Closer Job Security Chances
Chasing Saves frustrates many owners as injuries and poor performance piles up. Trying to accumulate Saves can be a tiring game with roster spots and FAAB wasted on arms who only keep their job for a week or two while piling up a 5.00 ERA and 1.50 WHIP. Last pre-season, I went through and examined reliever talent and how likely they were they were to keep their job. It’s time to give the 2018 bullpen arms their chances to make it a full season.
In the original study, I looked at the team’s initial closer and if he kept the role for an entire season. Some were traded or had a better reliever brought in who then closed. Most of the changes were caused by underperforming or injuries. Here the study’s findings.
| ERA Range | Whole Season | Poor performance | Injury | Other | Count |
|---|---|---|---|---|---|
| < 2.50 | 70% | 10% | 10% | 10% | 10 |
| 2.50 to 3.00 | 48% | 9% | 39% | 4% | 23 |
| 3.00 to 3.50 | 35% | 31% | 18% | 16% | 55 |
| >3.50 | 31% | 28% | 19% | 22% | 32 |
From these findings the following the two equations and table can be extrapolated:
Equation #1 (r-squared = .93): Full season closer chances = -0.2454*Projected ERA + 1.1888
Equation #2 (r-squared = .98): Full season closer chances = 2.137*Projected ERA^(-1.488)
| ERA | Equation #1 | Equation #2 |
|---|---|---|
| 2.00 | 70% | 76% |
| 2.50 | 58% | 55% |
| 3.00 | 45% | 42% |
| 3.50 | 33% | 33% |
| 4.00 | 21% | 27% |
| 4.50 | 8% | 23% |
The chances may seem low but here is a trio of last season’s sub-3.00 ERA projected closers who lost their jobs.
| Name | ERA | Equation #1 | Equation #2 |
|---|---|---|---|
| Zach Britton | 2.38 | 60% | 59% |
| Mark Melancon | 2.63 | 54% | 51% |
| Seung Hwan Oh | 2.92 | 47% | 43% |
While the projections for fulltime performance seem harsh, the results stack up. Two experienced injuries and the other one couldn’t repeat his 2016 season. All three were projected to be better than Craig Kimbrel who defied his ~45% odds and kept his job. Now, he’s the number two closer off the board after Kenley Jansen at 44th in NFBC drafts.
With the above information, here are the closer chances for making it through the 2018 season. I used our depth chart projections (combination of Steamer and Zips) and listed anyone projected for at least one Save.
| Name | ERA | SV | Equation #1 | Equation #2 |
|---|---|---|---|---|
| Andrew Miller | 2.41 | 3 | 60% | 58% |
| Craig Kimbrel | 2.49 | 32 | 58% | 55% |
| Kenley Jansen | 2.51 | 31 | 57% | 54% |
| Aroldis Chapman | 2.63 | 34 | 54% | 51% |
| Felipe Rivero | 2.91 | 31 | 48% | 44% |
| Dellin Betances | 2.96 | 1 | 46% | 43% |
| Zach Britton | 2.96 | 4 | 46% | 42% |
| Ken Giles | 3.03 | 31 | 45% | 41% |
| Mark Melancon | 3.04 | 29 | 44% | 41% |
| Corey Knebel | 3.05 | 30 | 44% | 41% |
| Roberto Osuna | 3.10 | 33 | 43% | 40% |
| Ryan Madson | 3.11 | 3 | 43% | 39% |
| Justin Wilson | 3.12 | 2 | 42% | 39% |
| Brad Hand | 3.18 | 35 | 41% | 38% |
| David Robertson | 3.19 | 4 | 41% | 38% |
| Keone Kela | 3.20 | 6 | 40% | 38% |
| Edwin Diaz | 3.23 | 33 | 40% | 37% |
| Drew Steckenrider | 3.23 | 2 | 40% | 37% |
| Carl Edwards Jr. | 3.26 | 1 | 39% | 37% |
| Jeurys Familia | 3.29 | 23 | 38% | 36% |
| Cody Allen | 3.30 | 32 | 38% | 36% |
| Scott Alexander | 3.30 | 5 | 38% | 36% |
| A.J. Minter | 3.30 | 3 | 38% | 36% |
| Sean Doolittle | 3.31 | 33 | 38% | 36% |
| Blake Parker | 3.33 | 14 | 37% | 36% |
| Tommy Kahnle | 3.34 | 1 | 37% | 36% |
| Raisel Iglesias | 3.34 | 31 | 37% | 36% |
| Brandon Morrow | 3.36 | 33 | 37% | 35% |
| Anthony Swarzak | 3.37 | 10 | 36% | 35% |
| Brett Cecil | 3.39 | 4 | 36% | 35% |
| Wade Davis | 3.41 | 33 | 35% | 34% |
| Archie Bradley | 3.42 | 22 | 35% | 34% |
| Luke Gregerson | 3.46 | 25 | 34% | 34% |
| Joe Kelly | 3.46 | 2 | 34% | 34% |
| Jerry Blevins | 3.47 | 1 | 34% | 34% |
| Chris Devenski | 3.47 | 1 | 34% | 34% |
| Tony Watson | 3.50 | 2 | 33% | 33% |
| Kyle Barraclough | 3.52 | 6 | 33% | 33% |
| Will Harris | 3.52 | 1 | 32% | 33% |
| Hector Rondon | 3.52 | 6 | 32% | 33% |
| Nate Jones | 3.54 | 2 | 32% | 33% |
| Dominic Leone | 3.56 | 2 | 32% | 32% |
| Alex Claudio | 3.56 | 24 | 31% | 32% |
| Hunter Strickland | 3.57 | 2 | 31% | 32% |
| Joe Smith | 3.57 | 1 | 31% | 32% |
| Kirby Yates | 3.57 | 2 | 31% | 32% |
| Joakim Soria | 3.57 | 6 | 31% | 32% |
| Mychal Givens | 3.58 | 3 | 31% | 32% |
| Andrew Chafin | 3.59 | 1 | 31% | 32% |
| Darren O’Day | 3.59 | 10 | 31% | 32% |
| Brad Ziegler | 3.61 | 30 | 30% | 32% |
| Arodys Vizcaino | 3.62 | 30 | 30% | 32% |
| Alex Colome | 3.62 | 30 | 30% | 32% |
| Dan Jennings | 3.62 | 5 | 30% | 31% |
| Aaron Loup | 3.63 | 1 | 30% | 31% |
| Steve Cishek | 3.63 | 4 | 30% | 31% |
| Blake Treinen | 3.64 | 32 | 29% | 31% |
| Jacob Barnes | 3.65 | 5 | 29% | 31% |
| Tommy Hunter | 3.65 | 3 | 29% | 31% |
| Michael Feliz | 3.66 | 1 | 29% | 31% |
| Trevor Hildenberger | 3.67 | 1 | 29% | 31% |
| David Phelps | 3.68 | 1 | 29% | 31% |
| Brad Brach | 3.68 | 21 | 29% | 31% |
| Tyler Lyons | 3.68 | 6 | 29% | 31% |
| Pat Neshek | 3.69 | 3 | 28% | 31% |
| Greg Holland | 3.70 | 10 | 28% | 31% |
| Liam Hendriks | 3.70 | 1 | 28% | 30% |
| Cam Bedrosian | 3.70 | 20 | 28% | 30% |
| Jose Alvarez | 3.70 | 1 | 28% | 30% |
| A.J. Ramos | 3.72 | 4 | 28% | 30% |
| Phil Maton | 3.72 | 5 | 27% | 30% |
| Carson Smith | 3.73 | 4 | 27% | 30% |
| Bryan Shaw | 3.75 | 2 | 27% | 30% |
| Josh Hader | 3.76 | 2 | 27% | 30% |
| Cory Gearrin | 3.77 | 1 | 26% | 30% |
| Andrew Kittredge | 3.78 | 3 | 26% | 30% |
| Luis Garcia | 3.82 | 1 | 25% | 29% |
| Kelvin Herrera | 3.84 | 28 | 25% | 29% |
| Brandon Kintzler | 3.84 | 2 | 25% | 29% |
| Shane Greene | 3.85 | 28 | 24% | 29% |
| Jim Johnson | 3.85 | 2 | 24% | 29% |
| Sam Dyson | 3.85 | 4 | 24% | 29% |
| Jake Diekman | 3.86 | 2 | 24% | 29% |
| Juan Nicasio | 3.86 | 5 | 24% | 29% |
| Matt Albers | 3.86 | 1 | 24% | 29% |
| Jeremy Jeffress | 3.88 | 1 | 24% | 28% |
| Edubray Ramos | 3.90 | 1 | 23% | 28% |
| Koda Glover | 3.90 | 1 | 23% | 28% |
| Addison Reed | 3.91 | 7 | 23% | 28% |
| Nick Goody | 3.92 | 3 | 23% | 28% |
| Ross Stripling | 3.94 | 1 | 22% | 28% |
| Ryan Tepera | 3.95 | 2 | 22% | 28% |
| Luis Avilan | 3.95 | 3 | 22% | 28% |
| David Hernandez | 3.96 | 6 | 22% | 28% |
| Taylor Rogers | 3.98 | 1 | 21% | 27% |
| Nick Vincent | 3.99 | 1 | 21% | 27% |
| Brian Flynn | 4.00 | 1 | 21% | 27% |
| Jared Hughes | 4.00 | 1 | 21% | 27% |
| Seung Hwan Oh | 4.01 | 2 | 21% | 27% |
| Sam Freeman | 4.01 | 1 | 20% | 27% |
| Jake McGee | 4.01 | 3 | 20% | 27% |
| Matt Belisle | 4.02 | 1 | 20% | 27% |
| Craig Stammen | 4.03 | 2 | 20% | 27% |
| Randall Delgado | 4.03 | 2 | 20% | 27% |
| Mike Montgomery | 4.03 | 1 | 20% | 27% |
| Sergio Romo | 4.04 | 1 | 20% | 27% |
| George Kontos | 4.05 | 5 | 19% | 27% |
| Matt Bush | 4.06 | 1 | 19% | 27% |
| Hector Neris | 4.06 | 32 | 19% | 27% |
| Paul Sewald | 4.10 | 1 | 18% | 26% |
| Tyler Olson | 4.10 | 1 | 18% | 26% |
| A.J. Schugel | 4.12 | 2 | 18% | 26% |
| Wandy Peralta | 4.13 | 1 | 18% | 26% |
| Daniel Hudson | 4.13 | 1 | 18% | 26% |
| Keynan Middleton | 4.14 | 2 | 17% | 26% |
| Yusmeiro Petit | 4.14 | 2 | 17% | 26% |
| Brandon Maurer | 4.17 | 6 | 17% | 26% |
| Jose Torres | 4.17 | 1 | 17% | 26% |
| Bud Norris | 4.17 | 2 | 16% | 26% |
| Fernando Rodney | 4.18 | 28 | 16% | 25% |
| Michael Lorenzen | 4.19 | 2 | 16% | 25% |
| Richard Bleier | 4.20 | 1 | 16% | 25% |
| Joaquin Benoit | 4.21 | 1 | 16% | 25% |
| Tony Cingrani | 4.21 | 2 | 16% | 25% |
| Junichi Tazawa | 4.23 | 1 | 15% | 25% |
| Donnie Hart | 4.24 | 1 | 15% | 25% |
| Juan Minaya | 4.28 | 29 | 14% | 25% |
| Danny Barnes | 4.31 | 1 | 13% | 24% |
| Daniel Stumpf | 4.34 | 1 | 12% | 24% |
| Alex Wilson | 4.37 | 3 | 12% | 24% |
| Chris Hatcher | 4.39 | 4 | 11% | 24% |
| Brandon Workman | 4.39 | 2 | 11% | 24% |
| Jose Leclerc | 4.40 | 5 | 11% | 24% |
| Brad Boxberger | 4.44 | 7 | 10% | 23% |
| Danny Farquhar | 4.47 | 1 | 9% | 23% |
| Huston Street | 4.49 | 2 | 9% | 23% |
| Kyle Crick | 4.49 | 1 | 9% | 23% |
| Adam Ottavino | 4.52 | 1 | 8% | 23% |
| Jose Ramirez | 4.52 | 4 | 8% | 23% |
| Jorge de la Rosa | 4.54 | 2 | 8% | 23% |
| Jarlin Garcia | 4.54 | 1 | 7% | 22% |
| Wily Peralta | 4.56 | 3 | 7% | 22% |
| Mike Dunn | 4.72 | 1 | 3% | 21% |
| Kevin McCarthy | 4.73 | 2 | 3% | 21% |
| Drew VerHagen | 4.90 | 6 | -1% | 20% |
| Carlos Ramirez | 4.97 | 1 | -3% | 20% |
| Warwick Saupold | 5.08 | 2 | -6% | 19% |
| Tim Lincecum | 5.18 | 2 | -8% | 18% |
One way this table can be used is to find backup arms who are ranked higher than the team’s projected preseason closer. An example would be Keone Kela (3.20 ERA, 40% keep rate) having a better chance of making through the season than the incumbent closer Alex Claudio (3.56 ERA, 31% keep rate) if he was given the job preseason.
Rostering closers is usually a necessary evil for fantasy owners as it’s tough to dump a category and still win. Playing the closer carousel can be exhausting but by knowing the odds for each closer making it through the season, owners can be a step ahead anticipating how certain bullpens may shake out.
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.
It is pretty insane that even some of the top closers have over a 50% chance of losing their job.
Seems Steamer is a little conservative in their ERA projections. Kenley just had back to back sub-2 ERA seasons. They are projecting 2.51. I’ll take the under, please. He’s not the only one. I know they try to get the projection system to predict ERAs as a league-wide population and therefore at the extremes of the curve they become a little conservative. Might be best served looking at last season’s ERAs.
Jeff, any chance you could show the percentages based on last year’s ERAs?
And by “become conservative” I mean force projections back to the median. Doesn’t help when the subset we want to look at (closers) is generally amongst the outliers of the general population.
That, and Steamer is assuming the massive increase in offense is going to persist. The 15th best team ERA was 4.50 last year, vs. 3.60 in 2014. Thanks to Manfred, we completely switched eras in a matter of three seasons.
Wait until MLB gets a cut of all US gambling revenue (the oxymoronic “integrity fee”) and has the full ability to dial offense up and down as it pleases… Giving the people who control the officiating and the equipment a stake in gambling revenue is possibly the dumbest idea I have ever heard. And I am someone who is 100% in favor of legalized sports betting.
Also does anyone really think Jansen would lose his job even if he “struggles”.
He would need to put up a performance that no one could survive. Something like what Fernando Rodney did at the beginning of 2017.
Oh, wait.
The arrow man cares nothing for your odds!
Have you tried other metrics (e.g. K-BB%) instead of ERA, given that RP tend to throw so few innings in a given season?
Most ERA projections are based off K and BB. I go with ERA as it’s easy to remember
If a closer goes on the DL but then returns to his closer job is he considered a success, ie, a closer who keeps his job all year?
After reading last year’s article, I think a DL stint marks a closer unsuccessful. As jeff pointed out per the data, a huge portion of lost closers comes from injuries. Only 10% of closers with a projected era below 3 lose their jobs because of performance–it seems better to focus on this number as the risk among top tier closers. It’s hard to evaluate the injury risk relative to other positions without more data.
I wonder how salary factors in to the equations (or rank of salary amongst relievers). Expensive closers have slightly longer leashes, while cheap relievers (league min, etc.) can be held back from saves due to rising arb costs for arbitrary save stats.
Because the initial data was collected from 2013-16 when the league ERA was lower I think we need to adjust either the equation or the projected ERAs to get a better prediction.
so felipe rivero is the best value then? odds aren’t much lower compared to chapman yet his ADP isn’t that close. i also have taken andrew miller the last few years knowing that if allen gets hurt, miller is a top 3 closer who cost you a ~200th pick.
Hey Jeff, thanks for this work, it’s really interesting to see. Just a couple thoughts:
–In your descriptive table leading off the analysis, there’s value in showing the percent that lost their job due to injury, but I think those instances should be removed from the calculation that lasted the season because it’s an exogenous factor that can’t be accounted for (in fantasy or real life). And to me the point, and value, of this work is to identify closers, or closer cohorts that lost their job due to poor performance.
–In that vein, when listing the entire population of people projected for at least one save, it would be helpful for you to list them by team. That way, in your example above we can easily surmise that Keone Kela is actually a better bet than Alex Claudio.
Thanks for the great work!
Hey Jeff…Great detail as always. Do you think Juan Minaya will be the closer and not Nate Jones?
Hey Jeff — thanks for the interesting work. I’m curious what you view as the takeaway. My understanding is that you lean pretty heavily toward the projections and quantitative analysis side of things in your fantasy prep and draft strategy. The Dodgers will probably save ~45 games this year. Would your research here really cause you to project Jansen for only 25 saves (45 times 55%) for valuation purposes?
That seems likely to lead to poor draft strategy. Assuming you wouldn’t go that far, how do you envision applying these findings to draft strategy?
Probably should update this study, since the median team relief pitcher ERA last year was 4.20. This is much higher than 2013-16 (3.70). I.e., 3.50 ERA RPs are becoming much more rare due to increasing offense.
It would be interesting to see the closer probabilities compared to starting pitcher full season probabilities or even position players.
Also familia is not the mets closer
He finished last year as the closer, so unless there’s a reason the Mets find for Ramos to take over this spring, Familia looks to be the favorite.