Mailing List

Sign up for email updates from Hot Corner Harbor any time there's a new post!

    Tuesday, March 28, 2023

    Predicting Today's Future Hall of Fame, 2023 Edition Bonus Update: Closers

    A few years ago, I decided to take a crack at applying my Future Hall of Fame methodology to Closers. Unlike with starting pitchers or position players, Wins Above Replacement traditionally doesn’t seem to line up with how Hall views relief pitchers. However, I found that Win Probability Added does a pretty good job of matching the voters, and went with that instead. For those who aren’t as familiar, Win Probability Added (WPA) is a context-sensitive value stat. A home run will basically count the same in WAR whether it happens in the first or ninth inning, but WPA is based on the real-time in-game probability of a team winning, and credits or debits the player with how much each of their actions swings those odds in their team’s favor.

    With Billy Wagner looking like a strong bet to make the Hall in 2024 or 2025, I thought this offseason might be an interesting time to revisit my closer predictions. You know, maybe take a look at which relievers could be following in his footsteps, especially since the role seems to be in a period of transition at the moment.

    It was supposed to be a pretty easy process; there hasn’t been a new closer inducted since the last time I did this in 2019 (following the inductions of Lee Smith and Mariano Rivera), so I shouldn’t need to do that much to update it. But after checking my results from last time I ran into some issues… I really don’t know how to explain it, but my numbers were just wrong. I have no idea how it happened, but the numbers don’t seem to line-up with anything. I tried a few things, referring to my notes and even reaching out to Baseball-Reference to see if they had changed anything in their site’s calculations, but I couldn’t recreate the results despite several attempts.

    Weirdly, though, the methodology was still right? I tried it again after re-calculating every number, and I was basically getting the same players in the same order, with relatively strong correlations to Hall voting. It was just… the WPA totals and Hall odds were different from what I had last time. I suppose I could just wait until Wagner gets inducted, since that would require a total re-calculation anyway, but I wanted to publish the corrected numbers at least once before then.

    I’m using the same basic framework as my Hitters and Starting Pitchers article, here: looking at the WPA for Hall of Fame closers by age to build a median trendline, then looking at how players above that line at each age have fared in eventual Hall elections. If you need a more detailed explanation, you can check out either of those examples and such.

    However, one big difference that I would like to expand on here: there just isn’t a large sample size of closers in the Hall to work off of, which is a big deal since this method is built off of precedence. There are currently 8 pitchers who qualify here, so each one has an outsized impact on our results.

    Like, a quarter of our cases are Hoyt Wilhelm (a knuckleballer who didn’t debut until age 29 but pitched for 21 seasons) and Dennis Eckerlsey (who spent the first decade of his career as a starter), each of whom is extremely unique. Each additional electee is going to significantly shift our standard here, and given the ways the role has changed so radically in its short lifespan, it is important not to get too attached to these numbers, and to remain flexible in your considerations of what a Hall Closer looks like. This method just isn’t as strong as the one for Position Players or Starting Pitchers.

    For similar reasons, I’m going to be going in a slightly different order for this one. I usually like to start with younger stars and end with the older ones, because that way keeps the more interesting and unusual names at the top of the discussion but allows me to end on longer write-ups of stronger candidates who will be facing the ballot sooner.

    In comparison, everything about closers feels less certain; stars regularly appear out of nowhere or suddenly flame out, the standard progression is kind of ridiculous and barely feels like a realistic guideline (more on that later), and to top it all off, we barely even know what Hall voters want beyond “A lot of saves, most of the time”. So instead, I’ll just start with the most veteran names and kind of work backward. There really aren’t any overwhelming favorites here anyway, so we might as well start with the most interesting discussions.

    Also, as a note, when I refer to a player’s age, I’m generally referring to last season; if I’m instead using their 2023 age, I’ll try to make it obvious. Now then, let’s queue up that entrance music and see what we’ve got!

    The Leaders

    Overall: 30.65 WPA Median; 100% of all players at this mark elected

    Right now, there are only five closers who are even halfway to the career median for Hall of Fame Closers, so we’ll start with them.

    The current active leader in WPA among relief pitchers is Kenley Jansen, which is actually a bit of a turn-around that’s happened here recently (more on that in a minute). Right now, heading into his age 35 season, Jansen is at 25.6 Win Probability Added, so making that last 5 WPA is probably out of the question for 2023 (Jansen has only ever had one season on that level, back in 2017). It’s plausible with two back-to-back good and healthy seasons, but three years might be the more reasonable scenario.

    And for those unfamiliar with WPA, it can go negative like WAR, but it’s much less stable year-to-year, so you can build up a big negative very quickly (especially in a role like closer, which is inherently tied to big swings in win probability). So this mark isn’t something Jansen can limp to, even by just being average; he’s got to remain at least competent through his age 36 or 37 season (and then either hang it up before falling apart or keep going to build up a little wiggle room).

    “Kenley Jansen is a Hall of Famer with two or three more good seasons” might sound extreme to you in the abstract, but it’s also worth keeping the more basic context in mind here as well: Jansen currently has 391 saves, eighth-most in history. There’s a strong chance that two-to-three good years moves him into third or fourth place on the all-time list. If he retires with more saves than anyone outside of Mariano Rivera or Trevor Hoffman, I imagine a lot of voters will come around to his case anyway. Of course there’s the standard disclaimers here, the role is volatile and he might also just suddenly melt down, even as soon as next year. But if he can stay solid for a few years into his late 30s, he does look a lot like a Hall of Fame closer.

    Now let’s move to someone who is deeply feeling that “closers can be volatile” truism: Craig Kimbrel. As I alluded to earlier, Kimbrel was leading Jansen in WPA as recently as one year ago, but he’s fallen down to 23.5 WPA. In fact, Kimbrel has actually been going in the wrong direction for a while now; outside of the start of 2021 (he was worth 2.8 WPA for the Cubs that season, before his disastrous trade to the White Sox), Kimbrel has actually lost 2.8 WPA from his career total dating back to the 2019 season. He’s basically just been treading water, at best.

    Granted, he is still ahead of Jansen in saves for the moment, 394 to 391. And that start of the 2021 season, aside from offsetting the rest of that stretch, was good enough that you can maybe dream of a longer-term rebound happening. But still, he’s the same age as Jansen, so all of those big caveats from Jansen’s section apply on top of the fact that Kimbrel has just been worse lately. It’s baseball, anything can happen, et cetera… but it’s hard for me to look at Kimbrel right now and not just put him a tier below Jansen, whatever that translates to in percentages.

    Of course, whatever you think Kimbrel’s chances are of making the Hall from here, they’re certainly better than the other three players in this category. At 20.5 WPA, we have David Robertson. He’s had a better last four years than Kimbrel, but that’s mostly because “one good year and three nothing-years due to injury” beats three and a half bad years. On top of that, he’s spent most of his career as a set-up man rather than a closer; in fact, his 310 career Games Finished is one of the lowest totals among relievers with over 20 WPA. In the modern game, a reliever this good not closing for so long is kind of shocking. Even if he can avoid injuries and put up a few more good years, you have to figure a lot of Hall voters will ding him anyway for those low save totals (157 and counting). On top of all of that, he’s three years older than Jansen and Kimbrel, so he’s on an even tighter schedule.

    After that, at just under 20 WPA, we have Aroldis Chapman (19.7) and Mark Melancon (19.4). Chapman is the same age as Jansen and Kimbrel, and while he was just a step behind them for a while, he had his own issues last season that he’ll have to rebound from if he wants to build a Hall case. But again, at least it’s just one bad year and not four straight rough seasons. Again, his odds probably aren’t great, but they definitely aren’t zero yet. I certainly think he has better odds than Melancon, who is Robertson’s age, did even worse in 2022, and over 50 saves behind Chapman on the leaderboard (315 to 262).

    The Young Stars

    Age 25: 5.2 WPA Median; 15.00% of all players at this mark elected
    Age 26: 8.0 WPA Median; 23.08% of all players at this mark elected
    Age 27: 11.5 WPA Median; 33.33% of all players at this mark elected
    Age 28: 12.7 WPA Median; 27.27% of all players at this mark elected
    Age 29: 13.6 WPA Median; 33.33% of all players at this mark elected
    Age 30: 17.2 WPA Median; 50.00% of all players at this mark elected

    Part of the reason that we’re skipping ahead here is that I feel like it’s easier to imagine younger pitchers stringing together a few peak seasons, something every pitcher from here on out will need since they’re all less than halfway to the goal. But it also helps that the sixth- and seventh-best pitchers by career WPA fall into this age group.

    Of course, both of those pitchers have their own issues, so there’s even still a ton of uncertainty here! First up is Josh Hader, at 13.9 Win Probability Added. Hader did not have a great start to 2022 in Milwaukee, got even worse following his trade to San Diego, but then seemed to return to form in the playoffs. If he can build off of that this year, his age-29 season, then he’s a lock to make it to the halfway point before this season’s end. On a side note, he only has 132 saves, which is actually a lot less than I expected? If he plays long enough to reach 30 WPA, I imagine that issue takes care of itself though.

    Next up is Edwin Díaz, who just turned 29 and sits at 11.6 WPA. Unfortunately, it looks like that won’t be changing much this year following his unlucky injury in the World Baseball Classic. We’ll have to see how he plays when he returns, but his hot start does mean he has some wiggle room. In fact, in contrast with Hader, Diaz already has 205 saves, third-most in history through age 28 and one of just five pitchers to pass the 200 mark by that age. However, that’s maybe not much of a predictor, given the other four names here (Kimbrel, Francisco Rodriguez, Huston Street, Bobby Thigpen) aren’t even matching the ~1 in 4 rate of the WPA leaders.

    There are other interesting names I’d like to at least mention here, even if they don’t have the resumes of Hader and Díaz quite yet. Devin Williams, Hader’s successor in Milwaukee, is next for under-30s at 8.0 WPA. He didn’t start quite as young as Hader did, and he didn’t really get many save opportunities until the trade, but he’s been strong so far as he heads into his age-28 season.

    And a little below him, we have division rivals Emmanuel Clase (4.8 WPA) and Jhoan Durán (4.6 WPA), who should both have an easy time meeting the age-25 median this coming season. Clase has the better early case, since he’s got more of a track record and has locked down the closer role, but both of these players still have an extremely long way to go.

    The Longer Shots

    Age 31: 17.8 WPA Median; 44.44% of all players at this mark elected
    Age 32: 19.0 WPA Median; 36.36% of all players at this mark elected
    Age 33: 20.25 WPA Median; 36.36% of all players at this mark elected
    Age 34: 20.6 WPA Median; 40.00% of all players at this mark elected
    Age 35: 24.1 WPA Median; 50.00% of all players at this mark elected
    Age 36: 25.5 WPA Median; 57.14% of all players at this mark elected
    Age 37: 26.85 WPA Median; 66.67% of all players at this mark elected
    Age 38: 27.05 WPA Median; 66.67% of all players at this mark elected
    Age 39: 26.25WPA Median; 66.67% of all players at this mark elected
    Age 40: 26.2 WPA Median; 66.67% of all players at this mark elected
    Age 41: 26.2 WPA Median; 66.67% of all players at this mark elected
    Age 42: 27.1 WPA Median; 66.67% of all players at this mark elected
    Age 43: 27.15 WPA Median; 66.67% of all players at this mark elected
    Age 44: 29.05 WPA Median; 80.00% of all players at this mark elected
    Age 45: 29.75 WPA Median; 80.00% of all players at this mark elected

    For those wondering how we bridge the gap from 17 WPA at 30 to over 30 overall, here you go. Wilhelm ends up factoring into the median for a while, which is why it keeps going up for so darn long (and also serves as a helpful reminder for just how little data we’re working off of here). I ultimately don’t expect any of these guys to build a Hall case in their 30s (even if I suppose anything is possible when it comes to closers), but I do want to be thorough and at least give them a shoutout. And besides, it’s not like it’s impossible, just very unlikely.

    The only active pitcher we have left with double-digit win probability is Sean Doolittle at 10.9. Doolittle will be 36 this year, and it looks like his career might be winding down, as he follows up an injured 2022 season by starting 2023 once again on the IL. Blake Treinen (9.8 WPA), who’s a year younger, is in a similar boat. Sure, he’s been better than Doolittle lately, but he’s also been more of a set-up man, and lags in traditional stats, too. Daniel Bard (8.5 WPA) is a nice comeback story, but he’s older than either of them and has even fewer saves.

    The two most compelling closers in this group are probably Raisel Iglesias (9.4 WPA) and Ryan Pressly (8.2 WPA), as both are a little younger (33 and 34, respectively), still pretty good, and still closing for winning teams in Atlanta and Houston. At the very least, that should give them a few years to rack up some stats, and if they can stay effective into their 40s, I guess it’s possible. I’d want several more years of evidence before calling either of them "likely", but given some of the other names I’ve mentioned or skipped over already, “possible” doesn’t seem that bad. And for the record, Iglesias sits at 157 career saves, while Pressley has 77.

    And… that accounts for every active relief pitcher over 8.0 career Win Probability Added. Will Smith is close (7.9), but I can’t imagine him picking up that much more as a lefty specialist at this point. Zack Britton is technically over halfway if you remove his early struggles as a starter (13.6 career with those seasons, 15.7 without), but he still hasn’t signed anywhere on the brink of his age-35 season following rehab from Tommy John surgery; I’d want to at least see a comeback season first. Sergio Romo (13.8) is retiring, as the number of players from the 2000s continues to dwindle. 34-year-old Alex Colomé has over 150 saves in his career, but a low WPA (6.7) even before mentioning that he’s fallen off steeply the last two years.

    I think that’s anyone with an even mildly interesting case? Unless you want to get into the even more dubiously-active closers like Greg Holland or Trevor Rosenthal. If not, it looks like our final tally is one player in Kenley Jansen who’s more or less a coin flip right now, depending on how his next few years go; another case in Craig Kimbrel who looks a little more like a dice roll; and then maybe four or five other names who maaaybe have the start of something brewing, but check back in five or six years.

    New Email List, since Blogger broke the last one!

    The old subscription service doesn't seem to be working anymore, so if you'd like to receive emails when a new Hot Corner Harbor post goes up, sign up here!

      We won't send you spam. Unsubscribe at any time.

      No comments:

      Post a Comment