ESPN Top 100

[Peter_Spaeth]

Originally Posted by Shoeless Moe

Yah I think you need to go 10 deep for Pitchers.

So my revised list would be:

1-Randy Johnson
2-Pedro Martinez
3-Roger Clemens
4-Bob Feller
5-Walter Johnson
6-Satchel Paige
7-Nolan Ryan
8-Sandy Koufax
9-Bob Gibson
10-Lefty Grove

So who just misses the cut, for me: Kershaw, Maddux, Seaver, Alexander, Spahn, Young, Mathewson & Steve Carlton

IMO you have Pedro too high, Feller too high, Ryan should not be anywhere near the list, Grove too low. My list would include Seaver Young Alexander and Mathewson and probably (not going to rekindle this) Spahn. So I'll go 15 deep.

W Johnson
Young
Clemens
Grove
Seaver
Randy
Paige
Alexander
Mathewson
Spahn
Pedro
Feller
Gibson
Maddux
Carlton

[molenick]

I get Bob's point, but just to be clear, I was trying to compare pitchers within the context of the Negro Leagues with the data available. I have no idea how I would incorporate those players (including Paige) into an all-time list with non-Negro Leaguers aside from a gut feeling. Other than that, I think Peter's list is reasonable. But I will say that the traditional stats for the players I mentioned look pretty good as well.

[Touch'EmAll]

I would go with these all time top 5 pitchers (career value, not peak), not necessarily in order best to worst:

Walter Johnson
Lefty Grove
Bob Feller
Seaver
Maddux
Pedro

Next Tier:
Mathewson
Koufax
Nolan Ryan - most sheer dominant, most fun to watch
Paige - really don't know, just gut feeling
Randy Johnson

[Shoeless Moe]

Originally Posted by Peter_Spaeth

IMO you have Pedro too high, Feller too high, Ryan should not be anywhere near the list, Grove too low. My list would include Seaver Young Alexander and Mathewson and probably (not going to rekindle this) Spahn. So I'll go 15 deep.

W Johnson
Young
Clemens
Grove
Seaver
Randy
Paige
Alexander
Mathewson
Spahn
Pedro
Feller
Gibson
Maddux
Carlton

It's very hard to exclude a guy with 7 No Hitters and 12 1-Hitters. When he was on, he was on. No one on the list comes close to that.

[darwinbulldog]

Originally Posted by Shoeless Moe

It's very hard to exclude a guy with 7 No Hitters and 12 1-Hitters. When he was on, he was on. No one on the list comes close to that.

Less so if that pitcher walked 2,795 batters. He did finish 2nd in the Cy Young voting that one time though.

[Peter_Spaeth]

Originally Posted by Shoeless Moe

It's very hard to exclude a guy with 7 No Hitters and 12 1-Hitters. When he was on, he was on. No one on the list comes close to that.

Yes, but he was often off. His control was not that good. His bb/9 was awful, almost 5. Read Bill James' discussion of him, it's IMO pretty enlightening. 29th ranked pitcher by JAWS. I think Bill James in 2003 ranked him around 20th or maybe even lower.

[Touch'EmAll]

I googled "Bill James top 100 All Time", up came the list and the highest ranked player that is a pitcher is Walter Johnson at #8. The second highest ranked pitcher is Satchell Paige at #17.

I honestly have never seen any meaningful accurate season by season/career stats for Paige. I figured they simply don't exist. Maybe I haven't scoured the internet deep enough to see what I am looking for.

So how does Bill James arrive at #17 for Paige (2nd best pitcher of all time)? From where is he getting his stats?

[Peter_Spaeth]

Originally Posted by Touch'EmAll

I googled "Bill James top 100 All Time", up came the list and the highest ranked player that is a pitcher is Walter Johnson at #8. The second highest ranked pitcher is Satchell Paige at #17.

I honestly have never seen any meaningful accurate season by season/career stats for Paige. I figured they simply don't exist. Maybe I haven't scoured the internet deep enough to see what I am looking for.

So how does Bill James arrive at #17 for Paige (2nd best pitcher of all time)? From where is he getting his stats?

Not saying he's wrong, but it's interesting that a statistician like James put so many Negro League players in his top 100 and relied mostly on anecdotal evidence.

[nat]

Bill never really was a statistician. (As he'd be the first to tell you.) His real strength as a baseball writer was his willingness to ask questions, and to be open to unexpected answers. He used numbers to answer questions when he could, but it would be a mistake to include him in the list of genuine statisticians (Tom Tango, Michael Litchman, and so on) who study baseball.

Somebody said: "it is the Negro Leagues that were probably filled with a significantly higher percentage of non-MLB talent than the white, segregated major leagues ever were."

One thing to keep in mind is that Negro League teams often had smaller rosters than AL/NL teams. Looking at the 1943 Kansas City Monarchs (grabbed a team from the middle of Paige's career): they only had eight players who got >100 at bats. Only four other position players managed to get even 10 at bats. They only had seven pitchers who appeared in more than one game. It looks like the entire team was the starting nine, a bench bat or two, and a couple spare pitchers.

Comparing them to the 43 Yankees. The Yanks had, by my count, 13 position players who appeared in a substantial number of games, to go along with 10 pitchers who made more than a cameo appearance or two. Almost twice as many players on the roster.

Negro League competition wasn't all that diluted, compared to AL/NL competition, because there weren't as many guys on each roster.

[Peter_Spaeth]

I thought Bill James was heavily involved in developing sabermetrics.

[BobC]

Originally Posted by nat

Bill never really was a statistician. (As he'd be the first to tell you.) His real strength as a baseball writer was his willingness to ask questions, and to be open to unexpected answers. He used numbers to answer questions when he could, but it would be a mistake to include him in the list of genuine statisticians (Tom Tango, Michael Litchman, and so on) who study baseball.

Somebody said: "it is the Negro Leagues that were probably filled with a significantly higher percentage of non-MLB talent than the white, segregated major leagues ever were."

One thing to keep in mind is that Negro League teams often had smaller rosters than AL/NL teams. Looking at the 1943 Kansas City Monarchs (grabbed a team from the middle of Paige's career): they only had eight players who got >100 at bats. Only four other position players managed to get even 10 at bats. They only had seven pitchers who appeared in more than one game. It looks like the entire team was the starting nine, a bench bat or two, and a couple spare pitchers.

Comparing them to the 43 Yankees. The Yanks had, by my count, 13 position players who appeared in a substantial number of games, to go along with 10 pitchers who made more than a cameo appearance or two. Almost twice as many players on the roster.

Negro League competition wasn't all that diluted, compared to AL/NL competition, because there weren't as many guys on each roster.

I was the one who said the Negro League stats were diluted down. I'm not arguing or disagreeing with your figures and statements, but you're using the wrong numbers.

I did some online research and am relying upon facts and figures I got off various sites. So if something I ended up using is wrong, I apologize, but blame the online sources.

So, since the National League was first formed in 1876, there have been 19,969 recognized MLB players through 2/17/2022, none of which appear to be solely Negro League players. And as of the end of 2020, MLB recognized approximately 3.400 Negro League player's stats as now being official MLB stats. These were taken from seven different Negro Leagues that operated during the period 1920-1948. And it appears that about 45 of those newly added Negro League player stats were for players that eventually made it into the majors, so I'll reduce the number of added Negro League player stats down to 3,355 (3,400 - 45) so as not to double count those players that did get into the majors also. Currently, depending on where you look, the black population in the US is at about the 12%-14% range. Back during the 1920-1948 period the Negro League stats were taken from, the black population in the US was even lower, at only about 10%. And finally, in 1956, Jackie Robinson's last year in the majors, the percentage of black players on MLB rosters was 6.7%. At the start of the 2020 season, the percentage of black players on MLB rosters had risen to 7.8%, still below the percentage of blacks overall in the US population.

Now without even adjusting for the increased number of teams and players over the past 60 years, starting with MLB's expansion that began back in 1961, if you take the overall total number of recognized MLB players in history and divide that by the number of years MLB has existed, you come up with an overall average of 136.8 new MLB players (19,969 / 146 Yrs) being recognized and added each year. Now if we do the same calculation for the Negro Leagues, we end up with them adding an average of 115.7 new MLB level players (3,355 / 29 Yrs) being recognized and added each year. The problem is, blacks accounted for only about 10% of the overall US population back then, and have historically represented an even lesser percentage of MLB rosters over a long period of time.

So based on those numbers and percentages, you would expect the average number of new black MLB level players being added from the Negro Leagues to be more like 13.7 players each year (136.8 MLB average X 10% black population), versus the 115.7 new MLB level players that were actually being added from the Negro Leagues annually. The 102 player difference (115.7 - 13.7) between these expected and actual average MLB level players being recognized and added each year by the Negro Leagues between 1920-1948 is most likely made up of players, the vast majority of which, that do not have MLB level talent, and are only playing at this level because Negro League teams needed to fill out their rosters with somebody. So as a result, it would seem logical to assume those Negro League stars that did have MLB level talent were able to feast on and pad and embellish their stats by playing a lot less MLB talent level players. Much, much, much more so than their white counterparts in the segregated major leagues. So to now take these Negro League stats and compare them straight across the board against everyone else in MLB is, I feel, totally unfair to all the regular MLB players, as they likely played against much higher overall MLB level talent throughout their careers, as opposed to their Negro League counterparts.

So if you want to go back and tell me again how the Kansas City Monarchs team having small rosters explains away the ridiculously disproportionate number of Negro League players that got their stats added onto MLB's records, and can do so with some actual facts, figures, and logical arguments, I'm all ears. Just promise you won't come back with some crap about how the black athletes just want to all play football and basketball now, and that's the best explanation you can give to explain how those Negro League players were all legit MLB level players back then. And so you know, I made sure to skew some of the numbers I was using so they'd actually go against the argument I was making. In other words, the numbers are likely even more ridiculously disproportionate than I was putting forth.

[molenick]

I think it is great that earlier rankings (and the ESPN list) included Negro League players, because we know some of them would have been stars if they could have played in what were then the Major Leagues. But I think after deciding that they should be included, people didn't have that much to go on except anecdotal evidence, some apocryphal stories, sketchy statistics, and in some cases earlier rankings (which themselves relied on anecdotal evidence, stories, etc.). Paige was the best pitcher, Gibson the best catcher, Charleston the best outfielder, Leonard the best first baseman, etc. It was just a matter of deciding how to rank them among the non-Negro League players.

Now that we have some statistics, in many cases they justify our "gut" rankings. In some cases, the stats are superhuman and clearly have to be taken in context...Josh Gibson's 162 game average .is 374/.458/.719 with 46 HRs, 166 Runs and 196 RBIs. I think we can agree he would not have done that in an integrated major leagues in the 1930s. But I think we can also agree that he certainly lives up to the hype (if not exceeding it). No one can look at those stats and say, I guess he really wasn't as good as they said he was.

But Paige's stats do not hit me in the same way. They are very good, his 162 game average is 14-9 with a 2.70 ERA and 169 strikeouts in 199 innings. His ERA+ is 152 and his lifetime WAR is 46.3 in 1695 innings (36.6 innings per WAR).

Then I look at Bill Foster: 162 game average 19-9 with a 2.63 ERA and 156 strikeouts in 253 innings. His ERA+ is 164 and his lifetime WAR is 47.0 in 1499.67 innings (31.9 innings per WAR).

Bullet Rogan: 162 game average 22-10, 2.70 ERA, 161 ERA+, 168 strikeouts in 275 innings.

Ray Brown: 162 game average 22-8, 3.02 ERA, 149 ERA+, 126 strikeouts in 271 innings.

Except for strikeouts, these players appear to be statistically better than Paige.

Now, there might be reasons for this. For example, Paige was used primarily as a reliever in the integrated Major Leagues, so his 162 game average was watered down for W-L. And pitching in the integrated Major Leagues also worsened his career ERA and W-L pct. There may also be statistics from games that have not been discovered yet.

But let's say Bill Foster was the traditional #1 ranked Negro League pitcher, and say there were stories about him taking the outfielders off the field when he was pitching, and throwing strikes over a bubble gum wrapper.

And then say we were presented with all these statistics we didn't have before. I think we would say, well, yeah, that just shows we got it right, like we do (or I do, anyway) with Gibson (and Charleston, Leonard, and others).

I don't think we would look at Paige's stats and say, no, really he should have been number one all along, it's obvious, can't you see that he has a worse W-L pct., higher ERA, lower ERA+, lower WAR, lower WAR7, lower WAR/162, and lower JAWS and that proves he was better than the guy we always said was number one.