About Us  |  Analysis  |  Stats  |  Stuff  |  News
Baseball Primer Sporting News Fantasy Baseball - Get in the Game Today!
Baseball Primer Sections

  You are here > Baseball Primer > The Hall of Merit

The Hall of Merit A Look at Baseball's All-Time Best

Start and McVey (July 11, 2003)

A place to focus the discussion for them.
--posted by Joe Dimino at 11:53 AM EDT


Discussion

Statements posted here are those of our readers. We ask that posters follow our submission policy. Please report any inappropriate comments.

Posted 7:35 p.m., July 11, 2003 (#1) - Jeff M
  I'll pitch in b/c it is fun. I'm not a friend or enemy of either and I'm not trying to sell one candidate over the other. Just putting down my thoughts. I currently have McVey a little bit ahead of Start.

1. McVey had a couple of dominant seasons; Start was very steady. Sounds like a peak argument, but I give McVey a nod, because for a couple of years, he was the best hitter in baseball. Some years he was nearly twice as good as an average player. Start was very good, but not quite as dominant in the big years. There's something to be said for being the best player in baseball, because not many can say it.

2. In the years in which they both played, McVey was better in all but one year in my opinion.

3. I have McVey as about 60% better than the league and Start about 40% better than the league over their careers. Start obviously played a lot longer, so this may be a wash.

4. Obviously we don't have Win Shares for the NA, but Start had two sub-par NA years and overall, his NA years were some of his worst. McVey's best years were in the NA. I think Cal's Win Shares for those years would push him ahead of Start in that category, particularly since Start was a mature player during the NA years and Cal was pretty young. Using WARP1 and adjusting for season length (ignoring the WARP2 timeline adjustment, which I don't understand), Cal would have 52.7 in the NA and Start would have 32.3 in the NA. Not sure how it would translate to WS, but it would be significant.

5. My Pennants Added calculations come out slightly different than Joe's, but Cal would be significantly stronger in this category if my assumptions in #4 above are correct about WS.

6. I see no significant defensive distinctions. In addition to 1b, Cal had 50 games at catcher and could really play anywhere. Start was the first to play off the bag at first.

7. Start is better overall in WARP1 adjusted for season length, primarily because he played longer. On a per season basis, McVey is at 8.6 and Start is at 6.8. Same argument as #3 above, I guess.

8. Start obviously had some prime years that are not documented. I can't make up stats for him, so when I get to my final grade of the players, I give Start a 10% bonus. That may be too low, or too high. Who knows? My main problem with pre-1871 is that I don't know who the competition was. I suspect it wasn't all that good. Although Cal also played before 1871, I don't give him any extra credit because he was just a kid in the pre-1871 years and I have no idea what he did. I don't give credit for the later west coast years b/c I don't give credit for playing in other leagues when there is a perfectly good major league to play in.

I think it is a pretty close call, but I take McVey by a nose.

Posted 8:53 p.m., July 11, 2003 (#2) - Jeff
  McVey played for the best team in America, the Cincinnati Red Stockings, in '69-'70. You gotta give him credit for that! Two more years as one of the elite!

Posted 2:46 a.m., July 12, 2003 (#3) - Cal McVey
  I think it is a pretty close call, but I take McVey by a nose.

You lay a hand on it and you'll feel my fist on yours!

Posted 7:29 p.m., July 21, 2003 (#4) - TomH
  McVey, the tall tower (Sears) of a great peak, versus Start, the wide office bldg (Pentagon) of a lengthy career; which is better in terms of area?
Per season: WSper162g
McVey 33.0 (1876-79)
Start. 25.2 (1876-86)
McVey was slightly better in the NA than his 4 yrs in the NL, and Start was somewhat worse; my guess for including the NA would be
Per season: WSper162g
McVey 34.0 (1871-1879) age 20-28
Start. 24.6 (1871-1886) age 28-43
So, McVey is 38% "better". But, if one wishes to argue (and I do) that WS baseline is too low, and that about 10 WSper162g would be a reasonable replacement level, then we have McVey at 24.0 and Start at 14.6 "WSper162g above decent-level replacement player". Now, McVey is 64% "better".
By WARP3, McVey is at 44 wins above replacement in 9 years (4.9 per year). Start is 50 WARP(3) in 16 years (3.1 per year). Again, McVey is 56% more valuable per year. This jives with McVey's advantage in EqA (300 to 269 when adjusted for timeline; larger if unadjusted), and his ability to play positions other than first base (hey, if I had Deacon White on my team, I might have moved Cal to 1B as well).

In their undocumented years, Start is generally recognized as keeping his prime in the ages 18-27 that he had later. Whether McVey was as good from ages 29-40 is questionable, although the fact that he played until age 40 at least suggests he had not lost too much. How much do I have to dock McVey's undocumented ears, relative to Start, to make up the difference?

Start's whole career is 26-ish years. McVey's is about 20. That's a 30% advantage for Start. If McVey is around 60% more valuable than Start per documented season, Start needs a large advantage for the UNdocumented portion of his career over Cal's to make up the difference. I could pretend it's something like this, using a ficitious 'wins above mediocre player' estimate:

Start 16 doc yrs @ 3 wins a year + 10 other years @ 3 W/yr = 16*3+10*3=78 wins
McVey. 9 doc yrs @ 5 wins a year + 11 other years @ 3 W/yr = 9*5+11*3=78 wins

Voila, they are even! I merely have to assume that Start was as good 18-27 as we was later, and McVey lost 40% of his value at later ages 40 versus his ages 20-28. Are those both possible? Sure. But I suspect the discount would not be as strong.

McVey is arguably the best hitter we have; his EqA and OWP match well with Stovey, Thompson, and Browning (discounting Pete a bit for the AA). He had a long career, and had defensive value as well. The more I look, the more I like. I am not down on Start; I hope he eventually gets honored in the HoM. But I've become in 4 short 'years' a true FOCM.

Posted 1:39 a.m., July 22, 2003 (#5) - Joe Dimino (e-mail)
  "although the fact that he played until age 40 at least suggests he had not lost too much. How much do I have to dock McVey's undocumented ears, relative to Start, to make up the difference"

I'm a friend of Cal's, but I would read too much into the fact that he played until he was 40. At the SABR convention a friend of mine purchased a minor league encylopedia type book (the one Lloyd Johnson wrote in the early 1990s) and anyone who was worth his salt went and played into his early 40s. It was amazing, I had no idea that guys like Dom DiMaggio went and played for several years after "retiring" from the majors.

I don't mean that to disparage Cal McVey, I give him credit for his post-NL play. But I don't think we can just assume his skills didn't diminish too much because he played until he was 40. Plus McVey was slipping the last few years in the NL, still good, but 151, 143, 132 OPS+; WARP3 8.5, 4.4, 2.2, 3.0, 1.4.

Again, I really like Cal, but we should be really careful with his post-NL credit.

Posted 9:15 a.m., July 22, 2003 (#6) - Chris Cobb
  A note on Cal's declining WARPs -- part of the decline is attributable to the fact that he pitched a small number of innings very badly in 77 and 79, for which WARP rates him a 11 and 13 runs below replacement level (he loses a third of his value in 1879 for pitching 14 innings??), and that seasons were _very_ short in some of those years, for which WARP3 may not fully adjust. Adjusted WARP1 1875-1879 without the pitching deduction (I remove the pitching bonus he got in 1876 also) looks like this: 21.3, 10.5, 10.2, 6.48, 10.0 . He has a bad year in 1878, which is partly attributable in WARP1 to his playing his only season at third base, and not playing it well. Otherwise his value is pretty constant across the NL years. It's not in any way a match for his monster years in the NA, but 10 adj. WARP is still a very high level of play.

Posted 9:29 a.m., July 22, 2003 (#7) - Chris Cobb
  Oops. McVey's adj. WARP 1 for 1875 should be 19.6, not 21.3.

Posted 10:10 a.m., July 22, 2003 (#8) - John Murphy
  A note on Cal's declining WARPs -- part of the decline is attributable to the fact that he pitched a small number of innings very badly in 77 and 79, for which WARP rates him a 11 and 13 runs below replacement level (he loses a third of his value in 1879 for pitching 14 innings??),

Which makes no sense, in my opinion. It's one thing not to give him credit for his pitching, but it's another thing to subtract it from his other accomplishments.

Posted 12:27 p.m., July 22, 2003 (#9) - jimd
  Which makes no sense, in my opinion. It's one thing not to give him credit for his pitching, but it's another thing to subtract it from his other accomplishments.

I don't understand this. This is a record of BP's estimates of the value of what McVey did on the field. Yes, the pitching was optional, but he did it. I understand that sometimes position players go into blowouts for the fun of it, and those results have no practical impact, but this isn't that kind of a situation. He pitched 176 innings. Because it was optional, I suppose one can pretend it didn't happen, judging it irrelevant to his HOM case, but that's an editorial choice, one that BP should not be making.

Posted 1:09 p.m., July 22, 2003 (#10) - OCF
  I have Start ahead of McVey, but neither one in the top reaches of my ballot. I know we're supposed to be honoring the best from each era without prejudice for or against any one time, but I'm developing many misgivings about the National Association.

Take 1875, the year in which McVey recorded his highest OPS+. Here are the runs scored per game, team-by-team:

10.13, 9.08, 6.71, 6.48, 5.51, 5.49, 5.00, 4.62, 3.82, 3.62, 3.46, 3.16, 3.00. (The 10.13 is the Red Stockings.)

The runs allowed, team-by-team:

3.99, 4.18, 5.22, 5.27, 5.37, 5.99, 6.03, 6.77, 8.45, 8.47, 9.86, 9.95, 12.07. (This time it's the 4.18 that's Boston. Hartford, with Candy Cummings and Tommy Bond, may have had better pitching than Boston with Spalding.)

The history of the NA seems to be the Red Stockings with their perennial All-Star team, a handful of other teams trying to compete with them, and a bunch of teams going 3-26 before dropping out of the league. What's the meaning of "average" player in that environment? How do you define "replacement level player" in a league infested with replacement level teams? Players on the Red Stockings derived a considerable benefit just from not having to play their teammates.

McVey was a good player, but you can explain the success of the Red Stockings without needing him as part of the explanation, and the year he spend on another team didn't change much.

The impression I'm getting is of a league whose talent pool wasn't deep enough to properly stock an 8-team or so "major" league. There probably were players with high level talent who simply never came to the attention of the NA teams; of the ones who did, they mostly got swept up by the best teams.

As for giving either McVey or Spalding much credit for play before 1871, they were teenagers. I do buy the idea of pre-1871 credit for Start, who may (but we don't know) have had his peak back then. I also don't see giving McVey credit after 1879. He left the majors. So he kept playing baseball somewhere - so did many others who left the majors. That leaves in McVey's case a 9-year career. That won't stand up to Start's longevity.

Posted 1:36 p.m., July 22, 2003 (#11) - jimd
  Under NA rules, any team willing to pay the entry fee could play. That highly democratic system didn't work too well, particularly in 1872 and 1875, when there were quite a few teams in over their heads. However, there were 7 solid teams (out of 13) in 1875, 6 of them played in the NL of 1876 (the lesser of the two good Philly teams was not invited); the upper half of the league was as good as the NL of the following few years, and the lower half didn't last very long, for the most part, before dropping out.

McVey refused to play in the NL under the reserve rule, which was new in 1880. He wasn't the only one who had problems with it; George Wright and Deacon White held out most of 1880 and I'm sure there were others less notable.

Posted 2:13 p.m., July 22, 2003 (#12) - John Murphy
  He pitched 176 innings. Because it was optional, I suppose one can pretend it didn't happen, judging it irrelevant to his HOM case, but that's an editorial choice, one that BP should not be making.

I'm not pretending it didn't happen. I have a problem with linear weight-type formulas (though WARP is not really the same as TPI).

I believe if you play, you have value. That value might be microscopic if your performance is poor, but it's still there.

Posted 2:14 p.m., July 22, 2003 (#13) - John Murphy
  McVey refused to play in the NL under the reserve rule

I didn't know that. Good info, Jim.

Posted 2:24 p.m., July 22, 2003 (#14) - OCF
  McVey refused to play in the NL under the reserve rule

So he's on the same list that has, among others, Amos Rusie and Curt Flood: Players who challenged the power of the owners and, for the most part, lost their personal battles.

Posted 3:14 p.m., July 22, 2003 (#15) - Andrew Siegel
  Not exactly, OCF. As I read the sources, McVey was much more comfortable in the "West" broadly defined than in the east anyway. He had moved from team to team a bit, hadn't found a situation that was really to his liking in the NL, and had other interests (family and financial) pulling him back West. The reserve clause was simply the last straw. His decision wasn't "I'm going to hold myself out of the major leagues in the hopes of getting them to drop this damn reserve clause" but rather "You know what, I'm going to sign with X team in Y league rather than with Z team in the NL, because it will be a better experience for me overall -- closer to home, more control over my own playing conditions, more gentlemanly, similar money, competition that's good enough to challenge me." McVey's decision was similar to the decisions players made all the time in the 1860s and early 1870s as to who to play for or to the decisions a number of players made to join the International League in the late 1870s. By modern standards, it seems strange to voluntarily play "minor league" ball, but that's not the way McVey likely thought of it. He was simply picking a league and a team that was best for him -- the significant but not overwhelming differences in quality between the NL and the top "minor" leagues was one factor on the table but often paled in comparison to other factors. I think there were probably dozens of guys like McVey making similar decisions. The only reason we know about him is that he initially chose to play a significant number of years with NA/NL teams before choosing a different venue.

Posted 3:51 p.m., July 22, 2003 (#16) - Howie Menckel
  Well put, Andrew.
We're trying to find the best baseball players of each era, and there is little reason to believe that McVey wasn't one of them in his time. I assume we won't get hung up on incomplete "major league" stats with the Negro Leaguers as they come along, so why do so with these guys?

Posted 4:04 p.m., July 22, 2003 (#17) - jimd
  The reserve rule began after the 1879 season as a "gentleman's agreement" amongst the NL owners; they submitted a list of N players (2 or 3; I forget exactly) to the league and the other owners agreed to not sign them. Some of the players were quite proud to have been designated as "reserved", but not after they received their next contract offer. The reserve lists grew during the 1880's until they covered the entire team. I don't know when the various leagues agreed to respect each other's reserve arrangements, or when the infamous "reserve clause" was inserted into the player contracts, but this whole concept is new and controversial throughout the 1880's, culminating in the Brotherhood of Baseball Players (Player's League) in 1890 and the compete defeat of the young players union.

I think Andrew has it right there. McVey just didn't like the way he was treated and went elsewhere. The "minor leagues" weren't like today; they were completely independent. The choice of playing in the NL vs the IA was maybe kinda like the choice to sign with a big market east/west coast club vs a small market team today; the IA markets were smaller, and so were the team salary budgets, but they were playing for a real pennant too, and signing the best players they could afford.

Posted 11:28 p.m., July 22, 2003 (#18) - Andrew Siegel
  Someone above has McVey at 35 WS/year for 9 years. That's 315 WS. Can anyone come up with any (position) player in the history of the game who earned 315 WS in a 9 year period who we are not going to elect? For that matter, can anyone name a position player who earned 300 WS in a 10 year period who we are going to skip over? If you credit those numbers, McVey put up an HoM career even without any credit for his pre-NA or post-NL performance.

Posted 12:32 a.m., July 23, 2003 (#19) - Chris Cobb
  Someone above has McVey at 35 WS/year for 9 years. That's 315 WS. Can anyone come up with any (position) player in the history of the game who earned 315 WS in a 9 year period who we are not going to elect? For that matter, can anyone name a position player who earned 300 WS in a 10 year period who we are going to skip over? If you credit those numbers, McVey put up an HoM career even without any credit for his pre-NA or post-NL performance.

As the "someone above," I should point out that at this moment I think the 315 WS number is most likely an overestimate, though not a gross one. I continue to work on an accurate conversion of WARP1 to win shares (when I'm not doing the work I'm paid to do :-) and I hope to have numbers on McVey that even the skeptical will credit in the next couple of days.

Posted 8:17 a.m., July 23, 2003 (#20) - TomH
  While awaiting Chris Cobb's detailed analysis, here's a simple sketch
McVey EqA NA McVey EqA NL
...1871 .365 ......1876 .309
...1872 .278 ......1877 .333
...1873 .336 ......1878 .304
...1874 .337 ......1879 .302
...1875 .361

On the 1B thread, McVey is listed at 33.0 WS per year for his 4 NL years...playing mostly 1B and some 3B. In the NA, he played mostly C, and hit better. He must be at LEAST 35 WS/year for those 5 years, maybe closer to 40. His WARP (adjusted for short schedule) are much better in the NA.
Question for the floor: Does anyone wish to make an argument that someone other than McVey is the highest quality position player we have? WS, WARP, OPS+/speed/defense all seem to give the same answer.

Posted 8:27 a.m., July 23, 2003 (#21) - Andrew Siegel
  I'm the biggest fan of McVey that we have (I'm the lone person who had him number 1 on the last ballot), but it is worth noting that some of his 1876-1879 WS are from his pitching. Since he didn't pitch much (if at all) in the NA, his NA WS probably aren't significantly higher than his NL WS and may even be slightly lower (even though he was clearly a better player).

Posted 11:11 a.m., July 23, 2003 (#22) - John C
  Tom H,
what do you mean by best? It seems he played well when he was in the league. But, it's only 9 years. In my mind, the short career limits his value, placing him below many on the ballot. He did have a great peak, but so did Pete Browning.

So, do you mean best by career value, or by talent, or by per season impact?

Posted 12:21 p.m., July 23, 2003 (#23) - John Murphy
  So, do you mean best by career value, or by talent, or by per season impact?

I think what Tom meant was that McVey was probably the best major league player at his position 7 or 8 eight times.

OTOH, Browning was the best three times (though close for a couple more).

Posted 12:25 p.m., July 23, 2003 (#24) - Howie Menckel
  JohnC,
I'll assume you weren't a big Amos Rusie fan, either?

Posted 12:38 p.m., July 23, 2003 (#25) - John C
  Okay, that makes sense. Was McVey really best at his position 7 or 8 times? I haven't looked at him that way yet. I thought he was not quite as good as Deacon White for several years in the NA. I'll go back and look.

Howie: Actually, I had Rusie in my top 2. With Rusie, it was a matter of confidence. I was confident that he belonged in the HOM, while I had not yet come to that conclusion on Spalding, Radbourn, or Galvin.

Posted 1:10 p.m., July 23, 2003 (#26) - TomH
  What I meant was, McVey's only documented 9 years appear to be better than anyone else's (not yet elected) best 9-year run anywhere in their career. They may even stand up to anyone else's best 9 non-consecutive years (pick any 9).

Posted 6:21 p.m., July 23, 2003 (#27) - Chris Cobb
  As promised here are career Win Shares for Cal McVey, as derived from WARP1

Year -- Total -- Batting / Fielding -- adj. To 162 games (from)
1871 -- 9 -- 7.8/1.2 -- 48 (30 games)
1872 -- 7.6 -- 4.7/2.9 -- 24 (50 games)
1873 -- 9.9 -- 8.5/1.4 -- 27 (60 games)
1874 -- 16.6 -- 14.7/1.9 -- 38 (70 games)
1875 -- 25.2 -- 20.5/4.7 -- 51 (80 games)
1876 -- 11.6 -- 9.7/1.9 -- 31 (60 games)
1877 -- 11.2 -- 10.7/.5 -- 30 (60 games)
1878 -- 8.8 -- 8.6/.2 -- 22 (60 games)
1879 -- 12.3 -- 9.9/2.4 -- 24 (84 games)
Total -- 121.2 -- 95.1/17.1 -- 295

For 1876-1879, WS itself records 54 WS for McVey (16, 14, 11, 13), which adjust to 136. However, 22 of those are pitching WS, which WARP does not accept. Less the pitching WS, McVey's raw WS are 45, compared to 43.9 derived from WARP. I did this calculation to verify the accuracy of the method, employed on a season-by-season basis. Expanded, the differences come to 114 WS, 107 WARP. I suspect the difference arises from the very low fielding runs WARP gives McVey, which make the conversion into a rough estimate.

Ways of looking at the total
295 -- WARP-derived adj. career WS -- this number should be within 5% of the total that would be arrived at by calculating all McVey's WS directly from the data.
302 -- WARP-derived adj. Career WS 71-75 + adj. WS 76-79
324 -- Pitching added to 302
314 -- No pitching added, but fielding WS in the 302 figure adjusted upward by 30% (my standard pre-1893 adjustment)

I am amused that the estimate I find most accurate -- 314 -- almost exactly matches the rougher estimate that I set out to refine -- 315.

Having given the numbers, I'll explain the method, for those who want to know.

Battting Win Shares

Rather than starting from the WARP1 total, I started from BRAR. WS and WARP are in much closer agreement about batting value than about fielding value.

WARP1 seasonal totals seem to be calculated by summing all batting, fielding, and pitching runs above replacement and multiplying that total by about .103. So BRAR/10 gives a fair approximation of BWARP. If WS and WARP are right, then this should be true Batting WS = (BWARP X 3) + Batting Replacement Level. As far as I can tell, it is. If you divide BRAR by 10 and multiply by 3 for any player season or career, you'll get a number just a bit below batting win shares. That leaves the problem of replacement level. Since replacement level isn't entirely constant, and BRAR and batting WS are not perfectly consistent, this number isn't entirely fixable, but I've found by some trial and error that 3.5 WS / 162 games gives results that are accurate within 5% in all but one case I have tried (that's Sam Thompson -- grist for the mill!). For McVey, I found that replacement level for his 76-79 seasons, as calculable by subtracting his BWAR from his BWS was 3.52, so I used this number in calculations of batting WS for his 71-75 seasons. I am very confident, therefore, that the batting win shares are accurate within 5%. (A while ago I calculated batting WS by hand for the 1875 Red Stockings, so I had a hard number to check that season against. The WS calculated off of BWARP came to 21.5; the WS calculated by hand came to 20.5 -- 5% discrepancy. Since I had the hand-calculated WS number for 1875, I used it).

Fielding Win Shares

Since win shares and WARP do not agree about fielding value, there is no way to predict consistently the ratio of FRAR and fielding win shares. Win shares gives much lower credit for fielding, so I figured that a straight ratio of conversion for McVey's fielding win shares would give a conservative estimate of what actually calculating his fielding WS would reveal. McVey's FWAR for 76-79 are 3.5 . His fielding WS for this period are 4.86. The ratio between them is 1.39. For McVey's 71-75 seasons, I divided his FRAR by 10 and multiplied that number by 1.39. This gives a much lower estimate of McVey's fielding value than WARP does, but I figured that it would be a fairly accurate representation of the numbers that the WS system would actually produce. In rating McVey myself, I (as I have indicated above) multiply his fielding WS by 1.3.

Pitching Win Shares

Win Shares gives McVey a fair amount of credit for his pitching stints, 76-79. WARP evaluates his pitching during that period as well below replacement level. These assessments are not reconcilable. Fortunately, McVey pitched hardly at all prior to 1876, so there's no need to worry about converting his pitching record in that period into WS. If you believe what WS has to say about pitching, you can give McVey additional WS. If you don't, you can leave those out. If you believe WARP, you can dock McVey's career WS by 4 or so to account for his hurting his team by taking the mound.

Posted 7:25 p.m., July 23, 2003 (#28) - jimd
  Nice work, Chris. I used a much cruder technique to estimate his Win Shares, and have 310-315 penciled next to his name. Good to have that verified.

Posted 9:19 p.m., July 25, 2003 (#29) - Chris Cobb
  Having translated WARP to WS for McVey and Sutton, next up of the NA players on the ballot is Joe Start. Before I list the numbers, I'll say that they are not especially impressive. However, there is reason to believe that the translation does not fully capture his value for two reasons. First, WARP does not like Start's hitting as much as WS does, so it's harder than usual to approximate battting WS accurately. Second, no player I've studied loses as great a percentage of fielding value going from WARP to WS. I think the fielding translation below accurately represents what WS would find, based on its fielding WS for Start's NL career, but WARP sees him as _much, much_ better defensively. I'll say more about these issues after posting the numbers themselves, for those who are interested.

Joe Start NA win shares

bWS/fWS = total (games) --> adj. WS (162)
1871 4.2 / .34 = 4.5 (35) --> 21 (162)
1872 1.9 / 1 = 2.9 (55) --> 9
1873 2.2 / 1.6 = 3.8 (55) --> 11
1874 9.1 / 1.7 = 10.8 (65) --> 27
1875 9.2 / 1.7 = 10.9 (70) --> 25

Total adj. WS 71-75 -- 93
Total adj. WS 76-86 -- 243

Grand Total 336
Fielding adjusted 349

Adj. To 162 games, Season by Season (no fielding adjustment)

21, 9, 11, 27, 25, 22, 30, 27, 23, 23, 31, 27, 17, 20, 22, 1

Comments

Start's NA years don't add a tremendous amount to his HoM case: he definitely played better in his decade in the NL, as WS sees it. However, this calculation seems likely to underrate Start significantly, 10-15%. The evidence:

On offense, WARP does not like Start as much as WS. For most players, the "replacement level" calculated to make up the difference between 3 X bWARP and bWS is 3.5 to 3.7 WS. For Start, I had to set it at 4.7 WS, and that accepted a 5% shortfall. If I set the number any to make up the shortfall, then Start's peak years would become exaggerated.

On defense, WS doesn't give Start much credit, and I don't know why. Most players have at least as many fWS as fWARP (that is they have at least 1/10 as many fielding win shares as they have fielding runs above replacement level). Even the players whom WARP rates excessively highly (at least that's how many of us see it), like Bid McPhee and Charlie Bennett, have a fWS to fWARP ratio greater than 1. Start's ratio is .68. The disagreement between WARP and WS on Sam Thompson's fielding has been much discussed, but as a percentage, it is less severe than Start's (Thompson's ratio is .85) So Start would be a prime candidate for a larger than usual adjustment to his fielding win shares. I can see an argument for adjusting them upward by 60% or 80%, at least for the NA years, if not for his whole career. Folks who like to analyze raw fielding stats might find Start's worth puzzling over as well.

A 10-15% increase on 93 win shares isn't going to transform your view of Joe Start, but, from doing the translations, that's my estimate of what an accurate adjustment would be.

Posted 3:26 p.m., August 2, 2003 (#30) - Paul Wendt
  OCF (#10):
Players on the Red Stockings derived a considerable benefit just from not having to play their teammates.

Total Baseball "park factor" accounts for the teammate effect. I don't know about the measures used here.

jimd (#11):
the lower half didn't last very long, for the most part, before dropping out.

Some of those teams played very unbalanced schedules. Maryland, the second 1873 team in Baltimore, played 4 of its 6 games against Baltimore. In 1875, Centennial and Atlantic played 14/14 and 40/44 games against the East; Keokuk played 12/14 against the West.

Posted 3:31 p.m., August 2, 2003 (#31) - Paul Wendt
  jimd (#17):
I don't know when the various leagues agreed to respect each other's reserve arrangements

They established joint reserve arrangements, which is not the same as respecting each other's arrangements.

Offseason 1882-1883, for the three leagues who signed the Tripartite Agreement. Four leagues signed the first "National Agreement" one year later.

Posted 1:04 p.m., August 4, 2003 (#32) - jimd
  Total Baseball "park factor" accounts for the teammate effect. I don't know about the measures used here.

BP/WARP claims to adjust for the teammate effect. To the best of my knowledge, Win Shares does not.

Posted 1:37 p.m., August 4, 2003 (#33) - Marc
  Then there's the fact that from 1859-70, Joe Start almost surely had the greatest "career value" of any player in the game. By 1870, George Wright was likewise almost surely regarded as having the highest active peak (he was the best player on the best team in America all four years from 1866-67 and 1869-70) and Al Spalding of the Chicago Excelsiors was also very highly regarded. But Start's team, the Brooklyn Atlantics, had won the championship of the New York area 8 times in 11 years from 1859-69 and aside from G. Wright's teams (the Washington Nationals 1866-67 and Cincy Red Stockings 1869-70) it is doubtful that any other teams in America could have competed with the elite New York teams before the NA.

Start's career reminds me of Ernie Banks'--top 3 for 11 years, then rapidly falling to a barely above average rating. The difference is that Banks stayed there, but Start suddenly was reborn and had a second prime, not as a top 3 or 5 player perhaps but as a top 20, a well-above average, player. Thought about this way, Start would rate well above Banks overall--a longer early prime plus a later prime that Banks didn't have. Among the company of slick-fielding, non-HR hitting all-time great 1Bs, I think of Start in the company of Sisler and Terry, and he would outrank both of them as well for a longer prime and a longer career, but at a very generally similar level--i.e. among the top 3 to 5 players.

I was not a FOJS at the beginning of this process, BTW, but exploring the 1860s has made me a big fan. His post-'71 record alone just does not get at the essential Joe Start.

Posted 10:10 a.m., August 6, 2003 (#34) - Marc
  Thanks to my friend Paul Wendt for an update re. Joe Start.

I guessed that Start began play with the Atlantics in 1859. Paul tells me of two sources that say Start joined the Atlantics in 1861 and 1862. He had played with the Enterprise previously while one source says he broke in with the Pheonix in 1859.

Comment: My basic point that he began organized play in '59 and was the star of the Atlantics for an extended period of time appears to be accurate. The details are fuzzy as Paul's two sources do not entirely agree.

Also, I had a source that said the Atlantics won 8 pennants. This is not entirely documented. Rather, one source shows the Atlantics winning 4.5 pennants 1859-'65 (i.e winning cleanly in '59-'61-'64-'65). In 1860, the Atlantics and Excelsiors (with Jim Creighton) split two games and then the third ended in a riot, the captains agreeing to a draw. Paul does not provide data concerning pennants after 1865, perhaps because there is no detailed source. We do know of course that after 1865 the Washington Nationals (1866-67) and Cincy Red Stockings (1869-70) had the best claims to being the best teams in the game.

Comment: I probably overstated both the Atlantics dominance AND Start's role in it. Rather than 8 pennants, I see that 4.5 can be documented and rather than 8, I see Start involved in only 2 of the 4.5. Nevertheless, the point remains, Start was the star (1861-62 through 1870) of one of America's elite teams.

The only players with a comparable career (>10 years) as of '70 are Harry Wright and Dickey Pearce (and maybe Al Reach though I cannot corroborate at the moment), and I still find Start a little more prominent than either among elite teams. As for peak value, Creighton remains my #1, G. Wright #2 and after that Start is, again, a step ahead of Harry.

Posted 3:35 p.m., August 6, 2003 (#35) - Paul Wendt (e-mail)
  The Atlantics may have been the best team in the New York area (modern New York City) 8 or 9 times in 12 years, 1859-1870.

Bill Ryczek, When Johnny Came Sliding Home: ... 1865-1870, covers his time period in great detail. (In writing to Marc privately, I relied on his sketch of 1958-1864 in chapter 1.)

Beginning 1866 or 1867, best in New York was no longer clearly best. From Ryczek's account, it is clear that the New York area teams conspired to keep the pennant at home, if it was a pennant. There were strong challengers in Philadelphia, Cincinnati (1869-70), and Chicago (1870). Conspiracy was mainly at the expense of the Athletics, Phi. By 1869, New York clubs were no longer clearly in control. The Red Stockings, Cin, did not jump through New York hoops to make their case. Read the book.

George Wright played for the Nationals, based in Washington, only in 1867. That was their tour year; I wonder how much they used their Washington base. In 1868, he was the regular shortstop for the Unions of Morrisania, now in the Bronx, who benefited from conspiracy, at least in scheduling, to win the championship and hold it until the next season. (The championship could be won in mid-season, by winning two of three from the champion.)

Marshall Wright's book on the NABBP covers 1857-1870. It is a few year old but we do not yet have a good critical review. I hope for one before the New Year.

Paul Wendt
Chair, 19th Century Cmtee, SABR

Posted 3:55 p.m., August 6, 2003 (#36) - John Murphy
  Paul:

Do you happen to know of a shortstop before 1867 who was comparable to Dickey Pearce (offensively, defensively or all-around) at shortstop?

Posted 4:20 p.m., August 6, 2003 (#37) - Marc
  BTW, the Chicago Excelsiors were the team of Al Spalding and Ross Barnes. I am not aware that the Excelsiors had a strong claim to national honors--that is, I don't know who they played. But they clearly were an elite club.

I wonder what the story is concerning George Wright's recruitment by the Washington Nationals for their barnstorm tour? Harry had moved from New York to Cincy to play cricket (!) in '65 (and it appears he played cricket and not baseball that year). He went back to baseball in '66. I do not know if George was in New York or Cincy in '66 but he clearly already had a reputation. He would have been comparable to Dickey Pearce as early as '66, if only by inference.

Posted 5:37 p.m., August 6, 2003 (#38) - Jeff M
  Anyone know how many wicked googleys Harry Wright had in 1865?

http://www.kottke.org/02/11/021109a_wicked_goo.html

Posted 5:52 p.m., August 6, 2003 (#39) - Jeff M
  Anyone know how many wicked googleys Harry Wright had in 1865?

http://www.kottke.org/02/11/021109a_wicked_goo.html

Posted 6:25 p.m., August 6, 2003 (#40) - Jeff M
  Anyone know how many wicked googleys Harry Wright had in 1865?

http://www.kottke.org/02/11/021109a_wicked_goo.html

Posted 6:40 p.m., August 6, 2003 (#41) - sean gilman
  3? :)

Posted 7:00 p.m., August 6, 2003 (#42) - Jeff M
  Correct!

I'm not sure why it was posted three times. I wasn't even at the computer for the second two.

Posted 6:53 p.m., August 8, 2003 (#43) - Paul Wendt
  I don't know a top shortstop who was a contemporary of Pearce.

Bill Ryczek, When Johnny Came Sliding Home: ... 1865-1870 lists Forest City Rockford IL as a major club 1868-70, with Spalding and Barnes the regular P and SS. According to 19c Stars, they were teammates on that club 1866-70, age 16-20 for Barnes.

Chicago 1870. That is the White Stockings, not the Excelsiors. Ryczek does not list Excelsior Chicago as a major club but he does have 11 index entries for that club.

Posted 11:16 a.m., August 9, 2003 (#44) - John Murphy
  So Barnes and (of course) George Wright are the only shortstops that could be argued were better than Pearce during the sixties. So Dickey had at least ten seasons where he was the Big Kahuna at his position. Mighty impressive.

Thanks for answering my request, Paul!

Posted 1:52 p.m., August 9, 2003 (#45) - Jeff M
  John Murphy wrote: "Do you happen to know of a shortstop before 1867 who was comparable to Dickey Pearce (offensively, defensively or all-around) at shortstop?"

Let me re-phrase the question. Do you know of ANY shortstops other than Wright, Barnes and Pearce before 1867?

My point isn't that Pearce is unworthy -- because I don't know the answer to that question. My point is that we don't seem to know much about the universe of shortstops during that time period. That's why I have a hard time with Pearce's claim to the HOM. It may be some indication of his greatness that we know his name, but I don't know that for sure.

Plus, for those who believe in having each position equally-represented in the HOM (I'm not one of them), shortstops are already very well represented.

Posted 2:30 p.m., August 9, 2003 (#46) - Marc
  Well, Doc Adams played shortstop at least 1849-1862. That's not too shabby of a "career" and his team was among the elites, such as they were.

Posted 3:10 p.m., August 9, 2003 (#47) - Paul Wendt
  Bill Ryczek, When Johnny Came Sliding Home: ... 1865-1870 (McFarland, 1998) lists Forest City Rockford IL as a major team 1868-70 and lists Excelsior Chicago IL in 1868 only. And so it goes. The list of major teams is embodied in Appendix B, "Team Rosters", which literally lists the regular players by position for each major team.

Number of major teams, 1865-1870:
18 20 17 15 16 19.

Listed in all six seasons:
Athletic, Phi
Keystone, Phi
Mutual, NY
Atlantic, Bro
Eckford, Bro
Star, Bro
Lowell, Bos

Listed in five of six seasons:
National, Was - all but 1868
Union, Morrisania (now in the Bronx) - all but 1869
Union, Lansingburgh (Troy Haymakers) - all but 1865
Harvard, Cam - all but 1868

Listings for NA1871 clubs:
1865-70: Ath. Philadelphia, Mut. New York
1866-70: Troy Haymakers
1868-70: F.C. Rockford, Oly. Washington
1869-70: F.C. Cleveland
1870: Chicago White Stockings
none: Boston Red Stockings, Kek. Fort Wayne

As far as I know, the designation of major teams was Bill Ryczek's innovation. I suppose that his judgment was based on the log of games played and reported in the New York newspapers, especially the Clipper.

So I can look up the names of many shortstops.

Posted 3:25 p.m., August 9, 2003 (#48) - Paul Wendt
  we don't seem to know much about the universe of shortstops during that time period.

That's an understatement, both in its tentative wording (seem) and in its scope (shortstop).

Posted 6:46 p.m., August 9, 2003 (#49) - Jeff M
  "That's an understatement, both in its tentative wording (seem) and in its scope (shortstop)."

Yes. Was trying to be diplomatic.

Posted 7:49 p.m., August 9, 2003 (#50) - John Murphy
  Well, Doc Adams played shortstop at least 1849-1862. That's not too shabby of a "career" and his team was among the elites, such as they were.

I have Doc Adams as a definite Pioneer selection (he and Cartwright would be my first choices), but I have no knowledge of how he played or who he played against.

I would stick with the professionals over the amateurs, but a persuasive argument could sway me. Your Harry Wright argument has almost done that now.

Posted 9:55 p.m., August 9, 2003 (#51) - Marc
  You're right, John, we know Doc Adams played a lot of baseball, we know he was a very smart man, but as best as I can determine "history" only says he played, not that he was a "great" player. History says differently about H. Wright (and Creighton and Pearce and Start). They were great.

Posted 11:16 a.m., August 12, 2003 (#52) - Philip
  Jeff M wrote: “Plus, for those who believe in having each position equally-represented in the HOM (I'm not one of them), shortstops are already very well represented.”

I do believe that each position should be representated fairly equally. However, it could be possible that throughout history this is not the case for shortstops. There may be a larger proportion of shortstops who deserve a place in the HoM. The following example will show this, taking into account the defensive spectrum (I hope you can all follow):

I assume a player will play the toughest defensive position he can play well, so that his offensive contribution will be maximized. This can be seen as the difference between absolute offensive value and replacement level (which becomes lower as we move along the defensive spectrum). Ideally, a leftfielder would move to second base if he could play this position just as well, and a second baseman would become shortstop if he could play shortstop just as well, thereby maximizing his value. However, once a player becomes a shortstop he cannot move to a more valuable defensive position, because such a position does not exist (assuming the skill sets for catcher and pitcher are too different to convert to). A-Rod and Nomar have the offensive value that would merit a place in the lineup at first base, but their value increases tremendously by playing shortstop. By applying the assumption that players move as far to the right of the defensive spectrum as possible, the shortstop position gets crowded with superstars, at least more so than other positions. This effect may not be very large, but it could justify a relatively larger representation of shortstops in the HoM.

Posted 2:57 p.m., August 12, 2003 (#53) - OCF
  To add to what Philip is saying, the times around the turn of the century saw at least three players - good, long-career players - start out mostly as outfielders, and with other positions in between, wind up as shortstops. The three I know of are Hans Wagner, George Davis, and Bobby Wallace. The only thing I find curious is that it took so long to figure out that these three could play shortstop. Philip used Rodriguez and Garciaparra as examples, but we know that the best example of the point he was trying to make is Wagner. On the other hand, I don't think we would have elected Glasscock had Pebbly Jack been a 1B/RF with the same offense.

Posted 10:46 a.m., August 13, 2003 (#54) - Marc
  It seems to me that this thread makes a somewhat artificial connection (McVey-Start) and misses the really obvious one (Start-Sutton). McVey and Start are not really contemporaries, Start having been in the game for about a decade already when McVey came along. Their strengths and weaknesses, from a statistical rather than a "tools" perspective, are difficult to compare. One short career, one long. One (ironically from a 21st century perspective, the 1B) consisting of much defense, the other of dominant hitting. And so on.

Start and Sutton are more easily compared. Long careers, their cases depending on a lot of defensive value, etc. etc. So:

Peak--Start was clearly one of the top 2 or 3 players in America in the mid-1860s. This is not "undocumented," not a matter of "faith." It is in the record. After the death of Jim Creighton and before the emergence of George Wright and perhaps Al Spalding in the barnstorming era (beginning in '67), Start, Harry Wright and Dickey Pearce were clearly regarded as the top players in that order. Start, whose team, the Atlantics, were the leading dynasty before the Washington Nationals and Cincy Red Stockings, was clearly regarded as the best of the three. H. Wright probably still was regarded as the best "career value" player active in the game, but Start was the top "peak value" player at the time. Now this may have been a fairly short window from, say, 1864 through 1866, but how many players are clearly the best for any longer than that? Not many. How many are even "among" the best players in the game for more than 3 years? Not 10 in the 19th century and of those 10 only Hamilton, Duffy and Caruthers are on this ballot with Start.

We now know that Sutton was among the very best players in the game for a 3 year peak from 1883 to 1885, though this fact utterly escaped those who saw him play. His 3 year WS total, adjusted for a fielding bonus and normalized to 162 games, was 133, the highest for any position player (tied with Charlie Bennett) of the 19th century. His 5 year peak was not very high relative to the 3 year peak, however, and as a result I would say Sutton was for 3 years only "one of" the best and in fact in the second tier of "the best" behind Hines and Gore, and along with Bennett, Glasscock and the emerging/arriving Brouthers and Connor. Despite his very high 3 year peak of 133, Sutton was part of a crowded field of players with similar WS totals, suggesting that the high numbers were in part a result from the overall conditions of the game at that time.

So for peak, I have Start clearly ahead of Sutton. If you want to put a number to Start's peak or you timeline, you may disagree, but I don't see it as a matter of faith, it is very clear from the record just discussed.

As for career, I will be more brief because Start's obvious superiority here is even easier to see. Sutton leads Start in "Bill James WS," 1876ff only and unadjusted for season length, 158-124, neither one being at all impressive. Thanks to Chris for providing NA WS and thanks also for adjustments for season length and the higher value of fielding vs. pitching, we think that Sutton 468 Start 409 better reflects their true career values 1871ff.

But, hey! Wait a minute. Start had already been active for 12 years prior to 1871. And not only that, he was the best player in America for a part of that pre-'71 period. It's not a matter of faith to believe that Start would have accumulated 60 more WS in the '60s. If you count WS from '71 and if you normalize to 162 (or 75 or regress to the mean or whatever), then you can hardly deny Start (or others) any value at all for '70...'69...etc. etc. Let's just say he gets 10 a year for 12 years, which is pretty conservative, now you're talking 527 WS.

Hamilton is #1 on this ballot for documented adjWS at 473. It seems crystal clear to me that Start is #1 in career value, Sutton #3, while for peak value Start is in the top 5 (acknowledging some uncertainty there) and Sutton probably in the second 10. So I end up with Start #1 on the 1907 ballot and Sutton #6.

McVey, meanwhile, is #3 overall because of a massive peak--top 5, like Start, to be sure, but also much higher than Joe's. For career value I have McVey at adjWS 336 not even counting 1869-70 for which he clearly should get 20 per year for a total of 376. Still below the best, so he settles in between Start and Sutton.

Posted 2:33 p.m., August 13, 2003 (#55) - sean gilman (e-mail)
  So if I'm reading you right Marc, in terms of career adjusted Win Shares you've got:

Sutton: 468
Start: 409
Hamilton: 473

before adjusting for the pre NA years.

I wonder how you reached those numbers, since the aWS numbers I'm using are from Chris Cobb for Start and Sutton (which are fielding adjusted) and from the Pennants Added thread for Hamilton (which I believe are schedule but not fielding adjusted). Anyway I have:

Sutton: 468
Start: 349
Hamilton: 393.3

Which is weird because our Sutton numbers are the same, but the other two are very different. . .

Posted 3:24 p.m., August 13, 2003 (#56) - Marc
  Sean, right now ya got me. Best I can see for Start is 244 (NL adjWS posted here a year ago) + 93 (Chris' NA number), so I will have to figure out where I got 409. Even then, 337 + 120 (very conservative for 1859-70) = 457 which puts Start among the leaders in career value.

As for Hamilton, as you said the 393 may not be fielding adjusted. The 473 is fielding adjusted by my own hand (and my own formula, so indeed it may not match other numbers).


Add A New Comment

Name

E-Mail (optional)

Homepage (optional)

Comments


Powered By Greymatter

Search this blog

You want baseball news?  We've got baseball news!
Weblogs

Hot Topics
The Hall of Merit

(comments - last post)

1907 Ballot (89 - 8:41am, Aug 16)

1907 Ballot Discussion (239 - 1:44pm, Aug 14)

Start and McVey (56 - 3:24pm, Aug 13)

All Time Negro Leagues All-Stars (89 - 6:36pm, Aug 12)

1906 Results - Spalding Elected (15 - 8:06pm, Aug 7)

1906 Ballot Discussion (208 - 2:42pm, Aug 7)

Pennants Added (updated for 1907 ballot) (33 - 10:39pm, Aug 4)

Al Spalding (35 - 9:34pm, Aug 4)

1906 Ballot (117 - 11:27am, Aug 4)

Thompson, Tiernan and Griffin (47 - 8:30pm, Aug 2)

Second Base vs. Third Base before 1920 (30 - 2:47pm, Jul 30)

The Hall of Merit Plaque Room (39 - 4:31pm, Jul 29)

1905 Results - Radbourn and Richardson Elected (12 - 3:15pm, Jul 22)

Length of Early MLB Seasons: Games Scheduled (0 - 1:29pm, Jul 22)

1905 Ballot (186 - 2:38pm, Jul 21)

Links to positional threads from last year (12 - 1:59pm, Jul 21)

Entries are in descending order of most recent post. Unread posts may be light blue.
Full Archive.

The Hall of Merit Sidebar (netscape 6)

Weblogs

An incomplete listing of baseball weblogs. Let me know if I'm missing one and we'll add. I hope you'll add clutch hits to yours as well.

Clutch Hits
Game Chatter
Transaction Oracle
Primate Studies
Sox Therapy
Sean's Outside the Box
Hall of Merit
Al's Baseball Tidbits
Jim's Lab Notes


Big Bad Baseball
Aaron's Baseball Blog
Ducksnort's(Padres)
Google News:Baseball
Lycos News:Baseball
Futility Infielder
AstrosDaily
Trade Rumors
Pappas' Business of Baseball
BaseballJunkie.Net
Baseball News
U.S.S. Mariner
Dan Lewis
Braves Journal
Minor Braves
No Pepper
Baseball Crank
Bravos Fan.com
ExposNet
Diamond Science
Steal Home
Indian's Report
EEEEEE (giants)
Replacement Level Yankee
Dodger Place
Batter's Box
Baseball 19 to 21
Baseball Writing
Baseball Talk
Eisenberg Sports
Diario del Cardenales
Mike's Baseball Rants
BaseballTruth.com
The Polo Grounds
Rookies Report
Digressions on Baseball
Sport&Nation
Musings From RSN
For Richer or Sporer
Only Baseball Matters
Newburg's BBall Page
Twins Geek
the dump's sportslog
Bryan's Baseball Banter
Baseball Musings
Uncouth Sloth
Julien's Baseball Blog
Mugs' Thoughts
The Cub Reporter
Cub Rants
The Clark & Addison Chron.
Northside Lounge
The [Untitled] Cubs Page
Bleed Cardinal Red
Baseballs and Redbirds
Dodger Thoughts
Dodger Blues
Gregg's Baseball
The BirdHouse (cards)
Boston Blog Sox
Royalties
Diamond Angle
STICKandMOVE
Angels News and Views
Darn Sox
(white) Sox News
Blue Jay Way
Mostly Baseball
Buck and a Half
Waiting for Boof
Zito Forever
@theBallpark
Pythag. Projections

Baseball Blogs Sidebar (netscape 6)





About Baseball Primer | Write for Us | Copyright © 2001-2003 Baseball Think Factory
User Comments, Suggestions, or Complaints | Privacy Policy | The Fine Print | Advertising