Baseball for the Thinking Fan

Login | Register | Feedback

btf_logo
You are here > Home > Hall of Merit > Discussion
Hall of Merit
— A Look at Baseball's All-Time Best

Wednesday, May 04, 2005

Relief Pitchers

Should we maybe start a thread for general discussion of relief pitchers at some point?

- that “definitely immoral” cat :-)

John (You Can Call Me Grandma) Murphy Posted: May 04, 2005 at 05:31 PM | 460 comment(s) Login to Bookmark
  Related News:

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

Page 2 of 5 pages  < 1 2 3 4 5 > 
   101. Backlasher Posted: June 01, 2006 at 06:58 PM (#2047271)
of course, that doesn't mean Steve is wrong :)

No, he's wrong for other reasons.

There is less tolerance for error in not only your closer's duties, but also through your setup man's duties.

And having "an option at hand" really means nothing. In most cases, there is not another option at hand. Hoffman, et. al. have a down year, you suffer.

You could play that game at SS too, and get an improvement over Edgar Renteria with Tony Graffanino.

Relief pitchers can have performance improvements b/c:

(1) They are pretty much guaranteed not to have to go through the lineup more than once; and
(2) They can exert more effort, because their outing is limited.

If you listen to some, they want to do away with the systematic mechanisms that make it so, just so you can carry a third catcher, and cheer wildly when the PA guy announces a pinch hitter.

But the role is a vital role, and anyone that has fulfilled it for a long period of time with success has made a worthwhile accomplishment.
   102. Daryn Posted: June 01, 2006 at 06:59 PM (#2047273)
most MLB managers would not only disagree (at least when it comes to closers), but vociferously disagree.

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


It goes a ways. :)

I think there is a fine line between astutely challenging the accepted truths of baseball to see if they hold any water and ignoring the psychology/intangibles/human history element of the game that has led the MLB managers in this case to see that great closers are anything but fungible. Mike Schooler -- fungible. Mark Davis -- fungible. Trevor Hoffman -- not.
   103. TomH Posted: June 01, 2006 at 07:00 PM (#2047276)
And even Rivera, brilliant as he is, failed as a starter. As did just about every relief pitcher in the universe, including the greatest among them.

1. Given Mo's ability to pitch 3 innings successfully when needed, I *highly* suspect that if he had been given more chances as a starter (even in the last few years), he would have been very successful.

2. Every first basemen in the universe failed as a shortstop and center fielder. They still have value.

What it comes down to is assessing the replacement level for relievers. Mo Rivera is much better than other modern closers / set-up men, and thus provides much value. Managers on other teams should thus be converting one of their 'failed starters' to a successful closer if the didn't have one, and if it were so simple as that. That they don't *strongly* implies that good relievers are Not somehow equated to failed starters. Yes, most relievers are failed starters, because in general teams need SPs more critically, but coaches are wise enough to see that certain types of pitchers (more often, those who have one great pitch) will be more successful in the pen.
   104. Backlasher Posted: June 01, 2006 at 07:09 PM (#2047285)
Young pitchers with certain skillsets are shepherded into reliever roles, but most of the great closers were never given a real chance (say 2 years) to start in the majors. Are you convinced Gagne couldn't have harnessed his stuff to be a great starter? Rivera? I know I'm not. They find early success as closers and get pigeonholed -- becaue, and we come full circle, major league teams appreciate how rare and difficult it is to get a closer who can truly dominate and truly last.


I'm not even sure that it's pigeonholing as much as it is being able to perform a valuable job.

But I agree, the point doesn't stand. Smoltz is certainly looking pretty good as a starter after being a closer. Other than that, the last person I even remember making the conversion was Rick Camp.

You can use common sense and figure out that closers are valuable, or you can use just about any alphabet soup statistic and figure out that closers are valuable.

And if you want to co-opt some fungibility argument about options and replacements etc., you can even argue that since fewer closers have had such sustained runs of success in comparison to starters they are even more valuable. That logic works both ways, but in the end its not that probative.

Elite closers were very great players and worthy of respect.
   105. Backlasher Posted: June 01, 2006 at 07:16 PM (#2047290)
Yes, most relievers are failed starters, because in general teams need SPs more critically, but coaches are wise enough to see that certain types of pitchers (more often, those who have one great pitch) will be more successful in the pen.


Yes, its a question of valuing the contribution in selecting the player.

But I would think in the Merit world, that is moot anyway. I presumed that utilization and talent acquistion, is pretty moot for this purpose.

All that matters is value. And if the reliever provided that value, then he gets credit for it.
   106. sunnyday2 Posted: June 01, 2006 at 07:23 PM (#2047293)
>Where does successful reliever fit on this scale? Or HOM reliever? What would a reliver need to do to get to the HOM starter level?

For 1978 the only answer necessary is: Do what Hoyt Wilhelm did.
   107. Chris Cobb Posted: June 01, 2006 at 07:24 PM (#2047294)
I agree with Steve Treder and Dr. Chaleeko in this debate, but it is surely worth noting that not every starter has the right talent or makeup to flourish as a relief pitcher. The successsful relief pitcher certainly ranks about the failed starter (nobody is thinking about having Jose Lima close ballgames for them . . . ), and the relief ace ranks above the successful starter. Pitchers of Wagner/Wettelend effectiveness are rarer than pitchers who are average starters, I expect.

What a reliever would need to do to get to the HoM starter level is indeed the question.

I think we should start off by accepting that there is no way for a career relief pitcher to get to the level of the all-time great starters: the pitchers who are effective as a good 1990s closer while throwing a starter's share of innings.

So the HoM relief pitcher has to be significantly more effective on a per/inning basis for his career than the typical HoM starter, whom I'd peg off the top of my head at a 3.80 DERA (that may be a little high or a little low: it'd be worth studying) and have one of the biggest careers of a reliever for his era. The greater effectiveness, when coupled with a long (and highly leveraged) career, would balance out the overall lighter workload of the reliever in comparison to the starter.

Wilhelm meets both of these criteria easily. So does Goose Gossage.

Other relievers who may meet or be close to meeing them, or who may meet them by the time their careers are over, include

1970s: Rollie Fingers, Bruce Hiller, Mike Marshall
1980s: Lee Smith, Kent Tekulve, Dan Quisenberry, Bruce Sutter
1990s: John Franco, Trevor Hoffman, Mariano Rivera
2000s: Billy Wagner

Pitchers with a mixture of starting/relieving, like Eckersley, Smoltz, Righetti, and Aguilera, don't fit under this model.

This post doesn't really add anything new, but my sense is that Wilhelm, Gossage, and Rivera are (or will be) clearly over the HoM threshold, and that the other 10 relievers that I've listed need to be examined carefully, because the in-out line will fall either above or within this group. We should focus on ranking these relievers against each other, and figuring out how to integrate that ranking with the rankings of starting pitchers and position players.
   108. DCA Posted: June 01, 2006 at 07:26 PM (#2047301)
All that matters is value. And if the reliever provided that value, then he gets credit for it.

The way I see it, this is important in determining how much value he provides. If any failed starter with good stuff can be expected to put up a 125 ERA+ in a closer role, then an actual closer who puts up a long run of 125 ERA+ seasons really isn't providing much value. Because there are plenty of those guys more or less freely available.
   109. sunnyday2 Posted: June 01, 2006 at 07:32 PM (#2047307)
Most generalizations about relievers are false, not because they are false, but because they are not differentially true. I mean, how many 1B are failed 3B, or 2B are failed SS, or RF who are failed CF. (Oh. Just saw Tom's post.)

And for every Rivera who is a failed starter (!), there are ten closers in the MLs who never started a professional game in their life.

Bottom line: If Hoyt Wilhelm had he 4th highest ERA+ in history in 2200 IP as a starter, where would he rate? Oh, and if he did that as a reliever...? Now, I'm more of a peak/prime voter, so I think he would rate higher if he had done it as a starter across (what?) 7-8 peak/prime seasons. IOW in terms of leverage one year at 21-6, 2.00 is better than three years each at 7-2, 2.00. But OTOH, in Wilhelm's case once you get past 8-10 peak/prime seasons you're now often comparing a 5-3, 2.50 reliever versus an 8-12, 4.50 starter.
   110. Mark Shirk (jsch) Posted: June 01, 2006 at 07:40 PM (#2047312)
Daryn,

The greatest starting pitchers in the world usually are so great because they have 3-4 plus pitches, Randy Johnson and a few other aside. However, as a reliever only 2 plus pithes are needed, in some cases only one. So taking a great starter and making him a reliever won't necessarily result in an even greater reliever because the utility of pitches 3 and 4 are next to zero as a reliever. Only pitches 1 and 2 really matter and there are plenty of failed starters that can excel if they only need two pitches and they get the chance to add 4-5 MPH on their fastball since they dont' have to save their arms for the next inning.

To me your argument seems to come down to MLB managers value closers and so should we. Of course managers value closers because the closer is the last one to pitch and his failure sitcks in the brain a little more, whereas a bad 4th inning by a starter does not. Thsi doesnt' make the job any harder or any more valuable perse, it just means the manager is only remembering what happens last. it is the PERCEIVED value of the closer, not his actual value that has made them $10 million dollar a year players, outside of a select few.

I must say that I really agree with Steve on their fungibility, if no one else could have done Keith Foulke's job the last few years he may still have it, instead both Papelbon and Timlin have proven they can do it.

Sunny,

While this isnt' really towards any of your points but last time I checked, a few days ago i think, Bonds was leading all LFers in Eqa at .332. If we could all work our kinks out like that!
   111. Backlasher Posted: June 01, 2006 at 07:41 PM (#2047314)
If any failed starter with good stuff can be expected to put up a 125 ERA+ in a closer role, then an actual closer who puts up a long run of 125 ERA+ seasons really isn't providing much value. Because there are plenty of those guys more or less freely available.

Again, that sounds like personnel selection. You could look at the 1990s and say, well maybe somebody bigger than 6'0" can play shortstop, therefore we are going to dock all those teeny tiny little guys who did play SS.

And if any failed starter could put up a 125 ERA+ then you would never have anybody in your bullpen that had lower than a 125 ERA+. So I don't think it works out that way.

Its likely that a RP will have some better output than starters b/c of the previously mentioned dynamics.

But to that degree, catchers would have longer careers and hit better if you make them all LFs (and not every C is going to be a LF).

Its probably reasonable to give SS,2b,C, CF some bump when you evaluate them compared to corners. That bump is built in and normalized when you compare them to other SS and 2b.

You can do that in reverse for your 1b, LF, and RF.

But as previously mentioned, many ML players are "failed SS or C" that are playing other positions. If you start making unrealistic changes then you may as well have the Hall Of SS.

And I asked Grandma about this before, and I'm a little surprised to see DER and other projection stats in your forum.

When you are enshrining someone for past performance all that matters is value. Woulda/coulda/shoulda isn't going to do much unless you are having the Hall of Talent.
   112. Daryn Posted: June 01, 2006 at 07:42 PM (#2047316)
Most generalizations about relievers are false, not because they are false, but because they are not differentially true. I mean, how many 1B are failed 3B, or 2B are failed SS, or RF who are failed CF. (Oh. Just saw Tom's post.)

An obvious but great point. As my Grade 11 Geography teacher used to say "To generalize is to tell general lies".

When it comes down to it, does every ballplayer fall into one of two categories:

a) a shortstop
b) a failed shortstop
   113. Chris Cobb Posted: June 01, 2006 at 07:42 PM (#2047319)
If relieving were so much easier than the great starters who become relievers would excel at an even greater level as closers. This didn't happen in the rare examples in the first 80 years of baseball history and it didn't happen with Smoltz -- he was about as good.

Once Smoltz fully recovered from his surgery and adjusted to the closer role, his per inning effectiveness was much higher than his per-inning effectiveness as a starter.

Smoltz top 5 DERAs as starter: 3.23, 3.29, 3.38, 3.43, 3.51
Smotlz's DERA 2001-2004: 3.67, 3.67, 1.52, 2.80.
His overall DERA for those 4 seasons: 2.92

Interestingly, WARP1 sees Smoltz's top performance as a closer (2003) as being of the same value as his excellent 2005 return to starting. Both earned him 7.4 wins above replacement. I see that as a reasonable assessment.
   114. Daryn Posted: June 01, 2006 at 07:47 PM (#2047322)
Randy Johnson and a few other aside.

But that's the point -- Wagner and Hoffman and Johnson and Rivera are outliers. You can't set them aside -- our HoM is filled with guys you'd have to set aside because they had unusual talents.

And Timlin proves the point of non-fungibility as does Howry as does Rhodes as do both Jones' -- lots of people can succeed as a closer for 1 or 3 years -- it is the rare great talent that can do it for 10.

And Papelbon has been a closer for 2 months. Only time will tell if he is Mariano Rivera, Bobby Thigpen, Tom Gordon or someone we don't even remember anymore.
   115. Backlasher Posted: June 01, 2006 at 07:49 PM (#2047323)
Randy Johnson and a few other aside. However, as a reliever only 2 plus pithes are needed, in some cases only one. So taking a great starter and making him a reliever won't necessarily result in an even greater reliever because the utility of pitches 3 and 4 are next to zero as a reliever. Only pitches 1 and 2 really matter and there are plenty of failed starters that can excel if they only need two pitches and they get the chance to add 4-5 MPH on their fastball since they dont' have to save their arms for the next inning.


And that is just false. There is leveraged value to performance. That's been proven, debated, ad nausium, no reason to do it again.

Closers provide value in output and in leveraged output.

Only pitches 1 and 2 really matter and there are plenty of failed starters that can excel if they only need two pitches

And there are plenty of 1b that can provide value because hitting and catching is all the really matter and range doesn't play a part. The can excel by focusing on hitting and not taking the pounding of playing a defensive position. And if they excel beyond a certain level, they are given credit.

And if a SS puts up offensive numbers that are above average for a corner, they are almost always in. If a starter puts up numbers better that are above average for a closer, they are almost always in. And if a closer puts up significant value numbers, then they should be in.

I didn't realize you were enshrining people based on Five Tools or Four Pitches.

if no one else could have done Keith Foulke's job the last few years he may still have it, instead both Papelbon and Timlin have proven they can do it.


(1) If no one else could have done xyz players job, why did someone do it.
(2) Foulke has always blown lots of saves, and has not proudced at a HoF quality reliever rate.
(3) Papelbon looks pretty good.
(4) Timlin has always struggled in a closer role.
   116. Daryn Posted: June 01, 2006 at 07:50 PM (#2047326)
Schilling closed poorly as a failed starter last year -- but the health issues cloud that already small sample size.
   117. Backlasher Posted: June 01, 2006 at 07:53 PM (#2047329)
Interestingly, WARP1 sees Smoltz's top performance as a closer (2003) as being of the same value as his excellent 2005 return to starting. Both earned him 7.4 wins above replacement. I see that as a reasonable assessment.


I don't know about WARP factors, but yes, that is a pretty reasonable assessment. Smoltz provided the same value closing as he did starting. He could provide more starting if he performed at the same level he did at closing, and no one would deny, but its also pretty hard to perform at that same level.

And that DERA isn't going to tell you much, because he did what he did. That WARPy thing, which seems to measure actual value, might lead you to that conclusion.
   118. Cblau Posted: June 01, 2006 at 07:54 PM (#2047330)
Bill James, in the BRJ article noted, found a 0.2 to 0.25 advantage for relievers, due mainly to partial innings. But the real ERA advantage for relievers is much greater than that. Witness-
Name ERA as starter ERA as reliever
Wilhelm (1957 on) 2.68 2.26
Fingers 4.32 2.73
Gossage 4.49 2.77
Eckersley 3.71 2.85
Smoltz (to 2005) 3.35 2.35
Rivera (to 2005) 5.09 2.09
Righetti 3.58 3.38
Guidry 3.32 2.56
Hiller 3.03 2.78
Marshall 4.95 2.97
LaGrow 4.42 3.79

And that's everyone I looked up. Try your own sample. I think DCA is right.

Daryn says in the 1978 ballot discussion that relievers are great at pitching. They aren't. If they were, they'd be starting, and pitching as many innings as their arms would bear. That's like saying Manny Mota or Smoky Burgess were great hitters. They weren't. If they were, they'd have played 9 innings every game, because there aren't enough great hitters to go around. If you have one, you don't leave him on your bench. Same with pitchers.

Daryn also makes the statement that MLB teams value closers highly. Well, in the deadball era, they highly valued players who stole a lot of bases even though they were thrown out 45% of the time. Study after study has shown that using pitchers as closers doesn't result in more wins. Eventually teams will wake up.
   119. karlmagnus Posted: June 01, 2006 at 08:00 PM (#2047340)
More relevant to Hoyt, Wakefield, a solid just-below-HOM starter with some ups and downs, did perfectly adequately as a closer when Gordon went down in 1999. However, the role did not make use of his ability to pitch deep into ballgames and save the bullpen, which is one of his core merits; therefore when a manager with a more-than-room-temerature IQ (Grady Little!) arrived at the Sox in 2002, he wass re-convereted to a starter.

I still like my add 50% to innings, subtract 10 points from ERA+ rule (maybe 100%/20 points for a pure closer like Rivera.) If the numbers look inadquate after that's done, they probably are inadequate. Gossage doesn't have enough ERA+ for one who pitched so few innings, IMHO, though he doesn't miss by much.
   120. Daryn Posted: June 01, 2006 at 08:00 PM (#2047341)
Daryn also makes the statement that MLB teams value closers highly. Well, in the deadball era, they highly valued players who stole a lot of bases even though they were thrown out 45% of the time. Study after study has shown that using pitchers as closers doesn't result in more wins. Eventually teams will wake up.

Just to be clear, I am not arguing that what MLB managers think is determinative of anything. I am just saying it is a factor to be considered, not to be disregarded as irrelevant just because the conventional wisdom has sometimes been wrong in the past. Conventional wisdom is often right -- it is fun and elucidating when we find out it is wrong, but that doesn't mean it is necessarily wrong.
   121. DCA Posted: June 01, 2006 at 08:03 PM (#2047342)
When you are enshrining someone for past performance all that matters is value.

And any measure of value includes, at least implicitly, an essentially arbitrary baseline. Which isn't to say that all baselines are equally valid, but there are quite a few places where one can draw the line, and no place that is obviously correct. For players with the same opportunity, the relative measurement and the absolute measurement lead to identical decisions. But without identical career opportunity the location of the baseline can change the rank order, in many cases quite significantly. And figuring out what is the short- and long-term replacement level performance for a specific role is important, perhaps even essential, to defining an appropriate baseline, which is absolutely essential for any value calculation more specific than a gut feeling.

And if any failed starter could put up a 125 ERA+ then you would never have anybody in your bullpen that had lower than a 125 ERA+. So I don't think it works out that way.

Probably not, that's just an extreme example that I made up. But I wouldn't be surprised if the replacement level closer was 125 or better ERA+ because of the substitution effect.
   122. Mark Shirk (jsch) Posted: June 01, 2006 at 08:07 PM (#2047351)
What kind of ERA+ does the average clsoer provide? Does anyone know this? Would it be helpful to create an ERA+ scale based on closers where say 120 is average (or 115 or whatever)? Maybe we should expand it a little to include modern set up men as well.

Also I am not even sure that there are 10 closers in MLB that have never started a game in their professional lives. If there are it isn't much more. I wonder what the real number is.
   123. Steve Treder Posted: June 01, 2006 at 08:15 PM (#2047361)
There are endless examples of rookies who are and prove to be better than veterans once given the chance. But when your great hitter struggles mid-career or at the end of his career (which may last several years, see Williams, B.) he is not replaced, but instead given an oppportunity to revert back to greatness while keeping his role. The solution is there but they just don't go to it because the urgency is not there.

Poor management decisions aren't what's at issue here. The normative pattern of management decisisons is. And the normative pattern of management decisions shows the best position players and the best starting pitchers with longer careers than the best relievers. Unless one believes that the typical management decision is flawed, there is no explanation for this phenomenon other than that replacements for the best relievers are found more quickly than replacements for the best players at other positions.

No assessment of value of which I'm aware asserts that closers are more valuable than regulars at other positions, including starting pitcher. To assert that closers get replaced more readily than players at other positions because teams can more readily tolerate failure at other positions than at closer is to disagree with the standard positional value assessement. Is that what you're intending to do here?

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

Experience shows us that, in general, when anyone falters, and a reasonable replacement is available, they get replaced. There is simply no reason to believe that if a closer were to be better than his reasonable replacements for 20 years, he wouldn't pitch for 20 years.

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

BECAUSE relieving is so much easier than starting, great starters simply don't become relievers until they're no longer great starters. For a team to take a great starter and convert him to relief would be asinine, and therefore it simply never happens. Smoltz was only moved to the bullpen because his injury status prevented him from starting.

If Rivera decided he wanted to start one year and went 17-9 with a 3.25 ERA would you suddenly say, now I believe he was great all those years. And, do you seriously believe, even with only the two pitches he has, he couldn't do it?

Well, when this happens, be sure and let me know, because of course it hasn't happened. The empirical record we have shows very, very few examples of great relievers being converted to starter at all, much less doing well at it.

And, yes I believe Rivera has been great all these years. He likely could have developed into a good starter, perhaps a very good one. But the likelihood of him being anywhere close to as dominant as a starter as he has been as a reliever is exceedingly small. He certainly wouldn't have put up the ERA and other rate stats as a starter that he has as a reliever. Do you disagree with this? If you don't disagree, then you simply have no choice but to acknowledge that starting is a more challenging role than relieving.

It is also an overstatement to call relievers failed starters. For the past 30 years, there have been two kinds of pitchers -- starters and relievers. Young pitchers with certain skillsets are shepherded into reliever roles, but most of the great closers were never given a real chance (say 2 years) to start in the majors.

It is something close to 100% accurate to call relievers failed starters. The percentage of major league relievers who didn't start in high school, college, and/or the minor leagues is fractionally small. Had they been excellent as starters, they would have remained starters. They were moved into relief because they weren't excelling as starters. This is true in something close to 100% of all cases.

Are you convinced Gagne couldn't have harnessed his stuff to be a great starter?

Of course it's possible that he might have. We can never know what might have happened. But it's far more likely that he wouldn't have developed into a great starter; his actual record is that he hadn't yet developed into an average starter, let along a good one. The number of great major league starters who through the age of 25 had yet to throw a complete game and had a career record of 11-14 with an ERA+ of 98 is not large.

They find early success as closers and get pigeonholed -- becaue, and we come full circle, major league teams appreciate how rare and difficult it is to get a closer who can truly dominate and truly last.

Gagne didn't find early success as a closer. He didn't become a closer until he was 26, in his fourth major league season. He became a closer because his team (correctly or not) gave up on him as a starter. Had Gagne been achieving success as a starter, he wouldn't have been made a reliever.
   124. John (You Can Call Me Grandma) Murphy Posted: June 01, 2006 at 08:17 PM (#2047364)
More relevant to Hoyt, Wakefield, a solid just-below-HOM starter

Wakefield is going to need a very long career (which is possible due to his #1 pitch) to make a real dent on HoM ballots when he's eligible. At this point, however, he has zero chance of induction due to his non-impressive peak.
   125. Steve Treder Posted: June 01, 2006 at 08:20 PM (#2047368)
And for every Rivera who is a failed starter (!), there are ten closers in the MLs who never started a professional game in their life.

No, there aren't. Nearly all closers started at least a few games in the minors, and something close to 100% of them started many games before they turned pro. The number of pitchers groomed as relievers from high school/college who succeed as major league relievers is very small.
   126. Mark Shirk (jsch) Posted: June 01, 2006 at 08:21 PM (#2047370)
BL,

My point wasnt' that closers don't have value. It was merely to refute Daryn's point that great starters woudl have to become some sort of uberslosers for the failed starter/great reliever thing to work. They don't and the reason is that many of the things that make a great starter great, 3rd and 4th pitches, abiliy to throw 250-300 innings (depending on era of course), etc. don't factor into making a starter great. It is almsot as if there is some sort of diminishing returns as to how great you can be if you switch from starting to closing. Sorry if I wasn't clear there.

I may try and take up trying to figure out how high a closer's ERA+ is on average in order to rescale the abnomrally high ERA+'s of closers. Anyone know if there is any record of starting/bullpen ERA splits? I owuld figure that a bullpen's ERA is a decent bit lower.
   127. Steve Treder Posted: June 01, 2006 at 08:23 PM (#2047376)
What kind of ERA+ does the average clsoer provide? Does anyone know this?

The data I compiled forthis article shows the average ERA+ of all major league closers from 1993-2004 to be 149.
   128. Mark Shirk (jsch) Posted: June 01, 2006 at 08:28 PM (#2047386)
Holy ####, that is much higher than I expected.

O fcourse as we argue this and I seem to be on the 'closer's ain't worth a damn' side, I must say that I still believe we should elect 4-5 closers, 5-7 when we figure in active guys.

I think that Wilhelm, Rivera, and Gossage are going to be elected early on. I would then take Quisenberry Eck, and Hoffman.

I don' think that 6 closers/relief ace types is too few or too many.
   129. Backlasher Posted: June 01, 2006 at 08:29 PM (#2047389)
And any measure of value includes, at least implicitly, an essentially arbitrary baseline. Which isn't to say that all baselines are equally valid, but there are quite a few places where one can draw the line, and no place that is obviously correct. For players with the same opportunity, the relative measurement and the absolute measurement lead to identical decisions. But without identical career opportunity the location of the baseline can change the rank order, in many cases quite significantly. And figuring out what is the short- and long-term replacement level performance for a specific role is important, perhaps even essential, to defining an appropriate baseline, which is absolutely essential for any value calculation more specific than a gut feeling.


I'm not sure I understand all of that, so I might not be responding correctly.

I don't see how we are having an issue with absolute or moderate baseline.

You have plenty of measures of value that you can use in the alphabet soup. That is an absolute scale.

If you want a comparitive scale, then you can compare them against other RPs.

As I said, like I told Murphy, I didn't see many of these before. I don't delude myself to determine that I can accurately assess Billy Grabinass in 1914 versus Hoss Murphy in 1919. These players I do know.

If its a scale issue, what grand scale do you have for assessing SS and C.

I presume you look at their value stats, and also make a determination as to how well they compare against SS and C of their generation and history.

What grand scale do you have to compare SP versus LF. I presume you have some alphabet soups to translate into some type of absolute value.

This is no different. You can use your WARP factors, win shares, subjective knowledge, some LI based stat, whatever it is you always use to derive the value component.

Then you can compare the closers versus all otehr people the same way.

Catchers have shorter careers too. I presume you take that into account against LF. I tried to look to see who you had enshrined, but I couldn't find it. I hope this isn't the Hall of Corners and Starting Pitchers.

You get the ability to get a performance boost when you shift right on your defensive rainbow. You get a performance boost when you go from starter to reliever. If you excel enough against your competitors when you have that performance boost, you have a job for a long time, rack impressive counting number, and impressive rate numbers.

Mo Rivera is a very special player no matter how you slice it. Maybe he could have been a starter, maybe he couldn't. Maybe he could have been a SS, maybe he couldn't. It doesn't matter what he could have done; what matters is what he did do.
   130. sunnyday2 Posted: June 01, 2006 at 08:30 PM (#2047391)
>>And for every Rivera who is a failed starter (!), there are ten closers in the MLs who never started a professional game in their life.

>No, there aren't. Nearly all closers started at least a few games in the minors, and something close to 100% of them started many games before they turned pro. The number of pitchers groomed as relievers from high school/college who succeed as major league relievers is very small.

I said professional.

Of course, anybody worth a damn started a game in Little League. Hell, I started a game in Little League.
   131. sunnyday2 Posted: June 01, 2006 at 08:32 PM (#2047392)
>I tried to look to see who you had enshrined, but I couldn't find it. I hope this isn't the Hall of Corners and Starting Pitchers.

Plaque Room.
   132. Steve Treder Posted: June 01, 2006 at 08:37 PM (#2047396)
I said professional.

Sure, but the point remains that the best pitchers in high school and college are the starters. At every level of competition a young pitcher climbs, from high school into college and into the minors, there's tougher competition to get and keep jobs, and at every one of those levels, the best prospects get starting assignments. Every reliever would have remained a starter had he been excelling as a starter.
   133. Backlasher Posted: June 01, 2006 at 08:52 PM (#2047409)
Poor management decisions aren't what's at issue here. The normative pattern of management decisisons is

Then you are in the wrong thread. That type of thing doesn't belong in the HoM. If you are going to be allowed to drag that in here, then I'm going to respond in here.

Unless one believes that the typical management decision is flawed, there is no explanation for this phenomenon other than that replacements for the best relievers are found more quickly than replacements for the best players at other positions.


And it shows that LF have longer careers and are easier to replace than SS. That has nothing to do with value.

On a broad scale its easier to find engineers than executives. That doesn't mean that engineers cannot be executives, or that a really good engineer doesn't provide more value to an organization than an executive.

So one can believe anything they wish, its moot. The premise does not support the conclusion.

Experience shows us that, in general, when anyone falters, and a reasonable replacement is available, they get replaced. There is simply no reason to believe that if a closer were to be better than his reasonable replacements for 20 years, he wouldn't pitch for 20 years.


Experience and science show us that some things have a useful life, and that varies based on wear and tear. We have already discussed how relievers THROW HARDER. We have seen butchers like Weaver that slag pitchers for no good reason.

The useful life of a reliever can be shorter than a starter for any number of reasons. Moreover, what you have asserted just isn't even close to being true. In the real world, there are less fault tolerant positions.

Aircraft pilots will get downed after a certain number of ejections. Air traffic controllers will get reassigned based on real time errors.

That is almost freshman level of algorithm analysis. Some systems have built into them failure not based on aptitude, but based on error or time failure. Hell, the whole field of quality analysis is based on this presumption. Whether Daryn is right about this applying to closers, I honestly don't know, but your rebuttal is not even tenable.

BECAUSE relieving is so much easier than starting, great starters simply don't become relievers until they're no longer great starters. For a team to take a great starter and convert him to relief would be asinine, and therefore it simply never happens. Smoltz was only moved to the bullpen because his injury status prevented him from starting.


That's not even logically consistent. Teams don't do x unless y. Smoltz did x because of z. As can be seen, Smoltz was likely capable of starting before last year. He was actually moved back to starter because of his own demands and a team need at starter (those starters are fungible, they just made one of out of Smoltz, and when Davies went down, right there was Travis Smith, LOL).

Well, when this happens, be sure and let me know, because of course it hasn't happened. The empirical record we have shows very, very few examples of great relievers being converted to starter at all, much less doing well at it.


Huh. Your empiracle record is pretty lean. You have Smoltz that did do that and succeed, and then you have, well you have.

Just discuss fairly and it would be real nice. Very few have been converted. That doesn't support any conclusion more. It equally supports, no need to convert Gagne because he provides more value at closer.

So, well, when you actually have some evidence, let me know when it happens.

It is something close to 100% accurate to call relievers failed starters. The percentage of major league relievers who didn't start in high school, college, and/or the minor leagues is fractionally small. Had they been excellent as starters, they would have remained starters. They were moved into relief because they weren't excelling as starters. This is true in something close to 100% of all cases.


I'd love to see you prove that one. We actually had a discussion on this recently, and the number of college closers getting drafted is increasing all the time. But hey if your going back to HS, why not go back to little league.

I guess most Primates are failed starters too. I started in little league. I just couldn't get my coach to understand the value I'd have as a closer.

We can never know what might have happened.

No, we now what has happened with Smoltz.

But it's far more likely that he wouldn't have developed into a great starter; his actual record is that he hadn't yet developed into an average starter, let along a good one. The number of great major league starters who through the age of 25 had yet to throw a complete game and had a career record of 11-14 with an ERA+ of 98 is not large.


I'm glad you are coming around a little bit. Now, you can see how someone can provide value as a closer, and how that would be diminished as an ACE RELIEVER.

But along the same token, can we do that with 1b and SS. I doubt that Willie McCovey would have made a great SS. We'll certainly never now. Maybe we should just discount all his performance.
   134. Backlasher Posted: June 01, 2006 at 08:54 PM (#2047411)
Hell, I started a game in Little League.

That's what I get for a long response. I missed this. I wonder how many major league players at any position never started a game pitching at all? at high school?
   135. Backlasher Posted: June 01, 2006 at 08:59 PM (#2047415)
Holy ####, that is much higher than I expected.


For the most part, that is what the position probably demands, so it includes both a performance boost, and a selection boost (Daryn's argument).

But I'd come back to my earlier point. Would you look at the average ABC of a SS and average ABC of a 1b and discount all 1b by that amount.

A "failed SS" can provide value at 1B.

A "failed starter" can provide value at closer.

Its the value that you are after.

Heck if we want to use the little league baseline, I played SS in little league too. I imagine most major league players did also.
   136. Max Parkinson Posted: June 01, 2006 at 09:03 PM (#2047418)
Shannon Stewart.
   137. Max Parkinson Posted: June 01, 2006 at 09:04 PM (#2047421)
The joke on Stewart's (lack of) arm strength is significantly less funny when a post appears between it and the setup.
   138. Kyle S Posted: June 01, 2006 at 09:13 PM (#2047427)
Most good to great closers began their professional careers as starters. Of course Gagne, Rivera, and Eckersley come to mind, but others I can think of include Billy Wagner, Francisco Rodriguez, Brad Lidge, Joe Nathan, Keith Foulke, Bobby Jenks, John Wetteland, Lee Smith, etc.

Of course, that's not to say that any crappy starter will become an ace closer; it's just a lot easier to go from a starter to a closer than the other way around, just (as BL says) it's easier to move down the defensive spectrum than up it.
   139. Dr. Chaleeko Posted: June 01, 2006 at 09:17 PM (#2047431)
When it comes down to it, does every ballplayer fall into one of two categories:

a) a shortstop
b) a failed shortstop


I mean this seriously: what about lefthand-throwing RF/LF/1B? Do we say they are failed SS because of natural selection? Or do we say that they are failed CF because they can't cover enough ground?

In other news...
The list of SP ERA vs RP ERA is interesting, but it doesn't really show anything. Why? Because all of those guys could really pitch. If you had a bunch of Randy Kieslers and Dave LaPoints showing similar splits it would be different. My guess is that they would show a split, but that it would be less dramatic because the great relievers were great. When you put players in any situation where they can accrue an advantage from it, the great ones will always gain more than the not-great ones (as a group).
   140. Dr. Chaleeko Posted: June 01, 2006 at 09:35 PM (#2047446)
Smoltz's 7.3 WARP as a SP and a RP does make sense. But there's more context. If most MVP candidates typically fall within 10-15 WARP, then Smoltz's legitimately outstanding relief season is placed into a larger context that suggests that even high, high quality seasons of RP aren't worth much compared to SP and especially position players.

Mariano's best WARP1: 9.3
Wilhelm's best WARP1: 8.0 (not including his two starting years)
Wagner's best WARP1: 7.9
Goose's best WARP1: 10.7
Fingers' best WARP1: 8.7
Hoffman's best WARP1: 8.8
Nen's best WARP1: 9.2
Quisenberry's best WARP1: 8.6
Sutter's best WARP1: 8.8
Willie Hernandez's MVP WARP: 9.1
Eckersley's best WARP1: 8.5
Lee Smith's best wARP1: 8.1
Marshall's best WARP1: 10.7

etc etc etc....

Here's the cream of the relieving crop, and only Gossage and Marshall come in over ten WARP (Goose did it twice), probably the only legit MVP candidates too.

So I think the issue of whether relievers are even all that HOMable is an open question. I know Hoyt's got the Koufaxian innings and ERA+ (which is an especially dubious argument for a knuckleball pitcher), but without the height of the peak, nor even a lot of seasons at peak level, I dunno.

Let's look at it this way. I'm not a WARP guy, but it's worth looking at:

Wilhelm's best ten years
8.8, 8.0, 6.9, 6.5, 6.1, 6.0, 5.7, 5.3, 5.2, 5.0

That's not very impressive. In fact that's substantially less impressive than pretty much every several backlog position-player candidate.

Minnie Minoso
11.1, 9.0, 8.6, 8.4, 7.9, 7.5, 7.3, 6.9, 6.0, 4.8

Joe Sewell
11.5, 10.8, 10.4, 10.4, 10.0, 10.0, 8.6, 7.3, 6.8, 5.4

Dave Bancroft
11.5, 10.6, 9.8, 9.4, 8.5, 8.2, 8.1, 8.0, 7.6, 7.3

Lefty Gomez
11.9, 11.1, 8.8, 7.2, 6.1, 5.6, 5.1, 4.4, 3.7, 2.1

Bucky Walters
13.4, 9.7, 8.7, 8.6, 7.0, 6.4, 5.9, 4.8, 4.6, 3.8

Most relivers, compared to the backlog would fail one of the Keltner questions badly: If this guy was the best player on his team, would the team win a pennant?
The answer is almost certainly no because the closer is normally not the best player on his team. When he is (like with the Devil Rays a few years back, or maybe some recent Padre teams) the team goes nowhere.

So I guess I'm just throwing this one back out there. If Hoyt Wilhelm or any reliever, even in his best years, consistently posts 80% of the value of the top players in the league (who are generally also the competition for the HOM), what's a good rationale for voting for relievers?
   141. Backlasher Posted: June 01, 2006 at 09:46 PM (#2047467)
Smoltz's 7.3 WARP as a SP and a RP does make sense. But there's more context. If most MVP candidates typically fall within 10-15 WARP

You will have to forgive me. I've been looking for a link to the Placque Room, but I can't find it.

So this will beg the question. Don't many of your 2b, SS, CF, and C also put up this same line.

I'm sure you have Hornsby and Wagner that are atypical, but I presume that you are also enshrining those at Defensive positions that don't have as much WARP factor. Hell, I don't even know if WARP factor is a good measure.

I know James tries to context adjust for closers, so you do have Win Shares that you could look at.

But on the generalized point, I probably agree. If Jose Mesa put up his performance longevity and output in the same comparitive rate as a 1b to 1b, you proably would be considering him for your HoM. (and you should). He doesn't even come close in this discussion.

But Mariano Rivera is Joe Morgan for all intents and purposes. And whatever you are doing to little Joe to compare him to someone like Kaline, you should be doing with Mo when comparing him to someone like Blyleven.
   142. Steve Treder Posted: June 01, 2006 at 09:48 PM (#2047472)
The list of SP ERA vs RP ERA is interesting, but it doesn't really show anything. Why? Because all of those guys could really pitch. If you had a bunch of Randy Kieslers and Dave LaPoints showing similar splits it would be different. My guess is that they would show a split, but that it would be less dramatic because the great relievers were great. When you put players in any situation where they can accrue an advantage from it, the great ones will always gain more than the not-great ones (as a group).

For the heck of it, I looked up both Keisler and LaPoint on Retrosheet. Actually, both had a higher ERA relieving than starting. But then I looked up the top 5 comps for each for which we have Retrosheet data. The average career starting ERA of these 12 pitchers (Keisler and LaPoint and 5 comps for each) is 5.45, relief ERA 5.35. A very slight relief advantage, just as you predicted.

Obviously it's a tiny sample, and doesn't prove anything. It does suggest an interesting avenue for study, however, that I think I might pursue sometime in the next couple of months.
   143. John (You Can Call Me Grandma) Murphy Posted: June 01, 2006 at 10:10 PM (#2047505)
You will have to forgive me. I've been looking for a link to the Placque Room, but I can't find it.

The Baseball Hall of Merit Plaque Room
   144. Backlasher Posted: June 01, 2006 at 10:22 PM (#2047521)
Thanks Grandma.

The fits my supposition. I just looked a couple of players at SS. They were picked semi-randomly, I just like their names.

Would you have voted in Jack Glasscock and Luke Appling if they were 1b. I doubt you would. Why shouldn't you treat the relievers the same way. (I really wanted to look at the Biz Markey guy, but I couldn't find him.)
   145. Chris Cobb Posted: June 01, 2006 at 10:23 PM (#2047523)
So I guess I'm just throwing this one back out there. If Hoyt Wilhelm or any reliever, even in his best years, consistently posts 80% of the value of the top players in the league (who are generally also the competition for the HOM), what's a good rationale for voting for relievers?

If (and it remains a big IF) WARP and win shares are correctly assessing the season-by-season value of relievers, then a peak seasonal performance by a relief pitcher will not match a peak seasonal performance by a position player. I think the matter of assessing the value of relief pitchers based on leverage needs more study before we fully trust either metric on this point.

However, if the metrics are right, then the rationale for voting for relievers can be two-fold:

1) career value. Wilhelm, at 97.7 WARP1, tops folks like Bunning and Koufax, and Gossage's 89.0 WARP1 is competitive with the lowest tier of elected starting pitchers.

2) effectiveness. Rate stats do have a place in the evaluation of player quality, and it is in the rate stats that relief pitchers shine. If you include a peak, prime, or career rate stat in your evaluation system, relief pitchers will make up ground that they use on aggregate measures of peak/prime quality.

3) seasonal peak value. This should't be ignored. By WARP1, Gossage, Marshall, and Rivera are not uncompetitive on this measure. Wilhelm isn't competitive here, perhaps partly because WARP is undervaluing him, but perhaps because he didn't have much of a peak. But Eppa Rixey isn't exactly competitive on 5-year peak value, and we elected him. Nor was Bid McPhee, whom we also elected.

My system looks at all three of these measures of merit, and, with no tweaking for relief pitchers taking place at all, it places Wilhelm in an elect-me slot.

It would be silly to say that relief pitchers are much better than starters because their ERA+ stats are better. It is likewise silly to say that starting pitchers are much better than relief pitchers because they throw more innings.

I firmly believe that an exacting assessment of leverage, effectiveness, and usage will show that the very best relief pitchers are of similar value to mid-to-lower-tier HoM starting pitchers.

WARP1, even with a fairly crude leverage assessment, suggests that this is the case.
   146. OCF Posted: June 01, 2006 at 10:27 PM (#2047525)
I know Hoyt's got the Koufaxian innings and ERA+ (which is an especially dubious argument for a knuckleball pitcher), but without the height of the peak, nor even a lot of seasons at peak level, I dunno.

The implied reference to Wilhelm's unearned runs is not really fair. Wilhelm had Koufaxian innings and RA+, spread over 20 years. That's fully charging him for the UER.

But I did say "spread over 20 years," so the second part of this statement, which is about peak, is at least a fair question.
   147. Steve Treder Posted: June 01, 2006 at 10:33 PM (#2047531)
I firmly believe that an exacting assessment of leverage, effectiveness, and usage will show that the very best relief pitchers are of similar value to mid-to-lower-tier HoM starting pitchers.

That's pretty much my gut sense of the situation. While theoretically a reliever could be as valuable as the best starter, no one in reality ever has; you'd have to allow something approaching zero runs and baserunners to do it, or perhaps have the durability to pitch effectively in 120 games a season, or something. Leverage is crucially important, but the role is just so specialized and limited in scope that it reaches a ceiling of practical value well before starters and position players do.

Which suggests that the very best relievers should be represented in the HOM, but only the very best: to elect a sample of relievers equal in size to the sample of starting pitchers or regular position players wouldn't be fair.
   148. Mark Donelson Posted: June 01, 2006 at 10:42 PM (#2047539)
The question Dr. Chaleeko asks has kind of been the elephant in the room on this subject for a while now. I'm reluctant to say that the very, very best of the relief aces aren't HOM-worthy—but why, then do their stats, the stats of the very, very, very best at this role, not match up to those of other HOMers at all? (And it's not just WARP. PRAA, WS, pretty much anything that's not ERA-related.)

I'm having a hard time even making Wilhelm an easy top-of-ballot choice. I want to, and I've been trying to look at all the numbers in any way that will put him where I think he belongs, and...you know, there's just something wrong with that.

So those who have Wilhelm as an easy #1, can you tell me why, in terms of anything but ERA+, which seems clearly biased in favor of relievers for the reasons already discussed here? Sunny, I've read your posts, and as a fellow peak voter I still can't quite see how Hoyt as much more than comparable to guys like Bunning. That'll get him in, of course, but not as a no-brainer.

I still intuitively feel that Wilhelm, Rivera, etc., are no-brainers--but shouldn't there then be SOME way to make the numbers show that? Or should we just say "our intuition is right, let's pick 3, or 6, or 9 of them" and only compare them with each other?
   149. Mark Donelson Posted: June 01, 2006 at 10:45 PM (#2047540)
Hadn't seen Chris Cobb's post yet, which does address some of these questions. Still, as a noncareer guy, there's not a ton there for me.
   150. OCF Posted: June 01, 2006 at 10:57 PM (#2047556)
BL: Biz Mackey, not Biz Markey. If you took all of our (post-1880) catchers and made them left fielders with the same offense and same career length, I think we would have elected at most one of them (Josh Gibson). Even the likes of Cochrane, Hartnett, or Berra needed a "catcher bonus" to be elected as easily as they were. And part of the reason Roger Bresnahan languishes as far down on the backlog list as he does despite some rather nice offensive numbers is that some fair fraction of that offense was accomplished as a centerfielder.
   151. Mark Shirk (jsch) Posted: June 01, 2006 at 11:07 PM (#2047567)
Appling and Glasscock's defensive prowess at SS makes them more valuable. So I guess the answer is at their offensive level they would not have made the HOM. But what does that prove when we are talking about relievers? I don't get it.
   152. John (You Can Call Me Grandma) Murphy Posted: June 01, 2006 at 11:15 PM (#2047578)
BL: Biz Mackey, not Biz Markey.

The Plaque Room is also divided into four parts due to space limitations, so you need to use the M-R link on the home page, BL.
   153. Backlasher Posted: June 01, 2006 at 11:36 PM (#2047599)
Appling and Glasscock's defensive prowess at SS makes them more valuable. So I guess the answer is at their offensive level they would not have made the HOM. But what does that prove when we are talking about relievers? I don't get it.


Their defensive value makes up for the lack of offensive stats. Reliever's leveraged value makes up for lack of innings.

You have some that are advocating penalizing players just because they are "failed starters". You are comparing SS against SS, with little if any real knowledge on the actual value you have defensively quantified in any way.

You should be comparing RP against RP with the value they provide. I think Win shares does this. I don't think WARP factor does.

BL: Biz Mackey, not Biz Markey.

I like my way better.
   154. Rob_Wood Posted: June 01, 2006 at 11:56 PM (#2047622)
Very interesting discussion on one of my favorite topics. My mental model of relievers vs starters is as follows. Requirements of a modern top reliever include being able to pitch several days in a row and many games in a season, can go at max effort/effectiveness from the very first pitch (often with hurried warm-up), premium on strikeouts and/or ground balls, can get the best major league hitters out if only faced once a game. Requirements of a modern top starter include being able to pitch deep into a game, several times through the lineup, and for 30+ starts a season.

Based upon the respective requirements, many starters can be very good relievers, if only for a couple of seasons. Few relievers can be very good starters. Few top relievers can be at the top of their game for many years not only based upon the wear and tear argument but also the fact that there are many others who can do their job almost as well next season. And, as has been raised above, the perceived value of a closer being so high managers are quick to make a change in the face of failure.

In my view, value is best measured relative to a replacement level (what is cheaply available). What is especially vexing for the evaluation of top relievers is that the replacement level is high in the short term (a season, say) but is much lower over the long term (ten years, say) since many pitchers can be a good closer for a season but very few can be a good closer for ten seasons.

This is just the opposite for all other positions! Since finding a good shortstop, say, for next season is very hard (there aren't that many good shortstops cheaply available every off-season). But given time, a team will develop or trade for a good shortstop so that, on average, they will field an average shortstop in ten years time and will have him for an extended period going forward from that point. There is little performance degradation to playing shortstop (maybe catcher) in the major leagues compared to that of top relievers.

If you have Mo Rivera, you can trade Wetteland (can't remember if he was traded or left the Yankees as a free agent) and you don't have to worry about the closer for the next ten years. This does provide the Yankees extra value since they can devote their resources elsewhere in player development, trades, etc.

In short I agree with the sentiment that the very best ever relievers can provide about the same value as a middle HOM starting pitcher. So maybe 3-6 relievers belong in the HOM, not much more than that.
   155. Steve Treder Posted: June 02, 2006 at 12:09 AM (#2047643)
If you have Mo Rivera, you can trade Wetteland (can't remember if he was traded or left the Yankees as a free agent) and you don't have to worry about the closer for the next ten years.

The Yankees let Wetteland go as a free agent following 1996, the season that Rivera had gone to the bullpen and just been ungodly lights-out as the long/setup man.

And of course you're right that Rivera's consistency and longevity has been a huge benefit to the Yankees. But it's always been the case that if Rivera ever did slump or suffer a major injury, the Yankees would have had a far easier time replacing him (likely not with quite his level of effectiveness, of course, but replacing him with at least an average closer) than they would have had if Jeter or Posada or Williams (in his prime) or Clemens or Mussina had gone down. That's no fault of Rivera's, but it is one of the ways in which the value of even the best closer is limited in a way that the value of the best players at other positions isn't.
   156. Dr. Chaleeko Posted: June 02, 2006 at 12:12 AM (#2047650)
WARP1, even with a fairly crude leverage assessment, suggests that this is the case.
Just for clarification, I thought WARP1, like WS, had a built-in leverage indicator of some sort? Chris, or Jimd perhaps, is that true? If these systems already account for leverage, I think there's going to need to be some very serious soul searching about how many relievers we induct---if any.

I've previously been on the wagon with RP, but I'm having a hard time believing in relivers as HOMers after a lot of discussion and data. Consider:
1) Closers exhibit a 149 ERA+ as a group (as cited previously in this thread), and that's the average closer, a guy a little better than Jeff Montgomery or Jose Mesa.
2) We have to ratchet leverage to even get these guys to even appear HOMable...
3) ...and their leverage is determined in good part by their usage which is more manager driven than any other position besides maybe pinch hitter/pinch runner.
4) Virtually as a rule, by the metrics we commonly use they don't produce enough peak impact to make them attractive, and they generally don't produce enough career value to be anywhere near realistic candidates
5) Their job is, by consensus on this thread, less demanding than starting pitching---though not undemanding
6) Their opportunities are not circumscribed by physical conditions like a catcher's are [sorry I don't do infielder credit], so there's no realistic "credit" scenario that can reveal "hidden" value or change an "unfair" circumstance
7) Even the best RP, as noted above, is only theoretically as valuable as a mid-tier HOM starting pitcher...at best.

That's a lot of compelling reasons to think that we are potentially overrating relievers as a group. I've previously been on the record as saying that the optimal number of relivers for us is likely 5 to 7. I'm beginning to wonder if that number isn't 0-3 or maybe 0-2. I could live very happily without Fingers, or Quis in the HOM. I'm not sure about Rivera, Wilhelm, and Gossage (and some others).

You should be comparing RP against RP with the value they provide. I think Win shares does this. I don't think WARP factor does.
BL, that's what I do. I compare everyone at their own position first and foremost. Then I start integrating my ballot thereafter. So Glasscock gets compared to Bert Campaneris before Raj Hornsby or Raj Clemens.

My point in bringing up the cross-positional comparison was simply that all other positions beside RP show similar patterns: the best seasons in 99% of all HOMable cases are worth 10-15 Wins...realistically 10-12. So even if though you may first rank the best relievers among themselves by WARP or by WS, the RP come up shy as a group of the general standards this institution has agreed upon by consensus.
   157. Joey Numbaz (Scruff) Posted: June 02, 2006 at 12:38 AM (#2047681)
"And even Rivera, brilliant as he is, failed as a starter. As did just about every relief pitcher in the universe, including the greatest among them."

That's ridiculous Steve. C'mon. He was never given a chance. In one of his 10 starts at the age of 25 he threw a 2-hit shutout against the White Sox on July 4, 1995.

Then they signed Gooden in 1996, which was completely unnecessary and off to the bullpen he went. He was so lights out (he finished 3rd in the Cy Young voting as a setup man) that the role seemed natural for him, and the team was strong in the rotation, so he replaced Wetteland.

Calling Rivera a 'failed starter' is hyperbole at best. There is little doubt in my mind that he could have become an incredible starting pitcher had he been given the chance.
   158. Joey Numbaz (Scruff) Posted: June 02, 2006 at 12:53 AM (#2047696)
BTW, if Aaron Heilman becomes a closer next week, are we going to call him a failed starter too?

Looking forward to discussing this over a beer or ten with you later this month Steve!
   159. Joey Numbaz (Scruff) Posted: June 02, 2006 at 01:37 AM (#2047750)
"And of course you're right that Rivera's consistency and longevity has been a huge benefit to the Yankees. But it's always been the case that if Rivera ever did slump or suffer a major injury, the Yankees would have had a far easier time replacing him (likely not with quite his level of effectiveness, of course, but replacing him with at least an average closer) than they would have had if Jeter or Posada or Williams (in his prime) or Clemens or Mussina had gone down. That's no fault of Rivera's, but it is one of the ways in which the value of even the best closer is limited in a way that the value of the best players at other positions isn't."


This is true, but the value of Rivera isn't Rivera over Gordon. Because when Gordon replaces Rivera, who replaces Gordon? Tanyon Sturtze. And who replaces Sturtze? Scott Proctor or Wayne Franklin or some other retread scrub. That's where Rivera's value lies - it isn't in him over the #2 guy, it's Rivera over the #5 or #6 guy.
   160. Joey Numbaz (Scruff) Posted: June 02, 2006 at 01:52 AM (#2047782)
"So those who have Wilhelm as an easy #1, can you tell me why, in terms of anything but ERA+, which seems clearly biased in favor of relievers for the reasons already discussed here?"

My revised system takes peak into account, using Pennants Added methodology. So big seasons get proportionally more value a 10.0 is more valuable than two 5.0's, etc..

Without any accounting for leverage, Wilhelm lands in the area with Wic Willis, Wilbur Cooper, Dutch Leonard, Ed Cicotte, Urban Shocker, Rube Waddell, Carl Mays. I don't have the latest updates with me (left the jump drive home by mistake), but from memory, he scored around .650 PA. The players above range from .676-.632. Wilhelm, as expect got less bang for his buck PA compared to Wins than ANY of the starters, because he didn't rack up as many innings in season, so he's already been docked there.

The last two electees, Drysdale and Bunning score at .848 and .843 according to this system. The other 3 that stand out from the pack of starters are Billy Pierce (.776), Tommy Bridges (.724), and Burleigh Grimes (.724). Bucky Walters would be included depending on how much credit you give him for his early career at 3B. He stands at .677 just on pitching. Then there are 13 between .621 and .680 - I think it's clear where the line for the HoM should be draw, right above that pack.

So back to Wilhelm. We haven't given him any bonus for leverage. Giving him a bonus for leverage to only say 1.3 would put him in line with Bunning and Drysdale - it wouldn't be a straight line adjustment, because as you increase his seasonal value, his PA goes up at a higher rate than 1.1.

Again, that's only giving him an LI of 1.3 - I think the case could be made to go to 1.5 without much opposition.

So I think the numbers do justify Wilhelm as an easy inductee.

One caveat, my system was designed for starters, it normalizes IP across eras, and assumes that pitching was just as important to run prevention in 1901 as it was in 2005. I don't think this impacts Wilhelm much though, as the normalization if anything reduces the innings of pitchers from Wilhelm's era, as they threw as many innings as anyone per season since the deadball era.
   161. Joey Numbaz (Scruff) Posted: June 02, 2006 at 01:57 AM (#2047792)
BTW, that system above also includes pitcher hitting, which it turns out can be REALLY important. It's enough (with bonus credit for color line and military service) to move Don Newcombe to be nearly equal with Virgil Trucks, who I really like as a pitcher, but who couldn't hit.

It's enough to vault Grimes over the pack I mentioned. I compare pitchers to how the average pitcher hit that season, since pitcher hitting replacement level is what the average pitcher hits (since no one is chosen to pitch for their hitting).

I can't wait to run George Uhle through the system. I haven't gotten to Ferrell or Lemon yet either, but I'm very comfortable, after looking at the ones I have done that we were correct to throw them over the top based on their hitting.
   162. jimd Posted: June 02, 2006 at 01:57 AM (#2047793)
Just for clarification, I thought WARP1, like WS, had a built-in leverage indicator of some sort?

See XIP, in the Advanced Pitching Statistics. It includes an adjustment that estimates leverage based on participation in decisions, including Saves but not Holds. Rivera has an 1.63 career IP multiplier with this adjustment, and he has usually been above 1.8 as a reliever.

OTOH, Wilhelm's IP multiplier is only 1.14. The relief model under which managers operated during those days did not attempt to accumulate most Saves in one "closer". There are other leveraged innings besides Save innings, but relief statistics such as Holds and not-yet-invented ones such as Ties-Kept and Keep-It-Close's, are not readily available for analysis and inclusion.

With Wilhelm, we are in the swamp.
   163. Joey Numbaz (Scruff) Posted: June 02, 2006 at 02:00 AM (#2047795)
Yeah Jim, it's tough - because while they weren't getting as many saves, they were winning 8-15 games a year which means they were still pitching in many high leverage situations, like coming into a fire in the 7th inning or coming into a tie game in the 8th after the starter was PH for.

What does retrosheet data from the 1970s say about LI's for Goose, Fingers, Hiller, Marshall, etc.?
   164. jimd Posted: June 02, 2006 at 02:19 AM (#2047830)
I'm also having a problem with Wilhelm. My system is based on both WARP and Win Shares, used to measure career, and on their seasonal totals which are used to determine impact seasons, used to measure both peak and prime. The problem is that both mega-measurements see Wilhelm as a wonderful role player, accumulating value over a long career, but not having the impact seasons needed to score well for peak and prime. (In this way he is similar to Chance and McGraw who both had excellent rate stats but problems staying in the lineup.) That is, not enough seasonal IP (or XIP) to leverage those nice rates into lots of runs saved, enough to compete with the other stars for that season. In both 1959 and 1964, he makes it "All-Star" level (top-32 and top-40 respectively), but that's it, and that's not anywhere near enough. And the career total is good but not exceptional, not enough to be ballot-worthy by itself.

Catchers have a similar problem, and I've worked out a catcher's adjustment. Relievers will need a similar one, but it's not yet clear to me how to calculate it, yet.
   165. jimd Posted: June 02, 2006 at 02:37 AM (#2047870)
The other problem is that the adjustment will change over time. As the relief model changes and the Hold stats become available, the mega-measurements will adapt and become better (maybe). Is this a one-of-a-kind adjustment for Wilhelm? Will it be easier to just wing it and treat him like some of the early NeL'ers, vote mostly on reputation and peer opinions, HOM voters' and Wilhelm observers?

Wilhelm may be the toughest vote since the 1920's.
   166. Brent Posted: June 02, 2006 at 03:03 AM (#2047919)
Here's a reminder that about 8 months ago we discussed a lot of these same issues. I'll point in particular to a post by Tango Tiger in which he says the adjustment needed to make ERAs for starters and relievers comparable is about 0.5-0.7. It took a while to find it, but its on the 1960 results thread # 124. Take a look at the thread, there's a lot of useful analysis.
   167. sunnyday2 Posted: June 02, 2006 at 03:06 AM (#2047924)
> So those who have Wilhelm as an easy #1, can you tell me why, in terms of anything but ERA+, which seems clearly biased in favor of relievers

By 0.30.

Wilhelm is #4 all-time on ERA+. Go ahead, dock him 0.30. He's still up there.

Now, rank Mark McGwire for me, but no fair mentioning HR.
   168. Daryn Posted: June 02, 2006 at 03:13 AM (#2047941)
Great discussion on both sides of this issue. It appears that the vast majority think that somewhere between 3 and 9 closers will/ought to end up in the HoM, for a variety of well thought out reasons. That seems fair to me. My original posts were directed at a couple of voters who wouldn't have any closers in baseball history on the top 40 of this weak backlog ballot. That's troubling to me -- almost as troubling as when we elected Hughie Jennings. :)
   169. Cblau Posted: June 02, 2006 at 03:14 AM (#2047943)
Re Rivera:In one of his 10 starts at the age of 25 he threw a 2-hit shutout against the White Sox on July 4, 1995.
Not quite. He only went 8 innings that game, and it lowered his season ERA (all as a starter to that point) to 6.65.
   170. Daryn Posted: June 02, 2006 at 03:17 AM (#2047947)
Not quite. He only went 8 innings that game, and it lowered his season ERA (all as a starter to that point) to 6.65.

That's about where King Felix is 10 games into his career -- but don't give up on him, I think he just might have a future.
   171. Tiboreau Posted: June 02, 2006 at 03:54 AM (#2048018)
Not quite. He only went 8 innings that game, and it lowered his season ERA (all as a starter to that point) to 6.65.

That's about where King Felix is 10 games into his career -- but don't give up on him, I think he just might have a future.


Actually, 12 games into his career Felix Hernandez's ERA was considerably lower: 2.67. Of course, 11 games into his second season it's much nearer Rivera's performance as a starter--5.89--but at this point he's 23 games into his career (with a 3.99 ERA).

;-)
   172. Daryn Posted: June 02, 2006 at 04:07 AM (#2048033)
That's embarrassing. I literally watched (and marvelled at) every game Felix pitched in 2005 and then totally forgot about them an hour ago.
   173. Chris Cobb Posted: June 02, 2006 at 04:12 AM (#2048040)
Mark Donelson wrote:

Hadn't seen Chris Cobb's post yet, which does address some of these questions. Still, as a noncareer guy, there's not a ton there for me.

Well, if your system takes a very narrow view of merit, it's going to miss some players whose merit doesn't fit the system. If the fact that your system misses Wilhelm and looks like it will miss relief pitchers in general is troubling to you, then broaden your system!

jimd wrote:

I'm also having a problem with Wilhelm. My system is based on both WARP and Win Shares, used to measure career, and on their seasonal totals which are used to determine impact seasons, used to measure both peak and prime. The problem is that both mega-measurements see Wilhelm as a wonderful role player, accumulating value over a long career, but not having the impact seasons needed to score well for peak and prime.

I suggest two possible approaches to adjustment.

(1) measure peak in terms of rate as well as seasonal totals. The usage pattern of a relievers is designed to maximize rate, so it makes sense to credit relievers for having talent that fits this usage pattern and benefits their teams.

(2) measure peak for relievers from a baseline of "average" that combines average usage level for non-starting pitchers with average effectiveness for non-starting pitchers. Determining what "average" would be tricky, but some reasonable estimates could be made.

The other problem is that the adjustment will change over time. As the relief model changes and the Hold stats become available, the mega-measurements will adapt and become better (maybe). Is this a one-of-a-kind adjustment for Wilhelm?

Yes, it certainly would change over time. I think that WARP and win shares will become much more reliable, at least for the relief pitchers with whom we are concerned, once the "closer" model begins to replace the "relief ace" model in the mid-1980s. Some closers will probably have overrated seasons because they racked up a lot of rather empty saves (although their weaker peripherals will pull them down), but the top relievers' leveraged usage is pretty accurately reflected by their saves. From 1970-1985, I think the relief aces will be a bit underrated, so I doubt that Wilhelm will be the only one adjusted.
   174. rawagman Posted: June 02, 2006 at 09:59 AM (#2048136)
I just caught up on about 100 posts about this, most interesting, issue.
I, personally, will not be comparing Wilhelm, or any other reliever, to any starter. The hybrids will be looked at as hybrids when their time comes, when we will all be wiser.

Regarding the idea that most closers are failed starters, what seperates them from failed starters who remain failing starters (Lima), failed startes who become little league coaches, failed starters who recover their starting/staying power, middle relievers, halfway decent relievers and ace relievers?

As to the question of whether it is easier to go from starter to closer with success than the reverse, witness Miguel Batista.
He has (so far) put up two seasons with ERA+'s above 130. Once as a starter, once as a swingman. After his second 130+ season (as a starter) the Jays signed him to be a starter. He put up a 101. The next season, they made Miggy their closer. He improved to a 109. Slightly better peripherals, as well.
Now he's back in Arizona (thank God) and starting. I haven't examined the numbers, but I don't think he's been blowing anyone away.

Conclusion, closing ain't easy. It also ain't starting - don't pretend like it is.
   175. sunnyday2 Posted: June 02, 2006 at 11:48 AM (#2048147)
One of the interesting aspects of this is the fact--I think it's a fact--that so few relievers put together a career similar in length to the best players at other positions. (Wilhelm is of course one who did, which is one of the reasons he will be high on my ballot. Failed starter or not, he was very very unique, one of a kind.)

But why do so few put together a career of normal length? And it's not just the classic current-day closer. There are hundreds of pitchers like the Twins' Jesse Crain: Last year 10-2 with an ERA of about 2.00 in a set-up/mop-up/bail-em-out mode. This year 0-2, ERA about 7.00. Maybe he'll get it together but more likely he has already had his brush with fame and from now on he is like the late-career Terry Mulholland--whether for 2 years or 12.

Why is this?

1) They (relievers) really are failed starters--i.e. not intrinsically very good--so that whenever they have a great season it is all just dumb luck? (I don't think this is it because enough of them put together 5-6-7 good years and that can't be dumb luck. But why not more?)

2) The wear and tear (physical? mental?) is that bad?

3) The hitters finally catch up? (This hypothesis has been advanced as a reason why starters have higher ERA--the hitters see them more, and catch up.)

4) Is there a fourth explanation that I'm missing?

Right at this moment I think it's like 1890s IF or catchers generally. It's gotta be wear and tear. I just don't see any other credible explanation. What am I missing?
   176. DanG Posted: June 02, 2006 at 01:22 PM (#2048172)
A few takes:

--I wouldn't use the term "failed starter" to describe closers. Good management uses talent with an aim towards maximizing value - putting players in a position where they can succeed. Bad management tries to force players to be something they are not (Sure, Santo can play secondbase, it's easy!). So many careers have been derailed by bad management failing to utilize talent effectively.

--As has been said, closers are often pitchers with limited stuff. Hitters can and do "catch up" to those most limited in this area. What I have not read here yet is that closers are also often pitchers with limited durability. You can't be a successful starter if your ERA skyrockets once you get beyond 50-60 pitches. But you can be a successful reliever.

--You find many cases where a former starter finishes up his career as a successful reliever. I think some starters are unable to make the transition due to the change in the demands of the position. Less time to mentally and physically prepare to pitch may be hard to adjust to after a lifetime of starting.

--In the end, the analysis must always aim to assess value, measuring how effective a player actually was. Downgrading a reliever because he can't be a starter is as misguided as downgrading another position because they can't be a shortstop. It's the old pitfall of assigning value based upon a hypothetical situation.
   177. DL from MN Posted: June 02, 2006 at 01:33 PM (#2048179)
4) They only have 2 pitches. If they temporarily lose command/control of the 2nd pitch, they've got nothing to fall back on.
   178. Dr. Chaleeko Posted: June 02, 2006 at 01:43 PM (#2048186)
Catchers have a similar problem, and I've worked out a catcher's adjustment. Relievers will need a similar one, but it's not yet clear to me how to calculate it, yet.

I couldn't disagree more on this. The issue with catching is that it is physcially exhausting and injury prone. The problem with relievers is they don't throw enough innings. One is an issue of intrinsic positional limitations, the other a question of usage. Charlie Bennett couldn't play a full schedule because the physical demands were too great. Mariano throws only 65-75 innings a year because that's how we do it nowadays.

In the end, the analysis must always aim to assess value, measuring how effective a player actually was.
DanG, I agree with this assessment, but that doesn't mean that we should elect someone who is a reliever. If he puts up the value, yes, but if they only put up good value for a reliever, that's not good balloting policy. Positional balance is extremely important, except when it comes to positions that aren't and never were equal to the others...namely RP. After all, there's a reason why closers don't make top-shelf money.

A possible model of positional balance viz positional importance:

c, 1b, 2b, 3b, ss, lf, cf, rf, sp
---------------------------------
dh, closer
---------------------------------
setup man
---------------------------------
ph, loogy
---------------------------------
pr, roogy?
---------------------------------
defensive-only replacement?, mop-up man
   179. sunnyday2 Posted: June 02, 2006 at 02:06 PM (#2048197)
>Charlie Bennett couldn't play a full schedule because the physical demands were too great.

Well actually we elected Bennett because he did play damn near a full schedule.

His experience shows that the platoon catcher was basically "the way we do things nowadays."

As long as humans pull the triggers, "the way we do things" is part of it. Our job is to determine how that fits in with everyting else we know.
   180. Daryn Posted: June 02, 2006 at 02:29 PM (#2048217)
5) There is an innings pitched issue. If a middle reliever gives up 5 runs one outing and three runs the next it brutalizes his stats for the year -- MLB teams look at stats and see a fungible player, not realizing that he was successful 80% of the time they threw him out there. A starter can skate by for 10 years being successful only 55% of the time they throw him out there because the stats normalize and there is a lower expectation of success. Relievers are expected to succeed at a much higher rate than any other player on the field and that leads to short careers which in turn highlights the greatness of the very few long careers.

Not once on this thread has someone made the usual comparison to placekickers -- but there are a lot of similarities (the perceived fungibility, the high rate of success expected and the limited necessary skill set being the three most important). Like relievers, the rare kicker who is successful day in and day out for 10 years or more is worth his weight in gold.
   181. John (You Can Call Me Grandma) Murphy Posted: June 02, 2006 at 03:07 PM (#2048263)
I, personally, will not be comparing Wilhelm, or any other reliever, to any starter. The hybrids will be looked at as hybrids when their time comes, when we will all be wiser.

I compared relievers with starters until the late forties, when the combination of their rate and counting stats started to eclipse starting pitchers.
   182. Steve Treder Posted: June 02, 2006 at 03:27 PM (#2048277)
That's ridiculous Steve. C'mon. He was never given a chance. In one of his 10 starts at the age of 25 he threw a 2-hit shutout against the White Sox on July 4, 1995.

Well, Joe, it may very well be the case that the Yankees didn't give Rivera the opportunity they should have for him to develop into a good starter. If they had, perhaps he would have developed into a terrific starter.

But these are simply historical facts:

- In Rivera's 10 starts in 1995, he allowed 64 hits (including 20 XBHs, 8 of which were HRs) in 50 innings, compiling a 5.94 ERA.

- In his 9 relief appearances that year, in 17 innings he allowed 7 hits (although 3 were HRs) and a 4.24 ERA.

- The Yankees decided to add Gooden to their rotation for '96, and move Rivera to the bullpen.

Now, if you object to the term "failed starter" as too harsh to describe this dynamic, I won't argue it. But to describe it as anything other than that:

- A 25-year-old rookie was provided a half-season or so's opportunity to demonstrate what he could do as a starter

- He had a good outing or perhaps a few good outings, but overall his record in that opportunity was quite ineffective

- Meanwhile he showed significantly better effectiveness in his relief outings

- So his team made the decision to stop trying him as a starter, and concentrate him as a reliever from then on

The details differ from pitcher to pitcher, of course, but some variation on this basic theme describes the early careers of the vast majority of relief stars. It just does. We might consider their team's decisions to have been faulty, and suspect that they would have been just fine as starters if only given more of a chance. But the early careers of the vast majority of star starters don't demonstrate early struggles to such an extent.
   183. Backlasher Posted: June 02, 2006 at 03:57 PM (#2048304)
Now, if you object to the term "failed starter" as too harsh to describe this dynamic, I won't argue it.

Good. But I seriously doubt that is the case.

The objection to "failed starter" has nothing to do with "harsh", it just has to do with being incorrect.

Its no more valid than calling Mike Schmidt a "failed shortstop"

or calling Ty Cobb a "failed power hitter."

Rivera was placed into a role where he had some intitial transition difficulties. If that counts against someone, I guess Greg Maddux isn't getting any votes or respect either.

While he was performing, someone noticed a role where he could provide unique value. He accepted the role, and thrived in it, and thrived in a way over an extended period of time that made him unique.

All that matters in this forum is the value that the relievers provided. Whether a term is "too harsh" doesn't even belong in this forum. Whether or not Rivera is a "Failed Starter" or any other reliever is a "Failed Starter" is pretty moot.

You just have this agenda about relief pitchers with tons of unsupportable and wild notions. You bring it to every conversation you can, and then when people show you its not supportable, you slip into the sophistry, defend it like a dog with a bone, vasillate your position, and get further and further away from the point.

Rivera is not a "failed starter". Because of his skill set, he may very well be able to provide more value as a reliever than as a starter. But calling that a "failure" is just incorrect. That is just having an irrational worship over something, in this case SP, and then juding the whole world from that standpoint.

These guys are engaged in a pretty real and serious debate. For the most part they have consensus on the value question, and they now have to decide if translation is necessary because of the inherent difficulty in the job. That is a big, serious, and legitimate question. Your longing for a certain type of baseball or worship of certain positions shouldn't be in that conversation.

Even if Mariano Rivera failed at everything he ever did in life, he didn't fail as a closer; and that has value whether you want to believe it or not.
   184. Daryn Posted: June 02, 2006 at 04:00 PM (#2048308)
Just one thing I wanted to add -- there is just no doubt that Steve's factual proposition that the vast majority of great relievers were failed starters at the major league level or before is a correct proposition.

The question is what to make of that -- is it relevant? I would say it isn't relevant -- being a reliever requires a different skill set physically and mentally and is a valuable/meritorious role.

If it were true that a vast majority of wide receivers or running backs were failed quarterbacks (and, of course, at the little league level there is some truth to this), it would be irrelevant to assessing the value of their performances. Perhaps a better analogy would be if it were true that the vast majority of point guards were failed centers, the great point guards would still be deserving of HoM recognition.
   185. Dr. Chaleeko Posted: June 02, 2006 at 04:07 PM (#2048313)
i don't think steve's "failed starter" nomenclature is wrong or right. i think it is probably often right, especailly in the era before the big strike. but there are very likely more paths to the role than there once were because post-Sutter it's been not only defined but overwhemingly adopted.

So now guys don't get a trial as a starter at the MLB level who probably would have before. Scouts and farm directors' eyes are trained to see what guys have the one/two pitch combo to make them effective in an inning or two. In a way, this is incredibly smart and efficient because it means that many pitchers who would have thrown poor starting innings aren't given those innings so much anymore.

On the third hand, there may also be a lot of guys whose careers are lengthened by medicine, and some of those guys may have been successful starters who've lost arm strength in the process.

I'm sure there's other ways too. Again, steve's not wrong at all, but I think he may be limiting himself to just one of the pathways to successful relieving.
   186. bunyon Posted: June 02, 2006 at 04:19 PM (#2048319)
I know a relief pitcher in college who has been a reliever since high school. I agree that in the past, even the recent past, it is fair to say that most relief pitchers are "failed starters" (that is, at some level they weren't effective starters and moved to the pen) but in the future (not that it matters here) that likely won't be the case.

If you go back far enough in age, every player is either a failed pitcher or shortstop (except MLB starters and shortstops, of course).
   187. Daryn Posted: June 02, 2006 at 04:21 PM (#2048322)
If you go back far enough in age, every player is either a failed pitcher or shortstop (except MLB starters and shortstops, of course).

And, of course, some lefties who never tried to pitch, who become centrefielders or failed centrefielders.
   188. Daryn Posted: June 02, 2006 at 04:26 PM (#2048329)
If there were a LOOGY who could get out every lefthanded batter he faced but could face no more than three batters a game and righties hit .400 off of him and he lasted 15 years with literally a .000 BAA for lefties -- is he a HoMer?

Typical season 50 IP .5 ERA, .2 WHIP, 8 saves, 4 wins.
   189. Howie Menckel Posted: June 02, 2006 at 04:28 PM (#2048333)
I think the Smoltz career is of interest here.
We ultimately will make comparisons on the order of Lee Smith vs Jim Kaat, etc. Hard to do.

But I think that Smoltz bolsters the case that most top-flight SPs can be as effective in a closing role as any of the top closers - whereas few if any of the closers (including, perhaps, Rivera) could have a typical Smoltz SP year.
No, one case doesn't prove anything too firmly, of course.
But it's hard for me to picture an All-Star level SP not being able to be an excellent closer. It's an easier job, frankly.
Being a longterm, above-average SP is a lot harder, I think. It's something to factor in down the road.

Now, I will have Wilhelm No. 1 or 2 on the next ballot (haven't decided yet), and several other relievers also surely will get my elect-me vote.
I'm just not going to get giddy over a guy with 250 saves and a 3.10 ERA over a guy with 200 Wins and a 3.60 ERA, to use a top-of-my-head comparison.
   190. Steve Treder Posted: June 02, 2006 at 04:43 PM (#2048350)
I know a relief pitcher in college who has been a reliever since high school. I agree that in the past, even the recent past, it is fair to say that most relief pitchers are "failed starters" (that is, at some level they weren't effective starters and moved to the pen) but in the future (not that it matters here) that likely won't be the case.

It might or might not turn out to be the case in the future. But in the empirical record of actual major league pitchers over the past 20 years (and to an even greater extent, over the past 40 years), which is the time frame in which the careers of all the relievers the HOM is going to be shortly considering, it is the case.

One can consider it to be relevant or not. I think it's inescapably relevant, as one of the ways in which we try to understand what relief pitching is, as a role in comparison with other roles: how the athletes who perform it have been selected, and what that selection process tells us about the relative scarcity of relief pitching talent, and what that selection process and talent scarcity tells us about how teams typically value the contributions of starters vs. relievers. It's an interesting piece of the puzzle, I think.
   191. Dr. Chaleeko Posted: June 02, 2006 at 04:47 PM (#2048353)
I wonder if at this stage of the evolution of relievers, they are in a sense analogous to the #2 pitchers on early, early pitching staffs---failed #1 starters if you will....
   192. Steve Treder Posted: June 02, 2006 at 04:54 PM (#2048358)
If there were a LOOGY who could get out every lefthanded batter he faced but could face no more than three batters a game and righties hit .400 off of him and he lasted 15 years with literally a .000 BAA for lefties -- is he a HoMer?

Typical season 50 IP .5 ERA, .2 WHIP, 8 saves, 4 wins.


It's an interesting hypothetical, but its very extreme lack of realism dooms it. Because, of course, such a pitcher wouldn't fact 3 LHBs a game. Once his performance record was established, any time his manager brought him in, he'd basically be handing a .400 BA plate appearance (or several) to the opposing team; sure, bringing him in would get Ortiz or Thome or whoever out of the game, but so long as Gabe Kapler or whoever is going to be a guaranteed .400 hitter as Ortiz's or Thome's pinch-hitter, bring it on.

The historical record shows us that the vast majority of LOOGYs face RHBs a majority of the time, precisely because opponents pinch-hit RHBs against them. Therefore, practically speaking, unless a LOOGY is somewhat effective against RHBs, he won't have a job very long, no matter how nasty he is against LHBs.

It's an illustration of the inherently limited value of extreme specialists. Any player with a glaring weakness that must be shielded from exploitation by opponents severely constricts his own team's capacity to deploy him, and requires his own team to enact elaborate support tactics around him. That, I would say, is the opposite of the concept of a highly valuable player, a major star, a HOMer.
   193. Mark Shirk (jsch) Posted: June 02, 2006 at 05:03 PM (#2048363)
Ty Cobb as a failed power hitter is kinda funny since when Cobb was coming up teams didnt' stress power at all.
   194. Backlasher Posted: June 02, 2006 at 05:05 PM (#2048366)
Just one thing I wanted to add -- there is just no doubt that Steve's factual proposition that the vast majority of great relievers were failed starters at the major league level or before is a correct proposition.


Its not a correct proposition wrt Mariano Rivera. Someone pointed out King Felix's stretch this season, but more important look at the start of Maddux's career.

As someone else mentioned, Rivera through a gem right before the conversion.

But the biggest issue is that its completely moot. If the person failed at everything else in life, but succeeded on the criteria you are studying, then it has no bearing whatsoever.

If you go back far enough in age, every player is either a failed pitcher or shortstop (except MLB starters and shortstops, of course).


And if you go back far enough, virtually every male (at least 30 or over) is a failed policeman, fireman, or astronaut by that criteria.

Failure usually requires an exhaution of opportunity and non-acceptable performance. That isn't the case unless you think that Rivera would have no other opportunity. The numbers shown to prove Rivera is a failure aren't far off from 5th starter level, and it was showing improvement.

But again, I didn't think that was the type of thing that you guys talked about in these Merit forums. None of it has anything to do with honoring the performance.

If there were a LOOGY who could get out every lefthanded batter he faced but could face no more than three batters a game and righties hit .400 off of him and he lasted 15 years with literally a .000 BAA for lefties -- is he a HoMer?

Typical season 50 IP .5 ERA, .2 WHIP, 8 saves, 4 wins.


That would seem to be a line that would be worthy of honor. I'd say it depends on how much value that performance has. In any of these cases, contextual usage is not uniform across the space of the career.

Here is the thing, prior to the innovation in bullpen usage, it was pretty easy to see value. Contextual usage didn't vary much between players. With the people that you were looking at, the law of large numbers would catch up to them, and the context of their usage would be similar throughout.

The toll on the players were similar except up the middle. You learned that and most often subjectively accounted for it in your evaluation, and it looks like in one voters case, used an objective translation mechanism (whther the translation is right or wrong is another question.)

Now you aren't facing that anymore. You are dealing with utilizations that are vastly different in context. That context creates both positive and negative influences, and greatly changes the value over the shape of their career.

So you are rightly deciding how to deal with that situation. And personally, I was enjoying reading the discussion.

The mini-problem that you know have is that you have someone who isn't going to help you resolve that problem. You have someone with an agenda about how players should be used, and an irrational promotion of certain roles.

And they are advocating, not discussing. They will:

(1) Show you all the things that would cause you to downgrade the performance. Many times this will be oversold, with little evidence, and an appeal to common sense.

(2) Ignore or try to argue you shouldn't do any of the things that would cause you to upgrade the performance. Many times this will be oversold, and it will make appeals on the absence of non-critical or non-relevant evidence.

(3) Interject the person's own agenda, rather than discussing the decision criteria.

Now, I watched that happen for years on the main board. If you want to overthink your matters of measurement by introducing dogma, its your perogative.

But in the end, I would implore you to make sure you use the same standard consistently. Don't let someone dupe you be talking about "irrefutable" and "empiracle records" when they have no evidence and the premise itself defies common sense. And don't let them send you off on a wild goose chase with "You don't know and I don't know" when you are making very rational inferences and there is tons of support for the proposition.
   195. Backlasher Posted: June 02, 2006 at 05:10 PM (#2048373)
Ty Cobb as a failed power hitter is kinda funny since when Cobb was coming up teams didnt' stress power at all.


Yes, that is the point. It should be equally humorous to label perhaps the person that excelled at his position more than anything else a "failed starter."

Because the environment does stress the need for a closer; the team wanted and needed for him to close; and that is what he did.

His "empiracle record" indicates he could have started. He had Maddux like first year numbers; performed at around a fifth starter level in his first few starts; and showed improvement.
   196. Backlasher Posted: June 02, 2006 at 05:13 PM (#2048375)
One can consider it to be relevant or not. I think it's inescapably relevant, as one of the ways in which we try to understand what relief pitching is, as a role in comparison with other roles: how the athletes who perform it have been selected, and what that selection process tells us about the relative scarcity of relief pitching talent, and what that selection process and talent scarcity tells us about how teams typically value the contributions of starters vs. relievers. It's an interesting piece of the puzzle, I think.


Good, you're on record. If that is your only objective, you will not need to advocate that position while these guys figure out translation mechanisms FOR VALUE.
   197. DL from MN Posted: June 02, 2006 at 05:14 PM (#2048377)
I think you have to place more value on a player who is a mediocre starter but great reliever over a player who is a mediocre starter simply because teams are winning more ballgames by converting mediocre starters into top closers. Therefore, the truly great relievers have a value equivalent to a #2-3 starter. The question boils down to whether or not you believe a #2-3 starter has HoM value if they can perform at that level for an extended period of time. I think the answer is yes. There are examples of this type of pitcher already enshrined (Early Wynn, Eppa Rixey) and more to be debated (Don Sutton, Phil Niekro). By this reasoning, it is the career voter who is more likely to vote for relievers, not the peak voter.
   198. bunyon Posted: June 02, 2006 at 05:14 PM (#2048378)
And if you go back far enough, virtually every male (at least 30 or over) is a failed policeman, fireman, or astronaut by that criteria.

Failure usually requires an exhaution of opportunity and non-acceptable performance. That isn't the case unless you think that Rivera would have no other opportunity. The numbers shown to prove Rivera is a failure aren't far off from 5th starter level, and it was showing improvement.

But again, I didn't think that was the type of thing that you guys talked about in these Merit forums. None of it has anything to do with honoring the performance.


To be fair, that first line is mine and I agree that it isn't really relevant. It was an off-hand remark made by someone who almost always just lurks on these threads. I wouldn't attribute anything I say to the HOM regulars.
   199. Mike Emeigh Posted: June 02, 2006 at 05:17 PM (#2048381)
Ty Cobb as a failed power hitter is kinda funny since when Cobb was coming up teams didnt' stress power at all.


And in fact, Cobb was a pretty good "power hitter" in the context of his era. From 1907-1918 he was never worse than third in the league in slugging percentage (and third only once), and in all of those seasons he was in the top three in ISO as well.

-- MWE
   200. Steve Treder Posted: June 02, 2006 at 05:19 PM (#2048383)
Ty Cobb as a failed power hitter is kinda funny since when Cobb was coming up teams didnt' stress power at all.

Well, they didn't stress home run-hitting power, because of course home runs were so difficult to hit. But they certainly did stress doubles- and triples-hitting power; being able to get the ball into the gaps, drive runners home from first and get oneself into scoring position at the same time, was highly valued. Power hitters were measured by their doubles and triples, and the RBI statistic was invented in this era, precisely as a means of identifying which guys were in fact better at driving in runs than others.

Cobb was, of course, a terrific power hitter. He didn't bat leadoff; he batted in the middle of the order and he was highly valued for his ability to garner doubles, triples, and RBIs, and in the context of the day he was an excellent home run hitter, too. Cobb led his league in doubles three times, triples four times, homers once, RBIs four times, and slugging percentage eight times.
Page 2 of 5 pages  < 1 2 3 4 5 > 

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
Edmundo got dem ol' Kozma blues again mama
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Syndicate

Demarini, Easton and TPX Baseball Bats

 

 

 

 

Page rendered in 1.1635 seconds
49 querie(s) executed