hot topics

There's a very nice biographical article by Bill James in "The Baseball Book 1991." Here is a quote from the article:

"Modern fans looking back at the astonishing numbers he compiled in the best minor leagues are always puzzled by his failure to 'get a chance' in the majors, so perhaps I should address that directly. There were several factors involved.

"- The Oakland owners did not want to sell him, and until the mid- to late-twenties did not have to sell him if they did not want to. (In fact, I'm not really sure when the Pacific Coast League teams agreed to the structural arrangements which required them to sell players to the majors after holding them for a couple of years. It wasn't before 1925.) The Oaks set a price for Arlett of $100,000--the same price Babe Ruth sold for in 1919--and, not surprisingly, no major league team would come up with the money.

"- Although Arlett had other pitches, he relied heavily on a spitball. The spitball was banned in 1920, but established pitchers were protected by a 'grandfather clause.' In the Pacific Coast League, Arlett was grandfathered, allowed to throw the spitball. Had he gone to the major leagues, he would not have been; he wasn't on the major league list. Thus, from 1920 to 1922, Arlett's prime as a pitcher, major league teams were reluctant to shell out the big money for a pitcher who couldn't bring one of his best pitches with him.

"- After Arlett moved to the outfield, it took him a couple of years to gain the same status as an outfielder that he had had as a pitcher.

"- Like Ruth, Arlett was enormously popular with his local fans, which made it both difficult and counter-productive for an organization to peddle him. Arlett was, incidentally, an extremely handsome man, with dark, curly hair and a dimpled chin.

"- Like Ruth, Arlett put on weight in mid-career, reaching 230 to 245 pounds by 1930. He lost some speed, and major league teams became somewhat wary of him because of this."

Later in the article, James concludes that Arlett was as good as some of the marginal Hall of Fame outfielders, including Chuck Klein, Hack Wilson, and Heinie Manush. He says he may have been as good as Goose Goslin, but probably was not.

Anyone with some PCL stats and some time on their hands want to tackle some kind of WS eqiuvalency for Arlett?

That last paragraph sure is a kicker.

There's a huge gap, in my opinion between Chuck Klein and Hack Wilson. Wilson is a Roger Maris to me, while Klein is quite similar to Larry Walker and a borderline candidate to me. Goslin is a no-brainer.

"- The Oakland owners did not want to sell him, and until the mid- to late-twenties did not have to sell him if they did not want to. (In fact, I'm not really sure when the Pacific Coast League teams agreed to the structural arrangements which required them to sell players to the majors after holding them for a couple of years. It wasn't before 1925.) The Oaks set a price for Arlett of $100,000--the same price Babe Ruth sold for in 1919--and, not surprisingly, no major league team would come up with the money.


In the Vets committee discussion, I pontificated that if the Joe D wanted to stay, teh Seals didn't *have* to sell him.

It's hard to tell from this that the Seals couldn't have set a ridiculous price and kept him.

The comment about when the NA set the price structure is a critical one.

The comment about when the NA set the
price structure is a critical one.

Then see http://mysite.verizon.net/brak2.0/ml_draft.htm

I was going to take a look at Arlett's PCL career and see if I could make some statistical sense out of it. His first big offensive season was 1924 when he hit 33 homers, 145 RBI and .328 for Oakland.

(The Minor League Register lists Arlett's 1923 RBI total as 191 but that seems to be a typo. 101 seems the real number.)

Anyway, I quit after just looking at 1924. Here are some individual RBI totals of PCL players that season.

Bert Ellison 188
Wally Hood 184
Jim Poole 159
Duffy Lewis (yes, that Duffy Lewis) 154
Ray Rohwer 154
Frank Brazill 148
Roy Leslie 147
Buzz Arlett 145
Dick Cox 141
Ted Baldwin 140
Charlie High 139
Johnny Fredericks 132
Jim McDowell 132
Brick Eldred 131
Cedric Durst 130
Les Sheehan 118
Eddie Mulligan 114
Howard Lindimore 114
(The Immortal) Pete Kilduff 108
Lefty O'Doul 101
Jim Blakesly 101

A lot of recognizable major league names, but few accomplished much.

What the heck. Here's 1929.

Ike Boone 218 RBI (55 HR)
Buz Arlett 189 (39 HR)
Fuzzy Hufft 187
Gus Suhr 177 (51 HR)
Wally Berge 166 (40 HR)
Johnny Vergez 165 (46 Hr) (Arlett's teammate)
Earl Webb 164 (37 HR)
Smead Jolley 159
Monk Sherlock 156
Mickey Health 156
Cleo Carlyle 136
Charlie Wade 133
Liz Funk (yes, that Liz Funk) 125
Red Wingo 124
Jim Keesey 124
Hank Severeid (yes, that Hank Severeid) 124
Slug Tolson 122
Bill Rumler 120
Ray Jacobs 118
Dave Barbee 118
Gordon Slade 115
Ernie Lombardi 109
Roy Carlyle 108

I think we need to set up a page for Smead Jolley as well as Arlett. :)

Arlett was born in 1899. His listed baseball height and weight are 6'4" 235 pounds.

Arlett became a full-time outfielder in 1923. He hit .341 for his minor league career of 2390 games with 432 home runs, 1786 RBI, 1610 runs, .604 SLG%, and 1137 extrabase hits. He drove in 100 runs 12 tiems, 140 8 times. He started as a pitcher in 1918 and went 108-93 with a 3.42 ERA using mainly a spitball.

From Futility Infielder and stewthornley.net for post Philadelphia career and coast.league.com for Oakland:
As a pitcher:
year gs Ws Ls InnsP Hts Run ERs Kks BBs ERA
1918 21 04 09 153.3 150 060 046 034 043 2.70
1919 57 22 17 348.0 315 172 116 079 112 3.00
1920 53 29 17 427.3 430 162 137 105 134 2.89
1921 55 19 18 319.0 371 180 155 101 115 4.37
1922 47 25 19 374.0 396 171 115 128 112 2.77
1923 28 04 09 125.0 182 106 080 034 047 5.76

1923 was the year he switched to the outfield full time b/c he hurt his arm - No ####.

As a hitter.
year Gms ABs RNs Hit 2B 3B HR RBI SB AVG. SLG.
1918 026 071 009 015 04 00 01 008 01 .211 .310
1919 056 140 014 040 07 02 01 019 02 .286 .386
1920 064 177 024 045 02 03 05 026 02 .254 .384
1921 064 128 012 028 05 01 03 014 01 .219 .344
1922 074 174 023 042 09 04 04 021 00 .241 .408
1923 149 445 076 147 31 05 19 101 09 .330 .551
1924 193 698 122 229 57 19 33 145 24 .328 .606
1925 190 710 121 244 49 13 25 146 26 .344 .555
1926 194 667 140 255 52 16 35 140 26 .382 .666
1927 187 658 122 231 54 07 30 123 20 .351 .591
1928 190 561 111 205 47 03 25 113 10 .365 .594
1929 200 722 146 270 70 08 39 189 22 .374 .655
1930 176 618 132 223 57 07 31 143 08 .361 .626

He missed some time in 1930 when he got into an argument with an umpire and the umpire smashed his mask into Arlett's face.

In 1932 with Baltimore of the International League, 54 HR 144 RBI.
In 1933, he hit 39 HR.
In 1934 with Minneapolis of the American Assoc.,
.319 41 HR 132 RBI in 116 games.
1935 with Minneapolis, .360 25 HR 101 RBI in 122 games.
1936 Minneapolis .316 16 HR 52 RBI 75games

I couldn't find his stats with Oakland for a long time. While looking, I found his grave in Minnesota: Lakewood Cemetery, Minneapolis, section 41, lot 562, plot 7.
Yes, his grave is online, but not his Oaks yearly stats.

Smead Jolley supposedly was an even worse fielder than Arlett. He did have about three seasons in the majors with the White Sox and Red Sox I think.

Here are his IL and AA stats, in a similar format:

Yr Gms ABs Hts Db T HR Rns RBI BBs SB AVE OBA SLG
32 147 516 175 33 4 54 141 144 112 11 339 457 733
33 159 531 182 40 3 39 135 146 113 20 343 458 655
34 116 430 137 32 1 41 106 132 92 8 319 439 684
35 122 425 153 26 2 25 90 101 60 6 360 439 607
36 50 193 61 10 4 15 55 52 41 1 316 436 643

Black Ink:
1932: Led in HR, Runs, RBI, Walks
1933: Led in Runs, Walks
1934: Led in HR

The OBA is just (W+H)/(AB+W), since I don't have HP and SF info. I figured OBA and SLG myself, and don't know who led the league--Arlett probably led (or was close to leading) the IL in OBA in '32 and '33.

Let me try this again, in better format (I hope):

Yr Gms ABs Hts Db T HR Rns RBI BBs SB AVE OBA SLG
32 147 516 175 33 4 54 141 144 112 11 339 457 733
33 159 531 182 40 3 39 135 146 113 20 343 458 655
34 116 430 137 32 1 41 106 132 092 08 319 439 684
35 122 425 153 26 2 25 090 101 060 06 360 439 607
36 050 193 061 10 4 15 055 052 041 01 316 436 643

Black Ink:
1932: Led in HR, Runs, RBI, Walks
1933: Led in Runs, Walks
1934: Led in HR

The OBA is just (W+H)/(AB+W), since I don't have HP and SF info. I figured OBA and SLG myself, and don't know who led the league--Arlett probably led (or was close to leading) the IL in OBA in '32 and '33.

Black Ink from Oakland years:

1920: Wins, IP, H, BB
1922: IP, H
1924: Triples
1926: RBI
1927: Doubles
1929: Doubles (I think - the record on the site is hard to read)

There is no walk info on www.coastleague.com.

I don't know where to find league scoring information and scoring information for the Oakland home park in order to create MLEs.

Jolley, a dubious-fielding outfielder, had an OPS+ of 112 in the majors, at the peak of his abilities. Since he was just about as good as Arlett in the pCL, it suggests that Arlett, too would have been a good but not great ML player. It also throws doubt on the conversions of minor league and Negro League stats to ML -- I am increasingly of the opinion that there is an inbuilt heavy positive bias in the conversion, and that in particular we are electing too many Negro League stars.

It also throws doubt on the conversions of minor league and Negro League stats to ML

How are you coming to that conclusion, karlmagnus?

We have had over 20 years of successful major league equivalents from Bill James to know that they work. As far as I'm concerned, it's not debatable anymore.

As fot the Negro League equivilents, are they 100% perfect? No, but that doesn't mean that they're not close. The introduction of park factors for the Negro Leaguers should help.

From modern minor league parks, to modern ML, conversions work in a general way at a plus or minus 10% level, with part of the error being random, part of it systemic but tough to find. However, as you go back or go to leagues where there wasn't a lot of movement to and from the ML, the conversions inevitably get more iffy. If A and B play in the same minro league, and A is better than B, then A would probably have been a better major leaguer too. However if A is a HOMer and B played in a minor league or the Negro leagues and his stats, when converted are similar to As, that does NOT in my view prove that B was a HOMER -- the errors are too great, and for the Negro Leagues especially, given the emotional involvements involved, they are all likely to be in the same direction. I do NOT believe that, with poorer facilities, fewer opportunities and more disadvantages, the Negro leagues produced twice the number of HOMers as the proportion of African-Americans in the population.

I think that we have a moral imperative to analyze the negro league stats under the best statistical models that we have and then to give them the same weight that we would anyone else's statistics (i.e., look at them in context, look at intangibles, take into account peak, prime, and carer, use subjective opinions of contemporaries as a check). Using the uncertainty principle reinforces the harm of the blacklist, b/c/ every negro leaguer excluded due to a discount for uncertainty is going to be replaced by a major leaguer (though not necessarily a white major leaguer--could be a second-tier player of color from the later years).

I don't feel the same way about minor league stats for two reasons:

(1) Using an uncertainty principle is morally netural--if you discount some white guys minor league stats for uncertainty and it drops him below the in/out line is place is likely to be taken by another white guy (or even a negro leaguer).

(2) The fact that a guy WAS in the minors is not a neutral fact--in most (though not all) cases it means that he could not get an acceptable job in the majors. While it is absolutely true that lots of major league quality guys were playing in the minors b/c/ bad talent evaluation, bad timing, etc., we should at least entertain the possiblity that the three major league teams that gave up on Cravath within a year of each other noticed something of relevance about his skills or performance that slips through the cracks of the statistics.

Karl, those are some awfully broad conclusions to draw from Smead Jolley's OPS+!

You could also throw Earl Averill into the mix: they were the same baseball age, and here's what they did when they played together in SF's outfield:

1926
Jolley/Averill
AB-575/679
AVE-.346/.348
HR-25/23
RBI-132/119

1927 (Lefty O'Doul was SF's other outfielder, five years older than Jolley and Averill)
Jolley/Averill/O'Doul
AB-625/754/736
AVE-.397/.324/.378
HR-33/20/33
RBI-163/116/158

1928 (let's add Ike Boone, whose team used the same home park--same age as O'Doul)
Jolley/Averill/Boone
AB-765/763/594
AVE-.404/.354/.354
HR-45/36/9
RBI-188/173/104

1929 (Averill went to Cleveland)
Jolley/Boone
AB-812/794
AVE-.387/.407
HR-35/55
RBI-159/218

Their OPS+ in the majors:
Smead Jolley (ages 28-31):112
Ike Boone (mostly ages 27-28, 33): 121
Earl Averill (ages 27-39): 133
Lefty O'Doul (mostly 31-37): 143

Buzz Arlett's OPS+ for this single major-league season (at the age of 32) was 138.

Boone's major league performance is the worst among this group of PCL outfielders of similar offensive accomplishments, playing in the same park at around the same time. Of course, this isn't comprehensive.

Karlmagnus is correct in that we don't have players jumping back and forth between leagues, but we do know what happened when the block from moving from Negro Leagues to Major Leagues:

Jackie Robinson
Larry Doby
Hank Aaron
Willie Mays
Frank Robinson
Don Newcombe
Satchel Paige (at an advanced age)
etc.

Given their performances, it's not a stretch to think that the top Negro League players of the 1930's and 1920's might have been able to accomplish similar performance if given the chance, or that if Satchel Paige was able to pitch effectively in the majors at 46 that he would have been a superstar in his prime...

Edit to my previous post: of course I meant Jolley had the worst performance.

Karlmagnus is correct in that we don't have players jumping back and forth between leagues, but we do know what happened when the block from moving from Negro Leagues to Major Leagues

For the record, the conversion factor of .87 that I use when translating Negro-League statistics into MLEs is based on comparisons of Negro-League performance to major-league performance for

Bob Boyd
Roy Campanella
Larry Doby
Monte Irvin
Sam Jethro
Jackie Robinson
Sam Thompson

I realized that voting starts tomorrow and I still wasn’t sure whether Buzz Arlett should rank number one or number 50 on my ballot, so I decided to do a very crude attempt at a major league equivalent (MLE) record for his career and calculate win shares (WS), based solely on the information that has been posted so far on this thread.

I’m well aware that there are some huge gaps in my analysis, which I’m hoping others can help fill. The biggest gap is the lack of information on the run environment in the minor leagues in which Arlett played. I believe that there is a book of old PCL statistics available, so if anyone happens to have a copy and can help fill in basic statistics like runs per game, either for the league as a whole or for the Oaks, I’d be appreciative. Also, I am using the method developed by Bill James in the 1985 Baseball Abstract to do the MLEs. I know there has been subsequent improvements in doing them, so methodological suggestions are also welcome.

Assumptions:

- Lacking any other information, I have assumed the run environment he faced was the same as the average of the two major leagues for each year. I am not aware of any information suggesting that the scoring environments of the PCL (1918-30), IL (1932-33), or AA (1934-35) were either higher or lower than the majors, nor am I aware of unusual park effects for Oakland, Baltimore, and Minneapolis during those periods, though I admit I am not well informed.

- I use the same adjustment for quality of competition that BJ used for modern Triple-A leagues – that is, 18 percent below major league level (in units of runs). My intuition suggests that the difference between the top minor leagues and the major leagues may have actually been narrower then, before the development of farm systems and with restrictions on the major league draft, but since I don’t have data to support my supposition, I am sticking with the standard 18 percent differential.

- Season length. I know that the PCL played a much longer season, but I don’t know exactly how much longer. To adjust for season length, I have adjusted all of Arlett’s counting statistics downward by 23 percent (= 1 - 154/200), except for 1918 when the PCL ended in mid-season due to the war. Again, if someone knows a better adjustment factor, I’d appreciate hearing about it.

- I am calculating MLEs for his entire career except for the partial season he spent in the Southern League in 1934 and the handful of games he played his last season in 1937. (And, of course, MLEs aren’t needed for 1931, which he spent with the Phillies). I know some voters have made the point that since most players pass through the minors for at least a year or two, and we don’t ordinarily give credit for these seasons, a couple of seasons should be trimmed off for purposes of HOM evaluation. I acknowledge the validity of this point, but go ahead and calculate MLEs for all his top-level minor league seasons, leaving it up to you to decide how much to trim off.

Pitching

This is pretty simple. I divided his ERA by .82 to adjust for quality of competition, and reduced his innings pitched by 23 percent (except 1918) for season length.

MLE Pitching
Year InnP ERA
1918 153.3 3.29
1919 268.0 3.66
1920 329.0 3.52
1921 245.7 5.33
1922 288.0 3.38
1923 96.3 7.02


Then I standardized to a run environment of 4.5 R/G and calculated pitching WS using the “short-form” method (the difference between his earned runs allowed and how many he would have allowed if his ERA had been 1.5 times the league average minus 1.00, which, in standardized units, is an ERA of 5.75).

MLE Pitching WS
1918 09.5
1919 14.9
1920 25.7
1921 07.4
1922 28.0
1923 -2.9
Tot 82.6

Batting

The BJ system is fairly straightforward.

Step 1. Adjust for the run environment (runs scored and allowed per game in the minor and major league environments). As already mentioned, lacking information I am assuming the two environments are the same.

Step 2. Adjust for the caliber of the competition, the assumed 18 percent difference in quality of play. Two factors based on this are m = .82, and M = sqrt (m) = .906.

Adjust the minor league statistics to ML levels using the following:

ML Hits = Hits * .98 * M
ML 2B = 2B * M
ML 3B = 3B * .85 * m
ML HR = HR * m
ML BB = BB * m

There’s also an adjustment for strikeouts, but since I don’t have the data and they don’t figure into the analysis, I skipped that step.

Walks are a bigger problem, since the PCL batting statistics in post # 8 by Kelly from SD do not include data on walks. I filled in walks data by applying the ratio of walks per out from Arlett’s 1931 ML season to each PCL season’s batting outs. I regard this estimate as conservative, since his minor league statistics for the IL and AA after 1931 suggest he was capable of drawing many more walks, as will be seen in a surge in the MLE walks estimates after 1931.

Step 3. Bill James next applies ML park adjustments, which I am skipping since I am comparing him to an “average” ML park of the period.

Step 4. The rest of the record is filled in using the minor league outs and games. Adding outs and MLE hits gives us MLE at-bats. I add the additional step of decreasing all PCL counting statistics (except for 1918) by 23 percent for the difference in schedule length. Batting average and slugging percentage are then easily calculated, and OBA can be approximated as hits plus walks divided by at-bats plus walks. I also add runs created, based on the basic formula.

MLE Batting
Year Gms ABs RNs Hit 2B 3B HR RBI BB AVG OBA SLG RC
1918 026 069 007 013 04 00 01 007 09 .188 .282 .290 006
1919 043 104 009 027 05 01 01 012 12 .260 .336 .356 012
1920 049 132 015 031 01 02 03 016 16 .235 .318 .341 014
1921 049 096 008 019 03 01 02 009 12 .198 .287 .313 009
1922 057 130 015 029 06 02 03 013 16 .223 .308 .369 015
1923 115 330 048 100 22 03 12 064 36 .303 .372 .497 061
1924 149 518 077 156 40 10 21 092 57 .301 .370 .539 103
1925 146 526 076 167 34 07 16 092 56 .317 .383 .500 101
1926 149 491 088 174 36 09 22 088 50 .354 .414 .599 122
1927 144 487 077 158 38 04 19 078 52 .324 .390 .536 102
1928 146 414 070 140 33 02 16 071 43 .338 .400 .543 090
1929 154 533 092 184 49 04 25 119 55 .345 .406 .593 128
1930 136 457 083 152 40 04 20 090 48 .333 .396 .569 103
1931 121 418 065 131 26 07 18 072 45 .313 .387 .538 086
1932 147 496 116 155 30 03 44 118 92 .313 .420 .651 136
1933 159 511 111 162 36 02 32 120 93 .317 .422 .583 126
1934 116 415 087 122 29 01 34 108 75 .294 .402 .614 103
1935 122 408 074 136 24 01 21 083 49 .333 .405 .551 091
1936 050 186 045 054 09 03 12 043 34 .290 .400 .565 042
Total 2078 6721 1163 2110 465 66 322 1295 850 .314 .391 .546 1450

Next, I convert the runs created to a standard 4.5 R/G environment, then calculate Batting WS using the short-form method (that is, runs created minus outs-divided-by-12, divided by 3; note – for pitchers, in the short-form method one does not subtract the outs-divided-by-12; I used the pitchers version for 1918-22).

MLE Batting WS
1918 02.5
1919 04.6
1920 04.8
1921 02.8
1922 04.6
1923 12.6
1924 22.4
1925 19.6
1926 30.7
1927 23.1
1928 20.9
1929 27.3
1930 19.4
1931 ----
1932 32.0
1933 32.5
1934 23.4
1935 20.3
1936 08.5
Total 330.8

Fielding
The short biography by Bill James in The Baseball Book 1991 says he “ran well, and had a powerful throwing arm.” On the other hand, I’ve seen a couple of comments on the Internet suggesting that Arlett may have been a below-average fielder. For 1931 BP shows him with -7 for FRAA. For now, I am sticking with the standard short-form value for fielding WS, which is one WS per 48 games in the outfield - you can adjust these estimates downward if you think he was well below average. For 1923 (the season he converted from pitcher to outfielder), I estimated his games in the outfield as total games minus games pitched. For the other years I used all his games, which is probably a slight overestimate due to games appearing as a pinch hitter.

Fielding WS
1923 1.9
1924 3.1
1925 3.0
1926 3.1
1927 3.0
1928 3.0
1929 3.2
1930 2.8
1931 ---
1932 3.1
1933 3.3
1934 2.4
1935 2.5
1936 1.0
Total 38.2

Total WS

Adding everything up, Arlett appears to have been a very valuable player. Even if one chooses to delete, for example, his 1918-19 and 1936 seasons, it still adds up to a player with more than 400 career WS and 6 seasons with more than 30 WS (2 as a pitcher and 4 as an outfielder).

1918 12
1919 20
1920 31
1921 10
1922 33
1923 12
1924 25
1925 23
1926 34
1927 26
1928 24
1929 31
1930 22
1931 16
1932 35
1933 36
1934 26
1935 23
1936 10
Total 449

I’m not sure whether I’m confident enough in these numbers to place Arlett number one on my ballot, but he definitely looks like one of my top 3.

Comments?

A couple more comments.

In an unusual way, the fact that Arlett was both a pitcher and a position player for the Oaks makes me more confident of these estimates for his PCL years. That is because any errors due to assuming the wrong PCL run environment would be at least partly offsetting. For example, if the actual run environment in Oakland were lower scoring than I've assumed, that would mean he wasn't quite as good a pitcher, but was actually a better hitter than I've calculated. Conversely, if the run environment were higher, the opposite would be true. The errors for the PCL years would tend to offset unless the run environment in Oakland relative to the majors changed between his pitching years and his years as an outfielder (for example, if they moved into a new stadiium sometime around 1923).

The numbers I am most suspicious of are for 1932 and 1933, when he played for Baltimore. The increase in MLE HRs suggests that a park effect may have been in play, though Buzz was still young enough that it could have been a simple (non-steroid induced) surge in home run power. Those seasons aren't that far out of line from the rest of his career to be implausible, but you may want to treat them with a little caution.

In 1933, Baltimore scored 962 runs in 166 games, which was 2nd in league.

They hist 180 HR's, which was by far first in league, so it does look like a hitter friendly park.

To add to what Gary posted, in 1933 Arlett was HBP 9 times and had 3 SH.

His Fielding for 1933:
G-147, PO-267, A-13, E-12, DP-1. Looks like only 2 regular outfielders had a worse fielding % than Arlett.

1929 PCL OF Fielding:

Jigger Statz, G-196, PO-517, A-22, E-17, Pct-.971
Smead Jolly, G-199, PO-319, A-37, E-12, Pct-.967
Buzz Arlett, G-115, PO-226, A-9, E-11, Pct -.965

There were about 4 regular outfielders with a worse fielding % than Arlett.

1933 Oakland Offense
G-202, Runs-1034. Oakland was 5th in runs and home runs, so appears to be a somewhat neutral park...

That should be 1929 Oakland, NOT 1933.

1928:
Oakland G-192, R-863, 5th in runs, 6th in HR's.

OF Fielding:
Buzz Arlett, G-154, PO-296, A-17, E-12, Pct-.963.
In bottom 3rd of regular OF for Fielding %.

Great stuff Brent.

I think a very conservative trimming would be to cut all of the pitching (based on Brent's post #2), and start his MLB career in 1924. If your estimates are correct, that leaves him as a 335 WS guy.

I'd really like to see actual numbers of scheduled games, and something that tells what the run environments of his leagues were, before I'm willing to vote based on this information though.

This is great stuff though, thanks for doing the work!

449 WS for Buzz Arlett (whose OPS+ of 138 in 1931 is perfectly OK for a poorly fielding outfielder, but not at all special) looks about as overstaed as the NL figures, and gives me further confidence that the translation process is fatally flawed, with at least a 20% bias to the upside. We've been over-rating NLers so far, and have probably elected a couple we shouldn't have; I for one will cease doing so going forward.

Brent, many thanks for the analysis. Without attempting to re-work all you did, one item immediately jumped out at me:

year WS
1926 34
1927 26
1928 24
1929 31
1930 22
1931 16
1932 35
1933 36
1934 26

The one major league season Arlett had was clearly the Worst by these projections. That makes me real queasy that he really would have racked up 30 WS per year had he played MLB the other seasons.

The one major league season Arlett had was clearly the Worst by these projections. That makes me real queasy that he really would have racked up 30 WS per year had he played MLB the other seasons.

Same here. I can't place him on my ballot just yet.

I think we can safely say that he was a very good player who should have had a long major league career, though.

TomH, very good catch. The REAL 16 surrounded by all those "conversion" 25s and 30s. I'm sorry, that Arlett conversion is now Exhibit A in my STRONG belief that we are WAY overconverting the Negro leaguers. My 15% discount of the conversions will henceforth become 20-25%.

Having said that, I've just started on '43 and put Oscar #1 -- there were truly great NL players, just nothing like as many of them as people are claiming. Beckwith in particular looks WAY overvalued.

The one major league season Arlett had was clearly the Worst by these projections. That makes me real queasy that he really would have racked up 30 WS per year had he played MLB the other seasons.

Two factors to bear in mind when evaluating this discrepancy:

1) Players jumping from one level of competition to another typically require an adjustment period of at least one season. Most Negro-League players who entered the majors directly from the NeL hit very poorly at first, but then made the adjustment and rose to the level we would expect in later seasons. It seems plausible to infer that we are seeing an adjustment season here, but with nothing else to follow.

2) Arlett played only 121 games that year (14 as a ph), which is much lower than in the surrounding seasons. If you project his 16 win shares to 146 games, it rises to 19, which is closer to his less productive PCL seasons.

I have no idea what to make of Arlett as a candidate at present, and I think a cautious approach is indicated until we have a) a better idea of the offensive level in the PCL vs. the majors and b) more data points for checking the conversion ratio.

I'd also be interested to see how the short-form projections that Brent is using compare to the projections one would come up with by matching Arlett's seasons to their closest major-league equivalents by raw stats and finding the ws from there. That's how I do the NeL projections for batting.

Speaking of NeL conversions, I'm home and gathering my scattered data on the subject, which I hope to post by the end of the day.

Karlmagnus, two words that you should keep in mind before judging all NeL to ML conversions on the basis of Buzz Arlett's 1931 season: SAMPLE SIZE.

Chris, I agree, and I need to be as careful with discounting as you are projecting, and not over-discount. All the same, the "buzz" about Arlett was that his 1931 proved him to be a near-HOF'er, so it's a little glib then to discount it as an off-season.

In my view, Arlett was a good ML player, but not close to HOF/HOM, but conversions tend to be substantially over-optimistic -- also WS and WARP are dodgy metrics anyway, and can be 25% out either way.

Charleston's in my PHOM, no question, but I will need a lot of convincing on Foster, Johnson and Lundy.

Bill Foster has a very strong reputation. I have no idea what the numbers will really show, so I don't know whether I'll be trying to convince you or not.

Few HoM voters will be convinced by Judy Johnson's case; I think it unlikely that he will do as well as Pie Traynor in the balloting.

Dick Lundy, from his raw stats, appears to be on the border. I don't know yet whether there's a convincing case to be made or not.

Brent--great work, really interesting. I just now realized I had enough information to come up with league averages for Arlett's years in the IL and AA:

Year-Ave-OBA-SLG-R/G
1932-284-350-425-5.21
1933-279-351-402-4.89
1934-291-351-413-5.44
1935-293-353-415-5.51
1936-295-350-431-5.59

(Again, the OBA is (H+W)/(AB+W).)

in 1932 Baltimore led the IL in runs (1036) and home runs (232). I don't have pitching totals, but looking at the ERAs for their pitchers, it looks to me like their team ERA was at best 6th in the league (more likely 7th). They finished a distant second (93-74).

In 1933, as KJOK pointed out, Baltimore finished second in runs and first in home runs, and my sense of their ERAs is that the team may well have finished dead last in the league in ERA. The team finished third, 84-80.

In 1934, Minneapolis led the AA in runs and home runs (and BA and doubles and walks), and their ERAs were not too great. Again, I don't have pitching team totals, but their ERAs do NOT look like the best in the league (maybe somewhere in the middle?). Minneapolis finished first, 85-64.

In 1935, Minneapolis again led the league in runs and home runs. At least three teams, possibly four, look to have had better ERAs. Again the Millers won the pennant, at 91-63.

In 1936, again Minneapolis led in runs and home runs (212-164 over Milwaukee), and their ERA may have been 7th or 8th. The team finished fifth, 78-76.

In short, both Baltimore and Minneapolis seem to have been definite hitters' parks during Arlett's time there, and both the IL and the AA had higher offensive levels than the majors.

Unfortunately, my PCL data is very spotty--only triple crown numbers for individual players, no team numbers. If anybody has guides for those years, maybe they would help...?

1929 PCL Offense:


1929

TeamGABRHSBHR3B2BSHRBIAVE
Mission2037087122222641561704339820911100.319
San Fransisco202720412132261641934541416510870.314
Hollywood2056892123321421111535038331411520.311
Los Angeles2056899121720921931804438826211040.303
Oakland202696110342099107154643722409560.302
Sacramento2047002100820697095513872489180.295
Portland2026779932195313997553362408610.288
Seattle20569509151955132103663151878230.281

TOTAL16285577487741683597211454182993186580110.302
</pre>

Thanks to everyone for your comments! I knew my first try at MLEs would have a lot of flaws, but I plan to do a second take (maybe tomorrow), incorporating the information you've been able to provide.

KJOK and Gary A: Thank you very much for the data. Of course, I'd always like more.

KJOK, do you happen to have runs allowed (or ERA) by Oakland for 1928 and 29? If not, would you have their W-L record? (With runs scored and allowed, I could do an inverse Pythagorean calculation to estimate runs allowed.)

Gary A, you mention the rank of Minneapolis in runs, do you have the actual numbers available? Again, even without runs allowed, I could do an inverse Pythag to estimate runs allowed.

Of course I'd really like a couple more data points on the PCL for the early 20s, but I'll work with whatever I can get.

I'll also lower my estimate of his fielding WS in my next run.

By the way, here's a comment by Bill James from an article on "Minor League Baseball Stars" in the 1986 BJHBA (p. 84): "The 'major league'writers put these players down by exaggerating their shortcomings as baserunners and fielders. One case in which this seems particularly glaring is that of Buzz Arlett, possibly the best of the minor league stars... Arlett, too, developed that bad-glove reputation, because it helped major league writers explain why he wasn't in the major leagues. In all likelihood he was a very good fielder... he had an excellent arm, and though a huge man and carrying a little extra weight that summer, he was quite mobile. There is no evidence at all that he deserved the bad-glove reputation that he was hung with." I don't believe BJ included that article in the NBJHBA, nor do I know if his WS analysis led him to re-think his conclusion. But while I'll agree that he may have been below average, I really don't think he was in the same class as Greg Luzinski.

We've got plenty of time to work on Arlett, since he's obviously not going in this "year" (nor should he!) But I think Arlett is a good example that several truly great ballplayers spent most or all of their careers in the high minor leagues, at least during the pre-1935 period.

For some reason the numbers won't format, so I'll upload the PCL totals for 1928 & 1929 to the HOM egroup if anyone wants to take a look at them...

KJOK, do you happen to have runs allowed (or ERA) by Oakland for 1928 and 29? If not, would you have their W-L record? (With runs scored and allowed, I could do an inverse Pythagorean calculation to estimate runs allowed.)

ERA isn't available unles you want to add all the team pitchers up individually.

In 1928, Oakland was 91-100
In 1929, Oakland was 111-91

Thanks!

If there was a minor league player discussed here that I would support, I think it would be Ike Boone, but I'm not gonna support him either. I just can't or at least don't believe right now that the MLs were that inefficient at evaluating talent even then. I mean, Arlett maybe coulda had a ten year ML career and it looks like maybe he coulda hit 300 HR and driven in 1000 and hit .333. But that to me is giving him the benefit of the doubt. I just don't see him as Chuck Klein or Hack Wilson, much less Al Simmons or Goose Goslin or Mel Ott or... or....

Minneapolis led the AA in runs scored every year from 1928 to 1937.

Here are Minneapolis team totals and league totals, 1934-36:

Year-Gms--ABs--Rns---Hits--Dbls--Tpls-HRs---BBs
1934-150-05448-1012-01676--0309--050--176--0603
AA---612-43546-6662-12685--2080--592--699--3981
1935-154-05430-0934-01601--0284--044--191--0573
AA---617-44088-6804-12905--2121--573--704--4108
1936-158-05758-1008-01703--0296--060--212--0527
AA---622-44671-6953-13159--2160--663--869--3781

Thanks for the data, GaryA!

Sunnyday2 wrote:

I just can't or at least don't believe right now that the MLs were that inefficient at evaluating talent even then.

I don't think efficiency in evaluating talent was the issue at all. I believe major league teams wanted Arlett (and other minor league stars), but in the early 1920s (the early part of his career) there were simply no mechanisms in place to force Oakland to surrender him to the major leagues. He was playing for an independent team in an independent league that valued him greatly and didn't want to give him up.

I wish copyright laws didn't prohibit me from posting the entire 1986 Bill James article on minor league stars that I mentioned earlier, but here is a quote:

"The minor leagues as they existed sixy to seventy years ago were something more like today's Mexican League, or perhaps Japanese baseball, except that rather than operating in another country in a foreign language, they operated in the hinterlands of the United States. They were independent. [emphasis in original] My experience has been that it is difficult to get people to internalize this concept to the point that they can stop coloring their perceptions of what happened then with modern notions of what a 'minor league' baseball team is, notions that have no relevance to 1922. My experience has been that if you tell somebody about the brilliant minor league career of Ox Eckhardt, who had a career batting average of .367, they will say, 'Wow, why didn't he ever get a chance?' No, no--it's not like that. He's not a player who might have done things if he'd had the chance; he's a player who did things. He played baseball. He made a good living playing baseball. His picture was on baseball cards; he was a local celebrity. The fact that he did these things in a league that was not the American League or the National League is important in its own way, but it doesn't make the things he did unreal or meaningless, as it would today; he did them."

For some reason the numbers won't format,

Don't use tabs in the layout. Format the table with spaces.