Introduction

 

        Comparing two players of reasonably equal Hall of Fame credentials, one of whom moves from team to team and the other of whom stays put for most of his career, the player who is easily identified with one team is not only more likely to be elected to the Hall of Fame, but MUCH more likely to be elected to the Hall of Fame.  A player who hopscotches from team to team may be reducing his Hall of Fame election chances by 50% or more by doing so.

            Last week on "Hey, Bill"  a reader (Phil Dellio) suggested that the knockaround, move-around, get-out-of-town nature of Gary Sheffield’s career might be impacting his Hall of Fame voting performance.   Actually, he was making a slightly different point, beyond that one, but anyway. . .that seems credible.  I may have suggested the same thing myself some time in the past, not sure, but it seems reasonably possible, so I responded that I would try to figure out how to study the issue. 

            I have now done that study.  Staying with one team for a longer period of time either directly results in better Hall of Fame chances for the player, or is allied with some other trait, some skill not identified and adjusted for in this method, which results in a quite significant increase in the Cooperstown chances of any player with a less than impeccable Hall of Fame resume.   Or, frankly, even a player WITH an impeccable Hall of Fame resume.  

 

GENERAL OUTLINE OF THE STUDY

            The breakthrough that made it possible to do this study was that I realized that I needed some way to measure the extent to which a player was identified with one team.   Without that, all you really have is one-team players and multi-team players, or something along those lines.   If all you have is one-team players and "not one-team players", you can’t do the study at all, because (1) there aren’t enough one-team players to work with, and (2) a grossly disproportionate percentage of one-team players are superstars. 

            You can divide players into two-team players, three-team players, etc., or you can divide them into groups of players who played 90% with one team, 80% with one team, 70% with one team, etc.   Those options are awkward and don’t really work.  John Smoltz, Gil Hodges, Juan Marichal, Norm Cash, Willie Mays, Hank Aaron, Warren Spahn, Hank Greenberg and Kevin Youkilis are actually NOT one-team players, but we automatically associate them with one team.   There has to be a way to measure the extent to which a player is ALMOST or essentially a one-team player, versus the extent to which he is essentially a hired gun. 

            Having reached the point of realizing that we needed a way to measure that, the solution was obvious.  I created what we could call a supporting methodology, which is "one team identification percentage."  I don’t want to get into a digression about how exactly that was done before I get to the main point, so I’ll explain in detail how that was done in Appendix 1, and present some output data from it in a second Appendix.  When I did this, by the way, I learned that Gary Sheffield just missed having the MOST scattershot career in baseball history for a player with 150 Win Shares—and had EASILY the most broken-up career for a player of his ability.  Sheffield has a one-team identification percentage of 14.4%.  Only one player in history is lower than that.   You won’t guess who it is, but I know that some of you will want to try, so I’ll leave that hanging for now, and then in one of the appendices I’ll tell you who the one player was whose career was even more broken up than Sheffield’s. 

            That problem solved, I estimated each player’s chance of selection to the Hall of Fame based on his career Win Shares.   I have done that before, of course, but having done it before I knew how to do it better than I had before, and with more recent data, so this time I did it somewhat better than I had before.  Again, I don’t want to get into a methodological digression at this point in the article, so I’ll explain how that was done in Appendix 3, and report on conclusions from that process in Appendices 4 and 5. 

            So at this point I had (a) a method to estimate each player’s probability of Hall of Fame selection, based on his career Win Shares, and (b) a method to state the degree to which each player had either a jump-from-team-to-team career, or a stay-with-one-or-two-teams mostly career.  A Gary Sheffield career, or a Cal Ripken career. 

            I then studied this data in two different ways.   To start with an example that runs counter to the general conclusion of the study, let’s take Lou Whitaker. 

            Lou Whitaker is credited with 351 Win Shares in his career.

            A player with 351 career Win Shares has a 70.4% chance of being selected to the Hall of Fame.  I’ll explain in an Appendix how this was calculated, but in simple terms, there are seven players between 347 and 352—Duke Snider (352), Max Carey (351), Lou Whitaker (351), Lou Brock (349), Barry Larkin (347) and Dwight Evans (347).  Five of the seven are in the Hall of Fame, and 5 out of 7 is 71%, so. . .that’s generally how we get there, only working with much larger groups of players.   

            Lou Whitaker played only for one team, the Tigers, so his "one team percentage" is 100%.

            We enter this, then, like this:

Player	1 team Pct	Expectation of Selection	Actual Selection	Input
Whitaker	100	0.704	0	Negative .704

 

            The fact that Whitaker has not been selected to the Hall of Fame is a Negative .704 marker for the group of players who played 100% with one team, if that makes sense. 

            But then we can add two contemporary and similarly qualified players, Ryne Sandberg and Barry Larkin.  They had 346 and 347 Win Shares, and both are 100% one-team players, and both went right in.  (Sandberg played very briefly for the Phillies, but earned no Win Shares there, so in terms of value he is 100% a Cub.)

 

Player	1 team Pct	Expectation of Selection	Actual Selection	Input
Whitaker	100	0.704	0	Negative  .704
Larkin	100	0.677	1	Positive     .324
Sandberg	100	0.671	1	Positive    .329

 

            Since no player’s Hall of Fame chances are either zero or 100%, and every player either is in the Hall of Fame (1) or is not (0), each player is either a positive or a negative contributor to the group that he represents. 

            In the study group. . . .I can see now that I’ll have to add an Appendix to explain who all is in the Study Group.   Appendix 6.   In the study there are 117 players who are 100% one-team players, a group including Terry Moore, Bobby Higginson, Ron Guidry, Bobby Feller and Roberto Clemente.    And Al Bumbry; Al Bumbry played one year for San Diego, but earned no Win Shares. 

            In this group of three players (Whitaker, Larkin and Sandberg) we have an expectation of 2.05 Hall of Fame selections, with an actual selection total of 2.00.  They’re on target.  But if we take all of those 117 players and figure the Hall of Fame likelihood of each one, based on his Career Win Shares, we have an expectation of 41 Hall of Fame selections (40.91).   In fact, 59 of them are in the Hall of Fame, 18 more than expected, or 44% more than expected.   Thus, if we take all of the one-team players in history, they have 44% more Hall of Fame selections within the group than would be expected based on those players’ performance numbers.  Lou Whitaker, Todd Helton and Stan Hack were left out when they could be in, but Bill Mazeroski, Travis Jackson, Bob Lemon, Phil Rizzuto and Earle Combs—all one-team players—are all in when they could have been left out.  On balance, the group over-achieves in Hall of Fame recognition compared to what they achieved on the field.   

            Then, if we take all of the players who are not QUITE one-team players but pretty close to it, players who are clearly and undeniably associated with one team, they ALSO over-achieve.   Gil Hodges, Frank Chance, John Smoltz, Hal Newhouser, Lou Boudreau, Tony Lazzeri, Earl Averill, Lloyd Waner, Nellie Fox, Chief Bender, Joe Sewell, Stan Coveleski, Jack Morris, Freddie Lindstrom and others all have less-than-50% chances of making the Hall of Fame, based on their career Win Shares, but all have one-team identification percentages over 85%, and are all in the Hall of Fame. 

            If we take all players in history with one-team identification percentages over 90%, they have an expectation of 43.5 Hall of Fame Selections, but actually have 63, making them +19.5 or +45%.

            If we take all players in history with one-team identification percentages over 80%, they have an expectation of 119 Hall of Fame selections, but actually have 143, or +24.

            If we take all players in history with one-team identification percentages over 70%, they have an expectation of 155 Hall of Fame selections, but actually have 180, or +25.

 

            Suppose that we start from the other end of the scale, from the Gary Sheffield end of the scale.  In the data given above, we started aggregating the data from the top end of the scale, from the Stan Musial/Kent Hrbek/Dave Concepcion/Kirby Puckett end of the scale.  This time, we’ll start from the Gary Sheffield/Royce Clayton/Doyle Alexander/Claudell Washington/Fred McGriff/Roberto Alomar/Dave Kingman end of the scale.

            Starting on that end, all players in history who have a one-team identification percentage less than 25% have an expectation of 10 Hall of Fame selections (9.76).  They actually have only three—Roberto Alomar, Deacon White and Dan Brouthers.   Players in this group who could be Hall of Famers but aren’t include Fred McGriff, Tony Mullane, Bill Madlock, Don Baylor, Rusty Staub and others.  And Sheffield, of course.

 All players who have a one-team identification percentages less than 33% (one third) have an expectation of 30 Hall of Fame selections (30.27).  They actually have only 20.   Players with a one-team identification percentage less than 40% have a expectation of 59 Hall of Fame selections.  They actually have only 39.  Players with a one-team identification less than 50% have an expectation of 96 Hall of Fame selections.  They actually have only 72.

            I told you that I had two lines of approach to analyze this data.  This was the first one.  To close off this line:  If you sort all players in the study by their one team identification percentage, you can draw a line at any point on that list—at ANY point, without exception—and the players above that line will have more Hall of Fame selections than predicted by their Win Shares, and the players below that line will have fewer. 

            Now, the other study; the other study is better, so I saved it for last.  This one doesn’t rely on the Hall of Fame expectation for any one player.  That doesn’t come into play here. 

            The other study sorted players by their one-team identification percentage.  This one sorts players by their Career Win Shares.  I sorted them into 10 groups.  All players in the study had at least 150 career Win Shares, while the #1 player, Babe Ruth, had 756.   So the ten groups were:

           

Group 10	400 or more Win Shares	51 players
Group 9	350 to 399 Win Shares	49 players
Group 8	320 to 349 Win Shares	47 players
Group 7	300 to 319 Win Shares	46 players
Group 6	275 to 299 Win Shares	69 players
Group 5	250 to 274 Win Shares	79 players
Group 4	225 to 249 Win Shares	123 players
Group 3	200 to 224 Win Shares	144 players
Group 2	175 to 199 Win Shares	187 players
Group 10	150 to 174 Win Shares	244 players

 

            Of the 51 players in Group 10, 47 are in the Hall of Fame, or 92%.   Of the 244 players in group 1, only 2 are in the Hall of Fame, or less than 1%.   This is the full chart of that data:

 

Group 10	400 or more Win Shares	51 players	47	Hall of Famers	92.2%
Group 9	350 to 399 Win Shares	49 players	42	Hall of Famers	85.7%
Group 8	320 to 349 Win Shares	47 players	28	Hall of Famers	59.6%
Group 7	300 to 319 Win Shares	46 players	25	Hall of Famers	54.3%
Group 6	275 to 299 Win Shares	69 players	23	Hall of Famers	33.3%
Group 5	250 to 274 Win Shares	79 players	26	Hall of Famers	32.9%
Group 4	225 to 249 Win Shares	123 players	19	Hall of Famers	15.4%
Group 3	200 to 224 Win Shares	144 players	13	Hall of Famers	9.0%
Group 2	175 to 199 Win Shares	187 players	11	Hall of Famers	5.9%
Group 10	150 to 174 Win Shares	244 players	2	Hall of Famers	0.8%

 

            In general, the data makes perfect sense, right?  The more Win Shares a player earns, the more likely he is to go into the Hall of Fame.  The only anomaly is that players with 275-299 Win Shares have not done meaningfully better in Hall of Fame selection than players with 250 to 274.   I wouldn’t worry about the anomaly; it’s just data. 

            OK, having established those groups, I then divided each group into a top half and a bottom half based on their one-team identification percentages—in other words, 10 High and 10 Low, 9 High and 9 Low, 8 High and 8 Low, 7 High and 7 Low, etc.    Group 6 is players with 275 to 299 career Win Shares—that is, good Hall of Fame candidates, but most of them actually NOT in the Hall of Fame.   There are 79 of those.   Group 6 High is players who had 275 to 299 career Win Shares and had a high one-team identification percentage.   This Group includes Bob Feller, Jim Rice, Bobby Doerr, Kirby Puckett and Bill Terry, and also Amos Otis, but I won’t mention that because I don’t want you to think I am stumping to put Amos Otis in the Hall of Fame, I mean, just because I named my dog after him.   Anyway, Group 6 Low is players who had equivalent careers, 275 to 299 Win Shares, but did so moving more rapidly from town to town.   This group includes Heinie Manush, Jack Quinn, Lave Cross, Kenny Lofton and Omar Vizquel. 

            There are 69 players total in Group 6, thus 34 players in Group 6 High, and 34 players in Group 6 Low.  (In Groups with an odd number of members, which is most of them, I eliminated the player who was in the center of the group in terms of his one-team percentage, so that the other two groups would be equal in size.)

            Of the 34 players in Group 6 High, 15 are now in the Hall of Fame, or 44%. 

            Of the 34 players in Group 6 Low, only 7 are in the Hall of Fame, or 21%.

            Among players who had 275 to 299 career Win Shares, a player is more than twice as likely to be selected to the Hall of Fame if he spent most of his career with one team than if he moved from team to team in his career.  The data for this group is reasonably typical of the study.   In all 10 groups except the bottom one, the percentage of Hall of Fame selections is notably higher among players who spent more time with one team.  Group 10 has only two Hall of Famers (Bruce Sutter and 19^th century player Tommy McCarthy), both of whom have relatively low one-team percentages, but still, that’s two players, so it doesn’t really mean much.   Here is the whole chart:

 

            Group 10 High         25 players, all in the Hall of Fame            100%

            Group 10 Low         25 players, 21 in the Hall of Fame               84%

 

            Group 9 High           24 players, 23 in the Hall of Fame            96%

            Group 9 Low            24 players, 18 in the Hall of Fame            75%

 

            The exception in Group 9 High is Lou Whitaker.  He is the only player with 350 or more Win Share who had a high one-team percentage but is NOT in the Hall of Fame. 

 

            Group 8 High           23 players, 16 in the Hall of Fame            70%

            Group 8 Low            23 players, 11 in the Hall of Fame            48%

 

            Group 7 High           23 players, 13 in the Hall of Fame            57%

            Group 7 Low            23 players, 12 in the Hall of Fame            52%

 

            Group 6 High           34 players, 15 in the Hall of Fame            44%

            Group 6 Low            34 players, 7 in the Hall of Fame            21%

 

            Group 5 High           39 players, 17 in the Hall of Fame            44%

            Group 5 Low            39 players, 9 in the Hall of Fame            23%

 

            Group 4 High           61 players, 16 in the Hall of Fame            26%

            Group 4 Low            61 players, 3 in the Hall of Fame              5%

 

            Group 3 High           72 players, 9 in the Hall of Fame            12.5%

            Group 3 Low            72 players, 4 in the Hall of Fame              6%

 

            Group 2 High           93 players, 8 in the Hall of Fame              9%

            Group 2 Low            93 players, 3 in the Hall of Fame               3%

 

            Group 1 High           122 players, none in the Hall of Fame     0%

            Group 1 Low            122 players, 2 in the Hall of Fame               2%

 

            And actually, in the bottom group, the two players who are in the Hall of Fame, Sutter and McCarthy, and actually very near the middle of the group in terms of one-team percentage rather than actually being "low".   They both have one-team percentages just under 50%. 

            Anyway. . . make what you will of it, but I was fairly astonished by the extent to which staying with one team for an extended period improves a player’s Hall of Fame selection chances.  I expected that the numbers would be higher for the stay-at-home group, but I had NO idea that the impact of that was as large as it appears to be.   I thought it would be hard to find in the data.   Many, many different things play into the Hall of Fame selection process.  The effect of this in most cases is to scramble the data so that patterns are hard to identify.  Stated another way, one can usually identify the MAJOR influences on outcomes by a study like this, but you can almost never identify the secondary inputs.  Something like a player being a coach post-playing career, or a player playing a big city as opposed to a small city, or a player winning Gold Gloves.. . that stuff is hard to identify, even though it may be significant, because the churning of the data caused by competing major influences will hide the minor influences. 

            But this, based on this study, has to be considered a major factor in the voting.  If a player has less than 300 career Win Shares, his chance of being selected to the Hall of Fame is consistently more than twice as great if he played most of his career for one team than if he split his career among three or four different teams.  It is more than twice as great in almost every sub-group once you get under 300 Win Shares.  

            Over 300 Win Shares, there is still a very significant advantage for the player who plays most of his career with one team, although it doesn’t measure as 2-to-1 because, well, it can’t.  If 80% of a group of players is in the Hall of Fame, the number has to be at least 60% in each part of the study, so it is mathematically impossible to have a 2-to-1 ratio. 

            The exact point at which a player becomes more likely than not to go into the Hall of Fame is 306 Win Shares.  Harold Baines, with 307 Win Shares, is in.   Al Oliver, about the same kind of a player but with 305, is not in.   Of course that’s not really a clear pattern, it’s just a way of thinking about it. 

 

            We should be careful in interpreting our result.  I have suggested here that playing most of his career with one team causes a player to do better in the Hall of Fame selection process.  That’s not necessarily true.  It could be that the players who stay longer in one city are more likely to go into the Hall of Fame because they are actually better players in some manner that is too subtle for the study; they have the same Win Shares, but one of them is actually better than the other.   Has more WAR or something, more World Series titles or better team-building skills.  Maybe these guys who move around just wear out their welcome after a year or two. 

            I would be skeptical of that because the differences are so large.  If we had a bunch of 39% to 35% advantages, 23% to 20%,  7% to 5%. . . if that was the data, I’d say "sure, maybe Win Shares is missing something in its evaluation of the careers."  When the ratio is more than 2 to 1—consistently—it would be harder to explain how that could happen for some "outside the box" reason.  But we don’t necessarily know what is cause and what is effect. 

 

 

Appendix 1

How the One-Team Identification Percentage is Calculated

 

I used the technique, which I have used many times before, of measuring by the sum of the squares divided by the square of the sum.   

            Suppose that a player has 10 Win Shares (or 10 WAR, or 10 games played, or 10 RBI, or 10 homers; it doesn’t much matter.)  Suppose he has 10, and all 10 are with one team.   Then his "one team percentage" is 100%. 

            (10 ^ 2)  / (10 ^ 2) = 100 / 100 = 1.000

            Suppose that he plays for two teams and has five Win Shares for each team; then his "one team percentage" is 50%:

            [(5 ^ 2) + (5 ^ 2)] / (10 ^ 2)  =  (25 + 25) / 100 = 50/100 = .500

 

            Suppose that he plays for three teams, and has four Win Shares for each team; then is "one team percentage" is 33.33%:

            [(4 ^ 2) + (4 ^ 2) + (4 ^ 2)] / (12 ^ 2) =

            (16 + 16 + 16) / 144 = 48/144 = .33333333

 

            It’s an easy, reliable method of measuring the extent to which a group of things is divided into sub-groups, vs. the extent to which they are all in the same group.  So then, John Smoltz played for three teams, but is 99.3% identified with one team, Gil Hodges played for two teams but in 98.5% a Dodger, Juan Marichal played for three teams but is 97.7% a Giant, etc. 

            Oh, there is one little problem with the study that I should acknowledge here.  In my Win Shares file, when a player plays for two teams in one season, that is listed just as "2 teams". . . .or 3 teams, or whatever.   Because of that, those seasons are counted as being not for either team, but for a "different" team, which is identified just as "2 teams".   That’s why Gary Sheffield figures at 14.4%; he was traded in mid-season a couple of times.   If I had the data for each split season, he wouldn’t be at 14.4%; he’d be at 14.9% or something, some number a little bit different.  It’s a little glitch in the data.  There’s always something.

            The one player in history who has a lower one-team identification than Gary Sheffield was 1890s catcher Duke Farrell.  Look him up.

 

Appendix 2

Some Results from the Method Described in Appendix 1

            As I think I mentioned, there are 117 players in the study who earned 150 or more Win Shares in their careers, all for one team.  This includes Brad Radke, Vern Law, Mike Scioscia, Terry Puhl, Bob Allison, Ryan Howard and 59 Hall of Famers. 

            60-some other players are very close to 100%, 90% or higher; these include Ty Cobb, Duke Snider and Eddie Mathews, but also Paul Blair, Johnny Logan, Hank Bauer, Bob Forsch, Eric Karros, Willie Mays, Hank Aaron, Norm Cash and Bob Friend. 

            The average one-team identification percentage is 58%.   It does not vary greatly between great players and lesser players.   Great Players sometimes stay with one team for 12 years, then go to another for 6; Lesser players make it 8 and 4, but the average percentage stays around 58% regardless of what caliber of player you are discussing.

            Carlton Fisk is almost exactly 50%, 182 Win Shares with the Red Sox, 186 with the White.   One team percentage of 50.006.   Another guy like that was Miller Huggins, 110 and 112, and Wally Moon (87 Win Shares with the Cardinals, 88 with the Dodgers, never played for anybody else.)   Jimmy Foxx, Vida Blue and Joe Medwick are around 50%, but with a different pattern, playing mostly for one team but played some with a third or fourth or fifth team. 

            Babe Ruth had 756 Career Win Shares, 574 with the Yankees, 180 with the Red Sox, two with the Braves.  That figures as a one-team percentage of 63.3%.  He had 76% of his value with the Yankees, and you might assume that, if he had 76% of his Win Shares with one team, he would have a one-team percentage of 76%, but that’s not QUITE the way the method works.  It works that way if you split it evenly between two teams or between three teams, but it doesn’t work exactly that way if the splits are not even. 

            Suppose that we used the rule that a player’s one-team identification percentage is simply the percentage of his games or the percentage of his value which is accumulated with one team—76%, for Babe Ruth.  But suppose that you had two players who each had 200 Win Shares in their career, but one player played for only two teams, like Carlton Fisk or Wally Moon, but had 100 Win Shares with each team.  The other player played for six teams, but had 100 Win Shares with one team, 30 with a second team, 25 with a third, 20 with a fourth, 15 with a fifth, and 10 with a sixth team. 

            Player A  100 + 100                                    =200

            Player B  100 + 30 + 25 + 20 + 15 + 10  =  200

 

            Obviously these two players are not the same in the extent to which their careers are broken up into separate parts—but if you used the "Babe Ruth is 76%" alternate, they would measure as being exactly the same, both 50% one-team players.  That’s why we don’t do it that way.

            The lowest one-team identification percentage for a Hall of Fame player is 19%, by Dan Brouthers.   Robby Alomar is second on that list, with 23%.  Other Hall of Famers with low one-team identification percentages include Deacon White (23%), Dave Bancroft (25%), Rick Ferrell (26%), Dennis Eckersley (26%), Goose Gossage (26%), Heine Manush (26%), Gaylord Perry (27%) and Hoyt Wilhelm (27%). 

            Bert Blyleven, mentioned earlier in this discussion, is at 29%, 13^th lowest for a Hall of Famer.  Cy Young is at 30%.  Nolan Ryan is at 33%, Randy Johnson at 37%, Reggie Jackson at 38%. 

 

 

 

 

Appendix 3

How a Player’s Hall of Fame Expectation Was Figured

 

            Lou Whitaker has 351 Career Win Shares.  One can easily observe that most players with 340 or more Win Shares go into the Hall of Fame, but how exactly did I determine that the figure of Whitaker should be 70.4%?

            There are 49 players in history with 350 to 399 Win Shares, of whom 42 are in the Hall of Fame, or 85.7%.   We COULD say that, since Whitaker is in this group, his expectation is 86%.

            But that’s not exactly right—actually it is not all that close to being right—because Whitaker is almost at the very bottom of his group.  The percentage for players between 320 and 349 Win Shares is only 59.6%.   The center of the 350-399 group is 374.5, while the center of the 320-349 group is 334.5, so Whitaker is actually quite a bit closer to the center of the 320-349 group than he is the center of the 350-399. 

            I assumed that the percentage of Hall of Fame players for any group of players represents the appropriate figure for the CENTER of that group, and then I interpolated between the centers of the groups.  The center for Group 3 is 334.5, and the percentage for Group 3 is 59.574, so 59.574 is the assumed percentage for a player with 334.5 Win Shares, although there is no such thing as a half of a Win Share. 

            The Center for Group 2 is 374.5 and the percentage is 85.714.   The distance between the centers of the two groups is 40 Win Shares.  Whitaker covers 16.5 of those 40.   If you divide 16.5 by 40 and multiply that by the difference between 59.574 and 85.714,  then add the result to 59.574, you get 70.357, I hope.  So I credited Whitaker with a 70.357% percent chance of Hall of Fame selection. 

            Since Whitaker hasn’t been selected, he scores at negative .70357.   But if you do this for all 1,039 players within the study, the aggregate plus or negative is zero.  Actually, it isn’t EXACTLY zero; it is some very small number, less than .50, because that’s just the way these things are; there’s always some little thing that keeps everything from balancing out exactly.

            Probably this.   The chart numbers jump suddenly very when a player crosses 300 Win Shares.   Because of that large jump, if I used the procedure outlined above (in re Whitaker), there would be a "fault line" at 309.5, 309.5 being the center of Group 4.   Using the procedure outlined above the numbers would jump suddenly forward between 309 and 310.  It is unrealistic to believe that a player’s chance of making the Hall of Fame increases dramatically when he earns his 310^th Win Share, so I didn’t want the chart to say that it does. 

            So in interpolating the data in that range of the chart, from 320 down to 275, I used a different method, to smooth out the curve in that area.   Occasionally theoretical data is better than real life data. 

 

Appendix 4

The Hall of Fame Expectations for Each Level of Win Shares

 

            One might assume that a player who has a very large number of Career Win Shares would be an automatic Hall of Fame selection, thus have a Hall of Fame Expectation of 100%.   You can see, on reflection, that this is not true.  Barry Bonds and Roger Clemens, perhaps the greatest pitcher and the greatest hitter of my lifetime, are not in the Hall of Fame, and Pete Rose, the All-Time leader in hits, is not in the Hall of Fame.   There are other things that could keep a very great player from being recognized.  If there was an OJ Simpson type candidate in baseball, he would not be elected.  In the current environment, a player who expressed clear racist hostility would not be elected.  It’s not 100% automatic, no matter what a player has done on the field. 

            How close does a player get to 100%, by his performance on the field?  The data says 92.2%, so that’s the number that I went with, not saying it will hold in the future.  My rule was than any player with 425 Career Win Shares or more has a 92.2% chance of being selected to the Hall of Fame. 

            By this research, there is a Hall of Fame probability associated with each number of career Win Shares.  These are fun or useful numbers to know.   They are relevant—not determinative, but helpful—in any Hall of Fame discussion.  Kyle Seager just retired with 205 career Win Shares.  What are the Hall of Fame chances of a player with 205?   8.1%.   Kyle Seager is not a Hall of Famer, but there are Hall of Famers who had less.  Albert Pujols is at 494, so he is quasi-automatic.  He’s automatic unless he does something bad.  Joey Votto is at 333.  What is the probability for a player with 333?   59.3%.   More likely than not, but not a lock.  Johnny Mize got to 338, despite missing three seasons due to World War II, but then had to wait almost 30 years to be welcomed into the Hall of Fame.   He was bitter about it. 

            This chart gives the Hall of Fame probability for a player with any number of career Win Shares from 150 to 425.

 

Win Shares	Probability of Selection to Hall of Fame
	+0	+1	+2		+3	+4	+5	+6		+7	+8	+9
420	91.6	91.7	91.8		92.0	92.1	92.2
410	90.3	90.4	90.5		90.7	90.8	90.9	91.1		91.2	91.3	91.5
400	89.0	89.1	89.3		89.4	89.5	89.6	89.8		89.9	90.0	90.2
390	87.7	87.8	88.0		88.1	88.2	88.4	88.5		88.6	88.7	88.9
380	86.4	86.6	86.7		86.8	86.9	87.1	87.2		87.3	87.5	87.6
370	82.8	83.4	84.1		84.7	85.4	85.7	85.9		86.0	86.2	86.1
360	76.2	76.9	77.5		78.2	78.8	79.5	80.2		80.8	81.5	82.1
350	69.7	70.4	71.0		71.7	72.3	73.0	73.6		74.3	74.9	75.5
340	63.2	63.8	64.5		65.1	65.8	66.4	67.1		67.7	68.4	69.1
330	58.6	58.8	59.1		59.3	59.5	60.2	60.8		61.2	61.9	62.5
320	56.5	56.8	57.0		57.2	57.4	57.6	57.8		58.0	58.2	58.4
310	53.6	54.7	54.9		55.1	55.3	55.5	55.7		55.9	56.1	56.3
300	45.4	45.9	46.5		47.2	48.2	49.2	50.2		51.1	52.0	52.8
290	36.4	37.4	38.4		39.5	40.5	41.6	42.7		43.7	44.2	44.8
280	33.2	33.2	33.2		33.3	33.3	33.3	33.3		33.3	34.3	35.3
270	33.0	33.0	33.1		33.1	33.1	33.1	33.1		33.2	33.2	33.2
260	31.5	32.2	32.9		32.9	32.9	33.0	33.0		33.0	33.0	33.0
250	24.5	25.2	25.9		26.6	27.3	28.0	28.7		29.4	30.1	30.8
240	17.5	18.2	18.9		19.6	20.3	21.0	21.7		22.4	23.1	23.8
230	13.7	13.9	14.2		14.4	14.7	14.9	15.2		15.4	16.1	16.8
220	11.0	11.3	11.6		11.8	12.1	12.4	12.6		12.9	13.2	13.5
210	8.8	9.1	9.4		9.6	9.9	10.2	10.4		10.7	11.0	11.3
200	7.5	7.6	7.8		7.9	8.1	8.2	8.3		8.4	8.6	8.7
190	6.4	6.5	6.6		6.8	6.9	7.0	7.1		7.2	7.3	7.4
180	4.5	4.7	4.9		5.1	5.3	5.5	5.7		5.9	6.0	6.1
170	2.4	2.6	2.8		3.0	3.2	3.5	3.7		3.9	4.1	4.3
160	0.8	0.8	0.9		1.0	1.2	1.4	1.6		1.8	2.0	2.2
150	0.5	0.5	0.5		0.5	0.6	0.6	0.6		0.7	0.7	0.7

 

 

Appendix 5

Hall of Fame Candidates

 

            This study rests on three key numbers for each player:

1)     His career Win Shares,

2)     The probability that a player with that number of Win Shares is in the Hall of Fame, and

3)     His one-team Identification Percentage.

 

This chart gives those three pieces of data for all players within the study who had 200 or more Win Shares, but are not in the Hall of Fame:

 

First	Last	Win Shares	HOF Probability	One Team %		First	Last	Win Shares	HOF Probability	One Team %
Barry	Bonds	704	92.2%	59.2%		Frank	Schulte	239	16.8%	77.9%
Roger	Clemens	437	92.2%	39.0%		Roger	Peckinpaugh	239	16.8%	48.7%
Gary	Sheffield	430	92.2%	14.4%		Will	White	239	16.8%	53.8%
Manny	Ramirez	408	90.0%	37.7%		Cupid	Childs	238	16.1%	34.6%
Tony	Mullane	399	88.9%	21.2%		Harry	Davis	238	16.1%	75.1%
Bill	Dahlen	394	88.2%	33.6%		Bobo	Newsom	237	15.4%	26.9%
Rafael	Palmeiro	393	88.1%	48.3%		George	Hendrick	237	15.4%	30.2%
Darrell	Evans	363	78.2%	33.7%		Wally	Moses	237	15.4%	43.5%
Rusty	Staub	358	74.9%	23.8%		Phil	Cavarretta	237	15.4%	94.3%
Bobby	Abreu	356	73.6%	41.6%		Willie	Wilson	237	15.4%	85.8%
Sherry	Magee	354	72.3%	61.9%		Edgar	Renteria	236	15.2%	27.7%
Lou	Whitaker	351	70.4%	100.0%		Joe	Adcock	236	15.2%	53.1%
Dwight	Evans	347	67.7%	94.4%		Shawn	Green	236	15.2%	37.1%
George	Van Haltren	344	65.8%	35.4%		Bob	Watson	236	15.2%	66.9%
Mark	McGwire	343	65.1%	49.4%		Doc	White	235	14.9%	70.1%
Fred	McGriff	342	64.5%	18.7%		Cy	Williams	235	14.9%	62.4%
Dick	Allen	342	64.5%	44.4%		Mel	Harder	234	14.7%	100.0%
Jimmy	Sheckard	339	62.5%	37.7%		Prince	Fielder	234	14.7%	52.0%
Jeff	Kent	339	62.5%	30.8%		Del	Ennis	233	14.4%	85.7%
Bob	Caruthers	337	61.2%	36.2%		Willie	Horton	233	14.4%	74.0%
Jim	McCormick	334	59.5%	41.4%		David	Justice	233	14.4%	42.2%
Will	Clark	331	58.8%	51.4%		Ben	Chapman	233	14.4%	35.1%
Bobby	Grich	329	58.4%	50.6%		Derrek	Lee	233	14.4%	41.9%
Tommy	Leach	328	58.2%	57.7%		Charlie	Hough	233	14.4%	45.0%
Dave	Parker	327	58.0%	44.8%		Juan	Gonzalez	233	14.4%	73.4%
Reggie	Smith	325	57.6%	35.9%		Dutch	Leonard	233	14.4%	37.0%
Jason	Giambi	325	57.6%	43.9%		Jesse	Tannehill	233	14.4%	47.9%
Willie	Davis	322	57.0%	71.2%		Dennis	Martinez	233	14.4%	36.1%
Vada	Pinson	321	56.8%	59.5%		Wes	Ferrell	233	14.4%	44.1%
Sammy	Sosa	321	56.8%	80.4%		Tim	Salmon	232	14.2%	100.0%
Graig	Nettles	321	56.8%	46.4%		Donie	Bush	232	14.2%	90.9%
Todd	Helton	318	56.1%	100.0%		Babe	Herman	232	14.2%	40.5%
Luis	Gonzalez	318	56.1%	38.6%		Andy	Van Slyke	231	13.9%	59.0%
Stan	Hack	316	55.7%	100.0%		Gene	Tenace	231	13.9%	44.6%
Jimmy	Ryan	316	55.7%	71.6%		Ray	Durham	231	13.9%	40.9%
Jack	Clark	316	55.7%	34.1%		Chuck	Knoblauch	231	13.9%	57.2%
Norm	Cash	315	55.5%	97.5%		Roy	Sievers	231	13.9%	36.7%
Lance	Berkman	313	55.1%	70.3%		Paul	Derringer	231	13.9%	54.1%
Jose	Cruz	313	55.1%	77.7%		B. J.	Surhoff	231	13.9%	39.7%
Bernie	Williams	312	54.9%	100.0%		Jack	Fournier	231	13.9%	31.3%
Willie	Randolph	312	54.9%	65.8%		George	Uhle	231	13.9%	56.4%
Keith	Hernandez	311	54.7%	43.8%		Kent	Hrbek	230	13.6%	100.0%
Johnny	Damon	307	50.2%	24.5%		Garret	Anderson	230	13.6%	94.1%
Al	Oliver	305	49.2%	42.9%		Ryan	Klesko	230	13.6%	48.3%
Jim	Wynn	305	49.2%	61.2%		Hardy	Richardson	230	13.6%	32.2%
Scott	Rolen	304	48.2%	29.5%		Ginger	Beaumont	229	13.4%	59.5%
Carlos	Delgado	303	47.2%	53.7%		Gene	Woodling	228	13.1%	29.0%
Ken	Singleton	302	46.3%	59.5%		Dizzy	Trout	228	13.1%	93.2%
Bobby	Bonds	302	46.3%	41.8%		Dode	Paskert	227	12.9%	42.2%
John	Olerud	302	46.3%	30.3%		Stuffy	McInnis	227	12.9%	41.8%
Jim	Edmonds	301	45.4%	49.0%		Ray	Lankford	227	12.9%	83.6%
Buddy	Bell	301	45.4%	34.8%		Al	Dark	226	12.6%	38.5%
Brian	Downing	298	44.4%	60.2%		Bill	Buckner	226	12.6%	34.3%
Frank	Howard	297	43.6%	52.8%		Tommy	Bridges	225	12.4%	100.0%
Steve	Finley	297	43.6%	25.1%		Urban	Shocker	225	12.4%	57.5%
Cesar	Cedeno	296	42.7%	74.3%		Lee	May	225	12.4%	33.8%
Mickey	Vernon	296	42.7%	60.8%		Larry	Jackson	225	12.4%	37.0%
Brett	Butler	295	41.8%	26.6%		Pete	Browning	225	12.4%	39.7%
Dale	Murphy	294	41.0%	82.3%		Jamie	Moyer	225	12.4%	33.9%
Mark	Grace	294	41.0%	84.4%		Dick	Groat	225	12.4%	48.6%
Kid	Gleason	294	41.0%	30.4%		Mickey	Lolich	224	12.1%	89.7%
Dave	Foutz	292	39.4%	40.6%		Rafael	Furcal	224	12.1%	42.9%
Fielder	Jones	290	37.1%	58.0%		Eric	Davis	224	12.1%	57.4%
George	Burns	290	37.1%	70.4%		Dolph	Camilli	224	12.1%	51.3%
Larry	Doyle	289	35.6%	76.3%		Willie	McGee	224	12.1%	51.1%
Tommy	John	289	35.6%	27.0%		Bill	Nicholson	223	11.9%	82.9%
Miguel	Tejada	288	33.5%	33.6%		Bob	Shawkey	223	11.9%	78.2%
Jack	Powell	287	33.3%	33.2%		Roger	Maris	223	11.9%	47.1%
Bob	Elliott	287	33.3%	44.4%		Topsy	Hartsel	223	11.9%	73.3%
Ed	Konetchy	287	33.3%	32.0%		Harvey	Kuenn	223	11.9%	50.4%
Bob	Johnson	287	33.3%	60.8%		Todd	Zeile	223	11.9%	23.5%
Brian	Giles	287	33.3%	33.7%		Charlie	Root	223	11.9%	100.0%
Kenny	Lofton	287	33.3%	44.0%		Billy	Nash	222	11.6%	67.2%
Toby	Harrah	287	33.3%	50.8%		Miller	Huggins	222	11.6%	50.0%
Jack	Quinn	287	33.3%	20.4%		Darrell	Porter	222	11.6%	30.8%
Amos	Otis	286	33.3%	98.6%		Freddie	Fitzsimmons	222	11.6%	62.3%
Jack	Stivetts	285	33.3%	52.5%		Ruben	Sierra	222	11.6%	41.3%
Chili	Davis	285	33.3%	31.7%		Bill	Hutchinson	221	11.3%	99.1%
Mike	Smith	285	33.3%	37.0%		Arlie	Latham	221	11.3%	48.2%
Sal	Bando	283	33.3%	70.0%		Chris	Chambliss	221	11.3%	33.7%
Charlie	Buffinton	283	33.3%	34.7%		Ed	McKean	221	11.3%	39.5%
Boog	Powell	282	33.2%	81.5%		Curt	Flood	221	11.3%	100.0%
Omar	Vizquel	282	33.2%	41.3%		Dom	DiMaggio	220	11.1%	100.0%
Ron	Cey	280	33.2%	67.2%		Mark	Buehrle	220	11.1%	66.4%
Fred	Lynn	280	33.2%	37.8%		Lon	Warneke	220	11.1%	48.0%
Julio	Franco	280	33.2%	32.0%		Andy	Pafko	220	11.1%	42.3%
Tony	Fernandez	280	33.2%	46.7%		Nomar	Garciaparra	219	10.8%	62.7%
Bert	Campaneris	280	33.2%	78.3%		Tim	Hudson	219	10.8%	46.1%
Steve	Garvey	279	33.2%	67.9%		Doc	Cramer	219	10.8%	30.4%
Ken	Boyer	279	33.2%	77.9%		Vic	Wertz	219	10.8%	29.6%
Lave	Cross	278	33.2%	22.8%		Richie	Hebner	219	10.8%	47.1%
Dixie	Walker	278	33.2%	53.4%		Charlie	Keller	218	10.6%	95.5%
Torii	Hunter	277	33.2%	42.1%		Art	Fletcher	218	10.6%	79.2%
Moises	Alou	277	33.2%	22.0%		Tino	Martinez	218	10.6%	41.2%
Andruw	Jones	276	33.1%	80.4%		Kirk	Gibson	218	10.6%	56.8%
Jim	Whitney	275	33.1%	56.9%		Tom	Brown	218	10.6%	14.4%
Robin	Ventura	274	33.1%	46.2%		Larry	French	218	10.6%	42.3%
Rocky	Colavito	273	33.1%	39.7%		Carl	Furillo	217	10.3%	100.0%
Adonis	Terry	273	33.1%	26.1%		Placido	Polanco	217	10.3%	27.1%
Heinie	Groh	272	33.1%	61.2%		Jeff	Heath	217	10.3%	51.6%
Aramis	Ramirez	272	33.1%	39.5%		Harlond	Clift	216	10.1%	77.3%
Jose	Canseco	272	33.1%	37.8%		Pete	Runnels	216	10.1%	46.2%
Cy	Seymour	272	33.1%	35.6%		George	Scott	216	10.1%	48.2%
Joe	Judge	270	33.0%	97.1%		Hooks	Dauss	215	9.8%	100.0%
Dave	Concepcion	269	33.0%	100.0%		Danny	Murphy	215	9.8%	92.8%
George	Foster	269	33.0%	55.0%		Tom	Daly	215	9.8%	62.3%
Ron	Fairly	269	33.0%	38.5%		Kip	Selbach	215	9.8%	20.9%
Tony	Phillips	268	33.0%	30.8%		Brady	Anderson	214	9.5%	97.2%
Bill	Freehan	267	33.0%	100.0%		Heinie	Zimmerman	214	9.5%	51.3%
Eddie	Yost	267	33.0%	68.2%		Tommy	Corcoran	214	9.5%	37.3%
Bobby	Bonilla	267	33.0%	30.7%		Rudy	York	214	9.5%	72.1%
Clyde	Milan	266	33.0%	100.0%		Chuck	Finley	213	9.3%	76.4%
Wilbur	Cooper	266	33.0%	88.6%		Frank	Dwyer	213	9.3%	50.2%
Herman	Long	265	33.0%	81.4%		Tip	O'Neill	213	9.3%	70.2%
Vern	Stephens	265	33.0%	49.1%		Raul	Ibanez	213	9.3%	38.8%
Chet	Lemon	265	33.0%	50.6%		Sam	Leever	212	9.0%	100.0%
Harry	Stovey	265	33.0%	43.7%		Theodore	Breitenstein	212	9.0%	49.8%
Bobby	Veach	265	33.0%	88.5%		Mike	Hargrove	212	9.0%	42.4%
Don	Mattingly	263	32.9%	100.0%		Jose	Cardenal	212	9.0%	29.3%
Augie	Galan	263	32.9%	40.2%		Frank	White	211	8.9%	100.0%
Silver	King	263	32.9%	38.0%		Edgardo	Alfonzo	211	8.9%	68.7%
Don	Baylor	262	32.9%	25.0%		Orel	Hershiser	210	8.8%	59.4%
Jim	Fregosi	261	32.2%	73.7%		Curt	Simmons	210	8.8%	44.5%
Jack	Glasscock	261	32.2%	22.5%		Bob	Boone	210	8.8%	42.7%
Ellis	Burks	260	31.5%	22.8%		David	Wells	210	8.8%	23.1%
Roy	Thomas	260	31.5%	80.9%		Milt	Pappas	210	8.8%	34.2%
Paul	O'Neill	259	30.8%	55.9%		Dave	Stieb	210	8.8%	100.0%
Ken	Griffey	259	30.8%	47.8%		Eddie	Rommel	209	8.7%	100.0%
Guy	Hecker	259	30.8%	48.1%		Rico	Carty	209	8.7%	43.8%
Jorge	Posada	258	30.1%	100.0%		Garry	Templeton	209	8.7%	49.1%
Larry	Gardner	258	30.1%	48.5%		John	Anderson	209	8.7%	19.8%
Gus	Weyhing	258	30.1%	32.8%		Sherm	Lollar	209	8.7%	73.8%
Rick	Monday	258	30.1%	35.3%		Bill	White	209	8.7%	53.9%
Buddy	Myer	258	30.1%	76.9%		Eddie	Joost	209	8.7%	58.4%
Bucky	Walters	258	30.1%	60.0%		Tommy	Henrich	208	8.5%	100.0%
Gary	Matthews	257	29.4%	26.6%		Brickyard	Kennedy	208	8.5%	89.1%
Carl	Mays	256	28.7%	27.3%		Ed	Morris	208	8.5%	35.2%
Luis	Tiant	256	28.7%	42.5%		Lonny	Frey	208	8.5%	55.9%
George	Mullin	255	28.0%	85.5%		Reggie	Sanders	208	8.5%	29.6%
William	Hoy	254	27.3%	15.6%		Harry	Steinfeldt	208	8.5%	49.7%
Paul	Konerko	254	27.3%	99.2%		Bob	Friend	207	8.4%	98.1%
Maury	Wills	253	26.6%	63.8%		Devon	White	207	8.4%	26.5%
Wally	Joyner	253	26.6%	35.0%		Tommy	Davis	207	8.4%	34.3%
Curt	Schilling	252	25.9%	30.8%		Jimmy	Williams	207	8.4%	50.7%
Darryl	Strawberry	252	25.9%	62.2%		Bret	Boone	207	8.4%	36.7%
Dick	Bartell	252	25.9%	29.4%		Deacon	Phillippe	206	8.3%	80.2%
Andres	Galarraga	252	25.9%	30.5%		J. D.	Drew	206	8.3%	29.3%
Mike	Tiernan	251	25.2%	100.0%		Kenny	Rogers	206	8.3%	42.7%
George	Gore	250	24.5%	46.8%		Chris	Speier	206	8.3%	44.4%
Fred	Tenney	249	23.8%	80.0%		Ed	Reulbach	206	8.3%	58.8%
Carlos	Lee	249	23.8%	32.4%		Rico	Petrocelli	205	8.1%	100.0%
Gary	Gaetti	249	23.8%	38.4%		Ron	Gant	205	8.1%	32.8%
Paul	Hines	249	23.8%	42.2%		Doug	DeCinces	205	8.1%	45.2%
Billy	Pierce	248	23.1%	81.7%		Jim	Perry	205	8.1%	43.8%
Lance	Parrish	248	23.1%	47.8%		David	Cone	205	8.1%	24.3%
Tim	Wallach	248	23.1%	74.2%		Tom	Zachary	205	8.1%	27.8%
Jim	Gilliam	247	22.4%	100.0%		Claude	Ritchey	205	8.1%	47.1%
Greg	Luzinski	247	22.4%	62.0%		Hippo	Vaughn	205	8.1%	73.6%
Marquis	Grissom	247	22.4%	25.1%		Bobby	Thomson	205	8.1%	52.1%
Pedro	Guerrero	246	21.7%	55.8%		Ryan	Howard	204	8.0%	100.0%
Mike	Griffin	245	21.0%	53.1%		Tim	McCarver	204	8.0%	52.0%
Magglio	Ordonez	245	21.0%	50.9%		Pink	Hawley	204	8.0%	29.9%
Dusty	Baker	245	21.0%	43.5%		Murry	Dickson	204	8.0%	30.7%
Jay	Bell	245	21.0%	44.7%		Tommy	Harper	204	8.0%	28.2%
Jimmy	Dykes	245	21.0%	67.0%		Bob	Allison	203	7.9%	100.0%
Wally	Schang	245	21.0%	25.7%		Elston	Howard	203	7.9%	90.5%
S Sam	Jones	245	21.0%	23.0%		Wally	Pipp	203	7.9%	72.1%
Jason	Kendall	245	21.0%	49.7%		Tony	Cuccinello	203	7.9%	26.8%
Carney	Lansford	244	20.3%	48.1%		Garry	Maddox	203	7.9%	50.3%
Babe	Adams	243	19.6%	100.0%		A. J.	Pierzynski	203	7.9%	31.2%
Tommy	Bond	243	19.6%	58.5%		Ted	Kluszewski	203	7.9%	78.1%
Joe	Kuhel	243	19.6%	50.2%		Earl	Whitehill	203	7.9%	51.2%
Al	Orth	243	19.6%	35.7%		Ken	Williams	202	7.8%	81.0%
Albert	Belle	243	19.6%	45.3%		Fred	Pfeffer	202	7.8%	53.1%
Mike	Cameron	243	19.6%	20.9%		Terry	Pendleton	202	7.8%	36.8%
Ken	Caminiti	242	18.9%	47.5%		Marty	McManus	202	7.8%	35.2%
Del	Pratt	242	18.9%	34.8%		Gavvy	Cravath	202	7.8%	87.0%
Kevin	Brown	242	18.9%	23.6%		Vida	Blue	202	7.8%	50.2%
Bill	Madlock	242	18.9%	23.7%		Frank	McCormick	202	7.8%	75.1%
Wally	Berger	241	18.2%	68.3%		Bill	Donovan	202	7.8%	63.3%
Felipe	Alou	241	18.2%	37.0%		Claude	Osteen	201	7.6%	56.7%
Frank	Tanana	241	18.2%	33.2%		Willie	Kamm	201	7.6%	54.5%
Alfonso	Soriano	241	18.2%	26.6%		Luis	Castillo	201	7.6%	50.7%
Matt	Williams	241	18.2%	51.1%		Lenny	Dykstra	201	7.6%	48.4%
Cecil	Cooper	241	18.2%	72.9%		John	Titus	201	7.6%	73.5%
Johnny	Callison	241	18.2%	76.0%		Patsy	Donovan	201	7.6%	36.8%
Dick	McAuliffe	241	18.2%	95.1%		Bill	Dinneen	200	7.5%	33.4%
Dolf	Luque	241	18.2%	72.8%		Mo	Vaughn	200	7.5%	59.3%
Davey	Lopes	240	17.5%	58.5%		George	Burns	200	7.5%	26.4%
Rick	Reuschel	240	17.5%	45.9%		Jim	Sundberg	200	7.5%	63.4%
Jerry	Koosman	240	17.5%	49.7%
Joe	Carter	240	17.5%	40.7%

 

 

Appendix 6

How I did some little things

 

            John Dewan and Sports Info Solutions after the 2019 season provided me with a list of each player’s Win Shares in each season in major league history.  The list has not been updated since then, but that’s not a problem because we are studying the effect of playing for multiple teams (or for a single team) on Hall of Fame selections, and anything that has happened in the majors since 2019 is not relevant to the Hall of Fame anyway.  The list is 91,000 and some lines long.

            I sorted the list by (1) the player’s name, and (2) the team, and made totals of the number of Win Shares by each player for each franchise. If the franchise moved, I counted that as the same franchise, and if a player played for a team, left and came back to the same team, that is still counted as the same franchise.   I used codes like "1" for the Angels, "4" for the Baltimore Orioles (St. Louis Browns, 1901 Milwaukee), etc.   As I mentioned before any two-team player is coded just as "99" for that season, so that functions as an independent franchise.  A flaw in the study. 

            Then I made one-line summaries of each players’ Win Shares with different teams, T1 (team one), T2 (team two), etc.  The player who posted value for the most teams was Matt Stairs, who had value for 11 different franchises.  I think he played for more teams than that, but Montreal and Washington are the same franchise, and maybe he had zero value for one of them, I don’t know. 

            Anyway, with only one line per player, that cut the number of lines in the data to something like 19,000.   Then I eliminated everybody who had less than 150 career Win Shares, since players with less than 150 career Win Shares are not viable Hall of Fame candidates, although I think there may be one player somewhere who stumbled drunkenly into the Hall of Fame with less than 150.  Anyway, that cut the number of lines in the study to something more like 1,300. 

            Then I started eliminating players who were not eligible for the Hall of Fame.  99% of those players are not eligible because they have played in the last five years, but I also eliminated anyone who was not eligible for the Hall of Fame because he was banned from baseball, like Pete Rose or Shoeless Joe.  Their data would pollute the study, rather than inform it. 

            I also eliminated the managers who were good players and could have been elected as players or as managers, you’re not sure, like Joe Torre, Red Schoendienst, John McGraw, etc.   A manager elected with 250 Win Shares.. . what does that tell you?   Eventually we got down to 1,039 eligible players in the study.

            There is probably a player or two who snuck into the study with 150 wins in 9 seasons or less, not technically eligible for the Hall of Fame, although that didn’t stop them from electing Addie Joss.  I should have eliminated anybody like that, but I didn’t think about it until it was too late, and I’m not sure how I would have found all of those, anyway.  

            Thanks for reading.