1. Pretty amazing that Koufax tops the '61-67 list, considering he literally wasn't even active the last of those seasons, but he was STILL the best pitcher in baseball. Domination, homes.
2. Along those lines: it would be easy to miss, but Fangraphs has Doc Gooden as the best pitcher in baseball from 1982-1988. He was not in the Majors in 1982. OR 1983. That's simply how magic he was in those first two years, in particular.
3. Bill James published a thing in the waning days of his Bill James Online site about coming up with a score to assess the likelihood of a given player winning a Cy Young award at SOME point in his career. The all-time leader was Juan Marichal... who never won one. In a weird, weird quirk of his career, basically all his best seasons just so HAPPENED to align with someone else having a season the voters thought was even BETTER. What a bummer.
4. For those who are curious, if you're in the "we know neither WAR is PERFECT, so split the difference," the winners of the stretches Joe mentions would have been:
1962-68 - Gibson 40.0/40.2 vs. Marichal 43.9/37.5: MARICHAL
1972-78 - Phil Niekro 50.0/38.4 vs. Bert Blyleven 45.9/47.8: BLYLEVEN
1973-79 - Phil Niekro 51.9/36.5 vs. Bert Blyleven 44.3/45.3: BLYLEVEN
1974-80 - Phil Niekro 50.9/37.4 vs. Bert Blyleven 36.0/38.4: NIEKRO
1975-81 - Phil Niekro 44.7/31.7 vs. Steve Carlton 33.1/34.9: NIEKRO
1976-82 - Phil Niekro 41.1/31.1 vs. Steve Carlton 36.3/39.6: CARLTON
1979-85: Dave Stieb 40.0/28.8 vs. Steve Carlton 32.4/38.4: CARLTON
1980-86: Dave Stieb 38.5/28.0 vs. Steve Carlton 28.3/34.6: STIEB
1981-87: Dave Stieb 35.8/26.2 vs. Fernando Valenzuela 31.2/34.3: VALENZUELA
1982-88: Dave Stieb 35.5/26.4 vs. Dwight Gooden 29.2/32.5: STIEB (by 0.1!)
2002-08: Roy Halladay 38.3/37.3 vs. Johan Santana 42.5/36.9: SANTANA
2003-09: Roy Halladay 37.9/37.3 vs. Johan Santana 43.1/37.0: SANTANA
(Baseball-Reference WAR is listed first, and all this with the caveat that I only checked the two players Joe listed, whereas it COULD be true that a player who finished third by EITHER measure actually was the strongest by both COMBINED... but I'm not gonna work that hard.)
5. MAN does Baseball-Reference WAR just LOVE those 300-inning seasons. I don’t think it gets it right all the time, I really don’t. I just don’t think Phil Niekro was worth 50 WAR in seven years throwing junk. And there are lots of reasons to believe B-R WAR less than Fangraphs: it’s less stable, it’s possible that it’s double-counting park effects (with a separate park effect, in addition to the defensive adjustment), that it’s SO chaotic in the 1960s and ‘70s… so I take it with a grain of salt, and think that maybe 2/3 to Fangraphs makes more sense. Nonetheless, I think it’s really interesting how those results of a combined WAR show up.
FYI, for those "replacement level" doubters out there, this is a FABULOUS example of replacement level in action. You can see that Johan and Felix had a radically different number of decisions - 88 extras for Felix. What is that worth? Well, according to WAR, you can see that it was worth... almost nothing. Why? Because in those starts, Felix went 30-58. Meaning that Felix won at a .340 clip in those games... just ahead of replacement value (.294, according to the standard at both Baseball-Reference and Fangraphs). So his combined WAR value would be a little higher - a few wins. But not really anything significant, in spite of the obvious differences in the number of decisions.
So IF their W-L records reflected their true talent, this result wouldn't even be surprising. It's cool when the numbers work out so beautifully like this.
(YES - I KNOW that WAR is not based on W-L record... but it just shows up SO clearly here, and there are honestly worse approximations for a pitcher's total value than his W-L record. So don't @ me about how this is wrong; it's just showing something that "regular" and "advanced" stats functionally agree on!)
Joe or someone else is going to have to explain to me why it's inherently better to group your 7 best years together rather than spreading them out over 20 years. Especially with pitchers who are going to often have a stinker year in the middle due to arm problems (or a non existent one because of rotator cuff surgery). Is Catfish Hunter a better pitcher than Vida Blue because his best years were consecutive and Vida spread them all over the place?
To me, I think there are different conversations at play. If the argument is about a Hall of Fame peak, which is what Joe is discussing, order matters because a peak is inherently singular. If it's about a Hall of Fame career, obviously that is something different.
HOWEVER, I think this is goes toward a much, much bigger issue with modern Hall of Fame analysis and it's overreliance on WAR: While it is a universal number, the way a player achieves that total matters a whole lot.
For example, let's look at Scott Rolen and Graig Nettles. Both were star third basemen in their era and worth nearly identical bWAR over the course of their career (70.1 for Rolen, 68 for Nettles). Ever since Rolen's induction, there has been a ton of talk about how Nettles obviously belongs, too, which ignores that Rolen put up his value in 2,000 games compared to 2,700 games for Nettles, which clearly indicates that Rolen was the obviously more valuable player to his teams.
I think, from a team development perspective, there IS an argument to be made that bunching together top seasons is preferable. For example, if I have a player who's a once-in-a-generation talent, I might want to spend the most money on free agents and retain the most talent, centered on that players best year(s). If I could know in advance that I would have a player with a "smooth" aging curve or a "random" one, I'd rather have the smooth curve, because I can maximize my roster around that player in the knowledge that his mid-late 20s are the most productive years, and that his returns will be diminishing every year after that. However, if they're randomly distributed, I don't exactly know when to go "all in" on that player. It may be a small effect, and you might not consider it to be important in player evaluation (I literally don't use this consideration when I'm doing player eval., though I've considered it), but I think there is A reason. Whether or not it's good enough of a reason to care about... that's another matter.
It's not inherently better at all ... it's just another way of looking at things. Career WAR will look at all the seasons, so if you spread them out, that will come out in the wash. But I think that it is compelling if someone is the best pitcher/player in baseball over an extended period of time. Just another way of looking at things.
As for your example: Catfish Hunter is not better than Vida Blue because his best years were consecutive, but then he's not better at all -- he also has a lower career WAR, and he was also well behind Blue in the two seven-year blocks when their peaks overlap.
But Catfish is in the Hall of Fame because he won a Cy Young, won 20 a lot, timed his retirement perfectly for Hall of Fame purposes and was well-liked among the sportswriter crew.
Thanks for your reply Joe, I appreciate your explanation. I will say that Catfish also was a much better post season pitcher especially in the World Championship seasons and I am willing to give him credit for that.
I think so. The pitcher with closest similarity score in baseballreference.com is Luis Tiant, who was not clean shaven and, like Catfish, had his own unique duende, but, unlike Catfish, Luis didn't group his 20-win seasons together in a string, and he didn't play for teams as good as A's and Yanks. I think Bill James wrote something about Catfish and how the 20-win seasons were strung together had an effect on his election. Voters like the way his stat page looks.
Well, if he didn’t already go by Catfish I’m sure they would have called him James “Big Game” Hunter, so yeah probably still in….though he might have lost points with those Barbasol commercials
I'm totally on board with this. Also, I don't know if it's fair to give any bonus points for being a lefty, but by this criteria, it would make Santana one of the four most dominant southpaws since Koufax—after Carlton, Big Unit, and Kershaw? That seems pretty good.
I agree with the peak idea. Isn’t that what greatness really is anyway, being the best for an extended period? Longevity has its place, and it IS impressive, but being able to hang on and be a mediocre contributor for five years to compile stats is not greatness.
If you've listened to the latest Poscast and the number of names Joe 100% DOES NOT KNOW how to pronounce, the fact that he spelled one wrong is going to be no surprise...
My other random thoughts:
1. Pretty amazing that Koufax tops the '61-67 list, considering he literally wasn't even active the last of those seasons, but he was STILL the best pitcher in baseball. Domination, homes.
2. Along those lines: it would be easy to miss, but Fangraphs has Doc Gooden as the best pitcher in baseball from 1982-1988. He was not in the Majors in 1982. OR 1983. That's simply how magic he was in those first two years, in particular.
3. Bill James published a thing in the waning days of his Bill James Online site about coming up with a score to assess the likelihood of a given player winning a Cy Young award at SOME point in his career. The all-time leader was Juan Marichal... who never won one. In a weird, weird quirk of his career, basically all his best seasons just so HAPPENED to align with someone else having a season the voters thought was even BETTER. What a bummer.
4. For those who are curious, if you're in the "we know neither WAR is PERFECT, so split the difference," the winners of the stretches Joe mentions would have been:
1962-68 - Gibson 40.0/40.2 vs. Marichal 43.9/37.5: MARICHAL
1972-78 - Phil Niekro 50.0/38.4 vs. Bert Blyleven 45.9/47.8: BLYLEVEN
1973-79 - Phil Niekro 51.9/36.5 vs. Bert Blyleven 44.3/45.3: BLYLEVEN
1974-80 - Phil Niekro 50.9/37.4 vs. Bert Blyleven 36.0/38.4: NIEKRO
1975-81 - Phil Niekro 44.7/31.7 vs. Steve Carlton 33.1/34.9: NIEKRO
1976-82 - Phil Niekro 41.1/31.1 vs. Steve Carlton 36.3/39.6: CARLTON
1979-85: Dave Stieb 40.0/28.8 vs. Steve Carlton 32.4/38.4: CARLTON
1980-86: Dave Stieb 38.5/28.0 vs. Steve Carlton 28.3/34.6: STIEB
1981-87: Dave Stieb 35.8/26.2 vs. Fernando Valenzuela 31.2/34.3: VALENZUELA
1982-88: Dave Stieb 35.5/26.4 vs. Dwight Gooden 29.2/32.5: STIEB (by 0.1!)
2002-08: Roy Halladay 38.3/37.3 vs. Johan Santana 42.5/36.9: SANTANA
2003-09: Roy Halladay 37.9/37.3 vs. Johan Santana 43.1/37.0: SANTANA
(Baseball-Reference WAR is listed first, and all this with the caveat that I only checked the two players Joe listed, whereas it COULD be true that a player who finished third by EITHER measure actually was the strongest by both COMBINED... but I'm not gonna work that hard.)
5. MAN does Baseball-Reference WAR just LOVE those 300-inning seasons. I don’t think it gets it right all the time, I really don’t. I just don’t think Phil Niekro was worth 50 WAR in seven years throwing junk. And there are lots of reasons to believe B-R WAR less than Fangraphs: it’s less stable, it’s possible that it’s double-counting park effects (with a separate park effect, in addition to the defensive adjustment), that it’s SO chaotic in the 1960s and ‘70s… so I take it with a grain of salt, and think that maybe 2/3 to Fangraphs makes more sense. Nonetheless, I think it’s really interesting how those results of a combined WAR show up.
What an original and fun way to measure great pitchers. Thanks!
FYI, for those "replacement level" doubters out there, this is a FABULOUS example of replacement level in action. You can see that Johan and Felix had a radically different number of decisions - 88 extras for Felix. What is that worth? Well, according to WAR, you can see that it was worth... almost nothing. Why? Because in those starts, Felix went 30-58. Meaning that Felix won at a .340 clip in those games... just ahead of replacement value (.294, according to the standard at both Baseball-Reference and Fangraphs). So his combined WAR value would be a little higher - a few wins. But not really anything significant, in spite of the obvious differences in the number of decisions.
So IF their W-L records reflected their true talent, this result wouldn't even be surprising. It's cool when the numbers work out so beautifully like this.
(YES - I KNOW that WAR is not based on W-L record... but it just shows up SO clearly here, and there are honestly worse approximations for a pitcher's total value than his W-L record. So don't @ me about how this is wrong; it's just showing something that "regular" and "advanced" stats functionally agree on!)
I'm just excited to see/hear that Joe and Mike will be on a book tour again!
I've already readied myself for the disappointment that Minneapolis/St. Paul will, once again, but omitted from said tour...
Bill James for the Hall of Fame!
Joe or someone else is going to have to explain to me why it's inherently better to group your 7 best years together rather than spreading them out over 20 years. Especially with pitchers who are going to often have a stinker year in the middle due to arm problems (or a non existent one because of rotator cuff surgery). Is Catfish Hunter a better pitcher than Vida Blue because his best years were consecutive and Vida spread them all over the place?
To me, I think there are different conversations at play. If the argument is about a Hall of Fame peak, which is what Joe is discussing, order matters because a peak is inherently singular. If it's about a Hall of Fame career, obviously that is something different.
HOWEVER, I think this is goes toward a much, much bigger issue with modern Hall of Fame analysis and it's overreliance on WAR: While it is a universal number, the way a player achieves that total matters a whole lot.
For example, let's look at Scott Rolen and Graig Nettles. Both were star third basemen in their era and worth nearly identical bWAR over the course of their career (70.1 for Rolen, 68 for Nettles). Ever since Rolen's induction, there has been a ton of talk about how Nettles obviously belongs, too, which ignores that Rolen put up his value in 2,000 games compared to 2,700 games for Nettles, which clearly indicates that Rolen was the obviously more valuable player to his teams.
I think, from a team development perspective, there IS an argument to be made that bunching together top seasons is preferable. For example, if I have a player who's a once-in-a-generation talent, I might want to spend the most money on free agents and retain the most talent, centered on that players best year(s). If I could know in advance that I would have a player with a "smooth" aging curve or a "random" one, I'd rather have the smooth curve, because I can maximize my roster around that player in the knowledge that his mid-late 20s are the most productive years, and that his returns will be diminishing every year after that. However, if they're randomly distributed, I don't exactly know when to go "all in" on that player. It may be a small effect, and you might not consider it to be important in player evaluation (I literally don't use this consideration when I'm doing player eval., though I've considered it), but I think there is A reason. Whether or not it's good enough of a reason to care about... that's another matter.
It's not inherently better at all ... it's just another way of looking at things. Career WAR will look at all the seasons, so if you spread them out, that will come out in the wash. But I think that it is compelling if someone is the best pitcher/player in baseball over an extended period of time. Just another way of looking at things.
As for your example: Catfish Hunter is not better than Vida Blue because his best years were consecutive, but then he's not better at all -- he also has a lower career WAR, and he was also well behind Blue in the two seven-year blocks when their peaks overlap.
But Catfish is in the Hall of Fame because he won a Cy Young, won 20 a lot, timed his retirement perfectly for Hall of Fame purposes and was well-liked among the sportswriter crew.
Thanks for your reply Joe, I appreciate your explanation. I will say that Catfish also was a much better post season pitcher especially in the World Championship seasons and I am willing to give him credit for that.
Would Hunter be in the Hall if he were just a clean-shaven guy named James Hunter?
I think so. The pitcher with closest similarity score in baseballreference.com is Luis Tiant, who was not clean shaven and, like Catfish, had his own unique duende, but, unlike Catfish, Luis didn't group his 20-win seasons together in a string, and he didn't play for teams as good as A's and Yanks. I think Bill James wrote something about Catfish and how the 20-win seasons were strung together had an effect on his election. Voters like the way his stat page looks.
Well, if he didn’t already go by Catfish I’m sure they would have called him James “Big Game” Hunter, so yeah probably still in….though he might have lost points with those Barbasol commercials
I'm totally on board with this. Also, I don't know if it's fair to give any bonus points for being a lefty, but by this criteria, it would make Santana one of the four most dominant southpaws since Koufax—after Carlton, Big Unit, and Kershaw? That seems pretty good.
Compelling Joe!
I agree with the peak idea. Isn’t that what greatness really is anyway, being the best for an extended period? Longevity has its place, and it IS impressive, but being able to hang on and be a mediocre contributor for five years to compile stats is not greatness.
I love this comment
MLK Jr. agreed with you.
Yohan?
If you've listened to the latest Poscast and the number of names Joe 100% DOES NOT KNOW how to pronounce, the fact that he spelled one wrong is going to be no surprise...
He fixed the headline. I really thought "Yohan" might've been an inside joke or something!
Yoe Posnanski?
Yo, Posnanski! Fix the title!