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

Tuesday, January 02, 2018

Mariano Rivera

Eligible in 2019

DL from MN Posted: January 02, 2018 at 09:52 AM | 32 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.

   1. DL from MN Posted: January 03, 2018 at 10:42 AM (#5599984)
Best RP since Wilhelm or best ever?
   2. ERROR---Jolly Old St. Nick Posted: January 03, 2018 at 10:53 AM (#5599996)
Best RP since Wilhelm or best ever?

Best closer ever. It's hard to compare him with Wilhelm, a hybrid who was briefly a starter and used more for long relief than closing.

As a reliever, Wilhelm logged about 600 more innings than Mo, but less effectively, and in a career that spanned some of the lowest offensive seasons since the dead ball era. They're both easily worthy of the HoM, but they're two different breeds of pitcher.
   3. DL from MN Posted: January 03, 2018 at 11:55 AM (#5600069)
Pitcher PWAA WARP2
Rivera 33.1 56.4
Wilhelm 29.9 50.3
Gossage 28.4 44
Fingers 13.1 28.6
Hoffman 19.7 28.9

Numbers are a mix of Dan R WAR (pre 2005) and BBREF WAR.

Comparing to some starting pitchers

Tiant 24.4 56.6
Halladay 34.0 58.6
Santana 34.1 50.2

Koufax 30.1 47.5
Saberhagen 29.9 55.3
Walsh 29.9 52.1
Stieb 27.2 53.2
Dean 26.1 44.7
Cone 26.0 57.0
Appier 23.3 51.2
Ferrell 22.5 54.2 (but bat++)
Pierce 23.6 54
Pettitte 19.3 50.5
   4. Downing Almost Deserves It Posted: January 05, 2018 at 02:00 PM (#5601466)
A pro-Mariano Red Sox fan here.

Yes, we're looking at about 800 extra innings for Wilhelm (including playoffs), but they're not particularly good innings.

Overall, Mariano allowed 353 runs in 1424.67 innings with a below average defense (-0.17 RA9def per BBREF).

Overall, Wilhelm allowed 773 runs in 2256.67 innings with an above average defense (0.18 RA9def).

The difference is 420 runs in 832 innings, good for a 4.54 RA. And I'm not taking into account that the offenses Mo faced in the playoffs were likely even better than the offenses he faced during the regular season. Of course, the two pitchers are different, but they're both pitchers. If I could have the career of just one of them, it would be Mariano's every time, with a pretty significant gap between the greatest ever and someone who I consider third best (I prefer Goose to Wilhelm by a bit).

   5. OCF Posted: January 05, 2018 at 03:30 PM (#5601546)
My basic RA-based equivalent record doesn't really work well for relievers, and I don't trust it - but I ran it anyway.

Rivera: 111-31
Wilhelm: 158-92 (and a 47-61 record is roughly replacement level; by this crude measure, they're close)
Gossage: 124-77
Fingers: 111-78
Lee Smith: 91-53
Hiller: 88-50
Stu Miller: 108-80

OK, most of those are firemen rather than closers. I mostly haven't done this for closers.

Interestingly, Rivera has one decision per 9.04 IP. Which is well within the normal range for starters, even though you expect this to be wildly different, one way or the other. Of course, his 82-60 records is not remotely like his RA+ equivalent record.

I'm generally against electing closers, as it is hard to accumulate enough value in that role. But Rivera is different.
   6. Downing Almost Deserves It Posted: January 05, 2018 at 03:51 PM (#5601569)
#5

For those of us who are newer around here, might you briefly explain what the numbers mean?

I'm neither for nor against electing closers. I'm totally in favor of electing the pitchers with the most value. Rivera fits. I think Gossage barely does, Wilhelm doesn't quite, and nobody else is particularly close.
   7. OCF Posted: January 05, 2018 at 03:59 PM (#5601575)
I compute RA+, season-by-season, then use a Pythag (with an exponent that slides a little depending on run environment) to turn RA+ into W-L percentage. Then take the IP, assign one decision per 9 IP, and multiply that by the percentage to get a single-season W-L. Then add the seasons to get a career record.

For starting pitchers, the biggest weakness is that it doesn't incorporate defensive support. For relievers, I'm not sure it really captures what they do.
   8. Downing Almost Deserves It Posted: January 05, 2018 at 04:07 PM (#5601581)
#7

Ah, so those are records. Got it! I like that you equate to something we all understand, and, unlike W/L record, actually make the numbers mean something (not that even RP W/L record necessarily means a lot, but we can see a very clear difference between Rivera and Fingers, for example, in your numbers).

You say for relievers you're not sure it captures what they do. Okay, but at the very least, it passes the sniff test for me.

Thanks!
   9. 6 - 4 - 3 Posted: January 05, 2018 at 04:10 PM (#5601584)
I'm generally against electing closers, as it is hard to accumulate enough value in that role. But Rivera is different.

If you go with WPA, then top 5 among relievers (per FG):
- Rivera: 55.75
- Hoffman: 32.78
- Gossage: 31.69
- Nathan: 30.73
- Wagner: 28.40

Rivera is simply in a class by himself. And this doesn't take into account what he did in the postseason. WPA isn't a perfect metric, but it does the job with respect to differentiating Rivera from other historically great closers (along with one who really isn't).

Also, by WAA, he's at 32.7, which is near the lower bound of guys who are clearly deserving. He got there by always being above average (his only negative WAA was his rookie year, when he was mostly a replacement-level starter).
   10. Downing Almost Deserves It Posted: January 05, 2018 at 04:16 PM (#5601592)
(along with one who really isn't)


Care to elaborate? To me, the least historically great closer on your list is Hoffman. (I could accept either Nathan or Wagner too, I suppose).
   11. Mike Emeigh Posted: January 05, 2018 at 04:22 PM (#5601598)
Interestingly, Rivera has one decision per 9.04 IP. Which is well within the normal range for starters, even though you expect this to be wildly different, one way or the other.


There are 26 pitchers who have debuted since 1990, who have earned 200 or more saves, and who have relieved in at least 95% of their career appearances (they've started a total of 127 games). These pitchers have earned, on average, one decision every 9.03 innings.

Four pitchers - Joe Nathan, Bob Wickman, Ugueth Urbina, and Jeff Shaw - account for 97 of the 127 starts among this group of pitchers. If I remove those four guys from the analysis, the average goes to one decision every 9.09 IP.

There are seven active pitchers in this list: Fernando Rodney, Francisco Rodriguez, Huston Street, Joakim Soria, Aroldis Chapman, Craig Kimbrel, and Kenley Jansen. Those seven pitchers have been averaging a decision every 9.24 IP. The general trend is upward (i.e. more IP/decision for more recent relievers); Jansen and Kimbrel have 81 decisions between them in a combined total of 947 1/3 IP. Rodney and Chapman, on the other side, have 158 decisions between them in 1248 1/3 IP.

No real insight here, just throwing numbers around.

-- MWE
   12. 6 - 4 - 3 Posted: January 05, 2018 at 04:37 PM (#5601616)
Care to elaborate? To me, the least historically great closer on your list is Hoffman. (I could accept either Nathan or Wagner too, I suppose).

I was referring to Nathan. He had a truly excellent 6 year stretch, but not much more than that. He doesn't come to mind as being among the best closers ever.
   13. Downing Almost Deserves It Posted: January 05, 2018 at 04:43 PM (#5601619)
#12

Not unfair at all, but don't forget 2013. I'm okay clumping the undeserving together, and in my rankings Hoffman and Nathan are nearly indistinguishable.
   14. Howie Menckel Posted: January 05, 2018 at 07:25 PM (#5601722)
fwiw

Wilhelm was a full-time starting pitcher in only one season - 1959, at age 36, when 27 of his 32 appearances were starts (only 25 other total starts in his career). how'd it turn out? he led the AL in ERA at 2.19 and ERA+ at 173.

Gossage was a full-time starting pitcher in only one season - 1976, at age 24, when 29 of his 31 appearances were starts (only 8 other total starts in his career). how'd it turn out? he had a 3.94 ERA and 91 ERA+.

Rivera made the majority of his appearances as a starting pitcher in only one season - 1995, at age 25, when 10 of his 19 appearances were starts (there were 0 other total starts in his career). how'd it turn out? he had a 5.51 ERA and an 84 ERA+.
   15. Kiko Sakata Posted: January 05, 2018 at 07:45 PM (#5601725)
Wilhelm was a full-time starting pitcher in only one season - 1959, at age 36, when 27 of his 32 appearances were starts (only 25 other total starts in his career). how'd it turn out? he led the AL in ERA at 2.19 and ERA+ at 173.


One of my favorite baseball factoids: Hoyt Wilhelm led his league in ERA every season that he pitched enough innings to qualify (he had a 2.43 ERA in 159 innings, all in relief, in 1952). One of the odder "why did nobody led him pitch 200 innings every season?" in MLB history.

None of which has anything to do with Mariano Rivera, who will probably be #1 on my HOM ballot. Sorry.
   16. Jaack Posted: January 05, 2018 at 08:02 PM (#5601735)
The Wilhem vs Rivera question seems a little silly to me. Aside from the fact that they both were pitchers who mostly came in in relief they really can't be more dissimilar. On of them is clearly the best relief pitcher ever and the other is very likely number two (I think a case could be made for Gossage above Wilhelm bot not a stunningly good one. Gossage over Rivera is a non-starter though).

When it comes to voting this year, I think this leaderboard is worth a look. Now this only goes back to 1974, beacuse that's as far back as RE24 goes, but this is a counting stat that does not adjust for leverage (all innings are treated the same) that rates Rivera higher than Halladay. RE24 for starters isn't too far from basic RA - unless a pitcher enters or leaves mid inning, it's identical. For starters it penalizes them for leaving men on base at a more consistent rate - leaving a runner on first with one out is rated the same, no matter if the subsequent pitcher allows him to score or not. For relievers, it give them extra credit for getting out of jams caused by someone else, and it also affects them the same way as starters for leaving runners aboard. I don't currently incorporate RE24 for starting pitchers - since the data isn't complete for most of history and it's so similar to RA9, it's not really worth it - but it's worth a look to better evaluate some candidates.

Rivera and Halladay are neck and neck - Rivera with 346 and Halladay with 341. I'll probably have Halladay 1st and Rivera 2nd on my ballot but this makes it intriguing enough that it's not a slam dunk that Halladay will be first. That being said, when translated from runs to wins (REW), Halladay is infinitessimally better than Rivera (and Smoltz jumps them both).

Others notes from RE24:
-Tim Hudson looks pretty good - certainly worth a deeper dive when he shows up on the ballot. It's possible that Hudson's relievers allowed an inordinate amount of his runners to score, but I'm not exactly sure how to go about checking up on that other than just gamelog to gamelog.
-Andy Pettitte does not look good by RE24, and even worse by REW. He's in contention for my last couple of ballot spots at this point, but this is an issue worth taking note of. Pettitte may be the opposite case of Hudson, where his relievers stranded more of Pettitte's runners than you'd expect.
-Relatedly, Roy Oswalt does okay by RE24. Close to Saberhagen. I'm leaning towards Oswalt over Pettitte for now, but both are in play for my ballot.
-Among relievers, RE24 draws a clear line between Rivera and the rest, and another noteworthy one after Gossage and Billy Wagner, although it's worth noting the missing, pre-1974 years. Gossage only misses 2 poor years, while Fingers misses quite a few good ones, which drops him down to 16th among relievers. Wilhelm is obviously out of the picture entirely.
-REW on the other hand likes Hoffman a bit better, seperating him out from Joe Nathan and Lee Smith, although not quite elevating him to Wagner or Gossage. Overall I see Hoffman and Wagner as dead even, with Hoffman's decent postseason performance maybe pushing him past Wagner and his poor postseason peformances. Nathan I have in a tier below with Lee Smith and Bruce Sutter.

Will we be getting threads for anyone else - I know that personally I think that Helton, Pettitte, Oswalt, and especially Berkman are worthy of discussion. Squint and I can see Miguel Tejada too.
   17. SoSH U at work Posted: January 05, 2018 at 08:33 PM (#5601748)
I think a case could be made for Gossage above Wilhelm bot not a stunningly good one.


I really don't know what that case is. To me, Rivera's an easy HOM/HOF choice, Wilhelm is an arguable one, and while Goose is the best of the rest, he's nowhere close to worthy.

Do the HoMers voters still consider Fingers and Goose worthwhile HoMers, or has there been any reconsideration of their position on short relievers?
   18. Howie Menckel Posted: January 05, 2018 at 08:35 PM (#5601749)
For relievers, it give them extra credit for getting out of jams caused by someone else, and it also affects them the same way as starters for leaving runners aboard.

well, for a starter the latter means someone else pitches. but if a reliever loads the bases with none out, and then walks in the winning run, how does that formula handle that?

I think you're saying that it "credits" still bases loaded and none out just like a starting pitcher would be stuck with, but just checking.
   19. Captain Supporter Posted: January 05, 2018 at 08:55 PM (#5601760)
Rivera made the majority of his appearances as a starting pitcher in only one season - 1995, at age 25, when 10 of his 19 appearances were starts (there were 0 other total starts in his career). how'd it turn out? he had a 5.51 ERA and an 84 ERA+.


Sigh. This stat gets bandied about, usually, although not in this case, by people looking to debunk relief pitchers. Is it that hard to understand that Mariano had not yet developed his cutter at that point? Somehow I think that if he added that pitch to when he was still a starter, it might have made a difference in his performance.
   20. 6 - 4 - 3 Posted: January 05, 2018 at 09:00 PM (#5601762)
Do the HoMers voters still consider Fingers and Goose worthwhile HoMers, or has there been any reconsideration of their position on short relievers?

My impression is that the consensus is now: Gossage, yes; Fingers, no.
   21. Jaack Posted: January 05, 2018 at 09:21 PM (#5601767)
well, for a starter the latter means someone else pitches. but if a reliever loads the bases with none out, and then walks in the winning run, how does that formula handle that?

I think you're saying that it "credits" still bases loaded and none out just like a starting pitcher would be stuck with, but just checking.


RE24 ignores score and inning - it only concerns the base states and how they change.

To give an example

Run expectency (RE) in an inning is generally about .5. The linked Fangraphs article has it at .493 in a 4.15 runs per game scoring envirnoment so we'll roll with that.

Andy Pettitte begins the inning. RE=.493
Let's say he gets a runner out, allows a solo home run, and then a double and is removed for say... Jesse Orosco. The base state with Pettite leaving is 1 out with a runner on second, which has an RE=.644, a difference of .151 runs from when the inning started. He also allowed a run to score, so he is in total -1.151 runs from average.

Orosco enters the game. RE=.644
He immediately allows a single, scoring the runner and is removed for Rivera. Now the base sate has changed from 1 out, runner on second (.644 RE) to 1 out, runner on 1st (.489). That's an improvement, so Orosco gets credited the difference - .155 runs. But since he allowed a run, you subtract that and get him as -0.845 runs.

Rivera enters the game. RE=.489
He gets the two remaining outs. He allowed no runs, so he gets credited the entirity of the RE - .489.

So the final leaderboard is
Pettitte: -1.151
Orosco: -0.845
Rivera: +0.489
Total: -1.507, or essentially, 1.5 runs worse than expected.

Pettitte is punished for leaving the runner on second and the homer, but credited for the one out he got. Orosco is punished for allowing the runner to score, but gets a little credit for give Rivera a slightly easier situation (runner on first instead of second). Rivera gets credit for getting out of the inning.

---
That being said, I'm not sure how it handles walk off situations. I think it just gives the last pitcher credit for whatever the base state would be afterwards - if he walks the bases loaded and then walks in a run with no outs, I think that would be a total of -3.282 runs - the run he allowed, plus the 2.282 runs you'd expect to score in a bases loaded no outs situation. I could be wrong, but I think that's the answer to your question.
   22. Howie Menckel Posted: January 05, 2018 at 09:50 PM (#5601781)
I think it just gives the last pitcher credit for whatever the base state would be afterwards - if he walks the bases loaded and then walks in a run with no outs, I think that would be a total of -3.282 runs - the run he allowed, plus the 2.282 runs you'd expect to score in a bases loaded no outs situation. I could be wrong, but I think that's the answer to your question.

thanks, Jaack. I think we both can agree that the question matters. one reason for RP ERA being lower - aside from how easy it is to only pitch one inning - is that they also can't see the opposition "run up the score" when they are particularly ineffective. the one run scores, and the three runners left behind just leave the field with no outs, in my example. SPs must be jealous.
   23. Jaack Posted: January 06, 2018 at 01:03 PM (#5601902)
I looked at some Trevor Hoffman gamelogs - I can confirm that pitcher are dinged if they leave runners on base in a walk-off situation (A special thanks to one of the Alex Gonzalez's for providing the confirming walk-off double).

In general, ERA/RA9 do not function well at all with partial innings, whether they are due to entering an inning in progress, leaving an inning prematurely, or allowing a walk off run to score. This obviously matters more for relievers than starters, but it's something worth looking into for both I think.

RE24 seems to handle the job admirably, while being identical to RA9 for complete innings. The only problem is the completeness of data. Fangraphs has complete data going back to 1974, while bbref has partial data for 1930-1973, but each site has a different scale (bbref seems to have a lower definition of average, despite both sites having unified replacement level). I think I'm going to phase out RA9 completely for pitchers after 1974 in favor of RE24 converted into a WAR scale (REW is on a WAA scale). I see no reason to favor RA9 over it other than convention. For 1958-1973 the data is mostly complete, so it's probably safe to use there for most pitchers. 1930-1957 are incomplete enough that I'll hold off there for most guys, although I'll have to see the completeness of the data for any viable candidate in that era - Dizzy Dean, Paul Derringer, Bucky Walters, and Don Newcombe to name the main ones that concern me.
   24. Jaack Posted: January 06, 2018 at 01:28 PM (#5601922)
Scratch that, the difference between the fangraphs and bbref RE24 is park/run environment adjustments. That makes me more comfortable just using bbref for it.
   25. The Honorable Ardo Posted: January 07, 2018 at 12:35 PM (#5602355)
My impression is that the consensus is now: Gossage, yes; Fingers, no.
I'm okay with Fingers. I see him and Hoffman as comparable pitchers: other relievers had better rate statistics in short spells, but both displayed exceptional longevity in their roles ("70's fireman" and "one-inning closer"). Rollie's role, though, was more valuable enough than Trevor's to put him in and leave Hoffman about 30th in my ratings.

Rivera, of course, is the exception. He pitched 141 postseason innings - highly leveraged, in a high-scoring era, in front of bad defenses - and allowed 13 runs. That adds up to the largest postseason bonus I've ever given. I'm just deciding if he's #1 on my ballot or #2 behind Doc Halladay.
   26. vortex of dissipation Posted: January 07, 2018 at 09:37 PM (#5602517)
One of my favorite baseball factoids: Hoyt Wilhelm led his league in ERA every season that he pitched enough innings to qualify (he had a 2.43 ERA in 159 innings, all in relief, in 1952). One of the odder "why did nobody led him pitch 200 innings every season?" in MLB history.


One other pitcher almost did that, too, in about the same time period - Stu Miller. Miller only qualified for the ERA title twice. He led the NL in 1958. In 1959, he finished second by one point, 2.84 to 2.83.
   27. Howie Menckel Posted: January 07, 2018 at 10:03 PM (#5602520)
also, Wilhelm's 1952 all-relief IP ERA title was in his rookie year - at age 29

young for a knuckleballer, though
:)
   28. Rally Posted: January 08, 2018 at 09:51 AM (#5602587)
In 1982 Bob Stanley qualified and finished 2nd in ERA. Also led the league in park adjusted ERA+, though the stat had not yet been invented. Stanley pitched 168 innings and perhaps more amazingly did that with only 48 games. Not a single start, but he completed 33 of those 48. A typical Stanley appearance would be coming into the 5th or 6th inning and finishing the game. He had one appearance where he came in with 2 outs in the first, and finished the game.

He only pitched back to back once. On 8-18 he pitched 3 scoreless innings. Next day he pitched 2.2 innings, and gave up 4 runs. They gave him one day off, brought him back and he gave up 8 runs in 2 innings. Looks like they found his limit. He had 4 days of rest before he pitched again it looks like it worked, with a scoreless 6.1 inning performance.
   29. Rally Posted: January 08, 2018 at 10:11 AM (#5602597)
Rivera, of course, is the exception. He pitched 141 postseason innings - highly leveraged, in a high-scoring era, in front of bad defenses - and allowed 13 runs. That adds up to the largest postseason bonus I've ever given. I'm just deciding if he's #1 on my ballot or #2 behind Doc Halladay.


There is no postseason WAR on BBref but I can do an approximation. WPA is listed. For a reliever WPA is generally a close proxy for WAR. For a starter it's not because an average starter has WPA = 0. So for starters just add 2 WAR for each 180 innings. Mariano dominates the adjusted or unadjusted leaderboard like Barry Bonds dominates the intentional walks list.

Postseason WPA-WAR:
Rivera 11.7
Pettitte 6.6
Schilling 5.6
Smoltz 5.1
Glavine 4.8
Clemens 4.6
Lester 4.6
Hershiser 4.0
Bumgarner 3.7
***
Koufax 2.1
***
Morris 1.9

Next best reliever is Wade Davis at 2.7.
   30. DL from MN Posted: January 08, 2018 at 11:07 AM (#5602635)
I would add postseason to the WAA line, not the WAR line. Postseason "replacement" level doesn't really make sense.
   31. Eric J can SABER all he wants to Posted: January 08, 2018 at 11:18 AM (#5602644)
Postseason WPA-WAR

If you weight WPA by the importance of the game to winning the championship, Rivera still laps the field even though he blew a World Series Game 7.

Pitching championship probability added:
Rivera 1.870
Bumgarner 1.326
Fingers 1.207
Morris 1.027
Art Nehf 0.979
Allie Reynolds 0.859
Schilling 0.849
Smoltz 0.844
Herb Pennock 0.841
Gibson 0.821

Koufax is #11. Also of note from the Yankees of that time period, Mike Stanton (the permanent one, not the one who is now Giancarlo) is #12, at 0.810.
   32. Rally Posted: January 08, 2018 at 11:26 AM (#5602653)
I didn't know who Art Nehf was until the daily Jack Morris threads from a few years ago when he was in his final years of eligibility. Nehf did what Morris did, winning a championship clinching game with a 1-0 shutout. The run scored in the first inning, so he didn't have to go extras like Jack did. But this wasn't a game 7, it was game 8.

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
Dock Ellis
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Hot Topics

Syndicate

Page rendered in 0.6744 seconds
42 querie(s) executed