Monday, July 29, 2013

Automatic Ticket for Starting Pitchers to the Hall of Fame: What is the “New 300?” (July 29, 2013)

Is the 300-career win pitcher extinct?  With the only active 200+ win pitchers fading (Andy Pettitte, 252), on the shelf (Tim Hudson, 205 and Roy Halladay, 201), or in dire need of a reinvention (CC Sabathia, 200) you may have to go all the way down to Justin Verlander – at a mere 134 wins – to find the next potential contender.

One thing is clear: the "300-win" standard for "automatic" enshrinement in the Hall of Fame needs some fine-tuning.

It is obvious that the role of starting pitchers has changed dramatically from the days when 300 wins was the automatic ticket.  The three major and interrelated changes are the switch from four- to five-man starting rotations, the rise of specialized relief roles, and the increased focus on pitch counts.  All diminish the opportunities for a starting pitcher to stockpile wins and reach the fabled 300.  But by how much?

Before delving into that one, the cleanest answer is simply to abandon "wins" and instead use one of the new, fancy statistics that are impervious to “era effects.”  The two best are WAR and ERA+.  WAR purports to measure how many wins a player has generated in his career for his team versus that of a “replacement” player such as a journeymen pitcher or one just up from AAA.  ERA+ quantifies how much better a pitcher’s ERA is over his league for the duration of his career.

If one uses WAR, the “automatic" HOF entry level is 80.  There are 26 pitchers who have ever compiled WAR’s of 80 or better; 20 are in the HOF, and the other six are Pedro Martinez, Roger Clemens, Greg Maddux, Randy Johnson, Mike Mussina and Curt Schilling.  And if you are not a WAR fan, because, like me, you are not fond of measures that you cannot explain to another person (beyond conceptually), then you can use ERA+.  And there the answer is 128.  There are 25 pitchers who have an ERA that was 28% better than his league’s average over the duration of his career; 22 are in the HOF and the other three are Clemens, Johnson and Maddux.

But if you are a traditionalist and hate fancy statistics, you may prefer a downsized “wins” standard.  If 300 wins may be virtually unattainable and outdated, we should try to figure out a new standard.

The way I frame the basic question is in terms of "opportunities": with the five-pitcher rotation and the limitations on pitch counts, how many “win opportunities” is the average great pitcher left with versus his predecessors in the halcyon days of three days rest and routine complete games?

To determine this, I created two data sets of pitchers.  One was of the best 15 starters from the mid-1960’s to late ‘70’s, the guys who sought the full nine innings every time they took the mound.  The other is of their modern counterparts, the best 15 starters from the 1990’s and early 2000’s, the group that gave the best they had for 7 innings or so before turning it over to the short relief specialists.

I picked those pitchers who had at least 150 wins, 2,500 innings pitched, and who started at least 80% of their total games, and then I ranked them by ERA+.

The first group, the "60/70" set includes Seaver, Palmer, Gaylord Perry, Jenkins, Phil Niekro, Carlton, Rogers, Tiant, John, Koosman, Sutton, Blue, Lolich, Hunter and Holtzman.

The second group, the 90/00 set, includes Martinez, Johnson, Maddux, Brown, Saberhagen, Mussina, Appier, Cone, Glavine, Finley, Hershiser, Gooden, Candioitti, Langston and Millwood.  (I excluded Clemens for what I hope are obvious reasons.)

Here is a comparison of the career stats for the two groups.  I have put them on a “per season” basis; the first group complied 280 at least partial seasons, the second 264, not terribly different to begin with, about 1 season per pitcher.


Length of

W-L
Deci-


Comp.
CG/GS

Decisions/

Career
W - L
Pct.
sions
ERA+
Starts
Games
Starts
IP
Starts
60's/70's
18.7
14-11
0.559
25
113
31
11
35%
224
73.5%
90's/00's
17.6
12-9
0.591
21
121
28
4
14%
185
73.0%

The two groups are pretty similar.  As mentioned, both groups' pitchers had long careers, roughly 18-19 years for each on average.  The 90/00 group has a higher ERA+ and a higher won-loss percentage.  As expected, though, given the rise of five-man rotations, the second group had fewer starts, 28 per season versus 31.  And the modern group had far fewer complete games.  But I was somewhat surprised that the 90/00 group had virtually the same percentage of starts turn into decisions (73%).  One might have expected that with fewer complete games and pitch count limits, the 90/00 group would have had more “no decisions,” but that does not appear to be the case.

So what does this mean?  With about 31 starts per season over nearly 19 years, the first group had 574 starts to achieve 300 wins.  The modern group?  At 28 starts at 18 years, only 484.  If we take the career length variable out (that one season difference) and just focus on starts per year, the modern group had 10% fewer starts than the first group....that many fewer "opportunities" to compile wins.

So if the standard for automatic entry in the Hall of Fame was 300 wins for ‘60’s and ‘70’s starters, that would imply, though simple math, that a like standard for the modern era, taking into account the fewer starts modern strategies allow, would be 10% lower….or 270 wins!

So what does this mean for the current crop of recent or soon-to-be retired starters?  Here you go….using each of the measures we have identified.


Wins
ERA+
WAR
Standard =
270
128
80
Maddux
355
132
105
Glavine
305
118
67
Johnson, R.
303
136
104
Mussina
270
123
75
Moyer
269
103
50
Pettitte
252
116
59
Wells
239
108
54
Martinez, P.
219
154
86
Schilling
216
127
81
Rogers
209
108
51
Hudson
205
124
56
Halladay
201
131
65
Sabathia
200
122
54

By my standards, seven of these starters are "automatic," Maddux, Glavine, Johnson, Mussina, Martinez, Schilling and Halladay.  One might even argue that if Maddux, Glavine and Randy Johnson all exceeded 300 wins in this era, why bother changing the standard.  But Maddux and Johnson are two of the greatest pitchers of all time, by any standard, proven by their 130+ ERA+'s and 100+ WARS's.  Glavine is indeed the only "plugger" of the group, and a mighty good one at that.

Schilling is the surprise here, maybe.  Check out his postseason stats sometime as well.  You may not like him, or the way he runs video game companies, but he should be in the Hall.

Halladay is fascinating, especially given his current injuries.  One could argue that if he makes it back, every mediocre inning he pitches would diminish his ERA+ without a corresponding increase in WAR.  But right now, Halladay’s 131 ERA+ is equal to Sandy Koufax, and he has 36 more wins than Koufax.  He’s in for me even if he never pitches another game.

Mike Mussina just toes my new 270-win line, while justice is served by denying the ageless Jamie Moyer (269) by the same standard.  A comparison of their respective ERA+ and WAR is far more telling than their one-win differential.  Mussina is simply in a different class.

Those lovable Yankees Andy Pettitte and David Wells both fall short of automatic, but they will each get votes and Pettitte should make the Hall (remember, you don't need to be "automatic" to make the Hall, you just have to compile strong statistics across an array of measures);  both of them have excellent postseason records to buttress their cases.  If Pettitte returns for another season and regains some semblance of his form, he may approach the 270 mark.

Kenny Rogers falls short, while Tim Hudson needs to recover from his season-ending ankle surgery and put up more points for consideration of any kind, much less automatic entry.  And CC Sabathia needs to turn it around and learn how to pitch with a bit less velocity.  Having just turned 33, he certainly has the time and skill to make this transition, so look for him to rise in the coming years.  Who knows, he even has a shot at 270 wins!  (or 300?!)

Comments welcome!

2 comments:

  1. Why would Clemens not be included but you talk about Pettitte being borderline? If they both took PEDs...

    ReplyDelete
    Replies
    1. A good question and a good point. I mainly excluded Clemens because his career did not follow the normal arc, he clearly had a second, and even third hump which seemed to be related to steroids. It is suspicious enough that I did not want to taint the analysis. Pettitte seems to exhibit more consistency and arc in his career, which makes me think his PED usage was as limited as he claims.

      Plus the guts of the analysis did not use Pettitte's data; the output of the data was applied to his stats. I did not do that for Clemens because I do not think he will make the Hall given his vote count in his first year on the ballot. I also excluded Kevin Brown because he has already dropped off the ballot.

      Thanks!

      Delete

Leave a comment