Predicting Hall of Fame Pitchers Part I: or, The Voters Have Become Awful at Evaluating Starters

I’ve been meaning for a long time to write a follow up to my update looking at future Hall of Fame hitters. The pitchers presented an interesting finding though, and I couldn’t figure out how best to summarize it, so I let it sit. And before long, it just didn’t make sense to follow up; we were starting the season and everything. So, I figured I’d let it go into Hall of Fame weekend.

And now, finally, here we are. Once again into Hall of Fame season, thanks to the induction. All of the numbers are from before the season started, but the analysis is still good, so let’s go ahead.

For those who haven’t seen one of these articles before, here’s the down low: I looked at Hall of Fame pitchers and their career Wins Above Replacement (Baseball-Reference version) through each age 20 to age 35, calculating the median WAR for that age group. Then, I looked at how many pitchers in history had been worth that much through those same ages, Hall of Famer or not. Then, after taking players still on the ballot out of the equation, I figured out a simple percentage of how many pitchers at the Hall median at a given age would make the Hall. For pitchers, I made two other additional rules: they needed to start 10% of their games at any age group, and their careers had to start in the liveball era (1920 and later). Also, to keep modern players from influencing the results, I stopped the search prior to 2008 and kept out active players.

To walk through it, let’s use age 20. 14 Hall of Fame pitchers were starters by age 20. The median WAR of those 14 was 1.3, meaning seven of them had more WAR than that. 32 starters in history have been worth as much through the same age, none of whom are on the ballot. 7 out of 39 players works out to a 17.95% rate of induction.

As I mentioned, there are two factors that lead to lower percentages: not every deserving player makes the Hall, and the Veterans Committee might induct people at a later date. However, we sort of just need to live with those for the time being.

Below is all the data in chart form:

Age	Median WAR	# HOF at median WAR	# Non HOF at median	# non HOF still on ballot	% in HOF
20	1.3	7	32	0	17.95
21	2.4	10	78	0	11.36
22	4.8	12	59	0	16.90
23	6.5	14	79	2	15.38
24	9.6	14	80	2	15.22
25	12.25	15	77	2	16.67
26	18.1	15	44	2	26.32
27	24.55	15	25	2	39.47
28	27.6	16	24	2	42.11
29	34.6	16	14	2	57.14
30	38.4	16	13	2	59.26
31	42.4	16	12	2	61.54
32	45.5	16	11	2	64.00
33	51.6	16	5	2	84.21
34	55.6	16	4	2	88.89
35	59.9	16	2	2	100.00

You may notice what I did, the thing that froze me when I ran the calculations the first time: the BBWAA is terrible at voting in all but the most obvious of pitchers. This may explain why we went from the 1999 election all the way to the 2011 election with no starters being inducted. Or why Greg Maddux and Tom Glavine this year represent the first pitchers to make the Hall of Fame while debuting after 1970. Did we really go a decade and a half with no worthy starters appearing in the majors? That seems a little hard to believe.

If you need another way to look at how awful Hall voters have been at inducting starters, think about it this way: 60 players have been inducted to the Hall primarily as starters. Half of them debuted in the dead ball era. Baseball, at least on the Play Index, goes back to 1871. That means about five decades of dead ball (the ‘70s, ‘80s, ‘90s, 1900s, and ‘10s). Tom Glavine and Greg Maddux debuted in 1987 and 1986, respectively. Technically, guys who debuted in the 1990s have already reached the ballot, but we’ll just cut it off there for the sake of an argument.

That means the subsequent six-and-a-half decades, during which time there were multiple expansions (heck, the American League didn’t exist until 1901*) as well as growth in the base from which MLB draws it players (remember, there were no non-white players in the Deadball era, and almost no non-Americans), there have been just as many pitchers inducted. That’s just a little absurd. Compare this with the hitters: 147 of them have made the Hall, but only 30 of them debuted pre-1920.

*Technically, the number of teams in the 1800s fluctuated wildly. Still, I think it’s safe to assume that there were, on average, more teams playing in the majors in any given year in the live ball era.

Why exactly is this? Well, the obvious reason would be that voters don’t understand pitchers. I mean, think about it; up until maybe a decade and a half ago, the premiere way to evaluate pitchers was still the win. And remember, this was within the front offices themselves, pre-Moneyball. It’s fair to say that sportswriters haven’t been as quick to adopt new thinking.

So, writers have still by and large stuck to the win as the currency of the pitcher, a relic from the days of when pitchers threw every second or third game, usually a full 9 innings, wrapping up a win or loss that was largely influenced by themselves.

Another issue, I think, is a failure to understand and adjust for context. Consider this: in 1901, there were sixteen teams. There were eight future Hall of Fame starters playing that year; every other team had a future Hall of Fame starter. That’s, like, our definition of an ace today.* And this wasn’t even one of those crazy eras, packed with Frankie-Frisch-led Veterans Committee picks (like 1928, with eleven starters making the Hall).

*At least, I feel like fifteen is a common number picked when people debate “How many aces are there today?”

Can you imagine that transposed to today? I mean, it’s not going to be exactly the same, because we have deeper rotations now, so the players aren’t necessarily racking up as many innings. But I can also see that as a good thing; you need to judge players by their context, and being one of the top 15 starters today is more impressive than being one of the top 8 starters in 1901, just on a percentile basis.

Either way, what would that look like, transposed to a modern situation? Let’s just use 2002 as an example, since it’s recent enough that we know the players, but far enough away that we have a fuller sense of most of their careers. Can we find 15 potential Hall of Fame starters from 2002?

Roger Clemens, Randy Johnson, Pedro Martinez, Mike Mussina, Curt Schilling (5), Kevin Brown, John Smoltz, Roy Halladay, David Cone, C.C. Sabathia (10), Andy Pettitte, Cliff Lee, Kevin Appier, Greg Maddux, Tom Glavine (15)

Maybe we can quibble around the edges; say, a Tim Hudson here, or a Johan Santana there. But that’s what 1901 looks like in today’s terms, with one Hall starter to every two teams.

But even if you cut it down to just eight, so that it matches 1901 in quantity alone, what would you get? Ignoring steroids for simplicity’s sake, you’d probably get Clemens, Johnson, Martinez, Mussina, Schilling, and Smoltz to go along with Maddux and Glavine for sure. And yet, Schilling and Mussina just polled in the 20-30% range, nowhere near induction level. We also have no idea how Smoltz will do next year; however, he looks really similar to Schilling, so I don’t see him exactly sailing in.

All of this is basically to say, the Hall has gotten really, really bad at inducting pitchers more recent than the Black Sox scandal. And that’s a little embarrassing. There’s not even some confounding variable, like steroids; they just really don’t seem to know what to do.

This has already gone on longer than I intended, so I’m going to cut it short here for now. Tune in sometime in the next day or two when I go over active players and find out which young aces have already surpassed these ridiculously high standards (because there are still a few of them).