My oh my, how closing times have changed

1988 Score #656 Steve Bedrosian HL / Saves Record

25 years ago today, on June 29th 1987, Steve Bedrosian recorded a save in a 6-5 Phillies win over the Pirates. It was the 19th save of his Cy Young-winning season, and it was the 12th straight appearance where he recorded a save.

Read the last part of that last sentence again…he got a save in 12 straight appearances. That doesn’t sound too unusual, right? Can you believe that at the time, he had just set the MLB record for most consecutive appearances with a save?

The card above actually celebrates that very streak. Read the back of the card, posted at the bottom here. At the time, Bedrosian’s record-breaking performance supplanted that of Sparky Lyle from 14 years prior.

In the 25 years that followed Bedrosian’s record, here are the longest streaks where a pitcher recorded a save in every appearance:

Rk Strk Start End Games
1 John Wetteland 1996-05-31 1996-07-14 24
2 Todd Jones 2005-07-19 2005-09-13 23
3 Lee Smith 1995-04-28 1995-06-25 19
4 J.J. Putz 2011-08-12 2011-09-25 18
5 Lee Smith 1993-05-24 1993-06-28 17
6 Randy Myers 1993-09-03 1993-10-03 16
7 Jose Valverde 2008-08-09 2008-09-10 15
8 Chad Cordero 2005-06-05 2005-07-02 15
9 Jose Mesa 2004-09-26 2005-05-10 15
10 Trevor Hoffman 2001-07-19 2001-09-01 15
11 Kazuhiro Sasaki 2000-06-20 2000-07-28 15
12 Jeff Shaw 1997-08-25 1997-09-19 15
13 Doug Jones 1988-05-13 1988-07-02 15
14 Mariano Rivera 2003-08-19 2003-09-19 14
15 John Smoltz 2002-06-03 2002-07-01 14
16 Rod Beck 1998-06-28 1998-07-26 14
17 Mariano Rivera 1998-06-02 1998-07-11 14
18 Jose Mesa 1995-05-20 1995-06-17 14
19 Jeff Montgomery 1994-07-08 1994-08-09 14
20 Ryan Franklin 2009-08-07 2009-09-05 13
21 Joakim Soria 2008-09-06 2009-04-22 13
22 Francisco Rodriguez 2008-04-14 2008-05-13 13
23 Francisco Rodriguez 2005-09-17 2006-04-10 13
24 Bob Wickman 2005-08-23 2005-09-23 13
25 Trevor Hoffman 2005-04-30 2005-05-29 13
Rk Strk Start End Games
26 Troy Percival 2003-06-08 2003-07-09 13
27 John Wetteland 2000-05-12 2000-06-05 13
28 Jeff Montgomery 1998-06-17 1998-07-24 13
29 Lee Smith 1993-10-01 1994-04-30 13
30 Rod Beck 1993-05-21 1993-06-21 13
31 Dennis Eckersley 1992-04-25 1992-05-29 13
32 Steve Bedrosian 1987-05-25 1987-06-30 13
33 Jonathan Broxton 2010-05-07 2010-05-30 12
34 Jason Isringhausen 2004-08-05 2004-09-01 12
35 Mariano Rivera 2004-05-26 2004-06-15 12
36 Joe Borowski 2003-09-04 2004-04-09 12
37 Troy Percival 2002-06-02 2002-07-03 12
38 Rod Beck 1996-09-28 1997-04-27 12
39 Mark Davis 1988-10-01 1989-04-29 12
40 John Franco 1988-07-05 1988-07-30 12
Provided by View Play Index Tool Used
Generated 6/29/2012.

You can see Bedrock ran his streak to 13 games, but since setting his record, it’s been tied or broken 31 times. Insane.

The way that closers get used now is so exclusively in save situations, that any guy who doesn’t get a blown save for a couple of months will rack up a streak that ties Bedrosian’s 1987 record.

I don’t even really know where to begin with the stupidity of how closers are used today. The dumbest thing of all is a road team saving a closer for when they have a lead. If it’s the bottom of the 9th (or 10th or 11th, etc) and the score is tied, managers just about never use their closers. They “save” him (ironically enough) for the next inning in the hopes that their team scores in the top of the inning so they can then bring in the closer to protect the lead. That means that the manager puts in a lesser reliever to pitch that inning, and if he gives up a run, the game is over while your closer is still sitting on his ass, having never come into the game.

I do miss the days of Goose Gossage, who routinely came in during the 7th or 8th inning to pitch 2 to 3 innings to close out games. It seems to make a lot more sense. I understand, though, that limiting a closer to 1 inning (and usually fewer pitches) means he’ll throw harder, batters have less opportunity to see the pitcher, and generally he will be more effective. But I can’t help but feel that managers have swung too far in the other direction, limiting the user of closers far too much.

When Joe Torre started using Mariano Rivera in the playoffs in the 8th inning, I had hoped that this would carry over to the regular season and that Rivera would be the first of a new breed of closers who would come in for as many as 6 outs. But this hasn’t happened, presumably because on the rare occasions when a closer blows a game in such circumstances because he’s lost an MPH or two off his fastball, managers feel like they would have been better off saving him for the 9th. But the current prevailing strategy has been shown to be monumentally flawed…

It takes time I suppose. I feel pretty confident in saying that in another 10 years time, closers will not be used in such restricted circumstances. Joe Madden will probably be the first guy to figure it out.

If you’ve read this far, you get a prize. Take a look again at the picture of Bedrosian on the front of the card at the top of this post. Notice anything unusual?

There is a Montreal Expo in the background who is not a baserunner, but an infielder. This means that the photo can only be from the 1987 All-Star game, and that must be Hubie Brooks.


My oh my, how closing times have changed — 34 Comments

  1. You know, I keep reading this idiocy over and over again, and I have to make a stand. Yesterday, in the Nats-Rox game, it gets to be the bottom of the 11th and the poor relief pitcher (Stammen) enters his 3 inning of work, and the “closer” (Clippard) who hasn’t pitched in the entire series, isn’t called in. Idiocy, right? NO NO NO NO!!!!!!

    Let’s look at the game and the team situation. The starter (EJax) is pulled in the 4th, and the almost-always-less-than-effective Tom Gorzelanny lets an inherited runner score, and at that point it’s 7-0. The Nats come back with five but Gorzy contributes another two to the Rockies, so it’s 9-5.

    Chances of the Nats winning are still pretty minute, and they have to fly back to Atlanta with no off-day and face record-breaking heat (not to mention it’s a day game and 93 degrees in Denver).

    But when by some Coors miracle the game is tied 9-9 he uses his excellent setup men (Mattheus and Burnett). Davey has now used all his relievers except Stamman, Clippard, and Chien-Ming Wang, a rehabbing starter who pitched the day before.

    So, Stammen pitches the 9th and 10th, game tied (10-10). If you bring in Clippard you are committed to using him not one inning but two. Or more. That is the key. Certainly Clipp could pitch two innings, he’s done it many times before. Then you go into the, what, 13th inning??

    Honestly, in that situation, I’d use a position player before I’d let Clippard pitch and have nothing left in the tank for Atlanta.

    I don’t even really know where to begin with the stupidity of how closers are used today. The dumbest thing of all is a road team saving a closer for when they have a lead. If it’s the bottom of the 9th (or 10th or 11th, etc) and the score is tied, managers just about never use their closers. They “save” him (ironically enough) for the next inning in the hopes that their team scores in the top of the inning so they can then bring in the closer to protect the lead. That means that the manager puts in a lesser reliever to pitch that inning, and if he gives up a run, the game is over while your closer is still sitting on his ass, having never come into the game.

    • There are certainly situations where saving a closer makes sense, even if he doesn’t get used, and things like schedule (doubleheaders, off days) play a big role in that. My comments have more to do with the opposite of what you’re saying, where managers use closers without really thinking about what makes the most sense, or because some arbitrary set of circumstances (i.e. save situation) exists.

      • OK. But when a good manager, maybe the best manager in baseball (he out-managed “that guru over there” and it wasn’t even close) doesn’t bring in his best reliever, then you have to inquire as to the reason.

        It’s not that managers don’t want to use their best relievers in a close game, it’s that they don’t want to use them in what, statistically, is a losing cause, namely an away game tied in the 9th or later.

        • The Nationals probably aren’t the best example because the difference between Stammen and Clippard isn’t all that large (though there is a larger difference between a fresh Clippard and tiring Stammen). But this game is definitely a poor example (even by anecdotal evidence standards) since the typical close game in the latter innings doesn’t start with the starter only getting 9 outs most of the bullpen being used to piece together the game. Clearly, these types of situations call for extending the few remaining pitchers to their reasonable limits.

          I can’t agree at all with your contention that you would use a position player before using your closer for at least one inning. On the other hand, since we’re specifically talking about the Nationals I think most of us would be interested to see Rick Ankiel take the mound again (note: I’m a Cardinals fan and this is genuine interest, not morbid curiosity).

          • I was sort of kidding about the position player (I have to check if Chien-Ming Wang was warming).

            As regards Ankiel, there have been chances, and questions from the fans, but it’s not happening. I wouldn’t be surprised if he has it in his contract or at the very least a verbal agreement. If you watch him play you can see that mentally and physically he is an outfielder and that asking him to pitch at this point would be no different than asking any other outfielder (e.g., Jayson Werth, who also has a good arm) to pitch–and it would damage Ankiel’s psyche. Anyway, Ankiel is playing injured so it’s moot.

            My point is that in any game, not just this one, you are asking the reliever who is conditioned to pitch one inning or maybe 1+ inning to come in and pitch at least one scoreless inning, and then another inning. If you score, then the “best” reliever (now possibly tired) is pitching his second inning, or you are bringing in the lesser reliever anyway to “close.”

            And there is no guarantee that you score. So, and this is sort of game-theory, do you reserve (“save”) your best resource for when you have scored and optimize your chance of winning the game, or do you utilize that resource in a tie situation where even his best effort doesn’t guarantee a win?

            The stats (up to 2007 on baseball reference) are that the home team wins extra innings 53% of the time. I am thinking of updating the stats to include games tied in the 9th and up to 2012 and writing a blog on this, because I’m not as quick as most analysts to think managers are stupid.

          • Rick Ankiel taking the mound would be must-see action. I would love nothing more than him entering as a pitcher and striking out the side. After all he has persevered through, he deserves it.

          • Tony La Russa stated he would have used Ankiel to pitch in the 20-inning game the Cardinals played against Mets in April of 2010 had he still been on the roster.


            Sorry, but I don’t have a link to the quote.

            It would be interesting to see if managers handled extended extra inning games differently depending upon whether the game took place in the 1st, 2nd or 3rd game of a series. Playing excessively long games might compromise your chances of winning subsequent games if you have a new opponent, but this effect would even out some in the first game of a series.

          • Sorry but ONE instance does not a tale tell.

            Yes there my be circumstances where it does not make sense to bring in your best pitcher in a tie game (like where he would be due to bat in the next inning in the National League and others). But normally you are far better off using your closer in a tie game in the 8th inning than in the 9th inning with a 3 run lead. There is overwhelming statistical proof of this.

            I also don’t agree that “stretching” a pitcher from 1 inning to 2 or even 3 is any big deal (unless maybe they’re just coming off an injury or something). Starters go 3 innings or longer after just a couple of weeks in the spring. If you can do that after 2 months, you don’t belong in the majors. Up until the 80’s a lot of relief pitchers pitched 140 innings or more in a season. That may be a bit much but many, many relievers who had long, productive careers- Rollie Fingers, Goose Gossage, Dan Quisenberry among many others- routinely pitched around 120 innings per year with no ill effect.

            And the best part is if you can routinely get 2+ innings out of your 3 best relievers than you can get rid of your 2 worst relievers and replace them with hitters or fielders who would have a far greater impact on producing a positive outcome.

      • About the only thing worse is when a game is like 2-1 in the 9th and they manager pulls out the starting pitcher once he reaches the magical 100 pitch limit. You know, when a guy’s arm will surely undergo irreparable damage if he throws another ball.

        Or when they pull a starting pitcher once the winning run is at the plate because “he’s pitched well and doesn’t deserve the loss.” Since when is managing to statistics a valid process??? That’s how I feel the Saves thing is… like the closer will pout if he doesn’t get his 30 saves and $10M incentive for finishing 50 games or whatever. Why is that the manager’s problem?

    • I disagree.
      Bring in Clippard.
      Win today. Win today. Win today.
      Tomorrow is tomorrow.
      You might win 11-0 tomorrow.
      And besides, if you’re stretched thin, you’ve got arms at AAA to meet you in Atlanta.

      • I agree with Voomo, for several reasons:

        — Of all the basic entry situations, a tie game on the road is the highest leverage. The value of a run prevented, when a run means a loss, is higher than in any other basic situation. That’s the main problem with modern closer usage: The basic underlying thinking is just wrong.

        — Clippard has not pitched since June 23 (and then he threw just 11 pitches then, after a day off). By not using him, Davey insured that Clippard would have at least 5 days between appearances. That’s certainly counter to the prevailing thought.

        — Clippard is perfectly capable of pitching 2 innings, having done so 14 times last year alone, allowing a total of 2 runs in those games. Sure, he might need a day off afterwards, so you wouldn’t have him for the first game in Atlanta. But you might not need him then. Win the game in front of you — not at any cost, but within reason.

        I would have pulled Stammen after he walked Giambi, before he faced Fowler (whom he walked in the 9th). He’d thrown 31 pitches by that point, including 3 batters with the winning run in scoring position. I want a fresh arm to maximize the K possibility.

    • Here’s a link to an interview with Cippard from Fangraphs author David Laurilia. They touch on Tyler’s views on the role of the closer:

      By the way, this is part of a great season-long series of interviews by Laurilia, who writes out of Boston. Click on Recent Stories and then you can click on his name from the list of authors to get an archive of his tremendous work. Here’s another gem, for all you Joe Madden fans on here, an interview with bench coach Dave Martinez:

  2. The one inning closer model probably makes more sense the better the team is. One reason to use a pitcher for shorter outings is that you’ll be able to use him more frequently (without increasing the presumed risk of injury or ineffectiveness from multiple innings in close proximity). The need to use your closer to protect leads multiple days in a row (or 3 out 4, etc.) is only important if you have these small leads frequently. It would be interesting to see a manager of a weaker team experiment with using his closer for longer stretches to preserve the infrequent leads that his team has in the 7th or 8th innings. Of course a lot of this is dependent upon the skill differential between the closer and the rest of the bullpen.

    Unfortunately it seems to be harder for weak teams to experiment with doing things differently. It seems like there has been quite a bit of criticism of the way Jim Tracy is employing his starting pitchers/bullpen. I don’t know if what he’s doing is a good idea or a bad idea, but I applaud him for trying something different with a Rockies team that doesn’t have a whole lot to risk in terms of playoff expectations. I assume Joe Maddon would receive more criticism for his various innovations (on and off the field) if he were managing a weaker team.

  3. P.S. Ankiel made a great play in Rox game…

    Bottom line, it’s an uphill battle for the away team to win a game that’s tied after the middle of the 9th. Davey uses Clippard–if available–to pitch the 9th at home in a tie game (not a “save” situation). His strategy is to bring Clip in for the 9th, and if that doesn’t work, he goes in descending order of quality of available short relievers for the tied innings, depending on when the pitcher’s spot comes up can be pinch hit for he tries to set it up so his last reliever available is the “mop-up” man.

    Yet I’ve never seen Clip used in a tie game away since he became the “closer.” So Davey clearly has a different strategy for those games. In fact, it’s the reverse. He brings in the “long reliever” (Stammen or Gorzy) for as many innings as the poor guy can stand, and then uses whoever else is available.

    As was pointed out, the other Nats “A” relievers (Stammen, Mattheus and Burnett) are pretty fine. So after he’s used his “B” and “C” relievers, he can pretty much take his pick. If he’s reserving one of them, and that one happens to be named Tyler Clippard, to lock in a win if the team scores, how is that “[t]he dumbest thing of all”?

  4. Replying to Evan at #16, I once asked Sean if he could do a database search to see if there was a significant difference in team records in games following an extra inning game, based on whether the opponent in the second game was the same opponent or a new opponent. It’s getting at the exact same thing you suggested. I forget what Sean said–I think he agreed it was an interesting question but didn’t have the bandwidth to address it.

    • That would be really interesting, since it could affect a manager’s strategy in the extra inning game, and Davey is a great one for “knowing when to fold.”

      In the meantime, I ran these stats (all 1990-2012) for games the home team wins when it’s tied middle of the ninth or goes extra innings. In other words, the home team pitches 8+ innings.

      All ——————————————-(27851/51721) = .539
      NL (because it’s different game, isn’t it?)—-(13401/24812) = .540
      Coors (because it’s REALLY a different game—-(711/1279) = .556
      Coors and 90-99 degrees (because I could)—– (55/93) = .591

      • NatsLady:

        I haven’t run a huge number of stats, but home teams in almost any season, on average, win at a rate of around .540-.560, so your figures don’t indicate quite what you appear to be implying, it seems to me. If anything those figures suggest that “knowing when to fold” doesn’t really apply, not if the chance of winning is about the same, tied in the ninth, as it was when the first batter stepped to the plate.

        As for any effect on the following day’s game, that would be highly variable and certainly unpredictable in an immediate sense. Too many factors enter in.

        Or so I think.

      • I think the search referenced @19 is faulty for what it claims to be measuring. It seems to include ~2300 games/season, which makes sense if it actually was run as games where the home team has 8+ inning pitched. This wouldn’t pick up only games where it was tied middle of the ninth or later, rather it picks up every single game except for those shortened by rain or other. Road team with 8+ IP wouldn’t work either as it would include every non-shortened game won by the road team and every “walk-off win.”

        The win probability chart gives the home team ~65% (variation depends upon park factor) of winning when tied heading to the bottom of the ninth or an extra inning. So I assume it is safe to assume that the home team wins ~65% of the time in these situations. Though there may be some variation as I believe the WP number is based on the assumption that teams score in ~30% of innings.

  5. Great choice of card, Andy — Bedrock’s lone All-Star game, the 1987 contest was scoreless through 12 innings. Lee Smith pitched 3 innings, Tom Henke 2.2 IP. The only runs came from Rock Raines’s triple in the 12th, his 3rd hit in 3 ABs after replacing Eric Davis.

    P.S. What a shame that Raines didn’t start that game. In the midst of his 3-year peak, at the ’87 break he was hitting .346/.427/.520, 25 SB (2 CS), with 60 runs in 63 games. The fans chose Dawson instead.

  6. John A,

    You beat me to it…I was going to talk about how much times changed… in an extra inning All-Star game, Lee Smith pitched 3 innings!!! Not sure how long Davey Johnson thought the game would go as he left Smith in the game for 3 innings with his own Mets Starter (Sid Fernandez) in the bullpen. Then he effectively used Sid as the modern day closer, by bringing him in to save the game in the bottom of the 13th.

    Andy, that card is awesome – we have talked about how underrated that 1988 Score set is.

  7. “The dumbest thing of all is a road team saving a closer for when they have a lead. If it’s the bottom of the 9th (or 10th or 11th, etc) and the score is tied, managers just about never use their closers.”

    I’m guessing this came immediately to mind for you Andy, because your Red Sox did this very thing last night against the Mariners – and paid for it. In contrast, the Mariners stuck with a clearly tiring King Felix, who struggled through the top of the 9th, but preserved the scoreless tie with a gutsy effort despite running on fumes after 125+ pitches.

    • MY Red Sox? You don’t know me at all, Doug.

      I’m actually not even aware of scores from yesterday. I was working all night, and this saves this is something I’ve been working on for a while.

  8. Bedrosian’s winning the ’87 CY makes me wonder who would have won the CY if the same pitchers put the same stats up today. For example, relief pitchers are less valued today so doubt Bedrosian would win. Nolan Ryan would draw some votes due to league leading 2.76 ERA & 270 K’s, despite his 8-16 & zero CGs (I believe the GM had him on a pitch count that year). Ryan’s Astros teammate Mike Scott had a great first half & was the AS game starter but fizzled in the second half. Dodger teammates Bob Welch & Orel Hershiser would both be in the mix–Welch was the pitching WAR leader. Sutcliffe, the league leader in wins w/18 would get his votes. Rick Reuschel, who split his time in Pittsburgh & San Fran would be a contender too.

  9. Just my opinion……most relievers nowadays have one pitch; yes guys like Gossage and Sutter (to an extent) had only one pitch as well. But guys like Quiz had a bevy of pitches. EVERY reliever now basically has one pitch, with a few variations (like throwing a 2-seamer instead of a 4-seamer). Gone are the days when a reliever would be the fastball-curveball-slider guy, which IMO EXTENDS a pitchers shelf-life, not only in a game but in a season as well as a career. I beat a dead horse by saying this, but nothing gets me more annoyed with modern baseball then either, a) managers “saving” their closer for the 9th inning/inning where they have the lead, or b) managers playing yank-a-pitcher every inning even if they only throw 9 pitches in an inning.

Leave a Reply

Your email address will not be published. Required fields are marked *