Great discussion of evaluating relievers.

Tango uses his leverage index to give Sutter 1980 equivalent IP for his 1042 actual IP, and Gossage 2961 IP for his actual 1809 IP. WARP1 xip (based on decisions and saves) gives them 1536 and 2320 equivalent innings pitched, respectively.

So it looks like we should perhaps boost relief pitchers Warp1 by 28 to 29 percent (1980/1536 or 2961/2320). I had already been boosting them by 20 percent, for no particular reason other than it seemed about right. Any idea of what a similar ratio would look like for win shares? (For my pitcher evaluations I use an average of Warp1 and WS.)

Another consideration in comparing relievers to starting pitchers is that a starting pitchers innings may be concentrated in fewer seasons, giving them an advantage in peak value. For example, in 1934 Dean pitched 311 innings, going 30-7, and winning an MVP award that I regard as defensible. In contrast, I can't think of any season by a relief pitcher for which I would have supported an MVP award. I'd probably support a few Cy Young awards for relievers, but not an MVP.

I expect that we will support a few relief pitchers, but it will be interesting to see how it will sort itself out.

"But what if I said that we need to have a pinch hitter in the HoM? You'd say, well, so-and-so may have pinch hit .400 with an all-time high in pinch hits (I'm making this up) but I going 200 for 500 is just not enough career bulk, and I would have to agree."

I've thought of this myself. The difference to me is that no one is 'turned into' a PH, instead of being a regular because of a perceived advantage. There hasn't been one minor league star that I'm aware of, that was brought to the majors as a pinch-hitter.

There is perceived value in making a career as a reliever. People are paid salaries commensurate with some of the best players in the game to be relievers.

It's kind of like putting kickers or tight ends in the football Hall of Fame.

That's how I think of it anyway. I believe the best relievers of all-time should be in the Hall of Merit, even in if at first glance their IP totals appear to be a bit low. I also think the leverage of the innings they do pitch (generally accepted at around 1.5-2x 'starter' innings) should weigh into the argument.

But the leverage should be measured not againat 'replacement' for modern relievers, since if you lose your closer, you everybody else in the pen moves up a spot; the garbage man gets some middle relief time, the middle guy is a set-up man, the set-up man becomes the closer, and AAA pitcher gets the garbage innings with a leverage of 0.3 or something.

The 'leverage' of 1.5 to 2 should only be applied, IMHO, to runs saved above average, or even 'runs saved above a-little-better-than-average', which might take other overall LI down to 1.3 or so; higher for a guy with a super ERA+ like Mariano.

Brent wrote:

So it looks like we should perhaps boost relief pitchers Warp1 by 28 to 29 percent (1980/1536 or 2961/2320.

That Sutter and Gossage have about the same ratio between xip and leverage-adjusted innings here is probably due to chance. I would argue against simply applying a standard boost to relievers' WARP1.

Rather, we should try to get a sense, if we can, of how highly leveraged the reliever's usage was, check the actual xip/ip ratio used in WARP1, and then pro-rate accordingly.

Tom H wrote:

The 'leverage' of 1.5 to 2 should only be applied, IMHO, to runs saved above average, or even 'runs saved above a-little-better-than-average', which might take other overall LI down to 1.3 or so; higher for a guy with a super ERA+ like Mariano.

I don't see that replacement level for relievers is any different than replacement level for starters: if Randy Johnson goes down, many of the more highly leveraged innings that he would have pitched are going to go to the set-up man and the closer, while the replacement starter manages the first five innings or so: there's still going to be a trickle-down effect through the staff.

Anyway, the basic way in which WARP1 is calculated is to multiply DERA by xip to get runs saved above average: then the difference between pitching runs above average and pitching runs above replacement is added in. If what you care about is runs saved above average, just use xip and DERA and don't bother about replacement level. The point is, the leverage index should be applied to innings pitched, not to the measure of effectiveness, because the relief pitcher's overall ip is being held down by the team in order to make sure that the pitcher is available at maximum efficiency in highly leveraged situations, so it is the pitcher's ip that need to be adjusted to accurately reflect the pitcher's value to the team, which is what we need to measure in order to assess the pitcher's merit.

That's also why, although we recognize that relief aces and closers are used in ways that maximize their ERA+, we shouldn't dock that ERA+ in calculating value because the main strategic point of relief pitcher usage is to maximize ERA+ in highly leveraged situations. Because relief aces are given the opportunity to maximize their ERA+, it would be foolish simply to rank relievers versus starters on the basis of ERA+, but nobody here is going to do that, I expect.

The big RP advantage for me is this:
You enter with a 3-2 lead, allow single, single, walk, double.
Ballgame over, second and third, none out. For a starter, those fellas likely will score as well. But you are 'capped' at just allowing the two runs.
Extreme example, yes. But basically when you stink, you can't stink as much as an SP can.

I'm, for whatever reason, interested in the question of "How many relievers?" Here's a quick-n-dirty start to thinking about how many relievers might end up in the HOM.

1) There's around 250 league seasons in history.
2) Relievers have been prominent since, roughly, 1950.
3) So relievers have been seriously used in 110 or so of the 250 big-league seasons, 44%.
4) If the HOM should ideally include about 30% pitchers, and we're electing about 220 men by 2007, then it will include 65-70 pitchers.
5) Of those pitchers, all else being equal (and it's probably not), the HOM would induct about 30 pitchers from the post-1950 era.
6) Relievers account for about 25% of innings (roughly 2 per game).
7) 25% of 30 HOM pitchers post-1950 is 7.5.
8) BUT a HOM starting pitcher would be better than an average pitcher, so they probably are consuming more innings. Let's say that in a HOM starter's game, there's more like 1.5 innings of relief...three-fifths of 7.5 is 4.5 relievers.

So 4.5 relievers seems like a pretty reasonable starting point. Is it the "right" general total to work toward? Or is it too high?

As Chris said, The replacement level for relievers IS the same as starters.

But if we use 'leverage index', THAT should NOT be applied against 'replacement level', because of what I wrote previously. Agreed?

So 4.5 relievers seems like a pretty reasonable starting point. Is it the "right" general total to work toward? Or is it too high?

That's about the number I came to by following a similar line of reasoning, with a similar "high-end' estimate of 7.5.

Great minds, think alike...

Which is a scary thought for you, Chris, but very heartening for me!

; )

I'd suggest that replacement level for relievers is substantially higher than for starters. Not many guys can throw even 5 IP in the bigs, much less go through a batting order a third time and go 7-8 IP.

Just think about going out into the market (trade, free agents, whatever) to pick up a pitcher this winter. What are the odds of picking up a reliever, typical workload 1 IP in each of 50 games, who will compile a 3.00 ERA for you? Not bad. What are the odds of picking up a starter who will even go 5 IP in each of 30 games with ERA 4.00? Not good.

Like any pitcher, there's at least two dimensions to relievers:
1) usage
2) performance

Usage is captured in a combination of leverage index, saves, innings, and, perhaps, games finished.

Performance, on the other hand, requires different lenses than we're used to. Saves and blown saves might be in the mix, ERA and ERA+ probably are, and inherited runners and inherited runners scored probably should be as well.

Anyway, I decided to take a quick look at a group of about 60 of the best-known relievers to see what their mean ERAs were to see how much these stud relievers typically outdistanced their leagues by. All numbers are through 2004 only and DO include starting numbers.

The are the relievers I looked at:

abernathy
aguilera
beck
benitez
campbell b
carroll
eckersley
eichorn
face
fingers
fisher e
franco
garber
gordon t
gossage
granger
henke
hernandez r
hernandez w
hiller
hoerner
hoffman
hrabosky
jones d
kern
kinder
labine
lyle
marbury
marhsall m
mcdaniel
mcdowell
mcgraw t
mcmahon d
miller s
montgomery
murphy j
myers r
nen
olson g
orosco
page j
percival
perranoski
quisenberry
radatz
reardon
regan
righetti
rivera
smith d
smith l
stanley b
sutter
tekulve
wagner
wetteland
wilhelm
wood
worrell

Their mean career ERA+ was 126. Without starting innings, I'd imagine it would be closer to 130.

But that's not the whole story. I decided to break the group down by era this way:
pre-1970s firemen
1970s relief ace
1980s transitional relievers
1990s closers

Obviously there's overlap, but I did my best, and here's how it came out

pre-1970s firemen
name        era+
----------------
wilhelm    146
kinder      124
perranoski  123
radatz      122
hoerner     120
mcmahon d   119
murphy j    116
marbury     116
miller s    115
labine      111
mcdaniel    109
face        109
abernathy   106
page j      106
fisher e    101
regan        97
================
group mean  115

1970s relief aces
name        era+
----------------
hiller      134
tekulve     132
lyle        127
gossage     126
hrabosky    121
carroll     120
fingers     119
marhsall m  118
mcgraw t    116
kern        115
granger     113
wood        113
campbell b  111
================
group mean  120

1980s transitional
name        era+
----------------
quisenberry 146
eichorn    142
sutter      136
smith l     132
smith d     130
orosco      125
reardon     121
stanley b   118
hernandez w 118
garber      117
righetti    114
mcdowell    114
================
group mean  126

1990s closers
name        era+
----------------
rivera      190
wagner      169
henke       156
percival    156
benitez     152
hoffman     148
wetteland   148
franco      139
nen         138
montgomery  134
hernandez r 134
jones d     130
beck        124
worrell     122
myers r     122
olson g     122
aguilera    117
eckersley   116
gordon t    112
================
group mean  138

There's a pretty definite trend here toward increasing ERA+. Whether this is a result of people being converted to relief instead of banished to the bullpen or what, I don't know (there's probably a ton of little factors going on). Anyway, but Mariano's 190 is a little less staggering in context (thought it's still amazing!).

One more list. Same groups as above, this time, I'm listing the percentage difference between the individual pitcher's ERA+ and the group's mean ERA+.

pre-1970s
name    % of group
------------------
wilhelm     127
kinder      108
perranoski  107
radatz      106
hoerner     104
mcmahon d   103
murphy j    101
marbury     101
miller s    100
labine       97
mcdaniel     95
face         95
abernathy    92
page j       92
fisher e     88
regan        84

1970s firemen
name    % of group
------------------
hiller    111
tekulve    110
lyle        105
gossage     105
hraboski    101
carroll     100
fingers      99
marhsall m   98
mcgraw t     96
kern         96
granger      94
wood         94
campbell b   92

1980s transitionals
name    % of group
------------------
quisenberry 116
eichorn    113
sutter    108
smith l    105
smith d    103
orosco       99
reardon      96
stanley b    94
hernandez w  94
garber       93
righetti     90
mcdowell     90

90s closers
name    % of group
------------------
rivera    137
wagner    122
henke    113
percival    113
benitez    110
hoffman    107
wetteland   107
franco      100
nen         100
montgomery   97
hernandez r  97
jones d     94
beck         90
worrell      88
myers r      88
olson g     88
aguilera     85
eckersley    84
gordon t     81

Rivera and Wilhelm truly stand out from their peers looking this way. After them, Wagner seems to form his own second tier, then there's a large clump of All-Star relievers including Hiller, Tekulve, Quiz, Eichorn, Henke, Percival, and Benitez, all of whose ERA+s are 10% above their peers'.

Now, take all this with a grain of salt because guys like Eck, Gordon, Gossage, Fingers, Miller, Kinder, Wilhelm, Wood and others have lots of starting pitching in there to make things a little messy. My guess would be that the removal of their starting innings would lead to an upping of ERA, and a subsequent narrowing of the spread of the % of ERA+ tables. i think Retrosheet has that data for some guys, so someday I'll see what it can tell me.

That's great info Dr C. My personal HoM will have closer to 10 relievers than 5. I look at those lists and see more than five no-brainers. That said, I think the Hall should have 75+ pitchers, not 65-70. I can see my 2005 ballot having 10 to 12 pitchers on it as I diverge from the group. I already have Welch, Rixey, Ruffing, Grimes, Redding, Waddell and Griffith; and I am adding Ferrell next week. So that's eight. My goodness.

All-Star team for 2004 (60 selections, using WARP)
BOLD Top-7; CAPS Top-30; (lowercase) within 10%

DH Hafner Ortiz
RF GUERRERO DREW ABREU SUZUKI BERKMAN Sheffield
CF EDMONDS BELTRAN Kotsay Damon
LF BONDS CLee Ramirez Matsui (dunn)
SS TEJADA GUILLEN Jeter (wilson)
TB BELTRE ROLEN ARODRIGUEZ Chavez Lowell Mora
SB LORETTA Kent
FB PUJOLS HELTON (overbay teixeira wilkerson delgado)
XX Kendall IRodriguez Posada (lopez varitek)
RP RIVERA LIDGE Cordero Benitez Foulke Nathan Gagne
SP SANTANA RJOHNSON SCHILLING SHEETS LHERNANDEZ
-- CZAMBRANO MARTINEZ SCHMIDT WESTBROOK PAVANO
-- CLEMENS OSWALT Garcia Peavy OPerez Buehrle Escobar
-- Radke Glavine Drese (lilly davis rogers greinke sabathia)

Rivera ranked 19th on last year's combined list. A 30% boost would move him up to 3rd (behind Bonds and Santana; ahead of RJohnson, Schilling, Pujols, and Edmonds).

Was his season last year THAT good?

Best Pre-closer/fireman era relievers:

1910's
Doc Crandall

1920's
Firpo Marberry

1930's
Johnny Murphy
Clint Brown

1940's
Al Brazle
Al Benton

1950's
Bobby Shantz
Ellis Kinder

I want to point out that we shouldn't merely look at total league seasons when figuring out how many closers there will be. There were more teams, and hence more players, in the modern era. There will be more electees from the expansion era than from the 1880's. So a little more wieght should probably be put upon the era of the closer and the tweener, adding maybe another reliever or two. I think that 5-7 sounds about right.

On another note, what does everyone think of Trevor Hoffman. Seems to me that he was better than Smith, reardon, Franco and some other high saves non no-brainer guys. Right now I would expect to have him 2012 (or whatever) ballot. If I don't, I think my brother would disown me anyways. There are just some things more important than getting the HOM right. ;-)

jimd wrote:

Rivera ranked 19th on last year's combined list. A 30% boost would move him up to 3rd (behind Bonds and Santana; ahead of RJohnson, Schilling, Pujols, and Edmonds).

Was his season last year THAT good?

In a word, no. Rivera's 2004 WARP1 total provides a good example of why a standard 30% boost to relievers is inappropriate. The WARP xip formula comes much closer to capturing the leveraged value of the post-1990 closer than it does for earlier styles of relievers.

In 2004, Rivera actually pitched 78.7 innings. xip credits him with 136.8 innings -- what we might call a leverage index of 1.74. Since 2.00 is the theoretical maximum, the most that Rivera's 2004 WARP1 should be increased is 15%, and I would think that 5-10% would be a more appropriate amount, as Tangotiger has shown that Rivera's leverage index has generally been between 1.75 and 1.9.

TomH (# 98) wrote:

when we begin analyzing relievers, we'll also have to remember that it's easier to put up a gaudy ERA (or ERA+) as a reliever than a starter. Part of this is the 'fewer innings per outing' thing (no 'pacing' required), and some of it is the inherent reliever ERA advantage of not being responsible for inherited runnners on base when entering mid-inning.

Howie Menckel (# 105) wrote:

The big RP advantage for me is this:
You enter with a 3-2 lead, allow single, single, walk, double.
Ballgame over, second and third, none out. For a starter, those fellas likely will score as well. But you are 'capped' at just allowing the two runs.

These two examples show why a reliever's ERA is not directly comparable to a starter's. Does anyone know whether BP, in calculating DERA for relief pitchers, attempts to adjust for these effects?

Dr. Chaleeko (# 106) wrote:

If the HOM should ideally include about 30% pitchers, and we're electing about 220 men by 2007, then it will include 65-70 pitchers.

Although I agree with a target of 30 percent pitchers, at least so far, the majority of the electorate disagrees. If my count is accurate, we've elected 30 pitchers out of 116, or 26 percent. (I've counted Ward and Dihigo each as half pitchers.) If the present rate continues, we'll have only about 55-57 pitchers when the HoM reaches 220 inductees.

John,
In view of all the interesting technical issues that are being raised regarding the evaluation of relief pitchers, I wonder whether a separate relief pitchers thread might be warranted.

These two examples show why a reliever's ERA is not directly comparable to a starter's. Does anyone know whether BP, in calculating DERA for relief pitchers, attempts to adjust for these effects?

Although I can't say for certain, none of the descriptions of the process by which the WARP system goes from a pitcher's actual IP and RA through NRA and PRAA to DERA suggests that there is any adjustment made for the features of reliever usage patterns that may suppress their RA totals somewhat.

Following on the same theme, quoting Sean Gilman from a ways back on this thread:

Dean has almost 2000 IP with a 130 ERA+.

As of this season, doubling his IP for leverage, Rivera has about 1600 IP, with a 190 ERA+.

Those 60 ERA+ points would likely be the major difference in the electorate's view of those two pitchers.

But if Rivera had pitched a real 1600 innings instead of a leverage-expanded number, his ERA+ wouldn't have been 190.

Using the concept of leverage to expand his innings but letting him keep all of his ERA+ strikes me as a form of double jeopardy.

But if Rivera had pitched a real 1600 innings instead of a leverage-expanded number, his ERA+ wouldn't have been 190.

Using the concept of leverage to expand his innings but letting him keep all of his ERA+ strikes me as a form of double jeopardy.

Yes, but the VALUE of those innings theoretically really IS the number of expanded innings at 190 ERA+, even though if really had to pitch more innings per appearance, his ERA+ would be lower...

Here's a sudden thought. Dizzy Dean had 87 relief appearances and 30 saves. It seems like that part of his record ought to be leveraged. Let's say he threw 100 IP and a full 2.0 leverage (that's a guess, but I'm low balling the IP and high balling the leverage so maybe it averages out). Then let's say that his relief ERA+ was 200 (again guessing).

Now he's up from 1967 IP to 2067 IP and his 130 ERA+ goes up, okay, not very much. A point or two at best.

But how about his peak. 1934 30-7 2.66 (159) in 311.2 IP. 17 relief appearances, 11 saves. Say 25 IP at 200. Now he's at 337 IP at 165 (wild guess). His peak could get a little bit better and could possibly get competitive with some of the other really big peaks out there...

Could it?

Sunnyday2,

I'd agree that Dean's peak deserves another look in light of leverage issues. I think it would also be appropriate to look again at the pre-1960s relievers KJOK has listed to make sure that a lack of accounting for leveraging hasn't led us to overlook a deserving candidate.

Just for fun, best ERA+ from best to less best including only years of 10 saves or more (not necessarily a standard I would want to defend, just something for the moment). Also Rivera through 2003, don't have '04 handy.

Gossage 464-246-211-184-81-79-73-72-56-27-23-(2 yrs<100)
Rivera 265-260-236-230-192-70-62

Rivera 4 years >200, Gossage 3. And yes one of those was a short season (the 464 = 46.2 IP) but the other 2 encompass 274.2 IP. Rivera's 4 years > 200 = 272.2 IP.

7th best year (as of 2003) Gossage 173 Rivera 162, and Gossage has a 172 and a 156 remaining behind that.

Gossage has 11 years > 123 with a median of 179. He threw 10 years and almost 700 IP over and above that, dragging his career ERA+ down to 126. Another George Sisler?

In fairness Rivera also has another year at 237 and in a career high 107 IP but just 5 saves. So I'm not defending my method and not saying Gossage is better than Rivera. Just saying that Gossage had a hell of a prime and don't let the career number alone confuse you about that.

This is going to be explained more thoroughly in The Book, but the adjustment for RA for AL starter/reliever is 0.80. (ERA would be around 0.70). So, to put Mo on the same scale as Mussina, you add 0.70 to his ERA. This is based on starter/relief performance of 99-02. I don't know whether this is true in earlier years. (Once the book is out, you guys can apply the same process to the earlier years.)

For NL starters, since they have the benefit of the pitcher batting (which the NL reliever does not), the impact isn't as great.

Of course, since interleague play, the AL starter does get some boost because of this as well.

I recommend just giving a blanket 0.50 ERA boost for now, until you guys can come up with the proper timeline adjustments.

If we re-look at Dean based on leveraged innings, we're going to have to look at them all. To only re-examine one guy wouldn't be all that fair to the other candidates.

I would think that Marbury would benefit more than almost anyone from leverage-based analysis.

>To only re-examine one guy wouldn't be all that fair to the other candidates.

This is exactly what I said. To leverage one guy and not the other is unfair.

You say having Dizzy Dean inside the circle is more unfair. I disagree. I say having Dizzy inside the circle is less unfair. I say having Marbury inside is less unfair. Having Rivera inside is less unfair. Having Joe Blow inside is less unfair. Every player we can get inside is less unfair.

You say excluding one player is less unfair. That's not correct.

And fair to whom? We are trying to be fair to the idea of a HoM, and it just happens that Dizzy Dean is getting votes and pitched a lot of relief. Excluding him from further analysis is not less unfair.

Tango Tiger,

One question about the adjustment to relievers' ERA.

Are you saying that this adjustment needs to be made in order to make fair comparisons between starters and relievers on the basis of ERA alone?

Or are you saying that to assess a relief pitcher's value in terms, say, of wins above average, two adjustments need to be made:

a) an adjustment of innings for leverage and
b) an adjustment of ERA?

If you are making the second claim, could you explain why that adjustment is needed to assess value properly?

Sunnyday2,

I think you're extrapolating my thoughts out further than I meant them to go. I agree with you that further analysis is fair, all I'm saying is that he's not the only candidate whose case would probably benefit from it because his pattern of usage was somewhat common at the time.

If Rixey, Grimes, or Ruffing (or anyone else) were likely to benefit from this analysis, they should get it too as a matter of doing the best we can to identify the best candidates and make the best case for them that can be made.

Doc, agreed.

The same guy will have a lower ERA (and OBP and SLG) as a reliever than as a starter. Or, if your baseline is the average pitcher (or replacement pitcher), that pitcher will have a lower ERA as a reliever than as a starter. Your baseline is different.

Your baseline is different.

Right! Thanks!

0.70 SEEMS high, but I guess that is because it's adjusting for 2 different things?

1. For the fact that relievers come in during the MIDDLE of innings, resulting in fewer runs/earned runs CHARGED to them than starters, AND for given the same performance vs. batters.

2. For the fact that relievers are able to pitch "better" than starters because they can go more "all out" over a shorter time frame within the game.

IF that's true, then there would be 3 possible adjustments:

1. Adjustment to ERA for ALL relief innings pitched of something like 0.40 to account for #2 above.

2. Adjustment to ERA for relievers who are closers or "Loogy" relievers that come in during the middle of innings of something like 0.30.

3. Adjustment to the IP (or leverage adjustment) of 2.0 or less times IP for "closer" relievers.

Re: KJOK's post #132. I think #2 is much, much more important than #1.

And that means that it also affects starters from our time.

Pedro Martinez can go something closer to all out, becuase he knows he doesn't have to pitch past the 6th or 7th inning. (And I'm not talking specifically about Pedro - I'm talking about everyone, every starter). So his IP are lower than those of a starter from an earlier time. If he ever does pitch in the 8th, it's likely to be high leverage.

KJOK: Actually, I was only looking at #2 above. It's hard to believe, but a pitcher will have a component ERA that is about 0.70 higher, based solely on #2 (using 1999-2002 data). The data will all be layed out (though you'll have to wait about 5 months for it).

Now, if you want to have a different adjustment level for relievers with 1 IP per game, or 2 IP per game or 3 IP per game.... that's not a bad thing to do either. That's alot more work, though.

I suppose some sort of sliding scale, from 0.70 for 1 IP per game to 0.00 for 5 IP per game might be in order. I dunno, 0.80, 0.40, 0.20, 0.10, 0.00 for 1 through 5 IP per game? (Just speculating).

Speaking of inherited runners and ERAs and whatnot...

Thanks to BPro, I've assembled the career inherited runners (IHR) and inherited runners scored (IHRS) for most of the guys I did the ERA+ analysis for above. BPro's numbers only reach back as far as 1972, so some totals will be partials. For a frame of reference, somewhere around one-third of all inherited runners score (IIRC).

Career inherited runners/inherited runners scored totals for selected relievers
name             ir     irs    %irs
-----------------------------------
abernathy        54      16     30%
aguilera        279      81     29%
beck            279      81     29%
benitez         305      86     28%
campbell        577     178     31%
carroll         217      79     36%
eckersley       349      91     26%
eichorn         542     184     34%
fingers         671     189     28%
fisher           38      15     39%
franco          457     147     32%
garber          694     243     35%
gordon          266      83     31%
gossage         828     279     34%
granger         187      68     36%
henke           370     113     31%
hernandez r     360     122     34%
hernandez w     587     194     33%
hiller          432     138     32%
hoerner         222      78     35%
hoffman         316      63     20%
hrabosky        487     157     32%
jones           494     178     36%
kern            481     151     31%
lyle            690     240     35%
marshall        410     138     34%
mcdaniel        194      82     42%
mcdowell        465     148     32%
mcgraw          368     120     33%
mcmahon          46      14     30%
montgomery      304     114     38%
myers           401      92     23%
nen             203      59     29%
olson           290     101     35%
orosco         1057     261     25%
percival        213      53     25%
perranoski       27       9     33%
quisenberry     523     200     38%
reardon         523     157     30%
regan            11       7     64%
righetti        397     112     28%
rivera          234      70     30%
smith d         329     107     33%
smith l         489     127     26%
stanley         578     206     36%
sutter          445     137     31%
tekulve         770     216     28%
wagner          115      27     23%
wetteland       220      69     31%
wilhelm           9       1     11%
wood              4       3     75%
worrell         327     103     31%
====================================
Total         19134    5987     31%

Best individual seasons by IHR/IHRS among this group of relievers?

In 1998 Randy Myers stranded all 22 runners he inherited. Next best perfect IHR season is Greg Olson with 16 IHR stranded in 1998. Billy Wagner had two perfect years with 10 and 9 IHR in 2003 and 1998 respectively.

Other seasons of note:
-In 1990, 1992, and 1995 Dennis Eckersley allowed only 14%, 6%, and 5% of runners to score (in seasons with 29, 31, 19 IHR respectively).

-In 2004 Trevor Hoffman was under 10% when he allowed only 1 of 11 IHR to score. And in the two healthy years surrounding 2004 (2002 and 2005), he allowed only 10% of IHR to score.

-In 1990 Randy Myers allowed only 9% of 32 IHR to score.

-In 1994 Rob Nen prevented all but one of his 24 IHR from scoring, 4%.

-In 2001 Troy Percival allowed only one of his 14 IHR to score, 7%.

-In 1993 Jeff Reardon let just one of his 28 IHR score, 4%.

-In 1984 Lee Smith allowed a single runner of the 43 he inherited score.

The worst????
Well, there's a couple small-sample contenders (Joe Hoerner's allowing 10 of 11 to score in 1977 is pretty stinky) but for sheer volume of craptasticness it's probably The Quis's 1987 season. He allowed 26 of 43 IHR to score (60%). Yech!

++++++++++++++++++++++++++++++

One question worth asking about IHR is whether stranding inherited runners is, in fact, an actual skill, or if it's the luck of the draw. I don't know the answer, and I'm not exactly certain how to find it. In the next post, I'll put out the year-by-year data so that someone more statistically savvy than me can figure it out if they want to.

Would be interesting to know that. There seem to be some pitchers who go to pieces when a couple of hits are made off them, while others remain calm at all times. Also, some pitchers are at their best from pitch 1, whereas others take time to settle in. Timlin, for example, seems to take time to settle in, which is why he is so bad with inherited runners (but was that true throughout his career?) Good nerves and ability to "find the range" fast would appear a priori to be real abilities to a greater extent than mere "clutchiness."

This data will come in two posts to avoid the character limit.

Sorry that I didn't do more formatting, but as you can imagine, I'm not about to go and format all 595 lines....

name year ir irs %irs
-----------------------
abernathy 1972 54 16 30%

aguilera 1985 0 0 #DIV/0!
aguilera 1986 5 3 60%
aguilera 1987 0 0 #DIV/0!
aguilera 1988 2 0 0%
aguilera 1989 19 10 53%
aguilera 1990 38 8 21%
aguilera 1991 37 9 24%
aguilera 1992 40 7 18%
aguilera 1993 21 6 29%
aguilera 1994 15 6 40%
aguilera 1995 19 5 26%
aguilera 1997 22 9 41%
aguilera 1998 22 1 5%
aguilera 1999 18 7 39%
aguilera 2000 21 10 48%

beck 1991 14 5 36%
beck 1992 41 9 22%
beck 1993 35 7 20%
beck 1994 19 7 37%
beck 1995 34 9 26%
beck 1996 33 10 30%
beck 1997 15 4 27%
beck 1998 32 9 28%
beck 1999 9 3 33%
beck 2000 13 6 46%
beck 2001 21 7 33%
beck 2003 6 3 50%
beck 2004 7 2 29%

benitez 1994 4 1 25%
benitez 1995 46 16 35%
benitez 1996 16 2 13%
benitez 1997 57 12 21%
benitez 1998 42 14 33%
benitez 1999 30 6 20%
benitez 2000 26 13 50%
benitez 2001 13 2 15%
benitez 2002 13 2 15%
benitez 2003 30 9 30%
benitez 2004 23 8 35%
benitez 2005 5 1 20%

campbell 1973 21 6 29%
campbell 1974 72 24 33%
campbell 1975 59 19 32%
campbell 1976 79 25 32%
campbell 1977 71 14 20%
campbell 1978 28 13 46%
campbell 1979 44 16 36%
campbell 1980 21 7 33%
campbell 1981 28 9 32%
campbell 1982 31 11 35%
campbell 1983 45 19 42%
campbell 1984 13 4 31%
campbell 1985 39 6 15%
campbell 1986 20 4 20%
campbell 1987 6 1 17%

carroll 1972 41 12 29%
carroll 1973 31 12 39%
carroll 1974 35 14 40%
carroll 1975 38 12 32%
carroll 1976 24 9 38%
carroll 1977 48 20 42%
carroll 1978 0 0 #DIV/0!

eckersley 1975 15 7 47%
eckersley 1976 5 0 0%
eckersley 1986 2 2 100%
eckersley 1987 50 11 22%
eckersley 1988 32 11 34%
eckersley 1989 24 9 38%
eckersley 1990 29 4 14%
eckersley 1991 31 9 29%
eckersley 1992 31 2 6%
eckersley 1993 32 13 41%
eckersley 1994 21 5 24%
eckersley 1995 19 1 5%
eckersley 1996 21 6 29%
eckersley 1997 13 3 23%
eckersley 1998 24 8 33%

eichorn 1986 78 26 33%
eichorn 1987 97 27 28%
eichorn 1988 45 16 36%
eichorn 1989 34 14 41%
eichorn 1990 69 20 29%
eichorn 1991 54 15 28%
eichorn 1992 44 19 43%
eichorn 1993 49 20 41%
eichorn 1994 44 10 23%
eichorn 1996 28 17 61%

fingers 1972 44 6 14%
fingers 1973 69 23 33%
fingers 1974 78 27 35%
fingers 1975 75 19 25%
fingers 1976 89 31 35%
fingers 1977 46 10 22%
fingers 1978 53 10 19%
fingers 1979 26 14 54%
fingers 1980 68 21 31%
fingers 1981 38 10 26%
fingers 1982 35 8 23%
fingers 1984 19 2 11%
fingers 1985 31 8 26%

fisher 1972 24 9 38%
fisher 1973 14 6 43%

franco 1984 29 11 38%
franco 1985 34 10 29%
franco 1986 52 20 38%
franco 1987 23 7 30%
franco 1988 29 6 21%
franco 1989 17 4 24%
franco 1990 29 10 34%
franco 1991 25 8 32%
franco 1992 12 5 42%
franco 1993 12 6 50%
franco 1994 11 3 27%
franco 1995 21 5 24%
franco 1996 20 12 60%
franco 1997 18 3 17%
franco 1998 20 3 15%
franco 1999 8 5 63%
franco 2000 23 5 22%
franco 2001 18 8 44%
franco 2003 13 4 31%
franco 2004 17 4 24%
franco 2005 26 8 31%

garber 1972 60 22 37%
garber 1973 60 22 37%
garber 1974 70 25 36%
garber 1975 58 26 45%
garber 1976 24 12 50%
garber 1977 43 16 37%
garber 1978 25 7 28%
garber 1979 42 19 45%
garber 1980 42 13 31%
garber 1981 21 6 29%
garber 1982 33 9 27%
garber 1983 32 7 22%
garber 1984 42 20 48%
garber 1985 21 7 33%
garber 1986 47 14 30%
garber 1987 52 14 27%
garber 1988 22 4 18%

gordon 1988 2 0 0%
gordon 1989 33 16 48%
gordon 1991 17 6 35%
gordon 1992 19 5 26%
gordon 1993 23 11 48%
gordon 1997 7 0 0%
gordon 1998 38 8 21%
gordon 1999 8 2 25%
gordon 2001 4 3 75%
gordon 2002 8 2 25%
gordon 2003 37 7 19%
gordon 2004 37 6 16%
gordon 2005 33 17 52%

gossage 1972 28 7 25%
gossage 1973 14 4 29%
gossage 1974 55 19 35%
gossage 1975 99 27 27%
gossage 1976 4 2 50%
gossage 1977 59 19 32%
gossage 1978 63 19 30%
gossage 1979 33 10 30%
gossage 1980 51 11 22%
gossage 1981 27 6 22%
gossage 1982 41 13 32%
gossage 1983 74 36 49%
gossage 1984 33 10 30%
gossage 1985 17 6 35%
gossage 1986 26 13 50%
gossage 1987 27 10 37%
gossage 1988 21 10 48%
gossage 1989 21 9 43%
gossage 1991 53 22 42%
gossage 1992 13 2 15%
gossage 1993 29 13 45%
gossage 1994 40 11 28%

granger 1972 50 18 36%
granger 1973 35 11 31%
granger 1974 4 3 75%
granger 1975 63 18 29%
granger 1976 35 18 51%

henke 1982 10 3 30%
henke 1983 7 2 29%
henke 1984 21 10 48%
henke 1985 15 5 33%
henke 1986 51 12 24%
henke 1987 58 22 38%
henke 1988 38 9 24%
henke 1989 42 16 38%
henke 1990 32 6 19%
henke 1991 15 3 20%
henke 1992 7 2 29%
henke 1993 36 9 25%
henke 1994 22 10 45%
henke 1995 16 4 25%

hernandez r 1991 3 3 100%
hernandez r 1992 38 13 34%
hernandez r 1993 39 13 33%
hernandez r 1994 17 8 47%
hernandez r 1995 33 16 48%
hernandez r 1996 30 9 30%
hernandez r 1997 46 14 30%
hernandez r 1998 18 3 17%
hernandez r 1999 19 6 32%
hernandez r 2000 19 8 42%
hernandez r 2001 13 3 23%
hernandez r 2002 13 2 15%
hernandez r 2003 13 7 54%
hernandez r 2004 30 12 40%
hernandez r 2005 29 5 17%
hernandez w 1977 68 25 37%
hernandez w 1978 48 15 31%
hernandez w 1979 33 18 55%
hernandez w 1980 44 9 20%
hernandez w 1981 11 7 64%
hernandez w 1982 67 20 30%
hernandez w 1983 47 16 34%
hernandez w 1984 33 10 30%
hernandez w 1985 49 14 29%
hernandez w 1986 50 18 36%
hernandez w 1987 43 14 33%
hernandez w 1988 63 16 25%
hernandez w 1989 31 12 39%

hiller 1972 20 7 35%
hiller 1973 84 13 15%
hiller 1974 54 27 50%
hiller 1975 37 9 24%
hiller 1976 67 26 39%
hiller 1977 47 17 36%
hiller 1978 70 15 21%
hiller 1979 39 14 36%
hiller 1980 14 10 71%

hoerner 1972 35 14 40%
hoerner 1973 50 16 32%
hoerner 1974 41 10 24%
hoerner 1975 25 8 32%
hoerner 1976 60 20 33%
hoerner 1977 11 10 91%

hoffman 1993 45 6 13%
hoffman 1994 30 9 30%
hoffman 1995 18 6 33%
hoffman 1996 35 6 17%
hoffman 1997 44 14 32%
hoffman 1998 26 4 15%
hoffman 1999 22 3 14%
hoffman 2000 20 4 20%
hoffman 2001 34 7 21%
hoffman 2002 21 2 10%
hoffman 2003 0 0 #DIV/0!
hoffman 2004 11 1 9%
hoffman 2005 10 1 10%

hrabosky 1972 1 1 100%
hrabosky 1973 33 12 36%
hrabosky 1974 57 16 28%
hrabosky 1975 69 21 30%
hrabosky 1976 74 28 38%
hrabosky 1977 43 12 28%
hrabosky 1978 69 18 26%
hrabosky 1979 68 27 40%
hrabosky 1980 27 5 19%
hrabosky 1981 23 5 22%
hrabosky 1982 23 12 52%

jones 1982 3 2 67%
jones 1986 9 3 33%
jones 1987 50 26 52%
jones 1988 46 10 22%
jones 1989 40 9 23%
jones 1990 48 13 27%
jones 1991 13 9 69%
jones 1992 37 11 30%
jones 1993 20 9 45%
jones 1994 20 10 50%
jones 1995 15 6 40%
jones 1996 43 17 40%
jones 1997 18 4 22%
jones 1998 30 9 30%
jones 1999 55 21 38%
jones 2000 47 19 40%

continuing on...

kern 1974 0 0 #DIV/0!
kern 1975 6 2 33%
kern 1976 78 22 28%
kern 1977 67 21 31%
kern 1978 76 18 24%
kern 1979 91 25 27%
kern 1980 43 22 51%
kern 1981 26 3 12%
kern 1982 53 19 36%
kern 1983 2 0 0%
kern 1984 12 6 50%
kern 1985 3 0 0%
kern 1986 24 13 54%

lyle 1972 52 14 27%
lyle 1973 60 18 30%
lyle 1974 76 29 38%
lyle 1975 71 27 38%
lyle 1976 81 31 38%
lyle 1977 99 38 38%
lyle 1978 62 16 26%
lyle 1979 70 23 33%
lyle 1980 69 27 39%
lyle 1981 17 6 35%
lyle 1982 33 11 33%

marshall 1972 58 16 28%
marshall 1973 89 23 26%
marshall 1974 75 31 41%
marshall 1975 29 8 28%
marshall 1976 28 7 25%
marshall 1977 11 6 55%
marshall 1978 39 10 26%
marshall 1979 61 26 43%
marshall 1980 13 7 54%
marshall 1981 7 4 57%

mcdaniel 1972 22 14 64%
mcdaniel 1973 68 28 41%
mcdaniel 1974 31 19 61%
mcdaniel 1975 73 21 29%

mcdowell 1985 23 12 52%
mcdowell 1986 32 4 13%
mcdowell 1987 37 16 43%
mcdowell 1988 33 4 12%
mcdowell 1989 46 18 39%
mcdowell 1990 45 16 36%
mcdowell 1991 43 14 33%
mcdowell 1992 42 10 24%
mcdowell 1993 30 9 30%
mcdowell 1994 24 10 42%
mcdowell 1995 66 15 23%
mcdowell 1996 44 20 45%

mcgraw 1972 54 14 26%
mcgraw 1973 48 16 33%
mcgraw 1974 22 11 50%
mcgraw 1975 20 8 40%
mcgraw 1976 24 8 33%
mcgraw 1977 12 3 25%
mcgraw 1978 13 5 38%
mcgraw 1979 47 17 36%
mcgraw 1980 31 7 23%
mcgraw 1981 20 4 20%
mcgraw 1982 28 14 50%
mcgraw 1983 34 9 26%
mcgraw 1984 15 4 27%

mcmahon 1972 25 10 40%
mcmahon 1973 13 2 15%
mcmahon 1974 8 2 25%

montgomery 1987 6 2 33%
montgomery 1988 38 8 21%
montgomery 1989 31 14 45%
montgomery 1990 31 12 39%
montgomery 1991 41 17 41%
montgomery 1992 20 7 35%
montgomery 1993 30 8 27%
montgomery 1994 15 5 33%
montgomery 1995 20 10 50%
montgomery 1996 19 5 26%
montgomery 1997 24 10 42%
montgomery 1998 4 4 100%
montgomery 1999 25 12 48%

myers 1985 0 0 #DIV/0!
myers 1986 2 1 50%
myers 1987 38 6 16%
myers 1988 54 10 19%
myers 1989 54 10 19%
myers 1990 32 3 9%
myers 1991 38 8 21%
myers 1992 40 16 40%
myers 1993 24 4 17%
myers 1994 18 7 39%
myers 1995 37 10 27%
myers 1996 22 11 50%
myers 1997 20 6 30%
myers 1998 22 0 0%

nen 1993 11 4 36%
nen 1994 28 1 4%
nen 1995 17 5 29%
nen 1996 30 8 27%
nen 1997 20 10 50%
nen 1998 21 5 24%
nen 1999 15 8 53%
nen 2000 18 5 28%
nen 2001 17 5 29%
nen 2002 26 8 31%

olson 1988 6 1 17%
olson 1989 44 10 23%
olson 1990 37 12 32%
olson 1991 30 12 40%
olson 1992 31 5 16%
olson 1993 30 15 50%
olson 1994 0 0 #DIV/0!
olson 1995 1 1 100%
olson 1996 36 20 56%
olson 1997 24 9 38%
olson 1998 16 0 0%
olson 1999 23 9 39%
olson 2000 2 2 100%
olson 2001 10 5 50%

orosco 1979 12 2 17%
orosco 1981 5 1 20%
orosco 1982 48 12 25%
orosco 1983 37 10 27%
orosco 1984 36 11 31%
orosco 1985 46 9 20%
orosco 1986 41 9 22%
orosco 1987 38 8 21%
orosco 1988 52 8 15%
orosco 1989 58 18 31%
orosco 1990 50 13 26%
orosco 1991 60 21 35%
orosco 1992 64 14 22%
orosco 1993 40 11 28%
orosco 1994 33 9 27%
orosco 1995 66 15 23%
orosco 1996 46 10 22%
orosco 1997 66 16 24%
orosco 1998 47 15 32%
orosco 1999 73 16 22%
orosco 2000 8 1 13%
orosco 2001 32 12 38%
orosco 2002 47 6 13%
orosco 2003 52 14 27%

percival 1995 53 9 17%
percival 1996 28 4 14%
percival 1997 27 5 19%
percival 1998 24 5 21%
percival 1999 25 11 44%
percival 2000 3 2 67%
percival 2001 14 1 7%
percival 2002 19 10 53%
percival 2003 8 1 13%
percival 2004 10 5 50%
percival 2005 2 0 0%

perranoski 1972 18 4 22%
perranoski 1972 9 5 56%

quisenberry 1979 52 23 44%
quisenberry 1980 89 25 28%
quisenberry 1981 41 17 41%
quisenberry 1982 36 13 36%
quisenberry 1983 40 15 38%
quisenberry 1984 35 11 31%
quisenberry 1985 71 22 31%
quisenberry 1986 48 20 42%
quisenberry 1987 43 26 60%
quisenberry 1988 33 17 52%
quisenberry 1989 29 8 28%
quisenberry 1990 6 3 50%

reardon 1979 13 2 15%
reardon 1980 17 9 53%
reardon 1981 17 4 24%
reardon 1982 61 25 41%
reardon 1983 56 24 43%
reardon 1984 53 6 11%
reardon 1985 38 9 24%
reardon 1986 29 4 14%
reardon 1987 44 18 41%
reardon 1988 44 13 30%
reardon 1989 47 14 30%
reardon 1990 21 8 38%
reardon 1991 18 4 22%
reardon 1992 30 15 50%
reardon 1993 28 1 4%
reardon 1994 7 1 14%

regan 1972 11 7 64%

righetti 1982 9 3 33%
righetti 1984 50 17 34%
righetti 1985 71 24 34%
righetti 1986 49 9 18%
righetti 1987 42 13 31%
righetti 1988 53 15 28%
righetti 1989 31 6 19%
righetti 1990 15 4 27%
righetti 1991 32 5 16%
righetti 1992 17 5 29%
righetti 1993 23 9 39%
righetti 1994 5 2 40%
righetti 1995 0 0 #DIV/0!

rivera 1995 5 2 40%
rivera 1996 22 6 27%
rivera 1997 17 4 24%
rivera 1998 24 4 17%
rivera 1999 27 5 19%
rivera 2000 24 10 42%
rivera 2001 25 5 20%
rivera 2002 20 9 45%
rivera 2003 35 17 49%
rivera 2004 17 6 35%
rivera 2005 18 2 11%

smith d 1980 27 11 41%
smith d 1981 18 6 33%
smith d 1982 31 12 39%
smith d 1983 17 7 41%
smith d 1984 24 4 17%
smith d 1985 63 16 25%
smith d 1986 34 9 26%
smith d 1987 29 13 45%
smith d 1988 28 9 32%
smith d 1989 8 3 38%
smith d 1990 21 5 24%
smith d 1991 21 9 43%
smith d 1992 8 3 38%

smith l 1980 11 3 27%
smith l 1981 29 5 17%
smith l 1982 44 15 34%
smith l 1983 50 18 36%
smith l 1984 41 7 17%
smith l 1985 44 1 2%
smith l 1986 37 12 32%
smith l 1987 35 11 31%
smith l 1988 38 10 26%
smith l 1989 42 9 21%
smith l 1990 34 6 18%
smith l 1991 26 12 46%
smith l 1992 13 4 31%
smith l 1993 4 1 25%
smith l 1994 9 2 22%
smith l 1995 10 5 50%
smith l 1996 19 4 21%
smith l 1997 3 2 67%

stanley 1977 33 9 27%
stanley 1978 62 25 40%
stanley 1979 7 2 29%
stanley 1980 37 12 32%
stanley 1981 31 12 39%
stanley 1982 48 14 29%
stanley 1983 62 17 27%
stanley 1984 51 20 39%
stanley 1985 53 24 45%
stanley 1986 58 20 34%
stanley 1987 21 9 43%
stanley 1988 62 19 31%
stanley 1989 53 23 43%

sutter 1976 32 7 22%
sutter 1977 50 14 28%
sutter 1978 58 18 31%
sutter 1979 55 13 24%
sutter 1980 38 9 24%
sutter 1981 32 11 34%
sutter 1982 54 14 26%
sutter 1983 43 20 47%
sutter 1984 48 12 25%
sutter 1985 17 10 59%
sutter 1986 10 4 40%
sutter 1988 8 5 63%

tekulve 1974 12 4 33%
tekulve 1975 28 7 25%
tekulve 1976 46 19 41%
tekulve 1977 42 16 38%
tekulve 1978 69 14 20%
tekulve 1979 76 13 17%
tekulve 1980 61 20 33%
tekulve 1981 29 13 45%
tekulve 1982 68 16 24%
tekulve 1983 79 29 37%
tekulve 1984 59 13 22%
tekulve 1985 35 6 17%
tekulve 1986 47 10 21%
tekulve 1987 51 11 22%
tekulve 1988 50 16 32%
tekulve 1989 18 9 50%

wagner 1995 1 0 0%
wagner 1996 20 4 20%
wagner 1997 9 2 22%
wagner 1998 9 0 0%
wagner 1999 18 5 28%
wagner 2000 15 6 40%
wagner 2001 9 3 33%
wagner 2002 11 3 27%
wagner 2003 10 0 0%
wagner 2004 2 0 0%
wagner 2005 11 4 36%

wetteland 1989 4 0 0%
wetteland 1990 14 8 57%
wetteland 1991 5 3 60%
wetteland 1992 37 8 22%
wetteland 1993 33 6 18%
wetteland 1994 29 9 31%
wetteland 1995 24 7 29%
wetteland 1996 14 6 43%
wetteland 1997 19 5 26%
wetteland 1998 15 7 47%
wetteland 1999 12 4 33%
wetteland 2000 14 6 43%

wilhelm 1972 9 1 11%

wood 1973 0 0 #DIV/0!
wood 1977 4 3 75%
wood 1978 0 0 #DIV/0!

worrell 1985 20 7 35%
worrell 1986 65 15 23%
worrell 1987 66 18 27%
worrell 1988 50 16 32%
worrell 1989 30 8 27%
worrell 1992 8 2 25%
worrell 1993 13 3 23%
worrell 1994 14 9 64%
worrell 1995 26 10 38%
worrell 1996 24 10 42%
worrell 1997 11 5 45%

Apologies in advance if there are any inputting errors in what I've just posted. I found one and corrected it before I posted any of this information.

+++++++++++++++++++++++++++++++

Karl, I think there could be three possible explanations (among those which are under a pitcher's control, so hit-luck and which batter he enters against notwithstanding) for why a pitcher might perform well or not do very well with runners on base:
1) excellent control---which minimizes the advancement of baserunners beyond the slugging abilities of the opposing hitters

2) heavy ball---keeping the ball down leads to GIDPs and avoids long hits (including sac flies)

3) high k-rate---if you can't hit it...

Someone like Eckersley had all three of these and spun some amazing IHR seasons out of it. Hoffman too. Interestingly, neither of those two would be considered a "fastball" pitcher, relying instead on their assortment, especially Hoffman and his change. Strangely, though, Mariano allows a fairly high rate of IHRS based on his ERA+. This MIGHT be a function of his lousy defense.

Anyway, just speculating.

I would bet that a pitcher's "strand rate" can be estimated based on his OBP and SLG. I doubt it's a "skill".

The strand rate is HIGHLY dependent on the number of outs. It's also HIGHLY dependent on whether the runner was on 3b,2b or 1b.

Once you guys get close to the retrosheet years (1960 and later), someone (Joe? KJOK?) email me, and I'll run career LI. That should tell you how each reliever was used.

Since retrosheet now has detailed data back to 1960, which is pretty much the beginning of the "reliever" era, it might be instructive to look at reliever's actual "Batting against" line (OBP allowed, SLG allowed, Runs Created/27 against, etc.) as opposed to their ERA+?

yes, and their career LI, together!

Orosco, with an ERA+ of 125 looks to have added far fewer to others' ERA+ than Rivera, with an ERA+ of 190 (eyeballing the figures, and being lazy). The ideal analysis might be to take a group of oroscos, with good IHR and mediocre ERA+ and compare it to a group of Riveras, with wonderful ERA+ and poor IHP, and see if differences jump out at you. Pity we have only one year of Wilhelm in this context (2256 IP at 146 looks a clear HOMer, but what about his IHR rate?)

Tango,

We're actually at 1961 right now. Any chance you've got LIs for Stu Miller, Lindy McDaniel, Eddie Fisher, Eddie Watt, Ted Abernathy, Phil Regan, Ron Perranoski, and those kind of guys?

You really snuck up on me. I've yet to download the latest retro files. What's the date of your next election?

For the sake of posterity, don't forget the Relief Pitchers thread!

(I tried bumping it earlier, but the dang thing's been rejected so long that Hot Topics no longer recognizes it.)

1961 Election is 10/3, but those are guys who retired by 1955 I think, so we're not quite yet up to 1960 for any eligible electees, so I think it's December before we get to any eligible players who actually played in 1959, the first retrosheet available year (for AL)...

Ah, great, that's much more workable.

My problem with the 60s is that the run environment is much different. A 3-run lead is far safer in 1968 than in 1994. So, the leverage of the 94 situation will be much greater.

Then again, you will have far more close games in the 60s to compensate.

Nonetheless, if I do LI correctly, I have to account for this.

In case anyone's curious, here's the relief appearances, games finished, and saves for various candidates and potential candidates. Figure out how to leverage 'em as you see fit. No pre-1890rs because there were too few RA to make it worth checking on. From most to least RA:

Hoyt 251-175-52
Trout 199-134-35
Pennock 197-133-32
Luque 185-128-28
Haines 167-95-10
CMays 165-125-31
Harder 149-95-23
Cicotte 141-106-25
Rixey 140-97-14
Adams 127-89-15
Grimes 119-94-18
Newsom 117-71-21
WCooper 109-84-14
Warneke 102-63-13
Shocker 95-68-25
Ruffing 88-71-16
Fitzsimmons 87-64-13
Dean 87-76-30
Griffith 81-78-6
Waddell 67-53-5
Bridges 62-47-10
HVaughn 59-45-5
WFerrell 51-43-13
Willis 42-41-11
BWalters 30-18-4
Joss 26-22-5

Jack Quinn 312-216-57
Eddie Rommel 251-182-29

Any others?

That's candidates. I'd be somewhat interested in the already elected as background information for the lists in #149 and #150 - particulary 3F. Brown, W. Johnson, and Grove.

post-1893rs electees & (too many) candidats:

JMurphy 375-193-107
Firpo 364-271-101
Faber 186-134-28
Crandall 168-145-25
SSJones 160-115-31
Grove 159-123-55
Shawkey 154-113-28
SPaige 153-109-32
Dauss 150-121-40
3FBrown 149-138-49
Uhle 145-114-25
WJohnson 136-129-34
Derringer 134-98-29
Marquard 133-82-19
Zachary 124-84-22
Walsh 115-102-34
Newhouser 114-79-26
Lyons 110-92-23
Plank 106-74-23
Hubbell 104-82-33
Reulbach 99-78-13
Alexander 97-80-32
Vance 95-53-11
Young 89-84-17
Leever 89-75-13
Phillippe 84-71-12
Mathewson 84-78-28
McGinnity 84-73-24
Whitehill 68-46-11
Coveleski 65-52-21
JPowell 62-57-15
Chesbro 60-52-5
Mullin 59-50-8
Tannehill 39-38-7
Weyhing 35-35-4
Rusie 35-35-5
Van Haltren 25-24-4
Ruth 15-11-4
Sisler 12-8-3
BWallace 9-9-1

Some other relief totals for pitchers eligible in 1962 or earlier:

Konstanty 397-267-74
Heving 390-232-63
JaRussell 375-209-38
Kinder 362-255-102
Wilks 341-191-46
SyJohnson 332-201-43
Brazle 324-178-60
Ferrick 316-210-56
ABenton 288-171-66
DLeonard(1933-53) 266-171-44

Returning to a discussion that took place earlier on this thread, in # 89 jschmeagol asked:

How much should someone like Rivera, or a starter like Pettite, receive for postseason play?

Tango Tiger responded:

Of course World Series should count! Isn't it what baseball is all about? Personally, I would triple anything a player does in the playoffs.

OCF disagreed:

But there's this problem:

Lou Brock: 21 G, 92 PA, .391/.424/.651 plus 14-2 as a basestealer.

Ernie Banks: O G, statistics N/A.

If we triple Brock's record and add it to the appropriate seasons, then it makes a distinct difference to his value, especially his peak value. But is doing that fair to Banks?

sunnyday2 then wrote:

I agree with O here... as for post-season, the danger is that when Mariano becomes eligible, all of a sudden his post-season record will be used as XC, whereas nobody else's post-season record has ever really been discussed.

IOW in order for it to be fair to give Rivera a post-season boost, we really ought to calculate all the other post-season values, all-time, to make sure the list isn't getting skewed.

FWIW, I've been giving credit for post-season play ever since I started voted. Trying to do it systematically too--I look at how the team and the player did in each World Series he played and take a guess as to how many win shares it should be worth (usually zero, one, or two). Unlike Tango Tiger's suggestion, I don't triple them. I don't think it's ever been enough to make a significant difference in anyone's candidacy, though it might move a player up two or three places.

Regarding fairness to players like Banks...well, I long ago recognized that absolute fairness is unattainable in view of the diverse circumstances of each player's career. But I do think that players on really good teams face some offsetting factors that sometimes may hurt their statistics relative to players on poor teams.

In particular, on a top team like the Yankees or the Dodgers of the 1950s, playing time was at a premium, which made it more difficult to break into the lineup and easier to lose your spot when a player's performance began to fade. For example, both Keller and Rizzuto had excellent seasons in the high (triple A-equivalent) minors that would have earned them spots with two-thirds of the teams in baseball. In both cases, they were sent back to play another season before they were finally brought up.

So maybe win shares or postseason extra credit tend to slightly favor players on good teams, but I'm not sure that the effects aren't offset by reduced playing time.

Ultimately, I agree with Tango that championships are what baseball is all about. So somehow it has seemed to me that failing to give credit for postseason play would insult the principles on which the sport is organized.

I don't think fairness should matter.

Around the world, they speak highly of "caps" for players who
manage to play in these international tournaments.
Only in North America is the playoffs given so little credit.

Who cares that Banks gets little chance to pad (or drop)
his overall performance for lack of playoff PAs. Koufax and Mo
deserve the extra credit for performing when it counted most.

A really cool little (litte?) study for somebody would be to go back and list some of the really great post-season players. Every voter can decide how to use the info, but if we're going to give Mariano Rivera and Lou Brock (too late to worry about Frankie Frisch) post-season XC, we ought to know who else is in that particular mix.

And I wasn't talking about the "unfairness" of boosting Brock versus Ernie Banks, for example, though there is an element of "unfairness" there.

I am more concerned, frankly, about the "unfairness" of giving Brock or Mantle a boost, for instance, while forgetting even to look at and consider Clemente's or Snider's or Frank Baker's or Heinie Groh's post-season record or whomever. (Not that these are good examples.)

And what then of Temple Cup or NeLs...?

Anyway, it's not the selective opportunity of the players, it is a wholly selective consideration of post-season records by us voters, that's what would concern me.

I'd have a bigger concern over completely ignoring it, than choosing to look at it for a select group of players.

I'd have a bigger concern over completely ignoring it, than choosing to look at it for a select group of players.

FWIW, I think post-season play should be factored in. However, if you're elevating a good player into a great player with the boost, that's a little too much, IMO.

The pre-formatted table in the header of this thread still seems to display properly. Does that mean John will be able to post the results normally today? [crossed fingers]

I want to restate my position here just in case it got lost...

I am only giving credit to players with impressive post season play. If a player has been average or worse no credit is given, none is taken away. If a player never made the post season, no credit is given.

I think that this is fair because it doesn't give a player extra points over soemone who never got the chance unless that player did something special. So extra credit given to Whitey Ford, Mariano, Reggie, Gibson, etc.

I would, however appreciate some kind of rundown of players who excelled in the postseason. I know I have missed plenty of guys through the years.

Does that mean John will be able to post the results normally today? [crossed fingers]

I'm crossing my fingers, too, OCF.

It might well be that the player for whom the postseason credit question looms largest is not Rivera but Smoltz.

Perhaps 2005 regular season puts him over the top anyway - another full season as a starter with an ERA+ near 150. That gives him 2929 (regular season) career IP with an ERA+ somewhere near 126 or 127.

But his postseason record amounts to just about another full season: 39 games, 26 starts, 199.2 IP, 14-4 record with 4 saves. And what are the conditions of postseason? The weather is colder, which helps pitchers, but the level of competition is higher. Bbref, for obvious reasons, doesn't try to compute ERA+, but we can compare the raw stats.

Regular-season Smoltz, per 9 IP: 3.26 ER, 7.80 H, 2.71 BB, 7.89 SO.

Post-season Smoltz, per 9 IP: 2.70 ER, 7.26 H, 2.97 BB, 8.52 SO.

This is a test:

RK   LY  Player                   PTS  Bal   1  2  3  4  5  6  7  8  9 10 11 12 13 14 15
----------------------------------------------------------------------------------------
 1  n/e  Hal Newhouser            924   47  21  5  6  3     2  3  1  2     2  1  1      
 2    5  Hughie Jennings          443   28   6  2  4  3     2  1     2  2     1     1  4
----------------------------------------------------------------------------------------
 3    3  Joe Medwick              430   32   1  4  2  3  1  2  1  2  2  1  2  3  5  1  2
 4    8  Earl Averill             412   31   2  1  3  2  3  1  2  2  2  4  1  1  2  3  2
 5    4  Red Ruffing              403   27   3  2     5     3  4  2  1  1  2  1  1  1  1
 6    7  Wes Ferrell              375   26      3  2  3  2  3  2  1  4  1  1  2  1     1
 7    6  Biz Mackey               371   29      1  1  2  2  3  2  5  3  2  3  1  1  3   
 8   10  Clark Griffith           350   23   1  4  3  1  1  3     2  1  1  2  3        1
 9    9  Eppa Rixey               347   25   1  3  3  2  2  1  1  2     1  1     4  2  2
10   11  George Sisler            307   23   1     4  2  1  1     5  1  3     1  1  1  2
----------------------------------------------------------------------------------------
11   13  Cool Papa Bell           283   22      4        1  2  1  1  3  2  1  2     2  3
12   12  George Van Haltren       274   20      2  3  1  1  2  3     2  1        1  2  2
13   14  Jake Beckley             274   18   1  4  2  2              2  3        3  1   
14   15  Cupid Childs             226   19            3  2     2  1     3  2  2  3  1   
15   21  Pete Browning            225   16   2  1  1  1  1  1     1  1  1  2  2  1     1
16   16  Willard Brown            219   20      1     1     2  2     1  2  2  1  3  4  1
17   19  Hugh Duffy               210   15   1     1     3  3     1  3  1     1        1
18   17  Joe Sewell               209   16      1  2     3  1     2     1  1  2  1  2   
19   18  Bobby Doerr              206   15   1     1  1  1  2  3  1        3  1     1   
20   23  Cannonball Dick Redding  199   17      1     1  1  3  1  1  1  1        3     4
21   20  Mickey Welch             193   11   3  1     2  3                    1     1   
22   24  José Méndez              174   14         1  1  1  1  3  2           2  2     1
23   22  Dobie Moore              174   12      1     2  3     2  1        1  2         
24   28  Joe Gordon               164   17         1     1  1  1        1  1  3  1  3  4
25   26  Alejandro Oms            160   12      1  1     2  2        1  1  1  1  1  1   
26   30T Rube Waddell             148   13      1           1  1  2  1     3     2     2
27   25  Bucky Walters            141   10         2  2           1  3  2               
28   32  Charley Jones            134    9   1  1  1           2     2        1  1      
29T  35  Burleigh Grimes          122    9   1        1  1  1        2     2           1
29T  27  Tommy Leach              122    9      1        2        2     2  1  1         
31   29  Gavy Cravath             118    9   1     1     2           1  1     1        2
32   33  Edd Roush                115   11   1                 1  1        1  2  2  2  1
33   34  Wally Schang             113    8   1           2  1     1     1  1     1      
34   30T Roger Bresnahan          107    8      2                 1  1  1  1     1  1   
35   36  Bob Elliott               91    8      1              1        1  2     2  1   
36   38  Quincy Trouppe            90    9            1  1     1        1           2  3
37   37  Larry Doyle               79    6         1  1     1  1              1        1
38   39  Bill Monroe               61    6         1                       1  2  1  1   
39   47  Charlie Keller            55    5               1        2                 1  1
40   41T John McGraw               54    4               1     2           1            
41T  44T Tommy Bridges             53    5                        1  2     1           1
41T  43  Bob Johnson               53    5               1        1     1           1  1
43   41T Dizzy Dean                51    4               1     1        1  1            
44   49  Ernie Lombardi            49    4      1                          1  1     1   
45T  48  Sam Rice                  45    5                        1  1              2  1
45T  40  Jimmy Ryan                45    5                           1        2  1  1   
47   51  Luke Easter               43    3         1        1              1            
48   46  Dizzy Trout               40    3                     2     1                  
49   50  Ed Williamson             38    4                           1  1        1  1   
50   44T Ben Taylor                38    3            1                 1  1            
51   67  Vic Willis                28    3                        1           1        1
52T  53T Addie Joss                27    2                  1        1                  
52T  52  Dutch Leonard             27    2                     1  1                     
54   53T Chuck Klein               25    2                     1        1               
55   55  Carl Mays                 24    2                        1     1               
56   57  George J. Burns           19    2                                 1  1         
57   56  Bobo Newsom               18    2                                 1     1      
58   58  Pie Traynor               18    1         1                                    
59   59  Dom DiMaggio              17    2                                    1  1      
60T  60T Ed Cicotte                17    1            1                                 
60T  60T Fielder Jones             17    1            1                                 
62T  62T Tommy Bond                16    1               1                              
62T n/e  Johnny Pesky              16    1               1                              
64   70  Hack Wilson               15    2                                    1        1
65T  64T Dolf Luque                15    1                  1                           
65T  62T Bobby Veach               15    1                  1                           
67   66  Dick Lundy                14    2                                          2   
68   64T Leroy Matlock             14    1                     1                        
69   71T Frank Chance              13    2                                          1  1
70   68  Fred Dunlap               11    1                              1               
71   69  Lefty Gomez               10    1                                 1            
72T  71T Sam Leever                 7    1                                          1   
72T  71T Rabbit Maranville          7    1                                          1   
Ballots Cast: 49

This is a test:

RK   LY  Player                   PTS  Bal   1  2  3  4  5  6  7  8  9 10 11 12 13 14 15
----------------------------------------------------------------------------------------
 1  n/e  Hal Newhouser            924   47  21  5  6  3     2  3  1  2     2  1  1      
 2    5  Hughie Jennings          443   28   6  2  4  3     2  1     2  2     1     1  4
----------------------------------------------------------------------------------------
 3    3  Joe Medwick              430   32   1  4  2  3  1  2  1  2  2  1  2  3  5  1  2
 4    8  Earl Averill             412   31   2  1  3  2  3  1  2  2  2  4  1  1  2  3  2
 5    4  Red Ruffing              403   27   3  2     5     3  4  2  1  1  2  1  1  1  1
 6    7  Wes Ferrell              375   26      3  2  3  2  3  2  1  4  1  1  2  1     1
 7    6  Biz Mackey               371   29      1  1  2  2  3  2  5  3  2  3  1  1  3   
 8   10  Clark Griffith           350   23   1  4  3  1  1  3     2  1  1  2  3        1
 9    9  Eppa Rixey               347   25   1  3  3  2  2  1  1  2     1  1     4  2  2
10   11  George Sisler            307   23   1     4  2  1  1     5  1  3     1  1  1  2
----------------------------------------------------------------------------------------
11   13  Cool Papa Bell           283   22      4        1  2  1  1  3  2  1  2     2  3
12   12  George Van Haltren       274   20      2  3  1  1  2  3     2  1        1  2  2
13   14  Jake Beckley             274   18   1  4  2  2              2  3        3  1   
14   15  Cupid Childs             226   19            3  2     2  1     3  2  2  3  1   
15   21  Pete Browning            225   16   2  1  1  1  1  1     1  1  1  2  2  1     1
16   16  Willard Brown            219   20      1     1     2  2     1  2  2  1  3  4  1
17   19  Hugh Duffy               210   15   1     1     3  3     1  3  1     1        1
18   17  Joe Sewell               209   16      1  2     3  1     2     1  1  2  1  2   
19   18  Bobby Doerr              206   15   1     1  1  1  2  3  1        3  1     1   
20   23  Cannonball Dick Redding  199   17      1     1  1  3  1  1  1  1        3     4
21   20  Mickey Welch             193   11   3  1     2  3                    1     1   
22   24  José Méndez              174   14         1  1  1  1  3  2           2  2     1
23   22  Dobie Moore              174   12      1     2  3     2  1        1  2         
24   28  Joe Gordon               164   17         1     1  1  1        1  1  3  1  3  4
25   26  Alejandro Oms            160   12      1  1     2  2        1  1  1  1  1  1   
26   30T Rube Waddell             148   13      1           1  1  2  1     3     2     2
27   25  Bucky Walters            141   10         2  2           1  3  2               
28   32  Charley Jones            134    9   1  1  1           2     2        1  1      
29T  35  Burleigh Grimes          122    9   1        1  1  1        2     2           1
29T  27  Tommy Leach              122    9      1        2        2     2  1  1         
31   29  Gavy Cravath             118    9   1     1     2           1  1     1        2
32   33  Edd Roush                115   11   1                 1  1        1  2  2  2  1
33   34  Wally Schang             113    8   1           2  1     1     1  1     1      
34   30T Roger Bresnahan          107    8      2                 1  1  1  1     1  1   
35   36  Bob Elliott               91    8      1              1        1  2     2  1   
36   38  Quincy Trouppe            90    9            1  1     1        1           2  3
37   37  Larry Doyle               79    6         1  1     1  1              1        1
38   39  Bill Monroe               61    6         1                       1  2  1  1   
39   47  Charlie Keller            55    5               1        2                 1  1
40   41T John McGraw               54    4               1     2           1            
41T  44T Tommy Bridges             53    5                        1  2     1           1
41T  43  Bob Johnson               53    5               1        1     1           1  1
43   41T Dizzy Dean                51    4               1     1        1  1            
44   49  Ernie Lombardi            49    4      1                          1  1     1   
45T  48  Sam Rice                  45    5                        1  1              2  1
45T  40  Jimmy Ryan                45    5                           1        2  1  1   
47   51  Luke Easter               43    3         1        1              1            
48   46  Dizzy Trout               40    3                     2     1                  
49   50  Ed Williamson             38    4                           1  1        1  1   
50   44T Ben Taylor                38    3            1                 1  1            
51   67  Vic Willis                28    3                        1           1        1
52T  53T Addie Joss                27    2                  1        1                  
52T  52  Dutch Leonard             27    2                     1  1                     
54   53T Chuck Klein               25    2                     1        1               
55   55  Carl Mays                 24    2                        1     1               
56   57  George J. Burns           19    2                                 1  1         
57   56  Bobo Newsom               18    2                                 1     1      
58   58  Pie Traynor               18    1         1                                    
59   59  Dom DiMaggio              17    2                                    1  1      
60T  60T Ed Cicotte                17    1            1                                 
60T  60T Fielder Jones             17    1            1                                 
62T  62T Tommy Bond                16    1               1                              
62T n/e  Johnny Pesky              16    1               1                              
64   70  Hack Wilson               15    2                                    1        1
65T  64T Dolf Luque                15    1                  1                           
65T  62T Bobby Veach               15    1                  1                           
67   66  Dick Lundy                14    2                                          2   
68   64T Leroy Matlock             14    1                     1                        
69   71T Frank Chance              13    2                                          1  1
70   68  Fred Dunlap               11    1                              1               
71   69  Lefty Gomez               10    1                                 1            
72T  71T Sam Leever                 7    1                                          1   
72T  71T Rabbit Maranville          7    1                                          1   
Ballots Cast: 49

From 1969-2003, the runs scored in post-season, relative to regular season, by only looking at the teams in the playoffs, is 90%. A team that scores 5 runs will score 4 in the playoffs. A team that allows 4 runs will allow 4 in the playoffs.