When Closers Get Replaced Because They Suck
A few closers seem to be on shaky ground with their starts to the season, most notability Chris Perez and Sergio Santos. Both have started out horribly, but how horrible do they need to be to lose their jobs? Today, I am going to look at how bad a pitcher needs to suck to lose their closer role.
To get the list of players, I asked fellow Rotograph writers and the participants last night’s Fangraphs chat to list any closers that lost their job permanently because of poor performance. I didn’t want pitchers that went on the DL, became starters or eventually got their job back after a few games. For example, I didn’t use Joe Borowski from 2008. Even though he pitcher horribly to start the season, he went on the DL after appearing in only 6 games. I was not for sure if he lost the closer role because of injury or suckitude. Another pitcher I did not use was Joakim Soria in 2011. Even though he was removed from the role, partially by his own request, he got the job back and dominated for the rest of the year.
In all, I was able to get 12 pitchers that met the criteria:
Matt Thornton – 2011
Frank Francisco – 2010
Fernando Rodney – 2011
Chad Qualls – 2010
Kevin Gregg – 2009
Trevor Hoffman – 2010
Jason Isringhausen – 2008
Ryan Franklin – 2011
Roberto Hernandez (Fausto Carmona) – 2006
Billy Koch – 2003
Shingo Takatsu – 2005
Scott Williamson – 2003
I wanted to find was how bad the group’s ERA and results (Blown Saves, Losses, etc) had to be until they were relived of their duites. After looking at various time frames, I found that a closer is only as good as his last 5 games. Here is how each of the 12 pitchers did in their 5 appearances prior to losing their closer role:
| Name | Year | ERA | SD | MD | H | SV | BS | W | L | Good (SD+H+SV+W) | Bad (MD+BS+L) | Good% |
| Matt Thornton | 2011 | 7.71 | 0 | 4 | 0 | 0 | 4 | 0 | 4 | 0 | 12 | 0% |
| Frank Francisco | 2010 | 15.75 | 1 | 3 | 0 | 0 | 2 | 1 | 3 | 2 | 8 | 20% |
| Fernando Rodney | 2011 | 4.15 | 2 | 1 | 0 | 2 | 1 | 0 | 0 | 4 | 2 | 67% |
| Chad Qualls | 2010 | 19.64 | 1 | 3 | 1 | 2 | 0 | 0 | 2 | 4 | 5 | 44% |
| Kevin Gregg | 2009 | 7.94 | 2 | 2 | 0 | 2 | 1 | 0 | 2 | 4 | 5 | 44% |
| Trevor Hoffman | 2010 | 13.50 | 2 | 1 | 0 | 2 | 1 | 0 | 1 | 4 | 3 | 57% |
| Jason Isringhausen | 2008 | 12.27 | 2 | 3 | 0 | 2 | 3 | 0 | 2 | 4 | 8 | 33% |
| Ryan Franklin | 2011 | 12.27 | 1 | 3 | 0 | 1 | 3 | 0 | 2 | 2 | 8 | 20% |
| Fausto Carmona | 2006 | 13.50 | 0 | 3 | 0 | 0 | 3 | 0 | 3 | 0 | 9 | 0% |
| Billy Koch | 2003 | 9.64 | 2 | 3 | 0 | 0 | 2 | 3 | 2 | 5 | 7 | 42% |
| Shingo Takatsu | 2005 | 4.15 | 3 | 2 | 0 | 3 | 0 | 0 | 2 | 6 | 4 | 60% |
| Scott Williamson | 2003 | 1.42 | 2 | 2 | 0 | 3 | 2 | 0 | 0 | 5 | 4 | 56% |
| Average | – | 10.16 | 1.5 | 2.5 | 0.1 | 1.4 | 1.8 | 0.3 | 1.9 | 3.3 | 6.3 | 37% |
On average, the 12 pitchers had an ERA over 10 for the 5 starts. Also, the percentage chance of a “Good” result happening was only 40%. These pitchers nearly averaged 2 Losses and 2 Blown Saves in their 5 appearances.
Of the 12, there seems to be a few exceptions to the preceeding generalizations. The first is Fernando Rodney in 2011. He was having a reasonable season and then replaced with Jordan Walden. I have a feeling that the Angels planned on replacing him when the season started. Two others that seem out of place are Scott Williamson and Trevor Hoffman. Both of them had Good%’s over 50%. If a person goes back 7 starts for the pair, the Good % drops to 38% for Hoffman and 40% for Williamson. Also their ERAs jump to 18.00 for Hoffman and 6.14 for Williamson. The final exception was Shingo Takatsu. His game results weren’t bad for the entire season, but he did have a 9.64 ERA in his first 8 starts. He seemed to just be on a short leash.
As a whole, the last 5 appearances will be a good measuring stick to see if a closer has a chance of being replaced. Sometimes a person may need to look back a few more games to see if there is any history of problems closing.
With these numbers as a reference, here is a look at how Chris Perez and Sergio Santos measure up so far this season (3 games for each):
Name, ERA, Good%
Perez, 10.13, 60%
Santos, 15.43, 0%
Of the pair, Santos is definitely on thinner ice.
Hopefully, I have been able to shed a little light on determining when a person can get an idea of how much a team can tolerate a struggling closer before they replace him.
Jeff, one of the authors of the fantasy baseball guide,The Process, writes for RotoGraphs, The Hardball Times, Rotowire, Baseball America, and BaseballHQ. He has been nominated for two SABR Analytics Research Award for Contemporary Analysis and won it in 2013 in tandem with Bill Petti. He has won four FSWA Awards including on for his Mining the News series. He's won Tout Wars three times, LABR twice, and got his first NFBC Main Event win in 2021. Follow him on Twitter @jeffwzimmerman.
Santos has a longer leash than we all think. After all the Jays traded a top pitching prospect for him… 🙂
Well, I wouldn’t say a top pitching prospect.
My bad, he was the Jays’ top pitching prospect, but he’d easily be the #1 pitching prospect in a lot of systems. Nestor Molina was their #2 prospect at the time of the trade (behind Travis D’arnaud.)
Nestor Monlina is only a top pitching prospect for the White Sox. No one else.
No. Nestor Molina would have ranked #19 in the Blue Jays system according to Baseball America. John Sickels is the only one who had him inside the top 15 of Jays Prospects.
Although thats not the whole point of the article at all, so I’m not going to get into any debate that derails the thread.
Then as its posted, Santos has the highest WPA save of the season
Santos just improved his Good% this afternoon, though.
The maddening thing is that Thornton’s bad results in those games were largely the result of Juan Pierre’s defense.
I think the quirks of individual teams’ front offices and managers play far too big of a role in closer job security for these results to have much meaning.
No matter what your data suggests, I’d be very surprised if Sergio Santos was anywhere near thin ice going into this afternoon. Maybe you should include some kind of smarts modifier (use Fangraphs’ team rankings?) and weight the closer’s ZIPS ROS peripherals accordingly or something.
Good idea, but I think that each case has way too many additional factors (reputation, upside, service time, roster construction, manager tendencies) and way too small a sample to be something we can analyze objectively.
You obviously weren’t watching Rodney pitch in 2011. It was not reasonable and he deserved to get canned. Even after he got canned he still performed terribly in other non-closer relief situations and did nothing to win the job back.
Jeebus, man, you don’t mention Joe Borowski or Billy Koch in public; save that for when you and your therapist are having a little vergangenheitsbewältigung klatch.
What the f&@$ is a MD?
Meltdown.
See generally http://webmd.com/ (demonstrating that MD is commonly shorthand for doctor).
Meltdown guys.
The premise of this article immediately reminded me of the 1997 and 1998 Seattle Mariners.
In ’97 Norm Charlton and Bobby Ayala both lost their jobs for sucking. They were so bad, the Mariners traded prospects Derek Lowe and Jason Varitek for Heathcliff Slocumb so he could bring his slightly below average talents into the 9th inning.
In ’98 (after Slocumb lost the job in the offseason), Ayala got the job, stunk, lost it back to Slocumb, who almost immediately lost it to Mike Timlin. Finally, Timlin managed to hold on to the job and did well, saving 18 of 19 opportunities in the second half.
(Of course Timlin left that offseason, and the Mariners went through a year of “Ninth Inning Adventures with Jose Mesa” before finally stabilizing with Kaz Sasaki).
Neither of those two is getting replaced. If you want to know who’s going to get replaced, look no further than Joe Nathan. And I wrote about that BEFORE he blew the save tonight…
http://www.therotosaurus.com/2012/04/10/state-of-closers/
There will also likely be a few who get traded and lose their closer role as well, like Brandon League, who is VERY likely to get traded around the trade deadline…
Uehara and Adams are both better pitchers.
Tom Wilhelmsen is a strong saves candidate right now.
Well, The Kaiser is a strong saves candidate around the deadline, and right now is a strong reliever to get and stash.
correct. As I said, Adams will be closer sooner than later and you can pretty much bank on Wilhelmsen around the deadline, so make sure to pick him up around June
Quick correction -while he may now share the same name with the pitcher formerly known as Fausto Carmona, Roberto Hernandez had his very own career (with 326 saves of his own) using only one name!
Jeff –
(1) Not sure how you could say Rodney was having a “reasonable” season in 2011 when, in fact, he pitched only two times before he was replaced and the second appearance was God-awful, walking three to blow a save and hitting the strike zone about as often as Presidential candidates say something kind about each other. Of course Rodney was on an exceptionally short leash after two years in which he had shown a diminishing ability to strike out batters and a miserable ERA, FIP and xFIP for someone who comes in with no runners on in the 9th, not to mention a fireballer in Walden waiting in the wings.
(2) Your research is badly flawed. You chose as your sample to study closers who had lost their position but you failed to do something elemental: Compare them to closers who have not lost their role. I’m confident that among the latter group you would find 5-game stretches that were comparable, even worse, than the closers that lost their jobs. What distinguishes the two groups — those who lost their roles and those who did not — was the perception of pitcher BEFORE the five-game stretch. Rodney is a case in point. So too was Francisco, who never had the confidence of Ron Washington.
(3) Because your research is flawed you grossly overstate the chance that Santos will lose his position. The Jays’ brain trust believe in Santos, gave up a quality prospect to get him and are desperate for stability in that role. That’s why Alex Anthopoulos said after the second blown save that he was losing no sleep using someone as closer with Santos’s stuff.
It would be nice to see the arguments for the potential replacements before assuming a closer is on thin ice.
1. How are they performing so far this season
2. How they’ve performed in high leverage situations previously in their careers
3. Age factor (i.e. for Francisco Cordero)
That being said, great work on the analysis!
why do you keep calling them “starts”
“After looking at various time frames, I found that a closer is only as good as his last 5 games.”
No evidence stated about this?
I think, as others have alluded to, that this would be a better piece if you actually did some more situational research and discuss each closer with a paragraph. The chart, while useful, should be something like ‘exhibit B’.
Santos should not be in this discussion. 5 inning sample sizes? Are you kidding?