the game throwing at maximum effort, unlike starters who have to pace themselves a little more" . . . maybe this applies to only certain types of pitchers.

Probably wouldn't apply to knuckleballers, for instance. . .

A comparison of two pitchers: John Wetteland and Mariano Rivera. For Wetteland, I've taken an 8-year stretch, 1992-1999. That leaves off his first three years as a sometimes-injured starter/reliever, and his last year, out of two final years of somewhat declining effectiveness. For Mariano, a 10-year stretch, 1996-2005. That leaves out only a less-used rookie year, but includes his fabuluous year as a setup man in 1996 (the year the led Wetteland to seek employment elsewhere.) Rivera hasn't had a decline, yet.

Wetteland has 550.1 IP, or 68.6 IP/season for 8 years.
Rivera has 739.2 IP, or 74.0 IP/season for 10 years.

The following numbers are what they allowed per 9 innings. "Net BB" is BB - IBB + HBP. None of this is park-adjusted. There are some contradictory trends: Rivera pitching entirely in Yankee Stadium, Wetteland getting to pitch in the lower-scoring NL and the lower-scoring 1992, but also in high-scoring Texas. Overall, I'd say the effects are not huge.


Comparing the two, Rivera has an advantage in control and a strong ability to prevent the HR, but Wetteland's advantage in strikouts closes at least part of that gap.

Strikeouts are a defining feature of the top closers of our times, and Rivera's strikeouts are on the relatively low side by those standards. There's the intriguing possibility that after one blazing year (1996) he intentionally chose a style that resulted in fewer strikeouts, but more weakly-hit balls.

I have no problem with the idea that Rivera is the best of the bunch in the era of the closer (since about 1990). But how large is the margin by which he sticks up above the others from that time?

I have no problem with the idea that Rivera is the best of the bunch in the era of the closer (since about 1990). But how large is the margin by which he sticks up above the others from that time?

Depends on how much credit you give for him in the postseason (which is going to be extremely hard to ignore) and what his career numbers are when he retires. Personally, that margin looks pretty sizeable already.

Daryn - in the Wilhelm thread: "Given the small gradations in value of the current backlog, which will be even smaller 30 years from now, I expect that we will see an almost binary ranking of closers by the electorate -- they will either fall at the top of the ballot or not in the top 50. I just don't see someone saying that Lee Smith places 13th on their ballot -- I think it will be top 3 or not at all."

I can see Eckersley at a mid-ballot slot due to his hybrid career. John Franco might end up as a the Jake Beckley of relievers.

Just for grins (and knowing that you and I have no consensus on value) I ran Lee Smith through my system and plopped him on the 1978 ballot; he ranked 16th.

Here's where the top relievers would end up on my current ballot assuming they were all eligible now. I didn't consider any postseason credit in this initial analysis. This sanity check gives me some confidence I'm ranking Wilhelm correctly.

Eckersley - 2nd (hybrid)
Wilhelm - 3rd
Mariano Rivera - 6th
Goose Gossage - 9th
Lee Smith - 19
John Franco - 29
Trevor Hoffman - 30
Rollie Fingers - 43
---------
Tom Henke - 77
Billy Wagner - 78
Kent Tekulve - 82
Ellis Kinder - 83
Bruce Sutter - 89
Dan Quisenberry - 91
Sparky Lyle - 94

Eckersley - 2nd (hybrid)
Wilhelm - 3rd
Mariano Rivera - 6th
Goose Gossage - 9th
Lee Smith - 19
John Franco - 29
Trevor Hoffman - 30
Rollie Fingers - 43
---------
Tom Henke - 77
Billy Wagner - 78
Kent Tekulve - 82
Ellis Kinder - 83
Bruce Sutter - 89
Dan Quisenberry - 91
Sparky Lyle - 94


So, by the time they are up for election, that is 4 guys at the very top of the ballot, 4 tweeners and 8 nowhere near. Of your 4 tweeners, I think Franco and Hoffman will probably fall off many electors' radars, and Smith and Fingers may actually end up being tweeners. I don't like Fingers much, but others might. Man was he overrated at his peak. Must've been the 'stache.

Franco and Hoffman will fall off other voters ballots but Sutter and Quisenberry will get more support from other voters. Add in Fingers (who is probably a tweener for everyone) and there are about 6 guys who will hang around in the backlog.

Wetteland and Reardon slot between Kinder and Sutter

Talking about Wetteland versus Rivera. How many outs are each getting per appearance? I saw the difference of a small number of innings per year, but I'd be curious if the number of actual outs seemed bigger than the innings. And of course there's the matter of the leverage of those outs.

I also agree with someone who said above that ERA(+) is essentially flushable with relievers. If the <u>typical</u> reliever, say Bob McClure, gets a .30 ERA boost, what is your top bullpen dawg getting? Closers frequently have sparkly ERAs, not just a 4.20 in a 4.50 league, but closer to 3.00 than 4.00. Trick is that a HOMABLE closer/ace is going to be even better yet than that!

So to even get into the discussion, a reliever's probably going to be posting ERAs that are 65-75% of the league---and that's for openers. Obviously I'm just opining, no data in support, but I don't think I'm WAGging it too badly. Ron Shandler, for instance, warns fantasy players to avoid elite closers, and he defines elite as having really tremendous ERAs and peripherals.

All of which is leading me to say that by looking at ERA, we'll miss the point. Relieving is about getting out of tough jams or getting three to six outs. ERA's a non-starter (so to speak), we've got to think elsewise about this.

Doc, I don't follow what you're saying here. It sounds to me like you misunderstood this:

>I also agree with someone who said above that ERA(+) is essentially flushable with relievers. If the typical reliever, say Bob McClure, gets a .30 ERA boost, what is your top bullpen dawg getting? Closers frequently have sparkly ERAs, not just a 4.20 in a 4.50 league, but closer to 3.00 than 4.00. Trick is that a HOMABLE closer/ace is going to be even better yet than that!

The point is that a reliever who routinely enters the game mid-inning will have an ERA 0.30 below a starting pitcher with the same WHIP and/or other measures, essentially the same performance. The reason being that the first run or two or three that scores is attributed to the starter.

Closer A might be at 4.20 versus a starter with the same WHIP at 4.50. Closer B might indeed be at 3.00, but the point is that a starter with the same WHIP would likely be at 3.30.

No way does a reliever get "relief" all the way from 4.50 to 4 or 3.

And a closer or other reliever who almost never enters a game in mid-inning gets no advantage whatsoever.

So I'd be careful about docking relievers unless we know what their usage is. Or just use WHIP of OAV or OOB.

OK . . . here's what I'm thinking about for my system:

Do we have retroactive leverage indexes for the guys 1960+ from retrosheet? That would be ideal.

Short of that, I think I'm going to go with what Bill James did in Win Shares. He gives "Save Equivalent Innings", up to 90% of the innings the pitcher pitched. The formula is simple, add (3*SV)+Holds to each pitchers IP, with a max of .9*IP for "leverage credit".

It's crude but easy. I think it's pretty close, as no relievers LI can be higher than 1.9 based on this.

I could see docking RA by .20 or so for relievers, but I think I'd rather just use something like (RA*.5) + (ERC*.5) where ERC is component ERA.

Any thoughts?

I realize that ERC would have to be normalized to the league average - that is RA is going to be at least 10% higher than ERC, because RA includes the unearned runs. I'll take ERC, compare it league average and then adjust to a 'runs scale' based on win expectancy at that level - if I go that route. Just wanted to clarify that.

Of course if I'm going to give save equivalent innings bonuses for relievers, the starters that relieved occasionally will get them as well.

Here's another issue - I'm using my own version of translated IP for starters, to equalize the eras.

Do I need to do something like that for relievers too?

Maybe normalize to the where the top 5 in relief innings = the seasonal standard? Or innings of the top 5 in saves that year? Obviously normalizing relievers along the starters scale doesn't make much sense.

I guess my options are:

1) Use straight innings for relievers don't normalize at all.

2) Normalize relievers to where the IP of the top X (relative to league size) in Saves are the standard for that season.

3) Normalize relievers to where the top pitchers in relief innings are the standard for the season.

Of course, I'd tack the leverage bonus onto whatever final translated IP number I end up using.

Any thoughts as to what is best?

Jesse Orosco... c'mon

A problem with normalizing IP for relievers is that as the typical workload has decreased, the saves per appearance has gone up, and elite ERAs seem to be growing a bit. Any lev index (LI) that uses saves will show modern closers with higher LI, probably balancing out the lower IP.
So, I would use Joe's option 1, no normalization.

a few notes I just found at bb-ref.com:

the very first 40-save season was Quiz in 1983.

since then, we've had at least one 40-save pitcher EVERY YEAR (except for the strike season of 94). That's a pretty stark pattern!

>40 saves . . . . . . . . . . . . .. ____________
<40 saves _______________|
--------------------------time--------------------

was the save rule changed in 83?

also, from 1983 to 1991, a different man led in saves each year.

OK, so a while back I played around with the 1972-2005 relief numbers on BP. I was trying to figure out some assessment tools for relievers. The idea was to take the relievers' IHR, BQR, effectiveness, durability, etc, and plot them versus the league average in the same way we do with OPS (though I didn't park adjust). The next few posts show the results of these attempts.

IHRs+
Adjusted inherited runners scored (IHRs+) shows how many IHRs a RP had compared to his own leagues. Computed by dividing IHRs/IHR then normalized against the league average. Here the leaders among the key relievers of the period (NOTE: not all relief innings of all these relievers careers are accounted for due to the 1972 cutoff in retrosheet data at the time I was putting this together.) No guarantees of total accuracy or theoretical accuracy, it's just an idea for a stat I was trying out and thought I'd share.

IHRs+ LEADERS

RANK NAME IHRs+
---------------------
1 hoffman 1.39
2 myers 1.30
3 wagner 1.30
4 orosco 1.26
5 percival 1.24
6 smith l 1.23
7 eckersley 1.21
8 fingers 1.19
9 tekulve 1.19
10 righetti 1.15
11 benitez 1.14
12 nen 1.12
13 campbell 1.12
14 beck 1.11
15 aguilera 1.11
16 sutter 1.11
17 reardon 1.11
18 kern 1.09
19 rivera 1.09
20 hiller 1.09
21 henke 1.08
22 hrabosky 1.07
23 mcgraw 1.06
24 giusti 1.05
25 worrell 1.05
26 wetteland 1.05
27 smith d 1.04
28 hernandez w 1.04
29 mcdowell 1.04
30 marshall 1.04
31 gordon 1.03
32 franco 1.03
33 gossage 1.02
34 lyle 1.00
35 hoerner 0.99
36 garber 0.99
37 eichorn 0.97
38 hernandez r 0.96
39 stanley 0.96
40 carroll 0.96
41 granger 0.94
42 olson 0.93
43 jones d 0.91
44 quisenberry 0.89
45 montgomery 0.85

The big surprise to me is how poorly Quiz does. Wilhelm is not included in the group because there were only two seasons of his career in the retrosheet data at that time. More info to follow.

Leverage

This is simply the weighted annual average of the leverage index of these same 45 relievers.

LEVERAGE LEADERS

RANK NAME        LEVERAGE INDEX
---------------------------------
1    percival    1.93
2    hoffman     1.89
3    worrell     1.83
4    sutter      1.80
5    wagner      1.77
6    wetteland   1.74
7    franco      1.73
8    myers       1.69
9    smith l     1.68
10   rivera      1.65
11   eckersley   1.64
12   fingers     1.64
13   benitez     1.62
14   reardon     1.61
15   nen         1.61
16   hernandez r 1.61
17   aguilera    1.60
18   smith d     1.59
19   righetti    1.58
20   montgomery  1.56
21   henke       1.54
22   beck        1.52
23   hiller      1.50
24   gossage     1.50
25   jones       1.49
26   marshall    1.47
27   giusti      1.46
28   hrabosky    1.46
29   mcdowell    1.45
30   kern        1.43
31   orosco      1.43
32   mcgraw      1.42
33   gordon      1.41
34   olson       1.40
35   tekulve     1.39
36   garber      1.35
37   carroll     1.35
38   quisenberry 1.33
39   stanley     1.33
40   granger     1.32
41   hoerner     1.31
42   lyle        1.31
43   hernandez w 1.26
44   campbell    1.24
45   eichorn     1.08




Seems like there's a lot more closer-era guys toward the top and a lot more ace-era guys toward the bottom. Because leverage is already indexed to 1.00 I didn't adjust it for league usage patterns. I suspect I should have.
		

That's interesting that I'm probably underrating Trevor Hoffman. Add two more typical closer seasons from Hoffman and he's in my PHoM without considering leverage or postseason performance. With 2 more seasons he'll probably have the saves record. His 2006 is looking good so far.

Well I've screwed up the formatting for the first two lists....

One of hte things I wanted to know about releivers was how durable they were. Or to put it another way: how often were they available to pitch; how often would they answer the bell? For a reliever to be truly great, he's got to be able to pitch effectively more than twice a week....

So, APP+, or indexed appearances. It's actually kind of a misnomer, but bear with me. For each season I took a sample of the relievers in the league with most appearances whose cutoff for inclusion was based on the formula 4 x (# of teams), figuring the average bullpen in each year 1972-2005 to be 4 pitchers deep, first of all, and second that anything after the fourth guy is probably a tenusous slot. Coulda done top three, but I'm not sure it makes a huge difference or not. Anyway, for each season I took the number of appearances by the average fourth man in each bullpen as the basis for comparison. Then I indexed our 45 guys against that threshold to see how much more durable a guy was than other pitchers in a typical bullpen each in his career. This represents (I hope) how often he answered the bell compared to his compadres.



Not surprisingly, this time the moderns go to the floor, reflecting the decentralization of appearances in the modern pen as well as the "save-him-for-the-ninth" mentality, which inhibts expansive usage of an ace.

OK, last one coming up.

Ugh, I'm just having a bad formatting day. Sorry everyone.

So one more thing. We've looked at three things so far crucial to relievers
1) How well he cleans up messes (IHRs+)
2) How tough the messes are (LEV)
3) How often he's ready to pitch (APP+)

The only remaining issue is his general effectiveness. I wanted to try something a little different with effectiveness. What's most important to relievers is keeping guys off base and most important avoiding homers. I don't care how the do it (inducing DPs versus inducing strikeouts). So my homegrown idea for rendering effectivness is really a variation on BFP/INN:

(BFP - IBB) + (HR*2) / INN

This way the pitcher is not penalized for IBB (which relievers do a lot more often than starters it seems) but they are penalized more for homers since in higher leverage situations a homer is likely to be costlier than it is for a SP. Naturally, I compared each pitcher's "effectiveness" to his league (I did include starters in this) to create an index. As a note, "effectiveness" figures for the league have gone from 4.45 to almost 4.9 during the 1972-2005 period.



Seems to me the message here is fourfold:
1) This measure creates a narrow band of results
2) Within that narrow band, most aces fall within a few points of each other
3) But Rivera is obviously great by this measure, presumably by dint of his mastery of the strikezone and his tendency to keep the ball in the park.
4) Also, that contemporary bullpen aces are at the top of the list DESPITE playing in a homer-happy era.

OK, that's it. Hope some of this is useful or at least not meaningless junk stats.

One coment that I will make is that if you are a careerist, there's an argument that we've never seen a modern style reliever have the equivalent of a 300 win career or a 3000 hit career-either 15 great seasons or 20 very good seasons. Such a career reliever would likely have piled up close to 600 saves, for example. I realize that this is a tremedous requirement to put on a closer, but it's good for a rough sense of perspective.

I personally feel like Mariano Rivera's current career should be valued somewhere between Koufax and Pedro Martinez if he retired today-in the HOM, but nowhere near being a unanimous selection. I also feel like Rivera has the potential to add a lot to his already solid career (give Mo 5 more years even with some decline, and the careerist in me would be quite happy as well). If I was a HOM voter, I'd have a tough time adding any other pure modern closer to the HOM based on their current career (Eck, Smoltz are special cases that I still have issues with, Wilhelm isn't a modern closer but should be in, and Hoffman, Wagner, and Percival could get in by adding another 3-6 years at their current rate stats with Hoffman clearly being the closest to being in).

thok,

I think this may be a bit unfair to relievers because their margin for error is too narrow. Because relievers are mostly fungible, the threshold that management has for poor performance is very low. Closers are swapped out all the time, setup men change almost yearly. Blow a couple high-leverage games, and your role changes very quickly.

As a result of this, to keep the role of high-leverage reliever, a HOM-worthy closer must perform at a peak level of performance without interruption (unless he's extraordinarily well established). Whereas a team will wait out injuries from a third baseman, injuries to closers often dislodge them from their role. Look at Keith Foulke for an example of all these things. Jerry Manuel's caprice took Foulke out of the high-leverage role, and Foulke's own owies in 2005 and 2006 have led to his job being taken by young Papelbon.

So what I mean to say is that closers with long strings of uinterrupted, outstanding performance in the same role are probably retaining their role in a more difficult situation than many starters or position players who are given many more chances to fail. Therefore, I'm tempted to say that Rivera is more like Pedro or Randy than he is Koufax. Pedro and Randy now both own long primes full of super-effective seasons, whereas Koufax had "merely" five or six such years.

We might draw up a comparison like this in terms of relative career lengths:

Rivera = Pedro
Goose = Tom Seaver
Quis/Sutter = Jim Palmer
Wilhelm = Cy Young???
John Franco = Jim Kaat
John Wetteland = David Cone

That's a fun game, but it shouldn't be interpreted too literally.

I think this may be a bit unfair to relievers because their margin for error is too narrow. Because relievers are mostly fungible, the threshold that management has for poor performance is very low. Closers are swapped out all the time, setup men change almost yearly. Blow a couple high-leverage games, and your role changes very quickly ... So what I mean to say is that closers with long strings of uninterrupted, outstanding performance in the same role are probably retaining their role in a more difficult situation than many starters or position players who are given many more chances to fail.

But isn't the telling point here the fact that relievers are mostly fungible? Closers are indeed swapped out all the time, because their teams can do it. Starters or position players aren't, because their teams can't, which is another way of saying the task demanded of the starter or the position player is more difficult to perform.

Therefore, it isn't unfair to relievers to dock them for not attaining the career length of others. It's completely fair; it's a reasonable consideration of the empirical fact that relievers are performing a role in which replacements are more easily found. The greater career value delivered by others is real, and thus relievers shouldn't be given an adjustment to make it go away.

Before I attach any significance to Quiz' IHR+ I'd like to know what kind of sample size we're talking about here--not just Quiz but all of these guys. Modern closers just don't inherit anybody. If he inherited one and he scored, that makes him infinitely bad?

>contemporary bullpen aces are at the top of the list DESPITE playing in a homer-happy era.

And BECAUSE there's no decline yet?

Steve T,

You're right, of course, in one sense. But a HOMworthy closer is not fungible, and there are probably degrees of fungibility anyway. If the fifth best guy in your pen is fungible, is your #1? Or you're #2? I dunno. Depends on your #1 and #2 I guess....

But a HOMworthy closer is not fungible, and there are probably degrees of fungibility anyway.

Certainly there are degrees of fungibility. But the very fact that even HOM-worthy closers have rarely if ever achieved the career lengths of HOM-worthy starters and position players is a demonstration that even HOM-worthy closers are more fungible than the best players in other roles.

Closers are indeed swapped out all the time, because their teams can do it. Starters or position players aren't, because their teams can't, which is another way of saying the task demanded of the starter or the position player is more difficult to perform.

That's not necessarily the case. You take the position that there is a low tolerance for error because anyone can do the job. I say there is a low tolerance for error because the job is so important, and equally importantly, it appears to be so important that the consequences of failure are job threatening.

Hall of Merit hitters have down years and either keep their jobs without improving for several years or bounce back the next year -- they are allowed to do that, I posit, because their failures are masked by their 7 or 8 other hitting colleagues and the fact that hitting is not the only aspect of their job.

There are hundreds or examples of great players having down years, really bad years, and recovering. Closers are not allowed to do this. That makes the ten year closer equivalent to the 20 year hitter, in my view.

Sunny,

Happy to oblige! These are the IHR/IHRS sums for our group of 45 relivers (plus a couple others)based on the IHR totals on BP for 1972-2005. The grand total only includes this subgroup's IHR.



In addition, here's the yearly MLB average of IHRs for the whole period:

Interesting to note that both Wagner and Rivera have been in the league 11 years, but that Wagner's got half the IHR that Rivera does. He's been awesome with IHR on too. Orosco and Myers were both awesome with guys on as well. Makes you wonder if their handedness helped them. Probably not since it didn't help Franco or Hernandez....

>hoffman 316 63 20%

My new hero.

You take the position that there is a low tolerance for error because anyone can do the job. I say there is a low tolerance for error because the job is so important, and equally importantly, it appears to be so important that the consequences of failure are job threatening.

Of course the closer's job is important. But is it more important than the #1 starter's, or the shortstop's, or the cleanup hitter's? Is the cost of crappy performance in any one of those roles less than crappy performance in the closer's role? I see no evidence whatsoever that such is the case.

Teams are often forced to tolerate dips in performance in those other roles, not because they aren't suffering the cost, but because their options in addressing the problems are so limited. When my closer falters, it's very likely I've got at least one other reliever in my bullpen (and/or starters in my rotation) I can quickly insert into the role and very likely get improved performance. When my center fielder goes into a slump, it's far less likely that I've got a solution to the problem on hand -- because the skillset of being able to play quality major league center field and hit well is so rare.

Hall of Merit hitters have down years and either keep their jobs without improving for several years or bounce back the next year -- they are allowed to do that, I posit, because their failures are masked by their 7 or 8 other hitting colleagues and the fact that hitting is not the only aspect of their job.

No one's failures are masked by the successes of others; they may be mitigated, but their failures are costing the team potential wins nonetheless. But it is absolutely the case that hitting is not the only aspect of their job, and that's kind of the whole point: they aren't narrow specialists, who need to do only one thing well to hold their job. Their role requires a breadth of talent, that isn't easily replaced.

There are hundreds or examples of great players having down years, really bad years, and recovering. Closers are not allowed to do this. That makes the ten year closer equivalent to the 20 year hitter, in my view.

No, it doesn't. It simply demonstrates how much rarer the skill required to perform a starting pitcher or position player job is than to perform the relief pitcher role. If a closer was as much better than his job competitors as a 20-year hitter is, then he would keep his job for 20 years. The fact that his team finds someone better to do the job after 10 years does not render that closer's 10-year career as valuable as the hitter's 20-year career. Half as long is still half as long; the 20-year guy has given his team twice as many seasons in which they haven't had to address that position.

No, it doesn't. It simply demonstrates how much rarer the skill required to perform a starting pitcher or position player job is than to perform the relief pitcher role.

Steve,

So I think what you're saying is that if relief pitchers have short careers then they must:
a) have a job that puts more strain on the body and causes quicker decline
or
b)be generally less talented than other pitchers, so their expiration comes more quickly and more implosively (i.e. no hang-on time, from prime straight to the bottom of poop chute.)

and that you say (b)...which is my answer too when I stop to think on it.

Here's the rub: I can accept (b) when I talk about Jeff Montgomery or Jesus Colome, but what does this say about guys with lengthy primes and/or careers such as Mariano, Hoyt, or Goose? Is it saying that they are really, really special in the universe of players? Or that they are really, really speical among relivers, but only as special as the most special #3 starter?

There are hundreds or examples of great players having down years, really bad years, and recovering. Closers are not allowed to do this. That makes the ten year closer equivalent to the 20 year hitter, in my view.

The new Radbourns and Clarksons?

>>And a closer or other reliever who almost never enters a game in mid-inning gets no advantage whatsoever.

>Wrong. There is an advantage in pitching shorter outings. The ERA advantage of relievers is not only due to partial innings, but having the opportunity to expend full effort on every pitch, and not facing the same batters multiple times.

Maybe.

But my point was: The study that attributes a 0.30 ERA advantage to relievers was based on inherited runners. Relievers who enter at the beginning of an inning don't have THAT advantage.

Maybe they have other advantages, but those other advantages haven't been quantified (and can't be) because there is no control.

>Teams are often forced to tolerate dips in performance in those other roles, not because they aren't suffering the cost, but because their options in addressing the problems are so limited. When my closer falters, it's very likely I've got at least one other reliever in my bullpen (and/or starters in my rotation) I can quickly insert into the role and very likely get improved performance.

How many ML teams have already switched closers this year and gotten an immediate improvement? Detroit, Oakland (to a committee), Seattle, Tampa, Texas; Atlanta (?), San Fran.

This doesn't answer the question why--my sense is both Steve and Doc are right. It's both because it's too important to screw around, and also because an option is already at hand.

As to the Doc's point--Seattle is still giving Richie Sexson a chance to get his #### together, while Everyday Eddie is already every-other-day; and the Giants are still letting some guy named Bonds work out the kinks while Worrell is out. Of course, in Texas both Nevin and Cordero are out.

So I think what you're saying is that if relief pitchers have short careers then they must:
a) have a job that puts more strain on the body and causes quicker decline
or
b)be generally less talented than other pitchers, so their expiration comes more quickly and more implosively (i.e. no hang-on time, from prime straight to the bottom of poop chute.)

and that you say (b)...which is my answer too when I stop to think on it.

Well, (a) is almost certainly true too, at least when compared with first basemen or corner outfielders. But I see no evidence that relief pitching is harder on the body than starting pitching is, and it probably isn't all that much harder on the body than catching or middle infielding, either.

Here's the rub: I can accept (b) when I talk about Jeff Montgomery or Jesus Colome, but what does this say about guys with lengthy primes and/or careers such as Mariano, Hoyt, or Goose? Is it saying that they are really, really special in the universe of players? Or that they are really, really speical among relivers, but only as special as the most special #3 starter?

But the whole point here is that virtually no closer actually has as lengthy a career as those we see from other players, isn't it? Hell, Rivera's only been in his job for a little over 10 years; Clemens and Maddux have been around twice as long.

And even Rivera, brilliant as he is, failed as a starter. As did just about every relief pitcher in the universe, including the greatest among them.

Virtually without exception, the very best young pitching prospects are deployed as starters, and they are kept as starters as long as they do well. Only if they falter as starters are they moved to the bullpen, where very frequently they thrive -- because relieving is less demanding than starting.

And if a relief pitcher struggles, he is out of a job. There is simply nowhere else he can go except to be demoted or released, and there are always plenty of other candidates who can be tried instead of him.

The number of great relief pitchers who later became successful starters is very small (Wilhelm and Wilbur Wood are among the very few), while the number of struggling starters who became successful relievers is very high (including almost all of the greatest relievers in history). Relief pitching is the easiest of pitching assignments, and thus the one in which talent is easiest to find.

So to now look at the question from another vantage point
---------------HOM starter----------------------------




---------------successful starter---------------------


---------------failed starter-------------------------
---------------reliever-------------------------------

Where does successful reliever fit on this scale? Or HOM reliever? What would a reliver need to do to get to the HOM starter level?

Relief pitching is the easiest of pitching assignments, and thus the one in which talent is easiest to find.

most MLB managers would not only disagree (at least when it comes to closers), but vociferously disagree.

of course, that doesn't mean Steve is wrong :)

When my center fielder goes into a slump, it's far less likely that I've got a solution to the problem on hand -- because the skillset of being able to play quality major league center field and hit well is so rare.

This is false. There are endless examples of rookies who are and prove to be better than veterans once given the chance. But when your great hitter struggles mid-career or at the end of his career (which may last several years, see Williams, B.) he is not replaced, but instead given an oppportunity to revert back to greatness while keeping his role. The solution is there but they just don't go to it because the urgency is not there.

There are literally countless times we here have correctly bemoaned the fact that a player who was ready to contribute was held back for months or years -- the same mindset, that position players can turn it around while failing closers must be immediately replaced, is at play in letting players have 20 year careers, but not letting closers have the same opportunity. The closers that make it to 10 straight years are special players who avoided a slump for an unheard of amount of time. The psychology of the closer and teams' reactions ot the closer role should not be ignored.


If a closer was as much better than his job competitors as a 20-year hitter is, then he would keep his job for 20 years.

Steve, you are not engaging in a disussion, you are just talking past me. This simply isn't true -- experience shows us that when closers falter they get replaced, and this doesn't happen to firstbasemen. You say it is fungibility, I say it is in large part a psychological reaction based on the importance of the job.

The number of great relief pitchers who later became successful starters is very small (Wilhelm and Wilbur Wood are among the very few), while the number of struggling starters who became successful relievers is very high (including almost all of the greatest relievers in history). Relief pitching is the easiest of pitching assignments, and thus the one in which talent is easiest to find.

This is generally true factually, but the conclusion you draw is erroneous. If relieving were so much easier than the great starters who become relievers would excel at an even greater level as closers. This didn't happen in the rare examples in the first 80 years of baseball history and it didn't happen with Smoltz -- he was about as good.

We are talking about outliers here -- sure there could be very few closers that would be great starters but 1) perhaps that doesn't matter and 2) there are outliers that can. If Rivera decided he wanted to start one year and went 17-9 with a 3.25 ERA would you suddenly say, now I believe he was great all those years. And, do you seriously believe, even with only the two pitches he has, he couldn't do it?

It is also an overstatement to call relievers failed starters. For the past 30 years, there have been two kinds of pitchers -- starters and relievers. Young pitchers with certain skillsets are shepherded into reliever roles, but most of the great closers were never given a real chance (say 2 years) to start in the majors. Are you convinced Gagne couldn't have harnessed his stuff to be a great starter? Rivera? I know I'm not. They find early success as closers and get pigeonholed -- becaue, and we come full circle, major league teams appreciate how rare and difficult it is to get a closer who can truly dominate and truly last.