Baseball for the Thinking Fan

Login | Register | Feedback

btf_logo
You are here > Home > Hall of Merit > Discussion
Hall of Merit
— A Look at Baseball's All-Time Best

Sunday, July 09, 2006

Bob Gibson

Eligible in 1981.

John (You Can Call Me Grandma) Murphy Posted: July 09, 2006 at 07:23 PM | 77 comment(s) Login to Bookmark
  Related News:

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

   1. John (You Can Call Me Grandma) Murphy Posted: July 09, 2006 at 07:27 PM (#2093069)
Was Gibson referred as "Hoot" frequently as a player?

Talk about a nickname that has lost it's currency over the years.
   2. Los Angeles El Hombre of Anaheim Posted: July 09, 2006 at 10:29 PM (#2093403)
I've read that his 1.12, pure luck.
   3. DavidFoss Posted: July 09, 2006 at 10:48 PM (#2093430)
I've read that his 1.12, pure luck.

Tee hee... *where* did you read that? :-)

Gibson had amazing peripheral numbers that year. His RA was 1.45 which BP lists as being a bit *unlucky* for his numbers (they said he was lucky on the hits though).

Whatever luck he may have had in ERA, he gave back in W/L record. A pitcher with that ERA should have gone 26-5 instead of 22-9.
   4. OCF Posted: July 09, 2006 at 10:50 PM (#2093432)
There was a thread on BTF Newsblog just last week. Here are four posts quoted from that thread:

4. OCF Posted: July 07, 2006 at 07:11 PM (#2091395)
Gibson becomes eligible for the Hall of Merit next "year" (discussion starting next week), and I spent a good fraction of the summer of 1968 with my ear glued to the radio listening to this accomplishment as it happened, so I'll have a great deal to say at that time.

As for 1968:

1. Gibson was, in fact, a great pitcher, and
2. He was at the peak of his abilities in 1968, and
3. 1968 was a the bottom of the trough of the "little dead ball days" of the mid-60's, a year of freakishly low run scoring, and
4. On top of all of that, Gibson's 1968 was a BABIP fluke in his favor.

Gibson's ERA+ was 258 that year. If you include his unearned runs, I have his RA+ at something more like 232, for an RA+-equivalent record of something like 27-7. It was a great, great year, but if someone wants to say that Gooden's 1985 was better, I won't object.

One thing I wonder about the BABIP aspects of that: is it possible that Gibson that year was, to a better-than-usual degree, able to keep the ball in the middle of the field and off the lines? Consider the defense he had behind him. On the corners he had a converted outfielder at 3B, an erratic and error-prone (albeit fast) LF, a sound, but old and slow, RF, and a 1B whose heart and concentration wasn't always into playing defense. But up the middle? Maxvill was a magnificent defensive SS (with a where'd-that-come-from 91 OPS+ in '68), Javier did a good DP pivot, and Flood was a fabulous CF except that he couldn't throw. It was an average-ish defense overall, but much, much stronger in the middle of the field.

----

23. OCF Posted: July 07, 2006 at 11:50 PM (#2091876)
Geez - can we un-hijack this? The article is about big pitching years in general and Bob Gibson's 1968 in particular. How did it get to be about the retirement of Mickey Mantle? We Cardinal fans (or ex-Cardinal fans, or whatever it is that I am) can have a few threads of our own, can't we?

Gibson's starts in June and July of 1968; every single one of them a complete game for him:

6/02: over Mets at NY 6-3; scoreless 8th and 9th.
6/06: at Hou; shutout 4-0.
6/11: at Atl; shutout 4-0.
6/15: vs. Cin; shutout 2-0. (vs. Gary Nolan; time of game something like 1:45)
6/20: vs. Chi; shutout 1-0. (Ferguson Jenkins's tough luck)
6/26: vs. Pit; shutout 3-0. (I remember him striking out Clemente to end the game with runners on)
7/1: at LA, versus Drysdale. Going after Drysdale's brand-new record of 6 consecutive shutouts. Dodgers score a run in the 1st on a WP (!) but Cardinals win game 5-1. (8 innings on a new scoreless inning streak after the previous one ended at 47)
7/6: at SF; shutout 3-0. (Juan Marichal's tough luck)
7/12 vs. Hou; win 8-1. The run scored in one of the later innings, after it was blown out.
7/21 vs. NY; shutout 2-0.
7/25 vs. Phi; shutout 5-0
7/30 at NY; win 7-1.

----

30. fra paolo Posted: July 08, 2006 at 07:22 PM (#2092304)
is it possible that Gibson that year was, to a better-than-usual degree, able to keep the ball in the middle of the field and off the lines?

I don't know if his distribution for 1968 was unusual for his career, but it does follow pretty closely the rest of the Cardinals pitchers. I went through the Retrosheet box scores for Gibson, and then I compared the fielding position of PO+A for Gibson as opposed to the rest of the team.

I took catcher PO+A out and got the following by dividing the positional total by the total for the team:

Position     Gibson     Non-Gibson
P             
,058          ,058
1b            
,305          ,315
2b            
,177          ,16
3b            
,088          ,091
ss            
,152          ,163
LF            
,062          ,057
CF            
,095          ,086
RF            
,063          ,07 


If you add up 2b+ss+cf for each category you get 42% for Gibson and 41% for everybody else. Would that one percent explain the big difference in BABIP?

   31. OCF Posted: July 08, 2006 at 07:32 PM (#2092315)
Interesting. I probably would have left 1B PO out of it as well. Not that strong a signal, even on the issue of GB/FB. For Gibson more LF plays, fewer RF plays, but also more 2B plays, fewer SS/3B plays. So probably not anything to my theory and the data is a little confusing even on R vs. L.
   5. DavidFoss Posted: July 09, 2006 at 10:55 PM (#2093437)
I have a feeling this is going to be a short thread because Gibson is just enormously qualified. 127 is a huge ERA+ for that workload in this era. He's basically Jim Palmer with more K's and in a better league. He's Tom Seaver without Tom's incredible longevity. Except for Seaver, he's better than any of the giant class of guys that followed 5-10 years after him (Carlton, GPerry, Niekro, Palmer, Ryan, Jenkins, Blyleven, etc).
   6. OCF Posted: July 09, 2006 at 11:15 PM (#2093454)
Gibson is listed as 6'1", somewhere around 190 or 195 lbs. In everyday society, that makes him a big man, but by the standards of major league pitchers, that's not all that big. Koufax was bigger. Drysdale was bigger. Veale, Maloney - those guys were huge in ways that Gibson wasn't. But he always seemed big standing on the mound. Part of it was that he was a terrific all-around athlete and looked the part. And part of it was the glare - the deep-set eyes boring right through the batter. He seemed to despise everyone who ever stepped into a batters box - or at least that's the message his face conveyed.

His motion was vigorous and on his follow-through he fell violently off towards first base. But I remember him being smacked in the rump by a line drive - and recovering to throw out the batter. After all, he was an athlete.

For a pitcher, he was a very good hitter. That's a long way short of being a good hitter, of course. He was used a little as a pinch hitter, but I wouldn't make much of that. But I sure wouldn't ever have walked Dal Maxvill intentionally just to get to Gibson.
   7. OCF Posted: July 09, 2006 at 11:25 PM (#2093466)
And he worked fast. Get the ball back from the catcher; right back on the rubber; get the sign quickly; go right into the motion. Probably nearly every pitcher from the mid-60's worked faster than nearly every pitcher now, but even compared to his contemporaries, Gibson worked fast. You didn't want to be late getting to one of his games.
   8. Howie Menckel Posted: July 10, 2006 at 01:26 AM (#2093624)
Yeah, if you're too young to have watched the glare, it's hard to describe what you missed.
Closest comp, to me, would have been Gossage.

"Game over" Gagne had a good stare, but that was just about you not getting a hit.

With the first two, you're not just worried about striking out - you fear for your life.
   9. Daryn Posted: July 10, 2006 at 01:44 AM (#2093640)
Dave Stewart had a nice glare. He has said he modelled it after Gibson.
   10. OCF Posted: July 10, 2006 at 01:50 AM (#2093651)
Two pitchers - both the same age (both born in the second half of 1935). Both started their careers with the Cardinals in 1959, and were with the team at least through 1963.

In the five years 1959-1963, the first pitcher had an RA+ equivalent record of 55-41, including years of 18-8, 14-9, and 15-13. The second pitcher had an RA+ equivalent record of 69-49, including years of 17-8, 17-8, and 16-12. Both pitchers were certainly effective, but the second pitcher more so: 69-49 has to be considered better than 55-41, and the second pitcher had a somewhat better 1963 season. Both were 28 years old at the beginning of the 1964 season - not young any more.

The first of my two pitchers? Bob Gibson. The second, the one with the slightly better record through age 27? Ernie Broglio.

Now, I'm sure the Cardinals saw Gibson as the more valuable property, and to some extent, that's justified: Broglio's strikout rates were good, but Gibson's were better, and in the long run, betting on the bigger power pitcher often pays off. But I've got two points to make here.

The first is that before "the trade," Ernie Broglio was a high quality pitcher and a highly valuable property. His sudden and total collapse was a shock and a surprise (and short of sending an MRI machine back in time to 1964, I'm not sure we'll get much of an explanation for it.)

My second point: Gibson was a late bloomer. How many pitchers with records through age 27 looking like that turn into all-time greats? One in a hundred? One in five hundred?

--

My overall RA+ equivalent record for Gibson is 265-166, which is significantly better than his actual record of 251-174. (He does have 9.14 IP/decision, which is high.) That puts him into some rarified company. I have him as the second best post-WWII pitcher we've seen, behind only Spahn - and there are some interesting peak vs. career dynamics when it comes to comparing him to Spahn. How about a 6-year peak? (That number isn't friendly to Gibson, because we have to include his broken-leg 1967 and 65-70, which is what I took, is only a tiny bit better than 64-69, so he'd prefer you took 7 years.) For the 6 years 65-70, I have Gibson's RA+ equivalent record as 123-62. Compare that to the last 6 years of Koufax: 123-58. And of course Gibson has a big "rest of career;" Koufax doesn't.
   11. Howie Menckel Posted: July 10, 2006 at 01:58 AM (#2093659)
Good one, Daryn, Stewart had a good game face - but it couldn't match Bob.

As OCF noted, Gibson was that rarest of athletes who exuded both power AND speed.
Randy Johnson frightened the hell out of LH batters, but if somehow they laid down a great bunt, RJ couldn't pounce off the mound like Gibson did.

Gibson even knocked in 19-20 runs three times. 5 HRs in both 1965 and 1972, and he gave off an aura at the plate even though he only hit .206 for his career (which for that era was excellent for a P, but he looked like a better hitter than that).
In 1965, he outslugged Javier, Groat, Boyer, and Shannon. In 1972, he outslugged every regular except HOM candidates Simmons and Torre.
   12. DavidFoss Posted: July 10, 2006 at 02:06 AM (#2093664)
As OCF noted, Gibson was that rarest of athletes who exuded both power AND speed.

What position did he play for the Globetrotters. At 6'1", its likely he was a guard, but it is hard for me to imagine that.
   13. Alex meets the threshold for granular review Posted: July 10, 2006 at 02:31 AM (#2093697)
He's Tom Seaver without Tom's incredible longevity. Except for Seaver, he's better than any of the giant class of guys that followed 5-10 years after him (Carlton, GPerry, Niekro, Palmer, Ryan, Jenkins, Blyleven, etc).


More than that, he's Tom Seaver without Seaver's incredibly quick startup period. Seaver was an ace straight out of the box. It took Gibson a while more to start up, but he and Seaver were pretty damn close at their best.

From 1961 to 1972 (I excluded 1967, which paints him in a more favorable light, so bear that in mind), Gibson had 3002.2 IP of a 139 ERA+. From 1967 to 1977, Seaver had 2979.2 IP of a 142 ERA+. Seaver's peripherals were, for the most part, a little bit better across the board (strikeouts, walks, hits, all just a bit better compared to the league than Gibson's were) - except for homers. Gibby murdered the long ball.
   14. Ardo Posted: July 10, 2006 at 03:13 AM (#2093753)
When I think of Bob Gibson, I think of only two other players: Jackie Robinson and Ty Cobb. All three men didn't just want to win; they wanted their opponents to die. And all three had the surpassing talent to make it happen.
   15. OCF Posted: July 10, 2006 at 03:20 AM (#2093765)
Gibby murdered the long ball.

I vaguely remember some long-ago article in which Gibson described two different fastballs that he threw. I think he was describing the difference between a 2-seamer and a 4-seamer, with the implication that he made rather more use of the sinker, but I won't swear to that. Or maybe he occasionally threw what we'd now call a cutter? Anyhow: fastball, slider. Keep it simple, work it in and out but throw strikes. Once in a great while a slow curve, but mostly just fasball and slider. And everyone always said he threw a "heavy" ball: hand-stingingly, bat-breakingly heavy - which probably means a lot of last-minute motion keeping it off the sweet spot of the bat.
   16. Jolly Old St. Nick Is A Jolly Old St. Crip Posted: July 10, 2006 at 03:33 AM (#2093772)
part of it was the glare - the deep-set eyes boring right through the batter. He seemed to despise everyone who ever stepped into a batters box - or at least that's the message his face conveyed.

When I think of Bob Gibson, I think of only two other players: Jackie Robinson and Ty Cobb. All three men didn't just want to win; they wanted their opponents to die. And all three had the surpassing talent to make it happen.


A nod of respect, please, for the late Salvatore Anthony Maglie, who put the fear of God into a few hitters himself. On the days the Dodgers were in town, the Braves would have gladly traded Warren Spahn for Maglie, told the Giants to keep the change, and thrown in a few hookers, too. He absolutely owned that Brooklyn team, scared them half to death.

But I remember [Gibson] being smacked in the rump by a line drive - and recovering to throw out the batter.

That play was in the last of the ninth inning of game 5 of the 1964 World Series, with one out, the Cardinals ahead of the Yanks 2 to 0, and the batter was Joe Pepitone. And if Gibson hadn't made that play, he likely would have lost the game, since the next batter, Tom Tresh, hit the first pitch for what would have been a game-winning home run, but instead McCarver won it with a homer in the 10th. That was the most amazing fielding play I've ever seen---the ball bounced off Gibson nearly all the way to the third base line and he was falling backwards when he threw to first---especially given the circumstances. It not only saved the game, but the Series as well, as the Yanks crushed the Cardinals in game 6.
   17. Dag Nabbit is part of the zombie horde Posted: July 10, 2006 at 04:11 AM (#2093833)
I have a feeling this is going to be a short thread because Gibson is just enormously qualified. 127 is a huge ERA+ for that workload in this era.

From 1961 to 1972 (I excluded 1967, which paints him in a more favorable light, so bear that in mind), Gibson had 3002.2 IP of a 139 ERA+. From 1967 to 1977, Seaver had 2979.2 IP of a 142 ERA+.

Gibson belongs in the HoM. No doubt about that. But I feel obliged to point out those Cards teams had some of the best up-the-middle defense of any team you'll ever see. That aids his ERA+. When I did my little defensive adjustment on my old sight, only Jim Palmer had better career defensive support in the liveball era.

Except for Seaver, he's better than any of the giant class of guys that followed 5-10 years after him (Carlton, GPerry, Niekro, Palmer, Ryan, Jenkins, Blyleven, etc).

Are you talking prime or career value?
   18. DavidFoss Posted: July 10, 2006 at 04:39 AM (#2093863)
Are you talking prime or career value?

RSAA lists him 2nd from 1946-1992 behind Seaver. 1992 was chosen to get all of Blyleven (who fares remarkably well with this high baseline).

I realize that RSAA is a very high baseline (average) and chosing a lower baseline will likely vault several career guys ahead of him. My only point was that 127 ERA+ mixed in with the 3800 IP is a combination only Seaver can match between WWII and the current crop of over-40 super-pitchers (Clemens/Maddux/Johnson).

I suppose it was a bit of an overstatement, but I'm not even a big Cards/Gibson fan and I was colored impressed. :)
   19. TomH Posted: July 10, 2006 at 02:51 PM (#2094080)
Quibbles about Bob Gibson:

1. Number of ERA titles: 1
2. Number of years, leading league in wins: 1

Having said that, he's qualified for the upper quarter of the HoM. Great peak, prime, fielder, hitter, World Series stats. I have 9 pitchers clearly above him (Walter, Rocket, Grove, Young, Maddux, Alexander, Seaver, Matty, Spahn), and another few that could possibly equal him (Satchel, Smokey Joe, Nichols, Unit, maybe Pedro soon?). That's pretty sweet company. I woulda left him off the all-century team, but that is more quibbling.
   20. John (You Can Call Me Grandma) Murphy Posted: July 10, 2006 at 03:03 PM (#2094091)
Having said that, he's qualified for the upper quarter of the HoM.

That sounds right, Tom. Definitely deserves to go in next "year," but I'm not ready to state that he's an inner-circle HoMer, either.
   21. Alex meets the threshold for granular review Posted: July 11, 2006 at 05:18 AM (#2095051)
Quibbles about Bob Gibson:

1. Number of ERA titles: 1
2. Number of years, leading league in wins: 1


True, but he had 2 ERA+ titles (258 in 1968, 149 in 1962), and in 1969, finished third at 164 - behind Seaver and Marichal who both had 166.
   22. Joey Numbaz (Scruff) Posted: July 11, 2006 at 05:42 AM (#2095068)
I posted this on the new eligibles thread, more relevant here (slightly updated):

Gibson is done - wow. He's what I thought Marichal would be.

1.136 PA - best pitcher we've seen in awhile. For comparison, he's as far ahead of Drysdale as Drysdale is ahead of Dutch Leonard, Dolph Luque, Don Newcombe (with about 5 years of extra credit) and Dizzy Trout.

He's as far ahead of Bunning as Bunning is ahead of Bob Friend, Curt Simmons and Larry Jackson.

He was a better hitter than Drysdale too (48 BRAR).

His 1968 stands up, second only to Dizzy Trout's 1944 among the non-deadball pitchers that I've done. Johnson has a bunch of better years and Pete Alexander's 1920 is ahead of it also, as is John Clarkson's 1889. I haven't done the HoMers that come alphabetical after "Ferrell" yet.

However the years that are ahead of it are there because of the innings those guys threw (even after adjusting for era), in terms of rate, the only seasons I've found ahead of Gibson's 2.19 aDRA (4.50 = average) are:

Walter Johnson 1913 1.96

That's it. That 1.12 in 1968 is every bit as good as it's cracked up to be.

Career wise, his defense helped him some (+.07 on his aDRA), his bullpen only saved him 2.9 runs over his career. He only threw 88 rIP, but they were at a 1.44 LI (69 rIP at 1.71 if you take out 1960).

His best 5 seasons (42.3 aWAR) are behind Johnson, Alexander, Feller and tied with Clarkson among the pitchers I've done. He is as far ahead of Marichal on 5-best as Marichal is ahead of Pierce.

Same for his 3-year run of 1968-70. His top 3 consecutive (27.8) is better than Dizzy Dean's (26.6), behind only Johnson, Alexander and Feller.

Johnson 2.276
Alexander 1.782
Feller 1.423 (with 3.75 years conservative WWII credit)
Gibson 1.136
Drysdale .933
Quinn .901 (with 1916-18 PCL credit)
Bunning .888
Pierce .839
Faber .838
Marichal .823
Ferrell .822
Bridges .814 (with about 2 years credit as a Sunday pitcher in 1944-45).
Coveleski .787
Grimes .773
Trucks .758 (includes 1944-45 credit)
3FBrown .758
Shocker .751

I'll have all of the HoMers done later this week, and I'll post a complete table like I did in the Marichal thread at that time.
   23. OCF Posted: July 11, 2006 at 05:48 AM (#2095071)
Looking backward, I have no problem taking Gibson over every post-WWII pitcher except Spahn. Gibson versus Hubbell might be an interesting debate but I don't want to go back that far. Looking forward ... he's obviously not Seaver, but I'm also running into a problem in comparing him with some other members of the next generation, notably Perry. (Gaylord, of course.) There was something in the way pitchers were utilized - the air-travel regularized schedule, the brief heyday of the the 4-man rotation, low scoring putting less strain on pitchers - that made possible seasonal IP totals that hadn't been seen in a generation or more. And so those pitchers piled up their innings, then switched over to the 5-man rotation and lasted forever. They've all got inning bulk and career length advantages over Gibson. And Perry even has his own outlier top year to nearly rival Gibson's. But Gibson's career extends into that time - he was in position to enjoy at least some of those advantages himself.

Returning to the backward look: the closest rival for that title of "best other than Spahn" is Robin Roberts. Now, Gibson was a workhorse, and a very healthy pitcher. As far as I can tell, his only significant injury was that 1967 broken leg - from which he recovered completely, and in time for the '67 World Series. But, workhorse that he was, his innings comparared to his contemporaries are nothing remotely like those for Roberts. When Roberts was pitching 300+ innings a year, he was dominating the league in that category.

I'll take Gibson over Roberts because Gibson was, for the most part, considerably more effective per inning. But you do have to at least think about that comparison.
   24. Alex meets the threshold for granular review Posted: July 12, 2006 at 03:43 AM (#2096412)
1.136 PA


PA?
   25. Cabbage Posted: July 12, 2006 at 04:21 AM (#2096430)
Gentleman, long-time lurker in the HOM threads, but a first-time poster.

Question about Gibson:

MAybe I'm missing something, but why am I not as impressed by Gibson as some of you are? I would rank him as a great peak pitcher certainly, but I'm stuck on trying to put him in a context removed from his contemporaries. I've learned quite a bit from reading these threads, so I'd love to hear some insight. I'm not sure I'd say Gibson had a better career than Carlton, but I suspect I'm more of a career guy anyway. Several of you have listed him as better than Carlton (who's really just an example I'm using), can you elaborate as to why?

Much obliged.
   26. Dag Nabbit is part of the zombie horde Posted: July 12, 2006 at 04:29 AM (#2096433)
Joe - could you (or someone else) give an explanation (or a link to a previous explanation) to what you're talking about. I have no idea what any of those acronyms you're using are. I see PA and think Plate Appearnacnes.
   27. DavidFoss Posted: July 12, 2006 at 04:34 AM (#2096438)
Pennants Added (definitely needs less ambiguous abbreviation)
   28. Flynn Posted: July 12, 2006 at 04:36 AM (#2096441)
My father said Bob Gibson was the best pitcher he ever saw. I don't know if you can really think he's wrong. You can disagree with him, but I don't know if you can think he's wrong.

It helped he was 13 in 1968.

Gibson was tough. He was a different kind of tough. Not a Ty Cobb or Sal Maglie, don't want to meet you in a dark alley kind of tough. But more like a snake. You don't #### with a snake, nor Bob Gibson.
   29. DavidFoss Posted: July 12, 2006 at 04:43 AM (#2096444)
Pennants Added comes from the 2002 BP Book. Page 470 "The Problem with "Peak"" by Michael Wolverton. The algorithm converts yearly values into a career total of pennants. It basically gives a bonus for peak performance that is based on mathematical logic and does not require that you provide a peak length.

There should be a couple of threads on the main page about Pennants Added with some out-of-date tables. Joe's got a new homemade base metric for his 'yearly values' which he hasn't fully explained, but I'm guessing the yearly-values-to-career-total algorithm is the same as in BP.
   30. DavidFoss Posted: July 12, 2006 at 05:29 AM (#2096464)
I'm not sure I'd say Gibson had a better career than Carlton, but I suspect I'm more of a career guy anyway. Several of you have listed him as better than Carlton (who's really just an example I'm using), can you elaborate as to why?

You wouldn't necessarily be wrong. Especially with your self-label of 'career guy'.

Gibson probably better peak but its close. Top ERA+ seasons:

BG-258-164-149-148139 -136-133-132-127-126
SC
-182-164-162-151-(150)-126-119-117-115-113 


(1981 strike year in parentheses)

Gibson's main advantage is in 'prime'. Carlton has a significant falloff after the top four and the strike year. Also note that Gibson's career is not "short".

The other advantage that Gibson may have is that Carlton had quite a few mediocre seasons mixed in with the greatness which weights down his career averages.

Gibson has 127 ERA+ in 3884.3 IP.
Carlton has 115 ERA+ in 5217.3 IP.

The difference ends up being a 90 ERA+ in 1333 IP which is really not that bad. Not as bad as I was expecting before I did the calculation. Does it add value and therefore Carlton a better pitcher? Maybe, but I can see why even a moderate peak voter could stick with Gibson. Plus, if you set a high baseline like RSAA does, then Gibson rank higher.
   31. Joey Numbaz (Scruff) Posted: July 12, 2006 at 08:15 AM (#2096537)
PA = Pennants Added. I guess I could used PenAdd or something, but a 6-letter abbreviation is a bit tough.

I've fully explained the system, piecemeal over the last few years, it's scattered - a bunch of it is in the Wilhelm thread, a bunch more in the 1977 Ballot Discussion IIRC - I need to do it again and put it somewhere easily referencable, but not right now.

************

I finally found a year besides Johnson 1913 with a better rate than Gibson's 1968 of 2.19 aDRA (that's adjusted DRA - my adjustments use the PythaganPat exponent, which makes a huge difference when working with individual pitchers). Basically it's R/9 converted to a 4.50 environment at the same Win Impact as in the actual run environment. It's adjusted for team defense (using the NRA/DERA adjustment from Baseball Prospectus) and after 1959 for things like bullpen support and performance out of the bullpen with inherited runners).

Sandy Koufax had a 2.06 aDRA in 1966. So Gibson's 1968 drops to 3rd on the adjusted scale, still holding up very well. With the pitchers I have left (HoMers from Lemon to Young alphabetically), I think Newhouser and Walsh are probably the only ones that might match it . . . won't finish tonight, but hopefully by tomorrow, Thursday night at the latest.
   32. Joey Numbaz (Scruff) Posted: July 12, 2006 at 08:19 AM (#2096538)
BTW, Koufax does pretty well on PA, .849, which places him between Bunning and Pierce. I couldn't wait to do him, as he's definitely a benchmark on the extreme peak side. It also includes Koufax's atrocious hitting.

I think that's pretty reasonable, and makes me feel pretty good about the peak/career balance of the metric.
   33. Joey Numbaz (Scruff) Posted: July 12, 2006 at 08:21 AM (#2096539)
When I'm done with the HoMers, I'll run Carlton, but only through 1980, since we are currently voting in January of 1981. No looking ahead :-)
   34. Joey Numbaz (Scruff) Posted: July 12, 2006 at 08:23 AM (#2096540)
Also note aDRA is adjusted for park and league, but I assumed that went without saying. But I'll say it, just in case . . .
   35. sunnyday2 Posted: July 12, 2006 at 12:38 PM (#2096571)
>Top ERA+ seasons:

BG-258-164-149-148- 139 -136-133-132-127-126
SC-182-164-162-151-(150)-126-119-117-115-113

As a peak voter I gotta take Gibson. This is not Norm Cash, where you've got one great year, but then it's year 9 before he catches up to the competition. In this case Carlton only takes years 3-4-5, but then Gibson take 5-7-8-9-10. Too much Gibson there, and for me the 90 for 1300 innings does not really factor in much at all, at least not against Gibson. Against some other pitcher whose prime is similar, yes, but not in this case. I don't get to the tie-breaker based on the top 10 seasons.
   36. Cabbage Posted: July 12, 2006 at 02:52 PM (#2096677)
Interesting. Thanks for the input fellas.

Frankly, I think I was just missing some of BG's "pretty darn good" seasons. He's got quite a bit more 'career' value than I originally thought. Certainly changes my thoughts about him.

I'll have to do a bit more pondering/number crunching.... and I probably need a more consistant system.... not that I have time to start participating here, but I'll keep reading. thanks for the thoughts!
   37. DL from MN Posted: July 12, 2006 at 03:31 PM (#2096734)
As an abbreviation for Pennants Added, how about FLAG
   38. Alex meets the threshold for granular review Posted: July 12, 2006 at 08:20 PM (#2097025)
'm not sure I'd say Gibson had a better career than Carlton, but I suspect I'm more of a career guy anyway. Several of you have listed him as better than Carlton (who's really just an example I'm using), can you elaborate as to why?


The elaboration has already been done, so I won't bother, but let me just say that I've never really been able to completely nail down who I prefer when it comes down to debates betweeen guys like Gibson and Carlton - on one hand, a guy with a great, consecutive peak/prime and lots of innings (Gibson), and another with a nonconsecutive peak with lots of 110-115 ERA+ seasons but a huge amount of innings (Carlton). It doesn't help that both of these guys pitched for my favorite team (briefly in the latter case...god, what a stupid, stupid, stupid trade...ARGH).

If Carlton's peak were consecutive I'd probably be a lot more tempted to rank him above Gibson. But it's hardly unreasonable to do so regardless, if you're a big career guy.
   39. OCF Posted: July 13, 2006 at 01:09 AM (#2097205)
... both of these guys pitched for my favorite team (briefly in the latter case...god, what a stupid, stupid, stupid trade...ARGH).

Well Rick Wise contributed two years of high-inning ERA+ 110 before being traded, in part, for Reggie Smith who had two years at OPS+ 157 and 137.

But still, ARGH.
   40. John (You Can Call Me Grandma) Murphy Posted: July 13, 2006 at 01:18 AM (#2097211)
My father said Bob Gibson was the best pitcher he ever saw. I don't know if you can really think he's wrong. You can disagree with him, but I don't know if you can think he's wrong.

Doc Gooden was the best pitcher that I ever saw for one year.
   41. Dag Nabbit is part of the zombie horde Posted: July 13, 2006 at 01:24 AM (#2097214)
   42. OCF Posted: July 13, 2006 at 02:15 AM (#2097246)
Gibson's 1968 of 2.19 aDRA ... Basically it's R/9 converted to a 4.50 environment at the same Win Impact as in the actual run environment. It's adjusted for team defense (using the NRA/DERA adjustment from Baseball Prospectus) and after 1959 for things like bullpen support and performance out of the bullpen with inherited runners).

Sandy Koufax had a 2.06 aDRA in 1966. So Gibson's 1968 drops to 3rd on the adjusted scale,


When you say R/9, are you talking about actual runs? Or are you projecting runs from DIPS data? If you're using actual runs, then you're doing something very like what I've been doing, only you're making two more adjustments - for defense and for bullpen issues. So I'm a little curious about how you arrive at this.

I have Koufax '66 as RA+ 183 in 323 innings. Using a 1.66 exponent, that gives him a .731 winning percentage and a record of 26.2-9.7.

I have Gibson '68 as RA+ 227 in 304.7 innings. Using a 1.56 exponent, that gives him a .782 winning percentage and a record of 26.5-7.4.

In other words, I still have Gibson with the superior season. I can't imagine that inherited runner issues made that much difference to either. So the remaining possibilities for the difference between us are differing calculations of the same things, and the defensive adjustment - and I suspect it's the defensive adjustment. I've seen repeated reference to those Cardinals as an oustanding defensive team, which leaves me wondering just how many good players it takes to make an outstanding team.

Maxvill was outstanding, and that's the most important position. But he's still just one player.
Flood was outstanding. He was fighting arm trouble in '67-'68 and couldn't really throw, but that doesn't make a large difference.

Javier I assume (with no data) was above average.
Maris I assume (with no data) was above average, but he wasn't fast.

McCarver was probably good, but how much difference does a catcher make, anyway?
Cepeda? He was an athlete, or had been before his various injury problems, but was he a defender?

Shannon must have been below average.
Brock? How much can raw speed make up for bad instincts and bad hands?
   43. Paul Wendt Posted: July 13, 2006 at 03:19 AM (#2097267)
My father said Bob Gibson was the best pitcher he ever saw. I don't know if you can really think he's wrong. You can disagree with him, but I don't know if you can think he's wrong.

It helped he was 13 in 1968.


I anticipate one of those famest lines from the history of film, something like
"I am your father. --OCF"

--
It's amusing that '+', the universal mathematical symbol for addition, is taken in sabermetrics so that it would be misleading to use 'Pen+' for Pennants Added.

It's a pain that the English words 'adjusted' and 'added' are so similar,
not entirely a coincidence but that would be a tangent here.
   44. OCF Posted: July 13, 2006 at 03:25 AM (#2097272)
I was 15 in 1968, but close enough.

Suffice it to say that if my son were posting here I would recognize his writing style.
   45. Joey Numbaz (Scruff) Posted: July 13, 2006 at 12:14 PM (#2097394)
OCF - runs are that basis, not DiPS adjusted (I'm very lukewarm - at best - to DiPS theory myself - I subscribe to the Mike Emeigh School on it).

Take R/9, adjust for bullpen support, inhereted runners. Then take that final R/9, and adjust for park and league. Park includes and adjustment for not facing your own batters (I take the B-R park factors).

Then convert that to the number of R/9 that would have the same WPct in a 4.50 R/G environment, using PythaganPat Exponent.

Then adjust for team defense (again not DiPS, quality of defense). For me this means taking the number from the last paragraph, dividing by NRA, multiplying by DERA. I trust the Prospectus defense quality adjustments enough to use them.

That gets you what I call aDRA (adjusted defense-adjusted runs allowed). To this I add a league adjustment for things like expansion, weak leagues within a two league environment (AA vs. NL, FL vs. NL, etc.), and wars. I just refined this methodology last night, and I've got my adjustments for all years from 1871-1891 done. I'll explain that (and give the results) in another post. It's not a timeline adjustment - just a relative to the year league quality adjustment.

Then I translate starting IP to make all league seasons equal. Basically I take the average of the league leaders in IP and make that equivalent to 258.3 IP. I take the leaders for any year based on the number of teams in the league. For example, in a 16-team league, I average the pitchers ranked 5-12 in IP and set that equal 258.3 IP. In an 8 team league I use 3-6, etc.. The 258.3 is based on the AL/NL average from 1901-2005. The actual average is around 264 IIRC. Prospectus uses the top 5 in the league = 275 IP. The actual average for the top 5 in the league over that span is 281, so I used the same proportion, so my numbers would be directly comparable.

For relief innings, I multiply them by Leverage Index. For years prior to 1960, I use Pete Palmer's method of estimating LI, which is quite good. The only tweak is that I set the upper boundry at 3.0, not 2.0. From the PBP data post 1960, it's quite obvious that starters used in occasional relief appearance can have LI's between 2.0 and 3.0 in any given year (though it's not the norm).

I then add the translated starter IP to the leverage adjusted relief IP to get tIP (Translated IP).

Then I figure runs saved above that of a 5.48 RA pitcher. In a 4.50 R/G environment, this is equivalent to a .404 WPct, or the same as if you considered 6 pitching WS per 220 IP replacement level. WARP uses 6.00, which is way to high, IMO. This is PRAR (Pitching Runs Above Replacement).

Then I take the pitcher's RCAP (Runs Created Above Position) from the Sabermetric Encyclopedia, and adjust those based on the ratio of sIP to tsIP (if he's pitching more or less often due to his era, he's also hitting more or less often). For pitchers with no sIP, I adjust based on team games, using 155.9 as the normal season (the average of all seasons from 1901-2005). This is BRAR (Batting Runs Above Replacement).

From there, I take PRAR+BRAR (which is also RSAR (Runs Saved Above Replacement)), and convert to wins. In a 4.50 R/G environment, it takes 9.558 runs to flip one game in the W/L column, so that's what I use - RSAR/9.558. This is aWAR (adjusted Wins Above Replacement).

Finally, I convert aWAR to Pennants Added (PA). That calculation is too much for me to explain right now. I believe the formula is in Baseball Prospectus 2002. Pennants Added gives more weight to bigger seasons, based on the average of all pennant winners, all teams, and the standard deviations of all teams too. Basically the player is added to every team in history and he gets credit for the percentage of teams that he would push over the top. That's not quite exactly how it works, but it is essentially what happens.

In terms of what counts as a pennant, I use the NL from 1876-the year before the current election year; the AL from 1901-the year before the current electoin year. That's why retired players numbers change a little every year. I consider pennant winners from 1876-1968, division champs from 1969 on, and wild card winners to be 'pennant winners'. Basically any team that goes to the post-season, or would have gone had there been one.

As the leagues get more compacted and the pennant winners get worse overall, there is more value in big seasons. In years where you had to play .700 ball to win, and the spread between the teams was huge, there's only so much impact a big year can have on a pennant race. As things get tighter, big years can push you over the top. So as we move on, peak becomes a little more important.

Hopefully that explains it :-)
   46. Joey Numbaz (Scruff) Posted: July 13, 2006 at 01:03 PM (#2097420)
As for the league quality adjustments, this is what I'm doing:

I'm looking at the top 5 in IP for each league/season. I'm taking their adjusted for season NRA (normalized runs allowed) from Baseball Prospectus, and comparing it to the adjusted for all-time NRA.

I'm using NRA, because that number doesn't take into account Prospectus' era relative fielding value adjustments, which I'm not a fan of.

This is a change, I had been using DERA and just eyeballing it. Because of this I was severely underestimating the deflation factors for things like WWII, Federal League, etc.. The NRA adjustments are higher, and make more intuitive sense.

While I don't trust their all-time adjustments, I do think they adjust for sudden league changes rather well, that's why I'm separating this out. I'm also dead set against time-lining, and this doesn't do that, IMO.

So I have those numbers . . . the number after the year is the all-time adjustment to NRA. Basically add this to NRA to adjust for era. Lower number means better league.

1871 NA .78
1872 NA .73
1873 NA .61
1874 NA .63
1875 NA .87
1876 NL .78
1877 NL .58
1878 NL .59
1879 NL .59
1880 NL .38
1881 NL .31
1882 NL .44 AA 1.65
1883 NL .42 AA 1.27
1884 NL .64 AA 1.21 UA 1.98
1885 NL .56 AA .88
1886 NL .55 AA .74
1887 NL .36 AA .66
1888 NL .42 AA .58
1889 NL .28 AA .58
1890 NL .53 AA 1.20 PL .43
1891 NL .27 AA .67
1892 NL .31 


I don't just add those numbers. That would be timelining. I only want to account for weaker leagues within their own time. Huge distinction.

Note the symmetry between the 1881 NL and the 1892 NL. League quality the same - everything in between - that's due to expansion.

So I only adjust when a league takes a step back. By 1874, the league had progressed to .63 in quality. In 1875, the league started falling apart and so did the play. Not because overall baseball got worse, but because the league wasn't getting as many of the great players as they did before. So I'm adding .20 to the aDRA of NL pitchers in 1875 and .10 in 1876. I'm eyeballing, since these numbers aren't precise anyway (they are simply an average of the 5 pitchers with the most IP in each league), and erring to the side of caution.

By 1877 we're back to normal, and play continues steadily improving, by 1881 the league is at .31. No adjustments for any of those years.

Then in 1882 we have a gihugic expansion. The AA comes in and it's basically AAA quality. I don't add the full 1.65 to aDRA, only the 1.65-.35, the established level of play for 1880-81. But note, the NL suffered a little too, so their pitchers will get a .08 adjustment for 1882.

In 1883 the expansion effect is still there, though it's washed out some. The AA adjustment is is down to .90, the NL is .06.

1884 we expand again, but really the NL takes the hit. The AA actually improves some, their hit is down .85, the NL gets bumped to .28, the UA is awful, their pitchers get a full 1.60 added to their aDRA.

1885 the UA disappears and both leagues move forward, again the AA more than the NL. The AA adjustment drops to .50, the NL drops to .20.

In 1886 the overall quality gets even better, but the AA is where the gains are made, they drop to a .38 adjustment, the NL is again at .20.

By 1887 NL has finally gotten back to it's 1880-81 quality, no adjustment. Now the AA adjustment is simply based off of it's difference with the NL, which is .30.

In 1888, the overall level of play is pretty much the same, but the AA actually improved at the NL's expense. So the AA adjustment drops to .22 and the NL gets a .06 adjustment. This it he closest the leagues will ever get.

In 1889 the NL takes back the AA's 1888 gains, and the same adjustments at 1887 apply, AA .30, NL nothing.

In 1890 we expand again. The AA gets killed. Their adjustment goes all the way back .90, the league is basically 1883 quality that year. The NL also takes a hit and their pitchers get a .25 adjustment, while the PL is the best league, it's not the quality of the 1881 (or 1889) NL, their pitchers get a .15 penalty. Notice again how everyone has it easier in the expansion years.

Now we get to 1891, the last year of the AA. The NL is the best it's ever been, no adjustment. The AA is better than 1890, but still much worse than the NL, so there is a .40 adjustment.

By 1892 the NL is the only league again, there won't be any adjustment going forward until the 1900 de-spansion and 1901 expansion of the AL.

I haven't had time to do the data entry yet for that yet, hopefully I get to that this evening. It took me about 2-3 hours to get this much done.

There is no immediacy needed for me to update the table since Gibson is a slam dunk, and there are no borderline guys this time.

I'm going to wait before I present any more pitcher numbers for pitchers who played after 1900 until I can get this relative league strength table done through our current election year, then I'll go back and apply these adjustments to everyone.

I still haven't figured out how to make these adjustments applicable to the batting runs of the pitchers. It's pretty important for guys like Caruthers and Mullane, who had a ton of hitting value and compiled it in the AA.

Here are how the 19th Century pitchers I've done look overall now, remember that batting runs are included but unadjusted for league quality:

Pitcher     PA  aDRA   tIP       top3
Clarkson  .774  3.73  2806.3 12.2
,9.3,8.7
Keefe     .654  4.03  3016.7  7.4
,6.0,5.9
Caruthers .598  4.09  1655.7  9.2
,8.3,7.1
Galvin    .590  4.18  3409.7  7.8
,6.9,6.0
Spalding
.577  3.76  1817.3  9.4,7.6,7.3
Welch     .546  4.19  2739.3  6.5
,6.4,5.6
McCormick .533  3.92  2338.7  7.7
,5.9,5.0
Mullane   .434  4.65  2760.7  6.7
,6.4,4.3
Bond      .400  4.12  1868.0  6.4
,6.2,6.1
*Spalding's offensive numbers are somewhat estimated and could significantly change when I redo them. 


I still need to do Cummings, Radbourn, Mathews, Foutz. Anyone else from the pre-1892 era that's worth looking at (remembering that it takes sometime to do :-) ?

Here's a table with the final adjustments from the text above (add the number to the pitcher's aDRA).

1875 NL .20
1876 NL .10
1882 NL .08
1882 AA 1.30
1883 NL .06
1883 AA .90
1884 NL .28
1884 AA .85
1884 UA 1.60
1885 NL .20
1885 AA .50
1886 NL .20
1886 AA .38
1887 AA .30
1888 NL .06
1888 AA .22
1889 AA .30
1890 NL .25
1890 AA .90
1890 PL .15
1891 AA .40
   47. Joey Numbaz (Scruff) Posted: July 13, 2006 at 01:08 PM (#2097425)
"Anyone else from the pre-1892 era that's worth looking at (remembering that it takes sometime to do :-)"


Answering my own question, Silver King and Jim Whitney will also be included.
   48. sunnyday2 Posted: July 13, 2006 at 01:59 PM (#2097467)
Note that the NL of Tommy Bond is almost exactly equal to the NL of Clarkson-Radbourne-Keefe-Welch. Obviously you've got peak versus career issues for these particular comps. But for peak voters, just note that Bond did not pitch in an obviously inferior environment. The NL--not to even mention the AA--was not consistently better in the period that Joe studied.

Secondly I would just add that Joe's adjusted IP reduce Bond's innings by about 50 percent. I mean, he pitched more innings in 50-60 games than Joe credits him for an adjusted 162 game season. I understand that there has to be some way to compare pitchers across eras, but I just can't quite get my head around pretending that a guy didn't pitch 500 innings when he did. Win Shares can go the other way when we assume that a guy who pitched every day could possibly throw 162 games, he couldn't. But this goes waaaaay the other way.
   49. Joey Numbaz (Scruff) Posted: July 14, 2006 at 05:35 AM (#2098565)
Marc, I see your point on Bond's IP, but the idea is that if everyone is throwing 500 innings, it's no big deal to throw 500 innings.

Also, fielding was much more responsible for run prevention in that era than it is in later later years (one of the reasons he could throw 500 innings, that and it was underhand), but I'm not docking Bond for that. I think it evens out.
   50. Joey Numbaz (Scruff) Posted: July 14, 2006 at 05:38 AM (#2098567)
Bond really only had 2 years where he was at the top of the league in IP, 1877 and 1878. He had a nice 5 year run from 1875-1879, but that's about it. IMO you've got ot be Spalding or Koufax to be a candidate with just 5 productive seasons. Bond wasn't anything special with the stick either, even for a pitcher. A little above average for a pitcher in his time, but that's it.
   51. Joey Numbaz (Scruff) Posted: July 14, 2006 at 05:45 AM (#2098572)
I meant a little above average as a hitter for a pitcher. He was more than a little above average as a pitcher, obviously.

Also he benefitted from great defenses behind, his career adjustment of 0.35 (which gets added to his aDRA) is one of the largest I've found. Spalding's is 0.57 (which should be obvious, he probably had more defensive support than any pitcher in history), Caruthers was at 0.37, Nichols 0.30, Brown 0.27, Mullane 0.29.

I know it seems biased towards the early eras, but that's basically because the spread between the good and bad defenses was bigger back then, and the smart front offices got the good pitchers and fielders (and hitters).

Dutch Leonard was -0.21 on the other end. His defenses killed him.
   52. Joey Numbaz (Scruff) Posted: July 14, 2006 at 06:19 AM (#2098587)
Also, it's a little misleading to say the NL of Bond's time was equal to the NL of Clarkson's.

A more accurate statement would be that at it's best, the NL of Bond's time was equal to the worst NL of Clarkson's time. But the late 80s NL was much better than the weakened NA/NL of 1875-76.

Not that it matters to me. I don't timeline. But I do think it's important to be fair when discussing the leagues in those terms.
   53. sunnyday2 Posted: July 14, 2006 at 12:44 PM (#2098657)
Joe, this is probably a small point, and I don't disagree at a micro level when you say Bond's NL at its best was comparable to Clarkson's (et al) at its worst. That is roughly correct. But I would have to believe there's a pretty significant margin of error in this sort of calculation. Not that it isn't valid, I think it is, and not that it isn't really necessary, it's that too. But I would say that within the margin of error, my statement was also correct--Bond's NL was roughly equivalent to Clarkson's (on average, throughout career). My point was to challenge the more macro belief that the NL progressed more or less evenly until 1901, which just isn't accurate at all.
   54. rawagman Posted: July 14, 2006 at 03:46 PM (#2098802)
I would have recommended looking at Larry Corcoran - 5 dominating years. But that's a bit short.
   55. Dizzypaco Posted: July 14, 2006 at 04:00 PM (#2098812)
Pardon my ignorance on this, since I'm not an expert on nineteenth century baseball, but isn't there a lot more than just a quality of league adjustment when envaluating pitchers from the early days of baseball? Wasn't it the case that hitters could call for a low or high pitch? That pitchers weren't allowed to throw overhand? If so, when were the rule changes?

And if so, doesn't this have huge implications on how to evaluate a pitcher?
   56. DavidFoss Posted: July 14, 2006 at 04:31 PM (#2098838)
Wasn't it the case that hitters could call for a low or high pitch? That pitchers weren't allowed to throw overhand? If so, when were the rule changes?

And if so, doesn't this have huge implications on how to evaluate a pitcher?


Yup. In the early 1870s, what you say is true. The batter could pick low (knees-to-waist) or high (waist-to-shoulders). The pitching restrictions were gradually removed and by 1884 or 1885, they were gone and pitchers pretty much dominated the late 1880s and early 1890s which prompted the pitching distanced being moved back in 1893. There were changes in balls per BB and strikes per K during this time as well. Someone like Paul Wendt would be able to give you the play-by-play as to when the rules were changed.

What you say about pitching in the 1870s is true. How much of pitching was really fielding? But on the other hand, with all the restrictions how easy is it for one pitcher to be better than another (especially consistently from year-to-year)? If a pitcher managed to work through the restrictions and still dominate, then that has value. But we only inducted Spalding from this era (and Monte Ward, too, though he shifted positions and had a nice career at SS too).

The 1880s was different. Pitchers ruled the league by then. We were correct to induct several of those guys.
   57. sunnyday2 Posted: July 14, 2006 at 05:03 PM (#2098863)
Thru 1893 I schedule adjust WS for pitchers to 154 X 2/3, on the assumption that nobody could throw 154 games nor even 77, and then I give half of the value to the fielders.

If fielding was so important--and of course it was--then I give you Ed Williamson!
   58. KJOK Posted: July 14, 2006 at 09:23 PM (#2099242)
"Anyone else from the pre-1892 era that's worth looking at (remembering that it takes sometime to do :-)"

Charlie Buffington
Wil White
Guy Hecker
Adonis Terry

??
   59. Joey Numbaz (Scruff) Posted: July 16, 2006 at 10:34 PM (#2101330)
Thanks KJOK.

Question for how to use the results . . . I've got a few options once I get into the AL/NL era.

Do I take the weaker league pitchers in any given season and add the difference between the leagues to their aDRA? For example, in 1916 the NL is the weaker league, .26, vs. the AL at .12. Using this method, I'd add .14 the aDRA of all NL pitchers.

Or do I take an average of the two leagues, and add the difference between the averages to the pitches. In the 1916 scenario, the average is .19, so I'd add .07 to the NL pitchers and subtract .07 from the AL pitchers.

I'm leaning towards the later.

I think it will work better in expansion years. Think of the 'average' of the two leagues as the 'baseline'.

Now take an expansion year, like 1901. The NL from 1897-99 has factors of .25, .23, .21.

So if I get conservative and use .25 as the baseline then all leagues will be subtracted from that, until the expansion washes out, meaning the overall level of the leagues gets back to .25. This happens from 1907-1910, although the leagues regress from 1911-15, before finally jumping the .25 hurdle for good in 1916 (save the war year of 1918, which saw MLB fall back to .29).

I'm really just thinking out loud here. But if anyone has any ideas, I'm all ears . . .

Here's the data as far as I've completed (through 1936)

1871 NA  .78
1872 NA  .73
1873 NA  .61
1874 NA  .63
1875 NA  .87
1876 NL  .78
1877 NL  .58
1878 NL  .59
1879 NL  .59
1880 NL  .38
1881 NL  .31
1882 NL  .44 AA 1.65
1883 NL  .42 AA 1.27
1884 NL  .64 AA 1.21 UA 1.98
1885 NL  .56 AA  .88
1886 NL  .55 AA  .74
1887 NL  .36 AA  .66
1888 NL  .42 AA  .58
1889 NL  .28 AA  .58
1890 NL  .53 AA 1.20 PL  .43
1891 NL  .27 AA  .67
1892 NL  .31 
1893 NL  .29
1894 NL  .28
1895 NL  .26
1896 NL  .31
1897 NL  .25
1898 NL  .23
1899 NL  .21
1900 NL 
-.02
1901 NL  .19 AL  .46
1902 NL  .41 AL  .17
1903 NL  .41 AL  .18
1904 NL  .38 AL  .21
1905 NL  .35 AL  .21
1906 NL  .31 AL  .35
1907 NL  .30 AL  .23
1908 NL  .22 AL  .28
1909 NL  .31 AL  .21
1910 NL  .29 AL  .22
1911 NL  .32 AL  .35
1912 NL  .26 AL  .34
1913 NL  .21 AL  .28
1914 NL  .19 AL  .27 FL .67
1915 NL  .21 AL  .24 FL .70
1916 NL  .26 AL  .12
1917 NL  .29 AL  .19
1918 NL  .28 AL  .30
1919 NL  .32 AL  .15
1920 NL  .29 AL  .19
1921 NL  .34 AL  .25
1922 NL  .40 AL  .18
1923 NL  .35 AL  .16
1924 NL  .31 AL  .13
1925 NL  .27 AL  .22
1926 NL  .22 AL  .11
1927 NL  .22 AL  .15
1928 NL  .17 AL  .19
1929 NL  .18 AL  .23
1930 NL  .15 AL  .28
1931 NL  .14 AL  .28
1932 NL  .03 AL  .19
1933 NL  .02 AL  .24
1934 NL 
-.01 AL  .17
1935 NL  .02 AL  .14
1936 NL  .02 AL  .19
1937 NL  .05 AL 
-.07 
   60. Joey Numbaz (Scruff) Posted: July 16, 2006 at 10:38 PM (#2101342)
I really wanted to get to the war years, but haven't been able to quite get them done yet . . . maybe tonight.
   61. Brent Posted: July 17, 2006 at 03:38 AM (#2101903)
1917 NL .29 AL .19
1918 NL .28 AL .30
1919 NL .32 AL .15


Did the AL lose more players to the war?

1924 NL .31 AL .13
...
1933 NL .02 AL .24


Why should I believe this?
   62. Sean Gilman Posted: July 17, 2006 at 05:49 AM (#2102017)
1881 NL .31
1882 NL .44 AA 1.65
1883 NL .42 AA 1.27
1884 NL .64 AA 1.21 UA 1.98


1913 NL .21 AL .28
1914 NL .19 AL .27 FL .67
1915 NL .21 AL .24 FL .70


Any theory on why these expansions would have such different effects? The AA in 1882 is terrible, yet has a bigger effect on the NL than the FL (a significantly better league, relatively, as I understand this metric at least) had on the AL or NL. And then the UA, the worst league of all, has the biggest impact of any of the lesser leagues. Maybe I'm misreading something. . . .

Seems to me the extreme volatility of these numbers is not a reason to trust them. We can't possibly measure that accurately the differences between leagues. Do I follow correctly, and this is just another derivation of the opaque competition adjustments Clay Davenport came up with and has never bothered to explain in any kind of reasonable detail?
   63. Paul Wendt Posted: July 17, 2006 at 06:37 AM (#2102037)
I agree that caution is appropriate.
The AL improved as much, 1933-1934, as the NL 1899-1900!?

If I understand correctly, however, the adjustment is implemented automatically, the hard work is elsewhere. Recall JoeD's parenthetical remarks "Anyone else from the pre-1892 era that's worth looking at (remembering that it takes sometime to do :-) ?" There must be a spreadsheet, time-consuming to set up for each pitcher, but it will be recalculate with revised league quality adjustments. Right?

--
So as we move on, peak becomes a little more important.

For everyone, whether he played in the 1890s, 1930s, or 1970s. So Van Haltren and Beckley fade from the picture, slightly, in favor of Childs and others with better peak or prime.

--
As for the league quality adjustments, this is what I'm doing:

I'm looking at the top 5 in IP for each league/season. I'm taking their adjusted for season NRA (normalized runs allowed) from Baseball Prospectus, and comparing it to the adjusted for all-time NRA.


There must be a better way, although I guess critics will focus more attention on weak points other than the Top5-IP dependency.

But nothing has urgency, right? It is all relatively easy to revise.

(It's late here, but I don't know the BP sabermetrics to be a good critic if it were early.)
   64. Mark Shirk (jsch) Posted: July 17, 2006 at 08:06 AM (#2102063)
Interesting how Joe's findings contradict my thought experiment in the ballot discussion thread about the value of a peak at different times in baseball history. Funny that Joe comes to his conclusions as somewhat of a career voter and friend to the likes of Jack Quinn and Jake Beckley while I came to mine as a peak voter and one of the best friends of Charlie Keller and Hugh Duffy.
   65. Joey Numbaz (Scruff) Posted: July 17, 2006 at 07:15 PM (#2102377)
Sean - on the AA vs. FL one thing is that the NL was an 8 team league, and the AA essentially doubled the size of the ML at the time. In 1914, MLB had 16 teams, and the FL only added 8.

Also, there really weren't that many great MLB players who went to the FL. A few, but you had guys just starting out like Quinn and Roush, and older guys like Brown and Plank. The leagues had just taken a step back from 1908-12, finally started getting better in 1913. The leagues would have been even better in 1914-15 (as evidenced by 1916) but the FL just caused them to stagnate.

My best guess as to the shift from AL to NL dominance from the mid-20s to the early 30s is that the AL stars were getting old and fading away, while the NL was adding plenty of new stars. Especially in the early 30s as Ruth starts to fade.

*********

Just to clarify what Paul said - when you look at Pennants Added through the lense of all of history to X point - and then move to X+1, the effect will likely be that the peak guys will nudge forward at X+1, because as the league compact, the importance of peak seasons goes up. If you are trying to place a player in a historically neutral context, to compare him with other players historically, I think this makes sense.

However, if all you care about is the value the player had at the time, you should probably be weighing career more heavily for the early players, and peak more heavily for the recent players. By 'more heavily', I mean more heavily than you otherwise would, not that you should switch from being a career guy to a peak guy.
   66. Joey Numbaz (Scruff) Posted: July 17, 2006 at 07:19 PM (#2102384)
Brent as to the war, I haven't checked. But the two big guys that I remember missing war seasons are Rixey and Alexander, both NL guys. I haven't looked at it comprehensively though.
   67. Joey Numbaz (Scruff) Posted: July 17, 2006 at 07:23 PM (#2102393)
One other thing - I don't really have any doubt that baseball of the 30s was better than that of the 10s and 20s, and probably significantly better . . . you've got a sport that has had the same number of teams for more than 30 years at this point. Population, nutrition, coaching, etc. have all driven the game towards being better, and there hasn't been any addition of teams to offset this. So the numbers definitely pass the smell test for me in that respect.
   68. Joey Numbaz (Scruff) Posted: July 17, 2006 at 07:24 PM (#2102394)
Not to mention the Depression - if athletes tend to come from worse economic situations, well just about everyone in the country was in a worse economic situation in the 30s.
   69. karlmagnus Posted: July 17, 2006 at 07:42 PM (#2102412)
Joe, that wouldn't do it, because of the lags involved, at least not till 1938-40 -- this is the same discussion I had re: Leever. The decision to become a pro ballplayer is taken at the time of drafting, which is normally at least 2-3 years before entering the ML, and maybe 5-7 years before getting close to a peak. Once you're say a AA ballplayer, there's no decision to be made about being called up; if you're asked to do it, you do it. Moreover, for any player of remotely HOM quality, you will stick at it until the majors open up to you, because that's where the money is -- the guys who get discouraged at 24 are the marginal talents.

Baseball had very lean years relative to the rest of the economy in 1891-93, because of the salary cap and contraction, so you'd expect a talent dropoff in 1896-1900 (thus our dip in HOMers in those years.) It had very fat years in 1929-31, particularly 1930, because baseball salaries were slow to react to the Depression (I think it was 1930 or 1931 when Babe, on being asked why he was paid more than President Hoover, responded "I had a better year.") Thus you'd expect a talent surge after 1934 or so. By 1933-34 times were tough everywhere including in baseball, the relative position had probably righted itself, so the talent surge probably washes out by 1941-42. Our 1930s glut is thus I think economically rational.

You'd also expect a low HOMer period following say 1964-68, a period in which baseball salaries hadn't increased but living standards had. Say 1972-77 for the dip. That would reverse after about 1980, as free agency and arbitration changed the economics forever.
   70. jimd Posted: July 18, 2006 at 12:27 AM (#2102678)
A few points:

Joe's table above represents adjustments to pitching.
Adjustments to batting (and fielding) are independent (though somewhat related).

The AA in 1882 is terrible, yet has a bigger effect on the NL than the FL (a significantly better league, relatively, as I understand this metric at least) had on the AL or NL.

Actually, the AA in 1882 had no effect on NL pitching (other than scooping up possible replacement parts).
The NL is in the midst of a transition in starting pitching models, from a "ace and backup" to a two-man rotation.
(Chicago has won the last two years with Corcoran and Goldsmith splitting the duties fairly evenly,
and other teams are now imitating, even if their number two was not particularly good.)

The signficant pitchers who played in the AA of 1882 and the NL of 1881 are:

Will White: Ace of AA Cincinnati, failed a tryout with Detroit in 1881 (0-2), former NL ace 1879-81.
Tony Mullane: Ace of AA Louisville, failed a tryout with Detroit in 1881 (1-4).
Harry McCormick (not to be confused with Cleveland ace Jim McCormick):
Backup for AA Cincinnati, failed a tryout with Worcester in 1881 (1-8), former NL ace 1879.

You can also look up Frank Mountain, Jack Leary, and Billy Taylor if you want the rest of the 1881NL-1882AA pitchers.

And then the UA, the worst league of all, has the biggest impact of any of the lesser leagues. Maybe I'm misreading something. . . .

It may have. There were 16 starting jobs in the 1884 NL, compared to 32-40 in the 1914 NL.
The Unions got both of Cleveland's starters,
Hugh Daily to Chicago, followed by Jim McCormick to Cincinnati mid-season; Dupee Shaw of Detroit to Boston;
Charlie Sweeney jumped Providence mid-season to St.Louis, leaving Radbourn to be
the "Old Hoss" all by himself down the stretch.
But that's only 3 of 16 starting slots in the UA.

The Federals were about as successful at raiding MLB.
10 pitchers qualified for ERA titles in both 1913 MLB and 1914 FL.
Mordecai Brown, Howie Camnitz, Claude Hendryx, Gene Packard, Tom Seaton, George Suggs from the NL.
Cy Falkenberg, Russ Ford, Bobby Groom, Al Schulz from the AL.

The impact is spread across both leagues though, plus the minors were more organized than in 1884, so replacements may have been easier to find. Meanwhile back in 1884, the schedule expanded from 98 to 112 games, so the existing NL and AA teams were now looking for third starters as a supplement to two aces. Pitching was being stretched thin as the majors went from requiring 8 starters in 1881 to 48 in 1886 (8 teams to 16; 84 game schedule to 140 games). OTOH, the effectiveness of the pitching also increased with overhand being legalized in 1885 (IIRC). Lots going on in the 1880's besides the rebel leagues.
   71. jimd Posted: July 18, 2006 at 12:30 AM (#2102685)
Oops. Should read:

Will White: Ace of AA Cincinnati, failed a tryout with Detroit in 1881 (0-2), former NL ace 1878-80.
   72. Sean Gilman Posted: July 18, 2006 at 07:41 PM (#2103661)
Will White: Ace of AA Cincinnati, failed a tryout with Detroit in 1881 (0-2), former NL ace 1879-81.
Tony Mullane: Ace of AA Louisville, failed a tryout with Detroit in 1881 (1-4).
Harry McCormick (not to be confused with Cleveland ace Jim McCormick):
Backup for AA Cincinnati, failed a tryout with Worcester in 1881 (1-8), former NL ace 1879.

You can also look up Frank Mountain, Jack Leary, and Billy Taylor if you want the rest of the 1881NL-1882AA pitchers.


Is this the entirety of the sample size for the translation?
   73. jimd Posted: July 18, 2006 at 09:08 PM (#2103774)
Is this the entirety of the sample size for the translation?

It's all of the first-order evidence that is available when comparing 1881NL pitching to 1882AA pitching. Of course less direct evidence is also available by comparing how 1881NL compares to 1882NL and 1882NL to 1883NL, etc., and 1882AA to 1883AA and 1883AA to 1884AA, etc., and then 1882AA to 1883NL and 1882NL to 1883AA, etc. so that one builds up all the within-league comparisons and the cross-league comparisons for the 10 year history of the AA. IIRC, that's how Davenport builds his league ratings, a modified version of the methodology of the original Cramer (sp?) study.
   74. Joey Numbaz (Scruff) Posted: July 19, 2006 at 08:27 AM (#2104415)
Hey jim - please clarify this for me.

AL is clearly stronger than the NL right now, it's very obvious, we all agree.

The pitcher differences, as I calculate from the Prospectus numbers, just checking a few guys (Clemens/Buehrle, ARod/Berkman, first guys I thought of) show that NL 2005 pitching at -.07, AL at -.04, meaning the NL was stronger - which doesn't make sense.

However, when I go to the hitters, I get what I expect, the NL EQA difference is +.001, AL is +.011 (for hitting the higher number indicates the better league).

So am I interpreting this correctly if I say that the AL is much better because the pitching is better? If the AL hitters get adjusted higher, that means they are facing tougher pitchers historically than is the norm, especially compared to the NL.

Likewise, since the NL pitchers get a slight bump over the AL pitchers, Prospectus is saying that the AL hitters aren't quite as good as the NL hitters? Obviously an 11 point EQA difference is much bigger than a .03 NRA difference (and it's just a couple of players sampled), so Prospectus is on track in saying the AL is better. But I just want to make sure I'm interpreting things correctly.
   75. Joey Numbaz (Scruff) Posted: July 19, 2006 at 08:34 AM (#2104419)
I guess the alternative would be to say that ARod's .338 adjusted for season EQA is in comparison to better hitters, so that's why it gets adjusted all the way to .349 for all-time, while Berkman's .317 is in comparison to worse hitters, so it only gets adjusted to .318.

And on the pitcher side it's saying that the NL pitchers were slightly better than the AL pitchers, so their pitchers get adjusted by -.07 to their NRA, while the AL pitchers only get adjusted to -.04.

Just from my own observation, I'd say this second interpretation make much more intuitive sense. I don't see much difference in the pitching between the leagues at all, especially since Pedro, Clemens and Pettitte were all in the NL in 2005. The AL's advantage over the NL is clearly on the offensive side, as a quick look at the All-Star Game lineups would show.

But it is important to note (as you did earlier) that the adjustments I'm giving here apply only to pitchers.
   76. jimd Posted: July 20, 2006 at 10:53 PM (#2105432)
Yes. The second post is the way that I would look at it.
   77. Joey Numbaz (Scruff) Posted: July 21, 2006 at 03:51 AM (#2106175)
Thanks jim!

I'm into the early 70s. Summary, no data now, not enough time . . . this 'theory' makes a lot of sense to me. As I've said earlier, this methodology used by Davenport nails all of the 'big ones' (AA, UA, PL, FL, WWII), so to me the burden of proof is on the doubters for the other years when they disagree.

Remember this applies only to pitchers. The background stuff of 1871-1925 or so is important, but the really interesting stuff comes later.

1871 NA - plays gets better until 1874.

1875/76 NA/NL - NA falls apart, NL gets started, competition suffers a bit.

By 1877 this has washed out, play gets progressively better, with a big leap forward in 1880.

1882 AA comes along. NL takes a minor hit, AA is very bad early, much worse relative to the NL than the WWII quality of MLB or the Federal League, for example.

1884 UA hurts NL AA stagnates.

1885-88 - AA makes gains, overall MLB quality improves. By 1889 both leagues combined are at about 1880 NL quality, but NL is still the better league. AA never reaches equality with the NL, gets closest in 1888.

1890 PL is the best league, AA takes the very serious hit. NL is worse than PL, but not by a lot, all three leagues worse than either league at any point after 1886. Seriously watered down baseball everywhere.

1891 AA is pretty bad - worse relative to NL than at any point since 1884.

Note the expansion to double the major league teams never really 'washed out' until the AA folded.

1892-1899 - 1892 NL is essentially the same quality as 1881 NL. NL makes steady progress, especially from 1897-99.

1900 NL - by far the toughest league ever to this point. NL of 1900 as strong relative to 1899 as 1899 was relative to 1889, with 33% more teams.

1901, NL far better than AL. But in 1902 the tables are turned and AL is the better league through 1905. Expansion doesn't wash out until 1907.

Overall quality of MLB from 1902-06 is equivalent to MLB of 1892-96. It doesn't get back to the level of 1897-99 until 1907.

1911-12 - MLB pitching takes a dip - perhaps the offensive explosion was simply caused by the fact the overall quality of pitching took a little dip?

1914 - Federal League doesn't have much impact on the two existing leagues in terms of overall quality - but is signficantly worse than ML overall. This tells me there were many good minor league players out there for the taking, since the FL wasn't God-Awful and MLB quality barely felt the impact (the few quality MLB players that were taken were replaced by minor leaguers of a similar quality). The FL quality is equal to that of the 1886-87 AA a little worse than the 1873-74 NA in absolute terms. Relative to MLB of 1911-13 it's about as bad as the 1891 AA (slightly worse, technically).

1916 - overall MLB as strong as ever (saving 1900 NL). AL still the better league.

1918 - slight dip in overall quality due to players leaving for WWI. This mostly impacts the AL for whatever reason (meaning I don't know).

1919-20 - play returns to quality of 1917, not quite 1916 level.

1921-23 - pitching a little worse - I think part of this might have to do with WWI - kids that would have been 18-23 in 1917 are now 22-29 . . . it's probably because a few pretty good pitchers that we never found were killed in WWI.

1924-31 - play steadily improves as population and scouting improve, and no teams are added. AL far ahead of NL pitching wise from 1919-27. By 1928 parity is reached, then from 1929-37 NL has the better pitching.

1932 overall quality of play takes another leap forward.

1937 another big leap.

1942 - AL is much more hurt by WWII than NL - makes sense, AL loses Feller and Greenberg for one. 1942 NL is the strongest league ever to this point.

Note huge gains have been made over the last 25 years. There has been no expansion of teams during this time.

1943-45 - overall quality of play gets much worse - 1943 the quality of play has regressed back to that of 1934. 1943-44 is knocked all the way back to that of the early 20s. Again - AL much worse, the 1945 AL had essentially the same pitching quality of the 1901 AL.

I know some might see this not being strong enough - how could the quality of pitching of the war years overall only be as bad as the early 20s? Shouldn't it be worse?

Well for one, you've still got guys like Bucky Walters, Hal Newhouser, Mort Cooper, Dizzy Trout - it's not like everyone left. And you've had 25 years of improvement, that's a pretty serious setback.

Now here's where it gets interesting, if you are still reading . . . 1946-51, we're basically back to 1937-41 levels, slightly worse but close enough.

But there's a huge drop from 1952-55. Why? Korea for one, but I think also as important an issue is that an 18 year old in 1942-44 is 25-33 during these years. Several of the guys that would have been great early 50s pitchers never got the chance because of WWII.

From 1956-60 pitching is improved but not quite to the 1946-51 level. I don't think integration was enough to offset the WWII/Korea impact. The NL is far better than the AL during this time - which makes sense since they integrated faster.

1961 AL expands and quality dips, but not as bad as you'd think. But when the NL expands in 1962 things are much worse. This expansion doesn't wash out until 1966-67.

1969 Expansion is even worse than 1961-62. I'm not done yet - but based off the few guys I've checked into the mid-70s, this expansion probably isn't washing out before the 1977 expansion.

OVerall - WWII really did a number on MLB quality in the "Golden 50s". By the time this washed out, MLB expanded by 50% over a 9-year period - while baseball was fully integrated by this time, it is also competing with the NFL and NBA for athletes and the plusses don't offset the minuses enough to allow a 50% expansion.

I think this lays out a pretty strong argument that is extremely logical for why the baseball of the late 30s is probably the strongest we've seen at least by the time of the early 70s. The AL is going to expand again in 1977 - I'm guessing baseball doesn't get back to the late 30s quality until the mid-1980s.

Again, when I say 'quality' I'm talking about the overall quality of play - obviously there are more good players in 1969 than 1939 - but not 50% more, and there are 50% more teams.

Integration was just one thing pulling quality higher - but there were three very strong forces (wars, expansion, competition from other sports) hurting the average quality of play. Those two forces appear to me to be stronger than integration, based on Davenport's numbers.

Anyway, I'm not penalizing guys based on the overall quality of play (except for cases of expansion and wars), but I thought it was interesting to document WARPs interpretation of it. I am going to adjust for the relative quality of league compared to other leagues those years.

I'm adjusting for de- or ex-pansion/war for 1875-76, 1882-91, 1900-06, 1918, 1942-45, 1952-55, 1961-67, 1969-?

I'm also adjusting for AL/NL league quality relative to each other from 1901-forward.

Out of town this weekend hope to have a new PA chart with these adjustments up sometime next week.

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
danielj
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Syndicate

Page rendered in 0.6477 seconds
49 querie(s) executed