37 Responses to “This Roy Halladay guy might be pretty good”

1. Alan Says:
   June 11th, 2011 at 10:02 am

   As an unabashed Halladay fan, I'm wary of counting this season before it's over. He's still nothing short of fantastic though.

2. Michael Sullivan Says:
   June 11th, 2011 at 10:24 am

   ah point, alan.

   So this list looks like a bunch of inner circle hall of famers (with a few yet to get there).

   One interesting thing is who isn't on it. No Seaver, no Cy Young, no Koufax. And of course none of the guys with 5 are really inner circle, but everybody ahead of that sure as heck is.

   Halladay is clearly the class of active pitchers for total career value, and the only one who seems nearly a sure bet for the hall at this point.

3. Andy Says:
   June 11th, 2011 at 10:27 am

   I noticed a general lack of guys from the 40s through the 70s, with the exception of the beginning of Carlton's career. In addition to no Koufax, there's no Bob Gibson, Whitey Ford, Hal Newhouser, etc.

4. Tim L Says:
   June 11th, 2011 at 10:30 am

   Mordeci Brown, Steve Carlton and Ed Walsh aren't inner circle? Not sure I agree.

   I do agree with your point about the number of great pitchers who never hit five. It's tough to do. Halladay is awesome.

5. Andy Says:
   June 11th, 2011 at 10:37 am

   I agree with Tim L about those guys with 5 being inner-circle, but of course that all depends on how big your inner circle is. I put mine at top 25.

   Taking a quick look at EloRater...Ed Walsh currently #24. Carlton is #10. Three Finger is #22. And FWIW Kevin Brown is #70.

6. Kingturtle Says:
   June 11th, 2011 at 10:50 am

   problem is this...in the 1960s, when the league ERA is under 4.00, like it was consistently from 1950 to 1972, a dominant ERA will less likely be in the 150 ERA+ range. When the league ERA is over 4.50, like it was in the 1920s, 30s, and 90s, 150 ERA+s are more likely.

   so i am more impressed with the guys on the above list who are from deadball eras.

7. Chuck Says:
   June 11th, 2011 at 10:59 am

   Thankfully there's no "Kevin Brown for the HOF" poll in our future.

   Not only has that ship sailed, it's already sunk.

   Kingturtle @#6..

   That's a great point.

8. Andy Says:
   June 11th, 2011 at 11:19 am

   Sean already ran a post about Kevin Brown and the Hall of Fame.

9. DavidRF Says:
   June 11th, 2011 at 11:23 am

   What a career Brandon Webb was having before he got hurt. I guess as an AL fan, I never noticed how many excellent seasons he had strung together. We often don't put careers like that into perspective until they're a little further along but 142 ERA+ is pretty hard to do even if it is "only" 1319.2 IP.

   Looking at the historical ERA+ leaderboards, I expected to see other injury-shortened careers sprinkled amongst the HOF-ers, future HOF-ers, active guys who have a shot at the HOF if they play long enough, and closers. But there's not that many. There's Jim Devlin (151) who was banned from baseball for throwing games in the 1870s. There's Smokey Joe Wood (146). Then you have to go down to guys like Breechen (134), Chandler (132) and Hahn (132).

   I guess Webb is trying a comeback. His first game in AA wasn't that hot, but you never know.

10. John Autin Says:
    June 11th, 2011 at 11:28 am

    @6, Kingturtle -- Excellent point about the relative difficulty of posting an outstanding ERA+ in a low-scoring context.

    This may be a stupid question, but ... Is there a simple explanation of why that is so? Is it just that a pitcher can't allow negative runs in a game, so the lower the league ERA, the harder it is for a pitcher to have an ERA substantially below league? Are there other factors?

11. Chuck Says:
    June 11th, 2011 at 11:48 am

    Andy @#8..

    I went through the poll list and didn't see it...how long ago?

12. Andy Says:
    June 11th, 2011 at 11:51 am

    It wasn't a poll, it was a post about him and others. I think it came out around the time of a HOF announcement, either this January or last.

13. PhilM Says:
    June 11th, 2011 at 11:53 am

    @10 J.A., I'll take a stab.

    When I wrote my study on relative team strength for SABR's "By the Numbers" newsletter some years ago, it turned out that run differential was the determining factor (actually standardized differential adjusted for league size, to be specific). If that "quality" assessment holds true for pitchers – and in the extreme, it could, as it did in 1870s when teams only had one real pitcher (think Tommy Bond) – then ERA+ will be different in different run contexts for the same run differential. A 3.00 ERA pitcher in a 4.50 league has a 150 ERA+, but a 2.00 pitcher in a 3.50 league is at ERA+ of 175. And a 4.00 pitcher in a 5.50 league has an ERA+ of 137.5.

    When I assess individual pitchers, I use ERA+ in conjunction with league ERA and plug them into negative binomial distribution, which helps to smooth out the wrinkles. The three pitchers above have roughly the same performance value, around a .700 winning percentage, which seems about right.

14. DavidRF Says:
    June 11th, 2011 at 11:56 am

    @2
    Re: Cy Young

    Cy Young should be on the list. There's probably a 1901 cutoff or something. Here are the guys with 19th century time:

    Cy Young - 6
    Kid Nichols - 5

    Also, Johan Santana should be added to the active list. He has 4. I understand its confusing keeping track of the guys who haven't pitched yet this year, but I think his comeback is more likely than Webb's, no? (comeback to MLB at least, not comeback to 150 ERA+).

15. PhilM Says:
    June 11th, 2011 at 11:57 am

    Now that I think about this more, it seems to be the opposite of conventional wisdom: if differential matters, it's actually easier to get a higher ERA+ in a low-run context. In my examples, the 5.5-run 137.5 ERA+ = the 4.5 150 = the 3.5 175. Now I'm not sure what to think. . . .

16. howard rosen Says:
    June 11th, 2011 at 12:17 pm

    I believe that the disparity between good pitchers and bad pitchers was greater during the dead ball era than it has been since, even during the offense crazy 30s, 90s and 00s. With crappy pitchers driving up the league ERA a great or even good pitcher could post a very high ERA+.

17. KB Says:
    June 11th, 2011 at 12:18 pm

    Somewhat off topic, but Cliff Lee is doing something right now I would suppose is rather unique.

    In 2008, Lee led HR/9, but not BB/9 or K/9.

    In 2010, Lee led BB/9, but not HR/9 or K/9.

    This season, Lee is leading K/9, but not HR/9 or BB/9.

    Lee has never led two or more of these statistics in a single season, but has led each one individually. Is there a way to determine when this was last done in the PI?

18. KB Says:
    June 11th, 2011 at 12:19 pm

    Check that. Lee led BB/9 in 2008 also. My mistake.

19. Jimbo Says:
    June 11th, 2011 at 12:31 pm

    He has a couple of seasons that just miss also, including his 22-7 Cy Young Season.

20. John Autin Says:
    June 11th, 2011 at 1:00 pm

    @13 / @15, PhilM -- I won't be able to follow the higher math, but it seems to me that "run differential" can be thought of in either absolute or relative terms. Considered absolutely, a given run differential will produce a greater swing in ERA+ in a low-run context, whereas treating run differential in relative terms is basically the same measure as ERA+. (But maybe I'm wrong to think that "run differential" can have either an absolute or relative meaning?)

    If I understand you correctly @13, you said that a 1.50 run differential (absolute) has about the same win value regardless of the run context. That surprises me. Would you say that applies to all run differentials, or more so at the extreme ends of the spectrum? (A 1.50 differential is pretty high.)

21. Anon Says:
    June 11th, 2011 at 1:06 pm

    @9 - DavidRF - Webb's numbers were further masked by pitching in Arizona which is a great hitter's park. Of course, ground balls play anywhere. I got to watch him a lot over those 5 years and at times it was just mesmerizing to see hitter after hitter pound the ball into the ground.

    Sadly, Brandon Webb barely qualifies as active any more.

22. DavidRF Says:
    June 11th, 2011 at 1:14 pm

    @20
    Using Pythagoras, scoring context actually drops out of the equation.

    Expected WPct is (ERA+/100)^2 / ((ERA+/100)^2 + 1)

    So, a 200 ERA+ translates to a .800 WPct in any era.

    Experts will probably note that the scoring context has an effect on the exponent. (2 or 1.82 or whatever it would be in an extreme era). I'm not sure how large of an affect that is.

    There was some stuff written about high ERA+ seasons a while back. People were curious about why Maddux & Pedro were able to outperform their leagues by so much more than Koufax & Seaver. High ERA+ seasons were indeed rare in the 60s and 70s but they were also rare in the 20s and 30s. They were common in the 10s but also common in the 90s. I don't remember the conclusions of the study.

23. John Autin Says:
    June 11th, 2011 at 1:19 pm

    Getting back to Kingturtle's point @6, after looking at all modern seasons of very high ERA+, I don't know if the data bears out the theory that an outstanding ERA+ is more common in a high-run context.

    I searched for seasons of 200+ IP and ERA+ of at least 170.
    There have been 116 such seasons:
    -- 39 came in the dead-ball era (1901-19).
    -- 14 fell between 1920 and 1941, a fairly high-scoring era.
    -- 4 came during WWII (1942-45), a fairly low-scoring era.
    -- 8 came from 1946-62.
    -- 8 came from 1963-69, the "2nd dead-ball era".
    -- 13 came from 1970-91.
    -- 26 came from 1992-2008, approximately the "PED era".
    -- 4 came from 2009-10, a fairly normal run context.

    This may not be the ideal breakdown by eras, and I haven't adjusted for the number of teams in the leagues. And the data will vary depending on the chosen thresholds of IP and ERA+. But I don't see a strong tendency for more very high ERA+ performances in high-run periods. In particular, consider those 39 seasons in the 19 years of the dead-ball era, with just 16 teams -- 0.13 per team per year. By contrast, there were 26 such seasons in the 18 years of the PED era, with (on average) over 28 teams in MLB -- roughly 0.05 per team per year.

24. John Autin Says:
    June 11th, 2011 at 1:23 pm

    @22, DavidRF -- Yes, that's more what I would expect -- ERA+ tracks W%.

    I was questioning the statement by PhilM that an absolute run differential of 1.50 runs -- e.g., 4.00 ERA in a 5.50 ERA context, 3.00 ERA in a 4.50 context, or a 2.00 ERA in a 3.50 context -- had the same win value.

25. Johnny Twisto Says:
    June 11th, 2011 at 2:36 pm

    I think PythagenPat has been accepted as the best Pythag calculator these days, and that has a floating exponent which changes with the amount of scoring.
    http://www.tangotiger.net/wiki/index.php?title=PythagenPat
    So in a real low-scoring season, you might use an exponent near 1.7, and in a high-scoring season you'd use one near 2.0. And a great pitcher can change his own scoring environment.

    So if the league average RA was 3.00, a pitcher with a RA+ of 200 has an expected winning % of .744. If the league average RA is 5.00, the 200 RA+ pitcher has an expected winning % of .775. At an RA+ of 150, expected winning % is .656 in the 3-run league and .678 in the 5-run league.

    So that's one reason (E)RA+ can't be absolutely compared across eras. Another is, as theorized above, whether it's easier to have a great ERA relative to league average when the average is higher. Intuitively, it makes some sense. But as JA shows in #23, a lot of the best ERA+ seasons occurred during the deadball era. But then you get into questions of quality of competition and standard deviations, as Howard stated in #16.

    Another reason I think we've had a lot of great ERA+ performances lately is the handling of pitchers. First, with starters pitching fewer IP, there's a greater chance of extreme performance just due to the smaller sample size. Also, by pitching fewer IP, they have fewer games when they go around the lineup four times. They're taken out of the game before they get tired or beaten up. And with bigger pitching staffs (both rotation and bullpen), there are now more bad pitchers dragging down (up) the league average.

    Second, teams don't take chances with possible injuries anymore. A little twinge, and the pitcher gets a start skipped, or is sent for an MRI, or is put on the DL. I think there were a lot more cases of pitchers battling through sore arms or outright pain in the past. In some cases, this probably just ruined their careers. In other cases, I think you have great pitchers who remained able to pitch, but went through stretches or even seasons where their arms just weren't right, and it hurt their numbers.

26. Albanate Says:
    June 11th, 2011 at 3:04 pm

    Seaver didn't make this list because of the arbitrary cut off of 150.--his ERA+ was 150 or better in 1969, 1971, 1973, 1977. But he also had seasons of 146, 143, 140, 137, and 136.

27. Dark Leviathan Says:
    June 11th, 2011 at 3:40 pm

    I love Halladay, and have a strange hope that he continues to pitch for the Phillies long enough, so that when enshrined in the HOF, he'll do so wearing a Phillies hat. (Of course a World Series ring would also help with that.)

    And just another observation: Halladay is leading in WAR right now in the NL overall, as he has 4 WAR, ahead of Kemp's 3.8. I feel that even if that continued, he would be overlooked as a possible MVP candidate.

    And one final thing of note, Hamels seems to be right behind Halladay is several categories. WAR, ERA, Wins (obviously could change once he has the same number of starts), K/9, and K/BB, is close in some other categories. Hamels is really rising up.

28. PhilM Says:
    June 11th, 2011 at 5:05 pm

    @20, 24 & 25:

    Those are great points, and my study was for teams, not individual pitchers, so I'm just hypothesizing here and pushing to see if the model works. I do like the contextual element in PythagenPat, since I've long been annoyed by the shortcomings of a one-power-fits-all "standard" Pythagoras. But both PythagenPat and my NegBin agree that a given ERA+ means more in a high-run context: the expected win percentage is higher.

    One conclusion that can be drawn from this is that a lower ERA+ in a high-run scenario can be roughly equivalent to a higher ERA+ in a low-run scenario -- which is achievable if the differential is preserved:

    30 decisions, ERA+ of 125
    120 ER, 150 LgER = 30 R diff
    ERA of 4, LgERA of 5
    .634 NegBin PCT (19-11)

    30 decisions, ERA+ of 133.33
    90 ER, 120 LgER = 30 R diff
    ERA of 3, LgERA of 4
    .653 NegBin PCT (20-10)

    30 decisions, ERA+ of 150
    60 ER, 90 LgER = 30 R diff
    ERA of 2, LgERA of 3
    .683 NegBin PCT (20-10)

    The higher ERA+ pitcher is still "better," but not by as much as ERA+ would suggest. I guess that's what I was getting at -- with more study, the differential may be more meaningful as indicator of pitcher quality than the ERA+ ratio.

29. Voomo Zanzibar Says:
    June 11th, 2011 at 7:11 pm

    Kevin Brown really has taken too much guff.
    He was as dominant as anybody there for awhile.

    Form what I've observed he gets (got) picked on for two things:
    1. Being the highest paid player in the NL for 4 years, and then
    2. Not being Walter Johnson when he played for the Yankees.

    He won 21 games in Arlington,
    and led the Marlins and Padres to pennants in consecutive years.
    I betcha a million imaginary dollars that nobody ever pulls off that trifecta again.

30. Johnny Twisto Says:
    June 11th, 2011 at 7:15 pm

    Phil, do you have a link to that article you wrote, and/or a description of NegBin?

31. PhilM Says:
    June 11th, 2011 at 8:24 pm

    @30:

    It's a little long in the tooth (2003, so my contact info is out of date), but I think the results are still valid. Begins on page 15:
    http://www.philbirnbaum.com/btn2003-02.pdf

32. Neil L. Says:
    June 12th, 2011 at 12:02 am

    When I look at some of the old-time ERA's and ERA+'s of pitchers on the list like Walter Johnson, Lefty Grove and Mordecai Brown.it's like they are from another universe.I find it hard to relate to the numbers.

33. Rich Says:
    June 12th, 2011 at 4:09 am

    As a fan of baseball in general, I'll be rather sad if Webb never makes it back. It's nuts just how good he was when he got hurt.

34. Norman Morrow Says:
    June 12th, 2011 at 5:18 am

    Nolan Ryan only had one season of ERA+ of greater than 150 and that in the strike disrupted season of 1981. I still wouldn't have traded him for Fregosi though.

35. John Autin Says:
    June 12th, 2011 at 3:28 pm

    @29, Voomo -- Good points about K.Brown. But I would add a couple other reasons for bias against Brown:
    -- He was named in the Mitchell Report.
    -- He had a prickly personality (perhaps related to the previous point). Wikipedia reports an incident in which a neighbor accused Brown of pulling a gun on him in a dispute about trash in his yard.
    http://en.wikipedia.org/wiki/Kevin_Brown_(right-handed_pitcher)

36. Andy Says:
    June 12th, 2011 at 3:29 pm

    If we're gonna start judging players by character, I have a feeling a whole lot of these fellows are going to fall short.

37. John Autin Says:
    June 12th, 2011 at 7:20 pm

    @36, Andy -- I was just describing the sources of anti-K.Brown sentiment, not endorsing them. I don't think you would deny that many baseball honors (HOF and major awards) have been influenced by whether or not the voters like a player personally.