Not "ick"...ECK!!!

I expect him to be #1 on my ballot.

Same here.

A roundup of information, issues, and opinions on Eckersley. I admit I come to him with a chip on my shoulder. I just never much liked him.

Regarding relief:
-Career ERA+ in relief of 136, a point better than Smith.

-His relief peak is 5-6 years, then he's mostly pedestrian, especially for a top-shelf reliever. Here's his ERA+ in relief, descending order:
606 237 196 160 137 130 129 107 104 100 97 90

I'd say the top four are legitimately excellent seasons, but 137 and 130 are not great for a closer candidate, and the rest are unimpressive.

Here's Wilhelm, Fingrs and Gossage's best 12
GG 465 246 212 193 180 180 173 173 156 132 127 123
RF 332 196 185 146 135 132 123 126 123 122 119 114
HW 236 195 192 190 176 176 173 170 160 152 141 133

Here's those guys from the same age as Eck's conversion onward:
GG 193 132 127 123 119 117 113 92 84 83 (plus one in Japan)
RF 332 196 143 123 82 79
HW 236 195 190 176 176 173 170 160 133 127 122 103

Eck is clearly the second best after Wilhelm post-32. His relief career alone, however, isn't as impressive as Gossage's or Wilhelm's.

-His peak occured with a dyanstic team: how did his team's defense help him?

-What's the leverage situation looking like? His career LEV is about 1.64, which is right around Fingers and Smith (1.60 and 1.68, respectively), higher than Gossage (1.50), lower than Sutter (1.80), and probably a smidge lower than the top-end modern closer.

-He's one of the first one-and-done closers, guys who were asked for three outs only (yes, I do know that he had more multi-inning appearances than modern closers). In fact, in a sense, he's almost as famous for his usage pattern as Sutter. Does the one-inning usage pattern affect our perception of him (as a modern closer) compared to backlog relievers or HOM'ed releivers?

Regarding starting:
-Is his starting career impressive? There's reason to say not really.
1) Through 1986 his ERA was 3.67 versus a league ERA of 4.07 (park adjusted), for about a 111 ERA+ in 2496 innings from 1975-1986 (if my calculations are correct).
2) Here's his best-worst ERA+ as a SP:
148 146 138 130 116 112 111 102 99 91 89 78

Compare to these SP (min 100 innings):

A) 130 130 129 128 120 119 111 89 89 85
B) 147 144 142 140 140 140 137 134 120 119 115 109 92

A) Don Newcombe (no ex cred.): Eck's better, but not a lot better, ex. cred might well tip it to Newcombe.
B) Tommy Bridges: Bridges mops the floor

3) He was not durable in a noteworthy sense as a SP. He finished in the top ten in innings three teams, each time with the 9th most innings.
4) Though he was on the ERA+ leaderboard four times (including the best figure in 1979).
5) He recieved Cy Young support in only two starting seasons, 1978 and 1979 (4th and 7th) and was 31st in the MVP voting in 1978.

-Any reason to see his defense as having been unsupportive?

Putting it together:
The case for Eck hinges on two things---
1. His relief peak was so darned high
2. His starting was enough shoulder to support his 475 inning relief peak

-In general, pitchers with 3285 innings and a 116 ERA+ have not generally been attractive HOM candidates without big years. Walters is in this vicinity, of course, and he has the big years. Eck's big years are not as impressive by ERA+ or durability.

-Eckersley's relief peak is excellent, but it's short compared to the other relief candidates we have seen, though it's fair to say that he didn't have a full relief career. Adding his SP and RP seasons, his peak is still not quite as long as the peaks of wilhelm and gossage, but about as long as Fingers'.

-He's not overly impressive as a SP. He has four seasons of 130+ ERA+ as a SP, which taken with his RP peak and the 129 ERA+ year as a RP give him an 11 year prime. In that 11 year prime, he pitches but 1406 innings, though with an ERA+ of 150. Still compare to a SP, that's more like a SP's best 5-6 years (at 280 or 230 IP/yr, respectively). But if we take his LEV times his peak-year relief innings, then we can approximate 1.70 x 475 ~ 810, added to his 871 peak starting innings to make 1675 or so innings. That's 280 innings for six years or 240 or so for seven years, including the LEV. So who has six or seven years with a 150 ERA+? Is it rare? I don't know, Ih'm hoping someone else can figure that one out, I've already spent a couple hours on this post!

Eckersley's a guy whom I've often seen decribed by sabermetric types as a HoF mistake, yet it looks like he's probably going to sail into the HoM. I'm not sure where I stand on him myself, but it does at first glance surprise me that people would rank him ahead of Molitor.

A career in four chapters, by RA+ Pythpat equivalent record:

I. 1975-1979: 80-48 (yearly average 16-10)
II. 1980-1987: 86-76 (yearly average 11-9)
III. 1988-1992: 31-9 (yearly average 6-2)
IV. 1993-1998: 18-17 (yearly average 3-3)

Chapter I is a nice peak for a starting pitcher - good rate stats, a little short of being a workhorse. Some comparison points with that 80-48 are some 5-year runs by a few other pitchers:

Tudor, 1982-86: 76-51
Scott, 1985-89: 80-53
Stieb, 1981-85: 93-49 (has something to do with why we elected him)
Tanana, 1974-78: 87-57
Viola, 1986-90: 86-55
Stewart, 1987-91: 78-65
Martinez, 1988-92, 76-51
Hershiser, 1985-89: 84-55
Gooden, 1984-88: 82-48
Saberhagen, 1986-89: 82-48

Chapter II is something I see as having positive value, but it won't much appeal to a peak-prime voter. It's the longest of my chapters, at 8 years, and a pitcher who could do that forever still probably wouldn't make the HoM.

Chapter III is five years of being a lights-out closer. It's the source of his legend. In rate terms he was great (including his walk rate), but it's modern closer usage - it's really not all that many innings, even over five years.

Chapter IV is, well, bleh. A closer with a 100-110 RA+ is a pretty useless creature. And he was that for six years.

Chapter III is indeed spectacular - but what kind of 5-year run did some other people put together? Henke? Wetteland? Gagne was even more spectacular but for just three years. By itself, it's not an HoM cases. Chapter I is very nice - it's a Saberhagen-level peak. Chapter IV is, well, bleh. I have the uncomfortable feeling of his overall case coming down to what you do with Chapter II - the OK-to-good, non-workhorse starter. What to do with that? I haven't decided.

You can If you're a career voter) do for Eckersley the same thing as I did for Caruthers: define him as 2/3 starter and 1/3 closer, then look at Starter-Eckersley (150% of starter figures) and Closer-Eckersley (3 times closer figures.)

That gives you:

Starter-Eckersley: 3744 IP @111, 249-186. Pretty well dead on the borderline, if below only a hair below

Closer-Eckersley: 2370IP@136. Longer and better than both Smith and Gossage, not quite Wilhelm (2254/146)

Then if Starter-Eckersley is lowish-borderline and closer-Eckersley is well in, but not Wilhelm, that suggests he's in with a little to spare, but by no means a slam-dunk.

BTW, Closer-Eckersley had 1,122 saves -- kind of an impressive number, no?

The real Eckersley, in case it's not obvious, being taken as 2/3 Starter-Eckersley plus 1/3 Closer-Eckersley, his HOM position thus being a weighted average of the two.

I supported Parisian Bob and I support Eck. But the fact that both are so far over the top in karl's system probably says mroe about the system than about Bob and Eck. (I once knew a guy named Bob Eck, just for the record.)

Eck will be #2 on my ballot in 2004 and close to being #1.

Bob Eck, however, is not close to making my ballot.

Eck has an argument for the "best pitcher born between years X and Y" but I forget which years X and Y were.

I have his starting career as worthy of the spreadsheet but not the ballot. I've compared him to Jimmy Key, Steve Rogers and Wilbur Cooper. His relief career is top 25 but not in the top 10 - comparable to someone like Jeff Reardon, Bruce Sutter or John Wetteland. The question I ask myself is "If I were a GM would I trade a HoM player to get Jimmy Key and Bruce Sutter?". My answer is yes.

"Course you couldn 't have Key and Sutter on your roster at the same time.

-In general, pitchers with 3285 innings and a 116 ERA+ have not generally been attractive HOM candidates without big years.

Wilbur Cooper territory

13. DL from MN Posted: August 20, 2007 at 10:22 AM (#2492112)
Eck has an argument for the "best pitcher born between years X and Y" but I forget
which years X and Y were.


I have his starting career as worthy of the spreadsheet but not the ballot. I've compared him to Jimmy Key, Steve Rogers and Wilbur Cooper. . . . "would I trade a HoM player to get Jimmy Key and Bruce Sutter?". My answer is yes.

I infer that the comparison turned out in Eck's favor. As a starting pitcher, he is about equivalent to Key, Rogers, and Cooper. How so?
It seems to me that Key has two four-year peaks about equivalent to Eckersley's three-year peak (1977-79 when his maximum innings and maximum ERA+ coincide). Beyond seasons 1-2-3, Eck works no more innings than Key, which is saying something, and he isn't nearly as effective.

How so? PRAA/PRAR.

I think the issue with OCF's chapter III is whether or not Eck's usage pattern is pumping up his ERA+. Here we have a guy with a 12-year career of about 111 ERA+ in a starting role. He goes gangbusters when switched into relief for like six years. Then he plops back into 110 land as he peters out due to age. Which pitcher is he? The 111 ERA+ starter? Or the 606 ERA+ reliever? Is this the ugly head of the "relief ERA bonus" rearing itself again? I think the one-inning relief role could be playing a very big part in this novel, and I'm frankly skeptical. 75 innings just isn't much of a sample, and while Eck may well have been excellent, 606 is so far outside the norm that in any other player people would howl about small samples.

Now to answer my own question....

Here's the NRA1/DERA1 for Eck's big six relief years:


FROM THE BP GLOSSARY: Note that if DERA is higher than NRA, you can safely assume he pitched in front of an above-average defense.

Eck's DERA is higher in 1988, 1989, 1990, 1992. It's only substantially lower in 1987, before the A's started winning. En toto he's 0.40 points of DERA higher than his NRA for the span, or .08 points a year. We can safely say he pitched in front of good defenses. FYI: that's an average of 1.7% of his ERA explained by defense (8pts / by 450pts since DERA and NRA are based on a 4.50 scale). Careerwise, his NRA+ was actually worse than his DERA+: 3.82 to 3.76.

Meanwhile, the DERA+ column is simply 4.50/DERA*100. Compare to the ERA+ column. Look at the non-606 years first. His ERA+ and DERA+ are generally close to one another. 27 points is the outlier for 1989, though the whole thing leans to his having good defenses as I just said. All told 27 points of difference (6%). But the 606 year is 361 points of difference. Screams fluke to me:
-it's a .29 difference between NRA and DERA in the direction of having stellar defense
-it's a 361 point difference between DERA+ and ERA+ suggesting that ERA+ may be wildly overstating his effectiveness.
-If an ERA+ in the neighborhood of 240 is the real number, it's still excellent, but it's not so otherworldly as the 606. In fact, it's exactly in keeping with Gossage and Wilhelm.
-It's merely 73.3 innings. Anything happened.


Someone asked about modern closers and their ERA+. Let's run the top six ERA+s (non-consecutive) of some high-quality modern relievers. Eck starts it for comparison's sake (through 2003 only):

Eckersley: 606 237 196 160 137 130
Rivera: 265 245 242 235 234 191
Henke: 232 209 197 182 182 181
Franco: 229 212 214 173 172 168
D Jones: 229 199 182 182 169 166
Wagner: 277 248 170 166 158 151
Wetteland: 304 248 238 179 156 149
Hoffman: 258 205 178 161 146 143
Montgomry: 281 201 187 161 143 142
Nen: 284 267 208 172 149 133
Percival: 243 226 212 179 132 129
Hernandez: 246 235 174 159 154 121
Olson: 276 225 196 157 143 119
Myers: 291 190 187 139 133 110

YR 1: Eck is first.
YR 2: Eck is fourth.
YR 3: Eck is t-sixth.
YR 4: Eck is tenth.
YR 5: Eck is eleventh.
YR 6: Eck is ninth.

Someone mentioned Wetteland above: he's better in every year except year 1. Eck is certainly in the top half among these guys. But to say his closer career is more HOM worthy than some, is that a stretch? If Eck's peak makes him potentially HOMable, then so does Wetteland's, right? Or Nen's? Would you HOM Wetteland or Nen if they had four years of 130+ starting pitching? How about Henke? How about Franco? How about Doug Jones? Beyond the flukish-seeming 606 (see above), what separates Eck's peak from these guys? (That's not even mentioning that not only is Mariano better, but he's got even more better seasons coming after 2003, so, for that matter do Wagner and Hoffman!)

The piece of evidence in favor of Eck's relief peak is that it comes after age 31, when some of these guys had washed up. But not Jones or Henke or Franco. Wetteland went out on top. Am I just so ornery and have such a dislike for Eck's hair that I'm missing something? (That's possible.) Why is Eck's peak special given the information above, and why are four 130+ years added to it making him a slam dunk?

A note about 1990, and about the fact that I have always used RA rather than ERA. In 1990, Eckersley allowed 5 earned runs and 4 unearned runs. It's not that he had any whole-career tendency to allow unearned runs - it's just that one small sample. But while his ERA+ was 606, I estimate his RA+ as 358 and his equivalent record as 7-1.

Since it's only 9 runs, and only 4 unearned runs, someone with more energy than me could look up the game logs for all of them and apportion blame, if you'd like. I'm not sure it would make much difference to me.

My RA+ numbers, year by year, for Eckersley, for the "chapters" listed above:
Chapter I - peak starter, 1975-79:
147, 107, 124, 138, 156
Chapter II - good but non-workhorse starter, 1980-87:
103, 90, 120, 78, 118, 139, 92, 144
Chapter III - lights-out reliever, 1988-92:
167, 273, 358, 136, 216
Chapter IV - bleh closer, 1993-98:
103, 93, 91, 118, 111, 98

But I realize now that I had 1987 misclassified; I was treating him as a starter when he was already a reliever than. I think I was fooled by the > 100 IP. I have that year as a 144 RA+, equivalent record of 8-4. So I should move that year into Chapter III. Revised totals:

Chapter II, 1980-96, 7 years: 78-72
Chapter III, 1987-92, 6 years: 39-14

A note about 606:

ERA+ isn't like many other rate stats in that the playing time component is in the numerator and not the denominator. That means that averaging ERA+'s needs to be done harmonically. If a guy has a 100 ERA+ one year and a 606 ERA+ the next, that averages to 172, not 303. During the 1990 season, Eckersley's career ERA+ jumped a measley two points.

Secondly, there is a diminishing margin of returns associated with a high ERA+ because you can have one win per game. Using the Pythagorean Theorem:

ERA+ SNWPct
100 .500
110 .548
120 .590
130 .628
140 .662
150 .692
160 .719
170 .743
180 .764
190 .783
200 .800
250 .862
300 .900
350 .925
400 .941
450 .953
500 .962
550 .968
600 .973

Since it's only 9 runs, and only 4 unearned runs, someone with more energy than me could look up the game logs for all of them and apportion blame, if you'd like. I'm not sure it would make much difference to me.

This one is easy. All four UER came in this game:

box

After Out-Error-Out, Eck imploded (2B-1B-BB-2B-IBB-Out) blowing a three run lead and losing the game.

Eck issued only four walks all year, one intentional. That's two of them right there. And the unintentional walk went to ... Julio Franco!

#20: That "diminishing returns" thing makes sense for a starter, but since Eck's value lies in his saves, not his wins, I don't think you can apply that chart to a reliever. Each run he saves is a run which didn't tie or lose the game for his team in a save situation, so I'd suspect the value would be more linear, without the long tail which you assume.

It's a very confusing question. A starter with Old Hoss Radbourn's stamina and an 0.00 ERA would have infinite value, because he'd win every game (presumably you'd just start him until your magic number was zero, about August 15, and then give him the rest of the season as vacation.) A closer with 0.00 ERA and stamina has only a finite value, because in only a limited number of games is 0.00 for 1 inning of any value.

Yeah, and there's also the question of the "reliever bonus" and how in the case of 606 it creates illusions due to
a) the small sample
b) the nature of the role allowing top pitchers to string together unusually high ERA+s that starters never get near.

The list of modern closers I've posted above should offer evidence that while closing is not necessarily easy, it offers advantages over starting in terms of upper-limit effectiveness and the ability to sustain that unusually high level of performance over several seasons.

A question that would be interesting to answer is how often a starting pitcher in the Eck era strings together 75 or so innings of 4 ER (or 9 RA) baseball. Given that Hershiser's record was 58 innings, it seems unlikely to me that many SPs do so, simply because they face lineups twice of thrice through, they fatigue as they do so, and they are generally removed when they become ineffective as opposed to having a one-inning (or so) limit.

Again, what does it mean when a guy is a 111 ERA+ SP (maxing out around 145 ERA+) then suddenly goes nuts for six years in relief?

Gibson strung together 99 innings - I think you can extend that to at least 101 - with 3 runs. Of coruse, that was (a) 1968, when runs were harder to come by, and (b) really unusual.

It is clear that there is a relief bonus in ERA+ (or RA+), and it comes mostly from being able to expend maximum effort on each pitch. If you focus just on the job title of "closer", it seems that the 1993-98 version of Eckersley was below average, and possibly below replacement level. That's a little unfair, of course: you can be below replacement level for the job of closer but above replacement as a 7th-inning reliever. Of course, famous and well-paid closers might not accept that as an assignment.

Was Eckersley as good as his ERA in 1990? Was Fingers as good as his ERA in 1981? In both cases, I have a quick answer: no. No, because no one is that good. (That's another way of saying "sample not large enough.")

From a DIPS perspective, Eckersley's 1990 is positively bonkers. In other years he had gopherball problems.

Gibson strung together 99 innings - I think you can extend that to at least 101 - with 3 runs.

3 in 101 innings is what I recall reading somewhen

Here's Eck's IHR data. I compiled this by hand from BP a long time ago, so it's possible there's errors in it. He rates out very well.


He allowed about 24 fewer IHR to score than the league average. Eck had what you might call a 127 IHR+ (just the league% divided by his %, times 100). Here's how he places among Eck-era closers...(this is through 2005---which is when i looked this info up):


*INHERITED RUNS SAVED ABOVE AVERAGE

Lee Smith is at 132/+41 by the by. One thing you see by looking this way is that some guys just don't get a lot of chances with IHR. Billy Wagner, for instance, appears to have simply been inserted into fewer games with IHR on board. Indeed, in 2004, Wagner inherited all of two runners, and in three other full seasons had just nine. Given his amazing effectiveness and a LEV through 2005 of 1.79, it's hard to believe that his managers felt that he couldn't handle IHR. I'd gather it's just how that particular cookie crumbled.

As you can see above, Eck's lowest season total was 13 in 1997 (for seasons in full relief), and otherwise was under 20 just once (19 in 1995).

Anyway, just thought I'd pass that along.

Sorry about the messy data. I took the tabs out of my excel-dump data, but I guess that wasn't good enough.

In '89 and '90, Eckersley walked almost and no one and yielded only about 0.6 WHIP; contrast about 0.9-1.0 in the two preceding and two following seasons. This is strong evidence that he really was a better pitcher in those two seasons.

He yielded 10 runs and 5 hr in 1989; 9 runs and 2 hr in 1990. (Unearned 0 and 4; earned 10 and 5)
These are small numbers.
Five and two are tiny numbers. Seven home runs divided uniformly between two seasons, modeling no real difference in home run rates, will divide 4 and 3 or 3 and 4, only 55% of the time. Five and two is statistically meaningless. (He pitched more innings in the second, 1990 season. So consider six and two, representing his observed home run rates. Eight home runs divided uniformly will divide 5-3, 4-4, or 3-5 only 71% of the time. Six and two is also statistically meaningless.)

In 1991 he yielded 26 runs and 11 hr (25 earned). Thirteen 1990-91 home runs divided 2 and 11 is statistically meaningful.

--
How sensitive is the ERA+ record to the timing of a single home run?

*Move one home run with two runners on base from 1991 to 1990 and his six-year record by ERA+ would be
137 160 237 378 148 196 ; bold is down and up from 606 130

*Move a 3-run and a 2-run homerun and his six-year record by ERA+ would be
137 160 237 303 163 196
(FWIW, I think 4 and 7 rather than 2 and 11 is probably close to the year-on difference in his vulnerability to the gopherball.)

*These thought experiments are more counterfactual suggested by the ones about dividing home runs uniformly between two seasons. His whole record implies that Eck yielded home runs with fewer than average number of runners on base, and solo home runs are most common anyway.

How sensitive is that?
I don't know. It depends how you, Eric Chalek and others, would respond to the hypothetical record.
137 160 237 303 163 196

Viewed through the favorite glasses, this modified Eck would beat Hoffman and Montgomery five seasons out of six instead of two!

Eckersl: 606 237 196 160 137 130
Hoffman: 258 205 178 161 146 143
Montgom: 281 201 187 161 143 142
new Eck: 303 237 196 163 160 137

*These thought experiments are more counterfactual suggested by the ones about dividing home runs uniformly between two seasons.

*These thought experiments are "more counterfactual" than the preceding thought experiments about dividing home runs uniformly between two seasons.

I presented this look at Eckersley in ballot discussion. Not "fair," perhaps, so Eck backers are welcome to set me straight:


Eckersley
1975-79 (age 20-24) - wow, ERA+s of 148, 146, and 138 this young suggest the start of something big.
1980-84 (age 25-29) - league-average pitcher (avg OPS+ of 99, in fact). oh well.
1985 (age 30) - good ERA+ of 130, but only hurled 169 IP.
1986 (age 31) - 201 IP of 89 ERA+. so much for a comeback. Is he done?
1987-92 (age 32-37) - Wow, great RP, including some off-the-charts all-time seasons.
1993-98 (age 38-43) - subpar numbers for a RP; maybe shoulda quit while he was ahead.

What this says to me, again, is how easy is to to be a modern closer.
If we vote him in ahead of contending SPs, are we saying they couldn't have done this in relief?
Ok, they wouldn't have been THIS good from 1987-9. But conversely, what would Eckersley have done if he stayed a starter? Probably nothing of note.

I might vote for Eck, because he has about half a HOM SP recipe, and a brilliant RP peak, and there isn't a ton of competition left (most of my faves already have been elected).

But I can understand the nose-wrinkling going on before he pops onto various ballots.
And I can't see him measuring up near Molitor at all...

What this says to me, again, is how easy is to to be a modern closer.

Not necessarily, Howie. It could mean that his arm as of 1987 was suited for just relief pitching.

I'm not sure that every average starter could be a world beater as a closer. If that were the case, I think there would be a long line of mediocre hurlers trying to switch over.

BTW, Eck is #1 on my ballot in 2004. Molly is extremely close, though.