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

Monday, October 11, 2004

Chino Smith

Does his career have enough value to overcome its brevity?

John (You Can Call Me Grandma) Murphy Posted: October 11, 2004 at 02:42 PM | 40 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. KJOK Posted: October 11, 2004 at 11:58 PM (#910713)
I have a feeling Chino is going to be the "John McGraw" candidate for the Negro Leagues - Probably led the league on OBP most years, but a relatively short career...
   2. sunnyday2 Posted: October 12, 2004 at 08:18 PM (#912302)
I don't know about the "John McGraw." I would rank him higher than that. Chino was MVP twice in his short career, McGraw was never that. I see Chino as more like Dobie Moore, for obvious reasons, which means like Hughie Jennings, except he was an OF. So maybe more of a Ross Youngs, only better.

But in the end, even a peak voter like me has Dobie Moore at around #25. Dobie only won 1 MVP, but was an All-Star 6 times to Chino's twice, and Dobie played SS. So Chino is clearly below Dobie, which means below Hughie and well, maybe down there with McGraw after all. I don't have McGraw in my top 50 and in fact not even on the waiting list for top 50. Chino might be top 50 and it is right that his named should be remembered and bandied about. What an awesome hitter--best other than Josh Gibson according to one report.

But not a ballot contender until after we've elected Dobie Moore.
   3. KJOK Posted: October 13, 2004 at 01:57 AM (#913063)
According to the Baltimore Afro-American, Smith's 1927 stats were:

G-34
AB-114
H-50
R-26
Ave-.439
   4. Mark Shirk (jsch) Posted: October 15, 2004 at 08:59 PM (#918295)
What are Smith's i9's? Right now he is between Poles and Hooper on my ballot (about 42 or so).
   5. Chris Cobb Posted: October 15, 2004 at 09:17 PM (#918314)
No i9s for Smith. Their project hasn't gotten to him yet.
   6. Michael Bass Posted: October 16, 2004 at 01:18 AM (#918635)
Just can't see him as a ballot candidate for me. Moore is just off my ballot, and I think I'd definately take him over Smith.
   7. John (You Can Call Me Grandma) Murphy Posted: October 18, 2004 at 12:54 AM (#921379)
Just can't see him as a ballot candidate for me. Moore is just off my ballot, and I think I'd definately take him over Smith.

I have to agree with that.
   8. KJOK Posted: October 20, 2004 at 01:29 AM (#927043)
Stats I could find on Chino Smith:

1924 Phil Giants No stats (Played 2B?)
1925 Brooklyn Royal Giants (Played 2B/SS)
Ave - .341
1926 Brooklyn Royal Giants
Ave - .326
1926 Cuba Winter League
Ave - .348
1927 Brooklyn Royal Giants
G-34
AB-114
H-50
HR-11 (Led League)
R-26
Ave - .439 (Led League)
1928 Brooklyn Royal Giants (played 2B)
G-9
AB-36
H-13
D-2
T-2
HR-1
R-9
BB-0
HBP-1
Ave-.361
OBP-.378
SLG-.611

FIELDING 2B
G-9
Def Innings - 76
PO-18
A-20
E-2
DP-2
PCT-.950

1929 Lincoln Giants
G-67
AB-255
H-118
2B-28
HR-24 (Led League)

1930 Lincoln Giants (Played RF?)
Ave - .492, .488 or .429 (Led League)

One strang fact - In all the boxscores I could find from 1925-1930, Smith ALWAYS batted 3rd. I have yet to find any games where he batted anywhere except 3rd...
   9. KJOK Posted: October 20, 2004 at 01:30 AM (#927049)
Missed 1929 AVE = .463 (Led League)
   10. Thane of Bagarth Posted: October 20, 2004 at 08:41 PM (#929120)
In all the boxscores I could find from 1925-1930 ...

KJOK, or anyone else who might know, is there any evidence in those boxscores that he ever walked?

In Riley's Encyclopedia it says "he had a good eye at the plate and rarely struck out." Later on it also says "he hit everything thrown to him" which, combined with no BB in 36 ABs in 1928 doesn't make me think he put up Bondsian OBPs. If I knew he was getting on base 50% of the time, even in such limited samples, I would give him serious consideration for my ballot.

Not that it's going to change anybody's rankings, but I thought the first sentence in Riley's entry on Chino is notable:

"This compact dynamo who, according to Satchel Paige, was one of the two greatest hitters in the Negro Leagues, was a scrapper, arousing the fans and intimidating pitchers as he shot through the world of black baseball like a meteor, with a career as brief as it was bright."
   11. KJOK Posted: October 21, 2004 at 04:28 AM (#930818)
I haven't found any walks for Chino, but without play by play walks are hard to find in most Afro-American boxscores which are like this:

Name R H PO A
Smith 1 2 4 1
...
Jones 0 0 5 3
TOT 4 8 27 9

then at the bottom of the boxscores
Base on Balls - Off Redding, 3, Off Mendez 2...

without any indiciation which batters walked. So, without AB info or specific batters walked, it's usually impossible to tell....
   12. KJOK Posted: October 21, 2004 at 04:31 AM (#930827)
In the one year we have "good" data, 1928, in the Eastern League, there were 778 Walks in 2,725 innings, so not many batters (not just Chino) were up there drawing a lot of walks...
   13. Mark Shirk (jsch) Posted: October 21, 2004 at 08:58 PM (#931680)
I have always wondered...

The Negro Leagues seems like a very low walk enviroment. Should we give their players a little boost in walks because of this, much like we would give say, Willie Davis, a boost in home runs?
   14. KJOK Posted: October 21, 2004 at 10:44 PM (#931847)
Good Question.
1928:

AL      42144   3817
NL      38462   3664
TOT     80606   7481   9.3
%

ECL     10795    779
NNL     20516   1440
TOT     31311   2219   7.1

Doesn't look like a lot of difference between MLB and Negro Leagues in 1928...
   15. KJOK Posted: October 21, 2004 at 10:46 PM (#931852)
Lost my headers - they are AB and BB...
   16. Mark Shirk (jsch) Posted: October 22, 2004 at 10:50 AM (#932561)
I don't know, 9.3% and 7.1% seem like a big difference. Isnt' that like 20 points of OBP or so?

Of course I also have no clue about historical norms so I really dont' have a point of reference.
   17. KJOK Posted: October 23, 2004 at 01:10 AM (#933539)
It's very roughly about .013 points of OBP, which is a little more than I was thinking, but stil wouldn't call it a BIG difference....
   18. Joey Numbaz (Scruff) Posted: October 25, 2004 at 09:58 AM (#936341)
I would call it a huge difference. That's 31% more walks in MLB than NLB. That's pretty big.

AVG MLB 650 PA - 60 BB
AVG NLB 650 PA - 46 BB

That's 22 or so points of OBP, which is a pretty huge difference. It's bigger than the difference between the historical norm of say 1965 vs. 1982, for example.
   19. KJOK Posted: October 26, 2004 at 12:34 AM (#937145)
That's 31% more walks in MLB than NLB. That's pretty big.

You're selecting a non-representative sample.

Take a look at 1933, 1934, 1949, 1935, 1936, 1932, 1938, 1916 & 1915. In all those years, the AL had a higher difference of walk rate vs. the NL than we're talking about here.

It is significant, but HUGE seems a little extreme.
   20. koufax Posted: May 30, 2005 at 05:57 PM (#1371666)
Answer to Sunnyday2
Chino Smith enjoyed a 6 year career in the Negro leagues, hitting a torrid .405.
He also played 5 years in the strong Cuban Winter League, against the best Negro league talent, the cream of Cuba, and major league players like Charlie Dressen, Eddie Brown, Jakie May, and Jess Petty, batting .335
In my books, 'Baseball's Other All-Stars' and 'King Of Swat', I developed formulas for converting a player's statistics from the Negro leagues, Cuba, or Japan, to their major league equivalent. The Japanese stats, which were developed in 2000, have since been confirmed by the experience of Ichiro Suzuki, Hideki Matsui, and others.
Based on his NL and CWL stats, Chino Smith could have been expected to hit about .349 in the major leagues with 39 doubles, 12 triples, and 20 home runs a year.
   21. andrew siegel Posted: July 29, 2006 at 01:21 AM (#2116148)
Columbia, SC's newspaper, The State, ran a long article about him last weekend. I never really took note of him until I saw the article, but this thread has some impressive numbers. Do we have updated numbers from the Hall of Fame project for him? Might we need some MLE's?
   22. Dr. Chaleeko Posted: July 29, 2006 at 04:41 PM (#2116678)
The HOFs pdf did not include Smith, but it did include league averages. On the other hand, this was before I was doing them, so Chris might be the man to ask here.
   23. burniswright Posted: December 07, 2007 at 08:21 AM (#2637766)
I'm a newcomer here, so I'm going to read all the Negro Leagues threads before I start giving a bunch of opinions. But I just want to emphasize, in this instance, that this is the player with *the highest lifetime BA of any player in any legitimate major league, ever*. No matter how you crunch his numbers, it's going to be hard to stare that one down.

As to brief careers, I must say there are a whole lot of conflicting theories out there. Bill James, for instance, feels that no one should be penalized for missing years spent in the military during WWII. OK, so we're conflating patriotic sentimentality with career value here. I'm not saying that isn't a legitimate opinion, I'm just pointing out how weird this discussion can get.

One point I would add is this: it might be useful to differentiate between voluntary acts of self-destruction shortening a career (Hack Wilson) and players who (I am quite confident of this) had absolutely no desire to get sick and die at age 30.

Among shortened NEL careers, I see Chino and Dobie Moore as roughly comparable. Moore's career was slightly longer, and he played a significant role on defense. But, as wonderful a hitter as he was, I would put him somewhere between 15th and 20th on the NEL list on offense. Smith, on the other hand, has to be right up there with Gibson, Charleston and Jud Wilson for peak value. That's really a consensus opinion among researchers.
   24. Bleed the Freak Posted: January 08, 2011 at 08:56 PM (#3726528)
http://www.baseball-fever.com/showthread.php?101933-Negro-Leagues-Hall-of-Fame-Round-one

Chino averaged over .400 in his admittedly brief Negro League career--while adding about 33 HR per 550 AB. He didn't walk like Rogers Hornsby, but I think, per my musings posts on him, that if you converted his data for those six seasons, he looks remarkably like a left handed version of Hornsby at the plate less about 60 walks a season. To me, that's one heck of a hitter.

The consensus has been that Chino Smith is Dobie Moore lite - shorter career, less demanding defensive position.

Does anyone here have Chino or in a personal hall of merit...or at least in the consideration set.

He seems to be a plausible candidate for peak voters in particular.

Are any MLE's available for his six years?
   25. Alex King Posted: January 15, 2011 at 02:19 AM (#3730426)
I’ve created some rough MLEs for Chino Smith, and they show him as having a strong case based almost entirely on his incredible peak. This post is very long, so I split it into 2 parts. I’ll present the MLEs in this post, and discuss the results in the next post.

The hard part here was finding credible Negro League stats. For 1925 and 1926, I used the batting averages listed above by KJOK (these were the only statistics I could find from that year). For 1927, I used KJOK’s numbers, listed above, for AB, H, HR, and BA. I used numbers from BR’s Bullpen article on Smith for his 2B. For 1928, I used KJOK’s numbers above. In 1929, I used KJOK’s stats for AB, H, 2B, and HR. Holway reports Smith’s SLG as .903 for 1929; I adjusted Smith’s 3B totals to create the correct number of total bases. In 1930, I referred to BR Bullpen’s article for Smith’s 2B, 3B, and HR, although the article is somewhat unclear:
He hit .492 against other top black teams, the top mark in the east, was tied for second in the east with 7 homers, tied for the eastern lead with Turkey Stearnes with five doubles and swatted the most doubles (17) in that part of the country.

I assume that “five doubles” is meant to be “five triples,” since 5 3B is far more likely than 17 3B. Smith’s batting average for that season is disputed; BR gives it as .492 above, Holway gives it as .488, Baseball Library reports a .429 average, and according to the Negro Leagues Museum he hit .468. I created 3 different MLE’s, reflecting the .488 BA, the .468 BA, and the .429 BA. I think that the .488 BA figure is probably the most credible, as it comes from Holway, while the .429 BA seems to have the least support behind it. Walk data is only available for Smith’s 37 AB 1928 season; however, Scott Simkus’s MLEs for the Negro League stat-o-matic set have Smith’s OBP-BA at 0.037 for his career. As this number sounds reasonable, I’ve used it for Smith’s OBP-BA for every year except 1928. Next, I estimated Smith’s triples in 1927 by assuming that he hit triples in 1927 at the same rate as the rest of his career. Last, I made a wild guess that Smith’s ISO was .200 in 1925 and 1926. In the context of the rest of his career, this seems a little low, as Smith recorded ISOs between .250 and .500 between 1927 and 1930; however, it’s likely that Smith was still developing as a hitter in 1925 and 1926 and would not have displayed as much power as he did later in his career.
Actual Stats

Year    AVG    HR    2B    3B    H    TB    AB    OBP    SLG
1924                                    
1925    0.341                            0.378    0.541
1926    0.326                            0.363    0.526
1927    0.439    11    16    2    50    106    114    0.476    0.934
1928    0.361    1    2    2    13    24    36    0.378    0.611
1929    0.463    24    28    4    118    230    255    0.500    0.902
1930    0.488    7    17    5    103    156    211    0.525    0.739 


I first calculated Smith’s unregressed MLEs by applying the standard conversion factors of 0.9 for batting average, 0.95 for OBP-BA, and 0.81 for ISO. Then I placed each season in the proper Negro Leagues context for that year, using the averages calculated by Eric Chalek. Next I regressed Smith’s MLEs, weighting each season’s statistics at 1/3 and the average of the three surrounding years at 2/3. However, for 1926 I only used Smith’s 1925 and 1926 stats for the “three-year average,” and for 1927 I only used Smith’s 1927 and 1928 stats for the “three-year average.” I did this for a couple reasons: 1. 1927 seems like a legitimate breakout year that should not be regressed to Smith’s 1926 stats, and 2. If I didn’t use such a cutoff, Smith’s regressed MLEs became far better than his unregressed MLEs.

In the next step I estimated Smith’s WAR. To calculate his batting runs, I used the PA*(wOBA – lg_wOBA)/A approximation of linear weights, where A is a constant, generally around 1.15 but varying from year to year. I also assumed Smith played full seasons, giving him 600 PA each year. I applied a park factor of 1.1 to Smith’s batting runs in 1929 and 1930, when he played in the Catholic Protectory Oval, a very good hitter’s park. I calculated this park factor from Gary A’s park factors, which have the Catholic Protectory Oval at about 1.2. However, this is a sample park factor coming from a very small sample of just 57 games, rather than a “true” park factor, and as a result it should be regressed to the mean. Thus, I decided to use a park factor of 1.1. Since Gary A has very little data for the Brooklyn Royal Giants’ park, I assumed that it had a park factor of 1.

For Smith’s positions, I used the BR Bullpen article, which listed him as a 2B in 1926 and a RF in 1927. KJOK has him as a 2B in 1928 and a RF in 1930; Holway has him as a RF in 1929. Smith’s positions in 1925 are disputed, with various sources listing him as a 2B, SS, RF, and 3B; for the purposes of my positional adjustments, I assumed he played equally at all four positions. I also assumed Smith was a -5 fielder at 2B and a league-average fielder in RF. Lastly, I credited Smith with 2.5 WAR in 1924; according to Holway and the Negro Leagues Baseball Museum, he played for the Philadelphia Giants that year, though no stats are available. 2.5 WAR represents an extrapolation of his career based on Smith’s 1925 and 1926 seasons.
Year    mle_BA    mle_OBP    mle_SLG    Bat    Fielding    Position    Pos Adj.    Rep    RAR    WAR    adj_WAR
1924                        2B                    2.5
1925    0.319    0.343    0.494    11    0    2B
/SS/3B/RF    3.5    18    33    3.2    3.4
1926    0.305    0.332    0.464    10    
-5    2B    5    18    28    2.8    3.0
1927    0.384    0.418    0.723    75    0    RF    
-6    18    87    8.7    9.1
1928    0.366    0.408    0.663    58    
-5    2B    5    18    76    7.6    8.0
1929    0.422    0.462    0.748    75    0    RF    
-6    18    87    8.3    8.7
1930    0.473    0.506    0.797    86    0    RF    
-6    18    98    9.1    9.6
Total                                        39.7    44.2

Year    mle_BA    mle_OBP    mle_SLG    Bat    Fielding    Position    Pos Adj
.    Rep    RAR    WAR    adj_WAR
1924                        2B                    2.5
1925    0.319    0.343    0.494    11    0    2B
/SS/3B/RF    3.5    18    33    3.2    3.4
1926    0.305    0.332    0.464    10    
-5    2B    5    18    28    2.8    3.0
1927    0.384    0.418    0.723    75    0    RF    
-6    18    87    8.7    9.1
1928    0.366    0.408    0.663    58    
-5    2B    5    18    76    7.6    8.0
1929    0.417    0.458    0.744    73    0    RF    
-6    18    85    8.1    8.5
1930    0.460    0.493    0.782    81    0    RF    
-6    18    93    8.6    9.0
Total                                        39.0    43.5

Year    mle_BA    mle_OBP    mle_SLG    
New reg bat    Fielding    Position    Pos Adj.    Rep    RAR    WAR    adj_WAR
1924                        2B                    2.5
1925    0.319    0.343    0.494    11    0    2B
/SS/3B/RF    3.5    18    33    3.2    3.4
1926    0.305    0.332    0.464    10    
-5    2B    5    18    28    2.8    3.0
1927    0.384    0.418    0.723    75    0    RF    
-6    18    87    8.7    9.1
1928    0.366    0.408    0.663    58    
-5    2B    5    18    76    7.6    8.0
1929    0.409    0.450    0.735    70    0    RF    
-6    18    82    7.8    8.2
1930    0.434    0.467    0.754    70    0    RF    
-6    18    82    7.6    8.0
Total                                        37.7    42.1 


Smith’s BA, OBP and SLG were put into an NL context following this example for BA: (BA/Negro_League_BA)*NL_BA. I made no attempts to adjust for park factors in 1929 and 1930.
   26. Alex King Posted: January 15, 2011 at 02:54 AM (#3730439)
Fantastic...My tables are messed up and the edit button just disappeared.

Here they are again, properly formatted:
Actual Stats

Year AVG HR 2B 3B H TB AB OBP SLG
1924         
1925 0.341       0.378 0.541
1926 0.326       0.363 0.526
1927 0.439 11 16 2 50 106 114 0.476 0.934
1928 0.361 1 2 2 13 24 36 0.378 0.611
1929 0.463 24 28 4 118 230 255 0.500 0.902
1930 0.488 7 17 5 103 156 211 0.525 0.739


Year mle_BA mle_OBP mle_SLG Bat Fielding Position Pos Adj
Rep RAR WAR adj_WAR
1924      2B     2.5
1925 0.319 0.343 0.494 11 0 2B
/SS/3B/RF 3.5 18 33 3.2 3.4
1926 0.305 0.332 0.464 10 
-5 2B 5 18 28 2.8 3.0
1927 0.384 0.418 0.723 75 0 RF 
-6 18 87 8.7 9.1
1928 0.366 0.408 0.663 58 
-5 2B 5 18 76 7.6 8.0
1929 0.422 0.462 0.748 75 0 RF 
-6 18 87 8.3 8.7
1930 0.473 0.506 0.797 86 0 RF 
-6 18 98 9.1 9.6
Total          39.7 44.2

Year mle_BA mle_OBP mle_SLG Bat Fielding Position Pos Adj
Rep RAR WAR adj_WAR
1924      2B     2.5
1925 0.319 0.343 0.494 11 0 2B
/SS/3B/RF 3.5 18 33 3.2 3.4
1926 0.305 0.332 0.464 10 
-5 2B 5 18 28 2.8 3.0
1927 0.384 0.418 0.723 75 0 RF 
-6 18 87 8.7 9.1
1928 0.366 0.408 0.663 58 
-5 2B 5 18 76 7.6 8.0
1929 0.417 0.458 0.744 73 0 RF 
-6 18 85 8.1 8.5
1930 0.460 0.493 0.782 81 0 RF 
-6 18 93 8.6 9.0
Total          39.0 43.5

Year mle_BA mle_OBP mle_SLG 
New reg bat Fielding Position Pos AdjRep RAR WAR adj_WAR
1924      2B     2.5
1925 0.319 0.343 0.494 11 0 2B
/SS/3B/RF 3.5 18 33 3.2 3.4
1926 0.305 0.332 0.464 10 
-5 2B 5 18 28 2.8 3.0
1927 0.384 0.418 0.723 75 0 RF 
-6 18 87 8.7 9.1
1928 0.366 0.408 0.663 58 
-5 2B 5 18 76 7.6 8.0
1929 0.409 0.450 0.735 70 0 RF 
-6 18 82 7.8 8.2
1930 0.434 0.467 0.754 70 0 RF 
-6 18 82 7.6 8.0
Total          37.7 42.1 
   27. Alex King Posted: January 15, 2011 at 03:16 AM (#3730452)
Double Post
   28. Alex King Posted: January 15, 2011 at 03:16 AM (#3730453)
Alright, these tables aren't getting any better...Here's the rest of the original post:


Chino Smith MLEs, version 2:

Smith’s BA in 1930 does look quite high in all three versions of the MLE. However, remember that his BA has been translated into the highest-scoring environment in the 20th century, and that the Negro Leagues hit significantly worse than the majors in 1930. Putting Hornsby’s 1924 into the 1930 context, following the same method as above, produces an average of .454 (Hornsby had a .424 average when league average was .283). Furthermore, Hornsby’s park, Sportsman’s Park in St. Louis, was roughly neutral that year, while Smith played in a very good hitter’s park. As a result, neither MLE version 1, with a 1930 BA of .473, or MLE version 2, with a 1930 BA of .460, can be immediately dismissed as unrealistic.

To address some of my skepticism about the MLEs, I compared all three MLEs to Hornsby to see how closely the MLEs support the Jim Albright’s assertion (from the link that Bleed gave above) that Smith was like Hornsby, but with fewer walks. Taking 60 walks per year off Hornsby’s total seemed like too many, especially since Hornsby only drew about 60 walks in many of his early seasons. Therefore, I decided to simply cut Hornsby’s walk totals in half. In order to create a comparison, I used batting runs. For Smith, I used the “bat” column from each of the 3 MLEs. For Hornsby, I adjusted his batting runs to reflect his newly lowered walk totals. First, I took his Baseball-Reference batting run and subtracted the linear weights value of half his walks (I used a LW value of 0.32, if you’re curious). I then added a pro-rated fraction of Hornsby’s batting runs, to account for the value he accrued in extra plate appearances when he would have walked. I ranked Hornsby’s seasons, and looked at his top 4 seasons, comparing them against Smith’s 4-year peak. Lastly, I considered each season on a rate basis and presented it as per 600 PA.
Smith 1   Smith 2  Smith 3   Hornsby minus half of walks
86   81   75   85
75   75   70   80
75   73   70   73
58   58   58   65 

Even in MLE #1, with the seemingly outlandish .473 BA, Smith’s peak does not tower over the peak of the reduced-walk Hornsby. This increases my faith in the credibility of versions 1 and 2, with the very high batting averages for Smith in 1930. It seems likely that Smith’s batting average in 1930 is more a product of an extremely hitter-friendly context than necessarily indicative of a bad MLE or bad data.
As I said before, I’d probably choose MLE version 1, with a .488 BA for Smith in 1930, over versions 2 and 3 (.468 and .429 BA, respectively). I still think that version 3, with the .429 BA, is the least credible of the three possibilities. However, all three MLEs indicate that Smith had a great peak, up there with the top backloggers on the 2012 ballot. By my WAR-based rankings, Smith version 1 has the best peak of anyone on the 2012 ballot; version 2 is just behind Ned Williamson, while version 3 is third behind Williamson and Wilbur Wood. In all three MLEs, Smith’s peak is comparable to those of peak-centric candidates Rosen, Gooden, Cicotte, and Wilbur Wood (though Smith lacks Gooden’s single-season impact). Smith’s peak is so good, in fact, that it overcomes his lack of career value to allow him to slot in around #15 in my rankings—higher than any of Rosen, Gooden, or Wood, and only slightly below Cicotte, who has quite a bit more career value.

It appears that Smith’s peak is as impressive as much of the anecdotal evidence indicates. I would encourage peak voters, or anyone voting for Gooden, Rosen, Cicotte, or Wood, to strongly consider Smith’s case. Moreover, I think anyone who will vote for extreme high-peak, low-career candidates should consider voting for Smith.
   29. Bleed the Freak Posted: January 16, 2011 at 02:04 AM (#3730831)
Thanks Alex for the MLE's, as this shows that Chino belongs in my consideration set for the Hall of Merit ballot and personal hall of merit, although at first blush, this would leave him a bit short of the ballot.

Discerning the figures above, you gave Chino limited but not poor defensive value, and I'm not sure how you incorporated baserunning merit.

Historian "James A. Riley, The Biographical Encyclopedia of the Negro Baseball Leagues" suggests that Chino was a good defensive outfielder and a fine base runner.

Would Chino's MLE's change much if your defensive rate is below that of a good outfielder/second basemen and a fine base runner.

Also, if Al Simmons AVG/OBP is a good indicator of what Chino's OBP would be, then his OBP may be understated too. George Sisler had an even worse OBP relative to AVG, and his OBP/AVG is higher than Chino's.

Your work is greatly appreciated Alex.

-Bleed

In addition, a more full-length analysis was penned by Jim Albright at the following link:

http://www.baseball-fever.com/archive/index.php/t-101776.html
   30. Brent Posted: January 16, 2011 at 05:26 AM (#3730872)
Interesting. Thank you for doing this work. While I have no doubt that Charlie Smith was one of the better hitters of his era, I do have some concerns that lead me to be skeptical of the claims for extraordinary peak hitting ability, especially for his 1929-30 years when he played for the Lincoln Giants.

First, there is the issue of park factors. Every source describes the Lincoln Giants' home park, the Catholic Protectory Oval, as a tiny bandbox. A recent article by KJOK at seamheads.com lists estimated dimensions of 180' down each of the lines and 358' at the deepest point, numbers that make Chuck Klein's Baker Bowl seem spacious. I would love to see more complete data on park factors. In view of the park's tiny dimensions and reputation, though, I would have retained Gary's estimated park factor of 1.21 rather than regressing it toward the mean.

Second, there are the Cuban League data, where Smith hit well in three winter seasons but was never one of the league leaders. For 1927/28, I have a data compilation made by Gary Ashwill that provides complete hitting statistics. Smith hit .336/.403/.462 in 135 PA in a league that averaged .293/.354/.404. That's an OPS+ of 128. (See the discussion of Cuban League statistics on the Alejandro Oms thread; in posts # 90 and 106, our experts suggest that the 1927/28 and 1928/29 seasons represented the high point of Cuban League play and that the level of competition then was about the same as in the majors). Smith was well behind the league's three leading hitters—Jud Wilson (.432/.540/.739, OPS+ 235); Oscar Charleston (.351/.472/.570, 174); and Martín Dihigo (.381/.448/.571, 168). Smith was in a group of hitters with OPS+ between 120 and 133 that included Dick Lundy, John Henry Lloyd, José Fernández, Ramón Herrera, and Alejandro Oms.

For 1928/29 I only have Figueredo's data, which lack BB and OBP. Smith's BA/SLG were .333/.533 in 135 AB in a league that averaged .302/.422 for the batters shown by Figueredo. (I note, however, that in contrast to Gary's compilations, Figueredo's data are somewhat incomplete; several players are listed as appearing on the roster but with a note saying "Statistics not available.") Again, Smith is one of the better hitters in the league, but was still well behind the league leaders, who that season were Oms (.432/.619) and Wilson (.397/.609). Other players with statistics that were somewhat better than Smith's included Agustín Bejerano (.378/.514), Chaney White (.394/.500), Charleston (.333/.561), and Cool Papa Bell (.325/.573).

In the winter of 1930/31, the Cuban winter season was cancelled due to a contract dispute with the management of the stadium. A "special season," lasting 14 games, was held in November in an older ballpark. Smith hit only .273 (9 for 33), with no extra base hits.

So overall, while Smith's Cuban League play confirms that he was an above average hitter, it does not buttress the argument that he had an extraordinary peak by historical standards.
   31. Alex King Posted: January 17, 2011 at 02:40 AM (#3731175)
Brent, thanks for the Cuban league averages. I’d noticed Smith’s Cuban League stats in the various articles I read about him, but I didn’t realize that more extensive stats, and Cuban League averages, were available on the Oms thread (I’d mistakenly surmised that Smith's lower averages were a result of lower run environments in Cuba). Do you also have Smith’s Cuban League stats for 1926 (listed at .342 BA by BR Bullpen), or 1929 (listed at .338)?

Bleed/29:

As Holway and Riley both describe Smith as a strong defensive outfielder known for throwing behind runners taking a turn around 1B, it probably makes sense to peg Smith as a slightly above average (+5) RF. The only data on Smith’s baserunning comes from 1929, when he stole 16 bases in 237 at bats, which equates to 38 SB over a full season. Based on major league baserunning and SB figures from this time period, it appears that Smith would be credited with about 6 baserunning runs based on his 38 SB. This is, however, a high total for a man not known primarily for his speed. As a result, I feel that 2 runs per year is a more accurate reflection of Smith’s baserunning contributions over the rest of his career.

Next, I redid Smith’s MLEs using the higher park factor that Brent advocates. I agree that in this case, regression to a mean of 1 is incorrect due to the tiny dimensions and notorious reputation of the Catholic Protectory Oval (what we really want to do is regress to the mean of all Negro Leagues parks 180 feet down each line, but that is basically impossible to do). I also noticed, upon re-reading the linked Holway article, that Smith’s 1930 stats come exclusively from home games, most of which were presumably played at the Catholic Protectory Oval (Holway does describe one game played at Yankee Stadium, however). Therefore, I used 1.21 as the park factor for 1929, and 1.42 as the park factor for 1930.

I also fixed an error in the MLEs, as I had been using the non-park-corrected statistics in computing three-year averages. I applied the park adjustments to the league averages by using basic runs created and inflating BA, OBP, and SLG equally such that runs created equaled the league average times the park factor. I also park-adjusted Smith’s MLE BA, OBP, and SLG.
Year      BA       OBP      SLG    WAR   Bat   OPS+
1924                2.5 WAR
1925    0.319    0.343    0.494    3.6    11    118
1926    0.305    0.332    0.464    3.2    10    118
1927    0.384    0.418    0.723    9.9    75    211
1928    0.359    0.401    0.650    7.8    54    180
1929    0.389    0.427    0.692    9.2    68    182
1930    0.418    0.447    0.705    8.5    69    182 

Total WAR: 44.6

Even after the park adjustments, Smith’s Negro League MLEs are still way better than his Cuban League stats. His 1927/1928 Cuban OPS+ of 128 is nowhere near his 1927 MLE OPS+ of 211 or his 1928 MLE OPS+ of 180. In terms of batting runs, Smith in Cuba is 26 runs above average, while in the MLE he is 75 and 54 runs above average. If we assume that the league and Smith both walked at the same rate in 1928/1929 as they did in 1927/1928, Smith’s 1928/1929 OPS+ is just 136, compared to MLE OPS+s of 180 in 1928 and 182 in 1929.

Why is there such a large difference between Smith’s Negro League and Cuban League statistics? I can think of a few possible reasons:

1. The park factor for the Catholic Protectory Oval is still too low.
2. The park factor for Smith’s time with the Brooklyn Royal Giants, which I’d assumed was 1, should be higher than that.
3. Smith was hurt by his Cuban parks. At least in 1927/1928, however, this is impossible, since, according to Gary A in the Oms thread, every game was played in Almendares Park. Brent, do you know the reputation of Smith’s home park in 1928/1929?
4. Quality of competition—Smith may have disproportionately faced the best pitchers in the Cuban League. In 1927/1928, according to Gary A, Smith primarily faced Bill Foster, Willie Powell, Manuel Garcia, Basilio Rosell, Adolfo Luque, Emilio Palmero, and Isidro Fabre. The pitchers on Smith’s Habana team were Oscar Levis, Cliff Bell, Martin Dihigo, and Sam Streeter. In 1928, Powell and Foster played well in the Negro Leagues, Fabre was good in limited time, Bell, Streeter, and Rosell were about average, and Levis, Dihigo and Garcia struggled in limited time. Luque had 119 and 111 ERA+s in his surrounding major league seasons. Levis and Dihigo, however, pitched well in Cuba and were probably better pitchers than their 1928 Negro League stats indicate. To me, it appears that Smith did not face disproportionately good pitchers in 1927/1928, though I'd like to know what Brent thinks about those pitchers. Also, Brent, do you know what the distribution of pitchers was like in 1928/1929? Gary A mentions in the Oms thread that Trent played for Cienfuegos, Luque played for Cuba, and Smith played for Habana along with Dihigo and Cliff Bell; however, Gary did not give full rosters.
5. Smith was particularly suited to his Negro League ballparks, and particularly unsuited to his Cuban league parks. I need more information about the Brooklyn Giants’ Park, and about Smith’s Cuban parks, to address this point.
6. Both datasets are properly adjusted, and the difference is merely a result of small sample size variation. Smith was neither as good as his Negro League stats indicate nor as bad (relatively speaking) as his Cuban League stats indicate. Probably the safest and most likely of these six possibilities.
   32. Brent Posted: January 17, 2011 at 08:34 PM (#3731425)
You raise some good questions. I'm afraid I can't fill in too many answers, but let me see what I can do.

Thanks for raising a question about missing seasons; I was careless and did miss the statistics for Smith's 1926/27 and 1929/30 Cuban seasons (it looks like he played a total five winter seasons in Cuba).

In the winter of 1926/27, Smith played for an independent league that split off from the main Cuban League; it lasted only one season. It was known as "Triangular" (because it consisted of three teams) and featured many of the better known Cuban and U.S. players because it offered higher pay. Overall, though, I'd say the quality of competition that season must have been lower than in the 1927/28 season because some good players stayed in the regular Cuban League. The league played its games at the University of Havana Stadium--I don't know anything about the stadium's characteristics, but if all the games were played there, the park factor would be indistinguishable from the other league characteristics.

Statistics for the Triangular League are very limited--Figueredo reports only AB, R, H, and Avg. Smith hit .342 (27 for 79) and scored 12 runs. The league leaders for batting average were Pablo Mesa (.433 in 127 AB), George (Tank) Carr (.416 in 125), Dick Lundy (.410 in 127), and Oscar Charleston (.404 in 151).

In 1929/30, Smith hit .338/.500 (batting average/slugging) in a league where the average (excluding most pitchers) was .286/.381. The two best hitters that season were Oms (.380/.572) and Wilson (.363/.569). Either Smith or Suttles (.314/.526) was probably the third best hitter in the league.

You asked about stolen bases. I have Cuban SB data for two seasons--in 1927/28 he stole 2 bases in 29 games; in 1928/29 he stole 7 bases—I don't have the number of games, but he had 135 AB. Unfortunately, there's nothing (even in Gary's compilation) on CS.

Regarding Cuban park factors, my understanding is that in 1926/27 all Triangular League games were played at University of Havana Stadium; in 1927/28 all games were played at Almendares Park; in 1928/29 three of the four teams (including Smith's Habana team) played home games at Almendares Park, and the fourth (Cienfuegos) played at Aida Park; in 1929/30, two teams (including Smith's Habana) played at Almendares Park, while the other two (Cienfuegos and Santa Clara) played their home games in those cities; and for the November 1930 "Special Season," games were played in Almendares Park. Almendares Park had huge dimensions, and the statistics show that players in the late 1920s Cuban League hit about twice as many triples as home runs. The only thing I know about Aida Park is that Cool Papa Bell hit 3 inside-the-park home runs in one game there (New Year's Day, 1929), so I'll assume it was also huge. I think it's safe to guess the park factors were fairly neutral throughout the league, though the parks themselves were about as different from the Lincoln Giants' home as could be imagined.

Pitchers in Cuba:
1926/27 (Triangular): Alacranes - Luque, Ping Gardner, Oscar Tuero, José Méndez; Habana Reds - Emilio Palmero, Merven Ryan, Justo Domínguez, Holsey Lee; Marianao (Smith's team) - Isidro Fabré, Oscar Levis, L. Morera, David Gómez.
1928/29: Habana (Smith's team) - Cliff Bell, Oscar Estrada, José Acosta, Oscar Levis; Almendares - César Alverez, Andy Cooper, Manuel García, L. Morera; Cuba - Dolf Luque, Basilio Rosell, Raúl Alvarez, Emilio Palmero; Cienfuegos - Heliodoro Díaz, Ted Trent, Charles Williams.
1929/30 - Cienfuegos - Heliodoro Díaz, Charles Williams, Harry Salmon, Sam Streeter; Santa Clara - Satchel Paige, Basilio Rosell, Ramón Bragaña, Lamon Yokeley; Almendares - Johnny Allen, Silvino Ruíz, Manuel García, Isidro Fabré; Habana (Smith's team) - Cliff Bell, Dolf Luque, Oscar Levis, Oscar Estrada, José Acosta.

I don't really know anything about park factors for the Brooklyn Royal Giants' Dexter Park, though from an aerial shot available here, it looks like the dimensions in right field were fairly short. It also isn't clear to me how many games Brooklyn actually played at home; Gary's comments on park factors suggest that they may have mostly been a traveling team.

My guess is that all of your factors (1 through 6) are reasonable conjectures. Regarding # 6, I'm always concerned about the small sample sizes that we have for Negro league players, which is one reason I like to add the information from the winter leagues. But those seasons are also typically about 150 PA, so the season-to-season variation is likely to be similar to the month-to-month variation in major league statistics—in other words, a lot! My experience in this project is that it's hard to approach the Negro league data as a true peak voter. If you don't regress the results, the data give peaks and valleys that really can't be compared to the major league data. On the other hand, as many HoM voters have noted, the regressed MLEs usually tend to smooth out the peaks, which is frustrating to peak-oriented voters.
   33. Bleed the Freak Posted: January 19, 2011 at 06:25 AM (#3732459)
I wish I had spotted this earlier, but, to confirm Brent's take on the ballpark dimensions, nearly six years ago Gary A mentioned a tid bit regarding Chino Smith, in the Dobie Moore thread, post 79:
79. Gary A Posted: May 30, 2005 at 10:10 PM (#1372680)
<i>Is there any more to Chino Smith than meets the eye? e.g. is his total career just his 5 (?) years in the NeLs? If so, despite his incredible peak, he does not have much of a case!?<>

Smith's NeL numbers were compiled under almost the best imaginable circumstances: he played in the higher-scoring east during a high-scoring era for the NeL, and his best years were in a great hitters' park (Catholic Protectory Oval). If Riley is right about his birth date (1903), then he would have played from age 22 to age 28, so he could be missing a good chunk of his prime (but also declining years).

My take on his Cuban League numbers, though, is that they're not as impressive as they might seem. The best hitters in that league when Smith was in it were Alejandro Oms, Jud Wilson, Oscar Charleston, and maybe Martin Dihigo; Smith is good, but he kind of fades into the pack a bit.

An interesting comparison would be to a contemporary, Pythias Russ. Riley doesn't give him an age, but he played from 1925-29; since he was a college man, it would seem likely he was born about the same time as Smith. Unlike Smith, Russ played under circumstances that suppressed his numbers, spending four of his five years with the American Giants in their extreme pitchers' park. Still, according to Holway, he batted .327, .268, .350, .405, and .386, before dying of tb in 1929. Holway's career numbers for him are .350 (313 for 895). I can't endorse the .405 (my research shows him at .346 for 1928), but he was still the best hitter on the American Giants that season.

The kicker is that Russ started out as a catcher, then converted to shortstop. His fielding numbers for 1928 are fine: fielding pct .940 (NNL ss .922), RF 5.58 (NNL ave 5.59), under circumstances (high-K pitching staff that allowed the fewest baserunners in the league) that tended to suppress chances for infielders. Considering park effects, I'd have to say he rivaled Willie Wells as the best player in the NNL that season.

The thing is, almost nobody remembers Russ, whereas there were several anecdotes about Smith, largely because of his temper. But after studying the 1928 season in some detail, I strongly suspect that Russ was a more valuable player, possibly by a lot.


This information further dulls my enthusiasm for Chino's record - I just wish he could have played out his career with a prime/decline phase, but sickness got the best of him.
   34. Alex King Posted: January 19, 2011 at 08:05 AM (#3732487)
I strongly suspect that a decent fraction of the difference between Smith's Cuban stats and his Negro League stats is due to his parks. Specifically, it seems as if Smith was perfectly suited for the Catholic Protectory Oval but struggled (relatively) in the much larger Cuban stadiums. One of his teammates describes Smith as hitting line drives that literally kept going, out of the ballpark--this is very doable when the fences are less than 200 feet from home plate! In the vast Cuban parks, however, those line drives would only go for singles or doubles. Thus, I suspect that Smith simply had more value to his Negro League teams than to his Cuban teams. If you're mostly concerned about Smith's value in the Negro Leagues and less preoccupied with his value in Cuba, it's probably unfair to hold Smith's Cuban stats against him (I would still suggest some regression to his Cuban numbers, however, as it is possible that some of the difference is actually reflective of a lower talent level).
   35. Brent Posted: January 21, 2011 at 03:18 AM (#3733784)
I guess my take on Chino Smith is that he's someone we should revisit when more complete Negro league data finally become available (hopefully sometime in the next 3 or 4 years). Until we can get things like better park factors and home/road splits, I just really don't think we can decide whether he was more like Chuck Klein or Rogers Hornsby.
   36. Yardape Posted: January 21, 2011 at 03:49 AM (#3733794)
when more complete Negro league data finally become available (hopefully sometime in the next 3 or 4 years).


At the beginning of last season, they announced that B-R was going to be adding some Negro League stats courtesy of Gary Ashwill and Scott Simkus. At the time, April was the target date. As far as I can tell, that hasn't happened. Does anyone have any idea when they will get it up? Or have I missed something?
   37. Alex King Posted: January 22, 2011 at 02:40 AM (#3734393)
Brent-–thanks for the SB data. Extrapolated to 150 games or 600 PA, Smith would had about 10 SB in 1926/1927 and 31 SB in 1928/1929 (coinciding with Holway's report of 16 SB in the 1929 Negro Leagues). This data seems to support my revised estimates of Smith's stolen base value, with one season in which he stole a lot of bases and others in which he offered a small amount of value through his baserunning.
   38. Alex King Posted: January 22, 2011 at 02:41 AM (#3734394)
With the complete Cuban League statistics for Smith, (thanks Brent!), I decided to create an MLE for Smith using only his Cuban statistics and the conversion factors provided in the Oms thread. First, for the purposes of this MLE, I’m ignoring Smith’s 1930 season due to its extremely small sample size (14 games, 33 AB). Next, for the 1926/1927 Triangular League, I estimated the league BA based on the top 3 hitters’ BA. I took the top 3 hitters that Brent gave for each of the other 3 leagues that Smith played in, averaged their BA, and compared it to the league BA, creating three ratios. I then divided the average of the top 3 Triangular League hitters’ BA by the average of the three ratios, producing an estimated league-average BA of .320. Next, I averaged the other three seasons’ ISOs to estimate the Triangular League’s slugging percentage. Finally, I estimated the league OBP by assuming that the Triangular League would have the same OBP-BA as the only league with OBP data, the 1927/1928 CWL. I repeated this step for the 1928/1929 and 1929/1930 CWL, as those leagues also lacked OBP data.

For Smith, I estimated his ISO in 1926/1927 by averaging his ISOs for the other three Cuban Winter League seasons. I used .037 for his OBP-BA in 1926/1927, 1928/1929, and 1929/1930—the same OBP-BA as I used in the Negro Leagues MLEs. I’m somewhat skeptical of Smith’s significantly higher OBP-BA in the CWL in 1927/1928 (.067), both because of the small sample size and because the anecdotal evidence, as well as Scott Simkus’s research, points to Smith being a very impatient hitter.

Lastly, I created WAR estimates and translated batting lines for Smith using the same processes that I outlined above for the Negro League MLEs.

Year        BA/OBP/SLG  cwlBA/OBP/SLG  mleBA/OBP/SLG  mleOPS+  mleWAR
1926
/1927  .342/.379/.505  .320/.381/.425  .301/.347/.442  117  3.5
1927
/1928  .336/.403/.462  .293/.354/.404  .314/.364/.452  125  4.0
1928
/1929  .333/.370/.533  .302/.363/.422  .314/.361/.488  128  4.3
1929
/1930  .338/.375/.500  .286/.347/.381  .328/.367/.528  127  4.1 

Overall, my take on Smith’s career is pretty similar to Brent’s. My best guess is that the huge systematic difference between Smith’s Negro League and Cuban League value comes from the dramatic contrast in parks, as Smith appears not to have been particularly hurt by the quality of pitchers he faced in the CWL. My viewpoint, however, remains mostly unverifiable without Negro League splits data. In any case, the Cuban league data is sufficient to bump Smith off the bottom of my ballot, and move him down to about #75 in my rankings.
   39. Brent Posted: January 22, 2011 at 09:35 PM (#3734816)
Yardape,

Regarding the promise of Negro league data for the b-r website, I noticed that Gary recently posted the following in the comments section of his agatetype site:

Sean has nearly all of the first installment, excepting only a revised & expanded version of the 1922 NNL, which I'm currently working on. He thinks that they will probably get the stats posted by Opening Day or so. That's probably a goal, rather than a firm deadline.

So maybe we won't have to wait too much longer.
   40. Yardape Posted: January 23, 2011 at 03:43 AM (#3734919)
Thanks Brent. I must have missed that at agatetype.

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
Don Malcolm
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Syndicate

Page rendered in 0.8009 seconds
49 querie(s) executed