Predicting Future Hall of Famers 2017 Follow-Up: The Disappearance of Below-Median Starters

In the previous post, I was discussing active pitchers above the median Wins Above Replacement for Hall of Famers by age when I noted something interesting at the end: there are only two starters* above the overall median who are eligible for the Hall but not inducted or still on the ballot. Kevin Brown and Rick Reuschel both cleared the median by less than a full Win. Other than them, everyone above the median has made it.

*As defined last time, this is pitchers who began post-1919 with 10% or more of their total games as a starter. See last time for the reasoning on that, although I'll be jumping back and forth in this one between "Live Ball" starters and "Total" starters; I'll make the distinction clear when those shifts happen.

This makes for an interesting contrast with the hitters. Limiting ourselves to just post-1919 debuts so that we're working with comparable sets of players, the median WAR for hitters is 65.1 (held by Craig Biggio)*, just a bit lower than the pitchers' mark of 67.95 (between Jim Palmer and Carl Hubbell). While there are only two pitchers above the median not yet in or still on the ballot, there are nine such hitters, ranging from Rafael Palmeiro (71.6 WAR) to Willie Randolph (65.5). And if we expand this to include players on the ballot, it becomes five starters to thirteen position players (plus the ineligible Pete Rose). And while those three pitchers all look likely to get inducted, the four hitters are much more uncertain, with Barry Bonds and Edgar Martinez looking likely to get inducted before they age out but Larry Walker and Manny Ramirez looking like long shots.

*Note that this is a little different than the median I used in the previous piece; that’s because I did not limit that median to post-1919 batters. However, I wanted the two to be on a level playing field here, so there’s a slight upwards shift as a result.

That's an interesting level of uniformity. Maybe this is a sign of those pitchers all being particularly obvious? But let's look at the flip-side; how have players below the median done? After all, the median is the middle point, not the end-all, be-all. We should be seeing at least some below-median names getting inducted, right? Well, that's the interesting part: we really aren't for pitchers.


Let’s look back at just the Hall of Fame elections since 2000. It’s a bit of an arbitrary endpoint, but it’s a starting point, and I’ll be moving the range around a bit. Since the turn of the millennium, Hall of Fame voters (between the standard Baseball Writers Association of America vote and the Veterans Committee) have added ten “below-median” position players to the ranks of Cooperstown: Mike Piazza in 2016, Deacon White in 2013, Andre Dawson in 2010, Joe Gordon and Jim Rice in 2009, Bill Mazeroski, Dave Winfield, and Kirby Puckett all in 2001, and Tony Perez and Bid McPhee in 2000. Now, some of those are exceptions; for example, catchers like Piazza usually get in with lower counting stats due to shortened careers from the wear of catching, or McPhee and White were pre-1900 players who got held to slightly different standards of their own. But at the same time, ten is a decent amount, and they run the range in WAR from the mid-30s to just under 65.

In the same timespan, only two pitchers who are below the median for starters have been inducted: John Smoltz in 2015, and Dennis Eckersley in 2004. And even then, both of those have major caveats: each spent substantial time as a closer, and closers (much like catchers) have historically been held to slightly lower WAR standards due to reduced playing time (not to mention that neither is especially far below the median, despite those years closing; Smoltz sits at 66.5, which is still above the hitters’ median, while Eck is at 62.5).

And there isn’t really a way to cut this off that makes the situation look much better for starting pitchers. Want to cut it off at the last “pure” starter to make it in despite having a below-median career value? You’d have to go all the way back to Jim Bunning’s VC induction back in 1996, but in the meantime, you’d be adding Orlando Cepeda (Class of 1999), Larry Doby (1998), and Nellie Fox (1997) to the other side. Going back another year nets you Vic Willis for the pitchers, but also Richie Ashburn for the non-pitchers. Hal Newhouser was inducted in 1992, but Phil Rizzuto was inducted two years after him. If anything, it seems like pitchers have gone from being at a slight disadvantage to a massive one.

There are a few other interesting ways to look at this. For example, of the 63 starters in the Hall of Fame, 29 debuted in 1919 or earlier. That works out to about 46.0%. In comparison, 50 of 153 hitters debuted in the deadball era, or only 32.7%. Now, some of that is probably that deadball pitchers are overrepresented; that tends to happen when one era favors one side of the game (high-offense eras seem to help hitters, in contrast).

But at the same time, five-to-ten extra below-median starters being inducted could help balance a lot of these disparities. It would bring the liveball era median WARs for Hall of Famers more in line with each other, help equalize the recent below-median election numbers, and help to somewhat bring the liveball era inductions closer to each other.

So what does this all mean? If I had to phrase it as simply as possible, I suppose I’d summarize it as “Hall voters now only induct the most obvious starters”. For as much as people complain at times about how they don’t want the Hall of Fame to “become the Hall of Stats”, you could write a pretty basic computer program right now that could sort starters into “Hall” or “not” and it would basically match what we’ve got. Voters used to be more willing to induct starters, and still induct position players like they more or less always have, but something has changed when it comes to voting for pitching.

Another interesting point worth considering is how these below-average pitchers have been fairing on the ballot. It seems like everyone not in the first-ballot group has been stalling out lately, after all. Even well-above-average guys like Clemens, Mussina, and Schilling have been languishing for years and years. Maybe voters have just gotten worse at identifying what makes a Hall of Fame starting pitcher? Let’s take a look at some of the top recent pitchers not yet in Cooperstown by WAR:

Name	WAR	First Year on Ballot	Years on Ballot	Peak %
Roger Clemens	139.4	2013	5+	54.1%
Mike Mussina	82.7	2014	4+	51.8%
Curt Schilling	80.7	2013	5+	52.3%
Kevin Brown	68.5	2011	1	2.1%
Rick Reuschel	68.2	1997	1	0.4%
Luis Tiant	66.1	1988	15	30.9%
Tommy John	62.3	1995	15	31.7%
David Cone	61.7	2009	1	3.9%
Bret Saberhagen	59.1	2007	1	1.3%
Chuck Finley	58.5	2008	1	0.2%
Frank Tanana	57.5	1999	1	0%
Jerry Koosman	57.2	1991	1	0.9%
Dave Stieb	57.0	2004	1	1.4%
Kevin Appier	55.1	2010	1	0.2%
David Wells	53.5	2013	1	0.9%
Billy Pierce	53.1	1970	5	1.9%
Wilbur Wood	52.1	1984	6	7.0%
Orel Hershiser	51.7	2006	2	11.2%
Kenny Rogers	51.1	2014	1	0.2%
Mark Langston	50.3	2005	1	0%
Dennis Martinez	49.5	2004	1	3.2%
Jimmy Key	49.4	2004	1	0.6%
Mickey Lolich	48.8	1985	15	25.5%
Dwight Gooden	48.2	2006	1	3.3%
Ron Guidry	47.9	1994	9	8.8%
Frank Viola	47.4	2002	1	0.4%
Milt Pappas	46.8	1979	1	1.2%
Brad Radke	45.6	2012	1	0.3%
Steve Rogers	45.3	1991	1	0%
Jim Kaat	45.3	1989	15	29.6%
Vida Blue	45.0	1992	4	8.7%
Jack Morris	43.8	2000	15	67.7%
Bob Welch	43.5	2000	1	0.2%

+ designates players still on the BBWAA ballot
* designates players above Hall of Fame median WAR for liveball-era pitchers

I don’t know that this necessarily “proves” anything; it’s a pretty small sample size that has been curated, and is pretty susceptible to noise. There’s a lot we can’t quantify, too, as Hall voting is still primarily driven by narratives (for example, how do we compare Dwight Gooden’s 1982 season to Bret Saberhagen’s 2 Cy Young Awards to Jim Kaat’s Gold Gloves to Tommy John’s surgery to Kevin Brown’s PED suspension to… etc.).

But if I had to draw a conclusion, it would be this: there’s been a definite drop in votes for marginal, starting in the early 2000s. Borderline below-median starters, if they had some notable extra factor, wouldn’t always get multiple ballots appearances, but a decent number of them would. And that makes sense; I might not think Ron Guidry or Mickey Lolich or Vida Blue deserves induction, but I can see why their cases would be worth discussing at least. And sometimes, those cases would build over the years, earning eventual induction (see Bert Blyleven, or Jim Bunning; Jack Morris will probably join them after an appearance on a Veterans Committee ballot).

But starting sometime after 2000, every borderline candidate stopped getting a second look. I can see only one below-median starter after that who appeared on multiple ballots, and it’s only barely at that (Orel Hershiser, with 2). Above-average guys who got overlooked (like Blyleven or Mussina) multiplied from a rarity to somewhat common, while below-average ones went from “small chance to stick around and build up a case” to “always one-and-done”.

What caused this? I have no idea, but my guess is that it’s a combination of things. You’ve got the standard factors affecting everyone, like steroid suspicions and the limited ballot size. And this is probably part of it, but it still seems like pitchers are getting hit harder than position players. What else is different?

One issue may be that closers are edging out the fringe starters, to an extent. As I noted earlier, five or so below-median starters would help pitches greatly, and there are technically five relievers in the Hall, plus players with multiple ballot appearances like Lee Smith (15 appearances starting in 2003), Trevor Hoffman, and Billy Wagner (3 and counting each, both starting in 2016). However, that still doesn’t quite balance out, as one of those relievers (Eckersley, who I’ve seen counted as both a starter and a reliever) still fell into my “starters” category, and two more (Rollie Fingers and Hoyt Wilhelm) were inducted over two decades ago, well out of the range we’re discussing here. So this might be part of the issue, but not the sole factor.

I would also guess that the changing nature of the starting pitcher is hurting things. As bullpens increase in prominence, injuries increase, and so on, we’re seeing lower career win totals, and wins have traditionally been the voters’ dominant factor of deciding which starters make it to Cooperstown. Excluding Smoltz and Eck (who, as part closers, seem to get their save totals considered in addition to their wins), the last few inductees to make it in without 300 wins have been Pedro Martinez in 2015, Bert Blyleven in 2011 (after 14 tries), Jim Bunning in 1996 (after 15 tries plus one Veterans Committee election), Vic Willis in 1995 (Veterans Committee again), and Hal Newhouser in 1992 (15 tries plus one VC election).

Strangely enough, though, this didn’t seem to be a major issue in the years before that, as there was something of a run on sub-200-win pitchers. Fergie Jenkins made it in 1991, Jim Palmer in 1990, Catfish Hunter in 1987, Don Drysdale in 1984, Juan Marichal in 1983, and Bob Gibson in 1981. It is worth noting, though, that only Palmer and Gibson were first-ballot picks; the rest each took 3 tries (except for Drysdale, who took 10). And immediately after that run, 300-win pitchers Gaylord Perry (1991), Tom Seaver (1992), Steve Carlton (1994), Phil Niekro (1997), Don Sutton (1998), and Nolan Ryan (1999) all were elected, which might have spoiled voters and given them an unrealistic expectation of what Hall of Fame win totals might look like.

None of these answers individually seems to explain everything, but taken together, they could easily each account for a small part of a larger issue. When you have a bunch of small concurrent upward shifts in voter expectations, that could make a big difference in who gets inducted and who doesn’t.

Overall, this is something of an interesting problem. Guys above the median are usually clear snubs, but the Hall of Fame isn’t just those players. Missing out on the other players leaves out a big chunk of the game’s history. The Hall voters missing those obvious names is a big problem, but I’d argue that it’s only half the problem, and the two halves are indicative of a larger underlying trend of not knowing what actually makes starters worthy of induction. Hopefully it’s a problem that can be overcome soon.