It is spring, and in the spring, in baseball, all things are possible. Some things are more possible than others, and in this realization there is the concept of "potential".
The concept of potential may be the largest issue that sabermetrics has yet to deal with in an intelligent way. In the spring of 2014, it seems entirely possible that Wil Myers might hit 35 homers and drive in 110 runs. The upper boundary of what he might do might be beyond that, but we’re trying to be realistic here. It does not seem unrealistic, as the season begins, to suggest that he might hit 35 homers and drive in 110 runs, and win a Gold Glove in Right Field. It could happen.
His teammate Sean Rodriguez also has a certain potential, which is much less than Wil Myers’ potential, but Sean Rodriguez might still have a good year. Realistically, Rodriguez might play 120 games, might hit .260 and might hit 12 homers. He hasn’t done any of these things yet (well, he did play 120 games one year), but it seems reasonable that he might. David Price might make 33 starts, strike out 225 batters, and might go 22-7 with a 2.50 ERA. He has come close to these numbers before; he might do it. Roberto Hernandez went 6-13 last year and 7-15 the last time he wasn’t hurt, but with some breaks, he might realistically go. ..what, 13-10? 180 innings with a 4.00 ERA? It could happen.
If we can make realistic estimates of what the 2014 potential is for all of these players, then why can’t we walk back in time and make estimates of what each player’s potential was? Oscar Gamble in 1975 hit .261 with 15 homers, 45 RBI, but what might he have done, realistically?
What difference does it make?
It makes a tremendous difference. It makes a tremendous difference because the issue of potential is tied into a hundred other discussions, such as managers, predictions, leadership and team chemistry. It makes a tremendous difference because if we don’t have measurements of potential, if we don’t have statistical statements to represent them, then we have no record after the fact what didn’t happen, which means that we only have one alternative in play. That creates linear logic where the truth requires non-linear reasoning. The truth requires remembering that that which happened was not the only thing that could have happened. The key sentence in this article: understanding the truth requires remembering that that which happened was not the only thing that could have happened.
Remember the year Sports Illustrated said that the Cleveland Indians were the best team in baseball, and then they lost 100 games? 1987, I think it was. What happened?
What happened was, a lot of players didn’t come close to living up to their potential. Tom Candiotti, 16-12 with a 3.57 ERA the previous year, dropped off to 7-18, 4.78. Greg Swindell, 5-2 after being called up in late August, 1986, and an 18-game winner in 1988, went 3-8 with a 5.10 ERA. Cory Snyder, who had hit .272 as a rookie in 1986, hit .236. Tony Bernazard, who had hit .301 with 17 homers, 73 RBI in 1986, dropped off to .239 with 30 RBI.
But in April, 2013, I thought that the Cleveland Indians would finish last. I looked at their roster; I saw Drew Stubbs, who had hit .213 the previous year with Cincinnati. I saw Ryan Raburn, a .171 hitter from Detroit, and Mark Reynolds, a .221 hitter from Baltimore, and Jason Giambi, a great hitter from a decade long ago, and Scott Kazmir, who had been out of baseball for a couple of years, and Ubaldo Jimenez, who had finished 9-17 with a 5.40 ERA in 2012.
If everybody on that team had the year you’d probably have expected them to have, they’d have finished last—but instead, they won 92 games. It is a tribute to their manager, sure, but what exactly did their manager do? He got players to play up to their potential.
I can’t predict who will finish where in 2014; I can’t, you can’t, nobody can—and the reason that we cannot is that there is a vast gulf between what players do in fact and what they have the potential to do.
My estimate is—and I’ll explain how I derived this estimate much later—my estimate is that players achieve, in a typical year, approximately 50% of their potential. It is much less than 50% if you look at career potential, the difference being explained by the fact that a great deal of potential has already been eliminated before the year starts. Kelly Johnson this year might have the potential to hit .270 and drive in 65 runs—but if we had looked at him years ago, his first few years in the league, we might have guessed that he would be a .310 hitter who would drive in 110. What we would perceive as his potential for 2014 has already been limited by his performance over the last three seasons: .222, .225, .235. He no longer seems to have the potential that he once did, but he still has some potential to have a better year.
Anyway, players in any season accomplish approximately 50% of what they have the potential to accomplish in that season; again, I’ll explain that later. But because this is true, it creates a huge gap between what a team potentially could do, and what they do in fact. Because there is this chasm between what players have the potential to do and what they do in fact, it is not necessary, for a team to surprise, that every player on the team have a maximum-output season. If you’re a manager and your team achieves 57% of their potential, you’re in great shape. You’re Terry Francona, 2013.
If we could measure potential, we would be far ahead of where we are in terms of evaluating managers, but that’s not really what I’m writing about. Getting a better handle on potential is a doorway toward evaluating managers, yes, but more important than that, understanding potential is essential to an intelligent discussion of team chemistry.
The Gil Hodges Debate
A few weeks ago, in the Bill James Online Reader Posts, there was an exchange between MarisFan61 and another poster about Gil Hodges’ role on the Dodgers. MarisFan61 was advocating for Gil Hodges, and the other poster was ridiculing his arguments. I am entirely in Maris’ camp here, so long as we’re not trying to put Hodges in the Hall of Fame. Hodges was not a Hall of Fame player. Otherwise I think his comments about Gil Hodges were exactly right, but before I get into that, I should stop to take my share of the responsibility for what I believe is the wrong side of this argument.
The other side of this argument was ridiculing "chemistry", ridiculing the notion that Gil Hodges contributed heavily to the success of his teams by his contributions to the team’s chemistry. That line of argument traces back, I know, to things that I wrote in the 1980s. In the 1980s, I also ridiculed chemistry and character in assessing baseball players; I wrote essentially the same things that I now disagree with. It wasn’t that I was wrong, exactly, but that I wrote about these issues with a lack of understanding and with a lack of clarity. By so doing, I held back the development of clear thinking in this area. I don’t want anyone to think that I am now attacking the other poster for saying things very much like what I myself once said.
Let’s start by quoting the debaters:
MarisFan61
About Hodges for the Hall of Fame:
We always look for things to enhance our knowledge of a player, hopefully objective things, but heck, we take what we can get. :-)
Something I pretty much assume as a valid thing: A player who later becomes a manager probably had valuable qualities as a player beyond what showed up in the numbers.
Admittedly, in order to believe in this, you have to believe that things like leadership, attitude, inspiration, role modeling, teaching..... whatever it is that manager-type people can contribute as players that go beyond what shows up in their own numbers ..... you have to believe that those things are important. You have to believe that they help other players do well, they help with teamwork, team morale and focus. I don't see how those things wouldn't be significant, and I think having such players on the team would tend to help. I don't at all mean that these players help to "manage" the team; I'm talking only about the general effect of such a player on the team.
I realize that even assuming all of that, it's not 100% valid to think that future managers have these qualities and these kinds of effects. But I'd bet it's at least 80% valid. I think it's almost certain that they tend to have these qualities more than the average player, and probably much more. Stealing a line from Bill (he said it when justifying an aspect of Win Shares, I think about defensive credit for catchers on good teams), it seems a lot more right to assume it than not to assume it.
So, especially when we're talking about a player who was on a successful team, I take the fact that he later was a manager as an intangible but probable plus on what he was as a player. And to me, it takes Hodges from 'borderline' to 'absolutely' -- not because I'm giving him extra credit for his managerial accomplishments but because of what his later being a manager tells me about him as a player.
Other Poster
"MarisFan61
Something I pretty much assume as a valid thing: A player who later becomes a manager probably had valuable qualities as a player beyond what showed up in the numbers."
I seriously think your 80% is a wild overstatement. Bobby Valentine? It's not like he's unique.
But beyond that, look at the Boys of Summer and the players who *didn't* become managers. You don't think they didn't also have some similar "valuable qualities"?
Duke
Jackie
Pee Wee
Campy
Amoros
Gilliam
Furillo
Erskine
Podres
Newcombe
Etc
This is the problem of trying to give unto someone like Hodges special Mystical Powers as a member of the Boys of Summer: it's a team full of players with similar Mystical Powers. How do we start dividing up things?
The 1955 Dodgers won 98 games, then 4 in the World Series. So how would you split up the credit for the Wins and Losses between:
Offense
Pitching
Defense
Manager
Coaching
General Management
Ownership
Mystical Player Powers
Fan Support
Dumb Ass Luck
Then for each of those things, how would you parcel them out for each player / manager / coach / front office person?
It's fine to talk about Mystical Player Powers, but how many wins did it add? Was Gil worth 2 wins a year just on Mystical Player Powers? Jackie another 2, Duke another 2, Campy (who everyone thought was their MVP when healthy) 3? Pee Wee 2?
Then how many for Alston being able to push all the buttons? 4?
When one actually tries to go down this path they quickly find that the Boys of Summer were an 80 win team without the stuff not accounted for in Offense, Defense and Pitching.
MarisFan61
Other Poster: About the Hodges 'leadership/intangible/whatever' thing:
We have different belief systems on this. :-)
Which means, I don't expect there's a high chance that the way I see this will have any meaning to you.
It seems that you assume that those qualities operate outside of the concrete things like offense, defense, pitching, etc. You assume it so much that you didn't feel you even needed to state it.
I see them as existing (when present) within those things. To whatever extent a player like (let's say) Gil Hodges helped his team with things like leadership, inspiration, role-modeling, and teaching, I say that those things were reflected IN the stats of his teammates, not just in addition to them.
Can I prove or demonstrate it? No. Does that bother me? Not at all. Sorry, but it's a thing that I feel confident enough about, just on the basis of everything else that we know about human functioning.
I mean the confidence only with regard to the general phenomenon and how it would operate; I wouldn't propose to be nearly so confident about any given player, Hodges or anyone else. But, as I put it (stealing a concept from Bill), I think that when we're talking about a player who later became a manager, it's a heck of a lot more likely to be right if we assume it than if we don't assume it. You're right that others on that Dodger team very likely had such qualities too. But Hodges' later future makes me think it's likely he was one of the top ones -- and that this had value. Since he's a guy who's a solid borderline Hall of Famer anyway, he doesn't need much extra value to put him across that border, and for me, this is enough.
Other Poster
So what you're saying is that Gil's leadership is why Jackie was such a good hitter, such a good defender, such a good base runner, and such a fiery competitor. That these weren't things he flashed at John Muir, Pasadena City College, and UCLA. Nor in the Negro League, nor in the minors before coming up. Gil Hodges is the reason that Jackie was X% better than he'd shown growing up.
Same goes for Duke, Pee Wee, Campy, Newc, etc.
The problem is that there were *many* players on that team who have similar "leadership" bullshit tossed around. And that's why I was asking you to quantify just how much of it comes down to Gil's Mystical Inspirational Leadership Qualites and how much of it comes down to those same qualities of everyone else... and then to the fact that the all were also really good Baseball Players.
To which you can't. Worse, you give onto Gil special credit while refusing to acknowledge that it because an extremely small "value" if you're forced to give similar "value" to everyone else on the club.
As far as the Manager nonsense, we all know there have been plenty of managers who were crap in terms of being inspirational leaders of both their teammates and later their players. Seriously, no one liked Bobby Valentine all the way back to Texas. There's very little record that in either his time as a player with the Dodgers or Angles that a slew of players thought he really helped them win because he had Mystical Inspirational Leadership Qualites. What they thought it that he was a great prospect, then he got hurt, then be became a crappy major league player.
You might think I'm pointing to Valentine because he's the only example. No, he's just the easiest. Go around the majors looking at all 30 managers and let me know which ones had during their playing days reputations for being Great Mystical Inspirational Leadership Quality Players. Why don't you start with John Farrell and his time as a player with the Indians, Halos and Tigers. He had the amazing Leadership ability to turn those teams into .450 clubs in his time with them.
OK, the manager/non-manager thing. . .that is the smallest issue in here, and we’ll get back to that later but it’s not our main focus.
Let us start with the concept of "Character". Character is a term which is on a continuum of terms like Leadership and Team Chemistry, but is broadest term of the lot. It is too broad to be useful in discussing athletes. Ex-athletes like to say that sports reveal character, that they test character, and that athletes embody good character. I used to ridicule those kind of remarks, but I don’t know that I was ever clear as to what exactly the problem is—and very probably, in ridiculing those remarks, I said things very much like what the Other Poster has said here.
The real problem with the concept of Character is that it is unworkably broad. "Character" could involve a great many different traits. It could imply honesty, responsibility in dealing with others, loyalty, good work habits, sensitivity, self-discipline, integrity, punctuality, modesty, courage, self-confidence, determination, and the willingness to put the needs of others ahead of oneself.
In my view, athletic contests do in fact test many of these elements of character, and do in fact bring forward persons who embody many of these elements. Athletes are in fact hard-working, in general, and self-disciplined, and certainly punctual, and, compared to those you meet in other lines of work, willing to put the needs of the group ahead of themselves.
The problem with saying that athletics are a test of character is that there are elements of character which are not tested in any manner by sports, and in these areas athletes are not only no better than others, but may be worse. It might be said, for example, that a man of good character would not take advantage of a vulnerable young woman, or that a man of good character would take his wedding vows seriously. I am not saying that athletes are worse than others on this account, but I’ve heard stories occasionally.
Did you ever think about this: that the term "intimidation" is a very positive term in sports, but the practical definitions of "intimidating" and "bullying" are almost identical? If you say that a pitcher is intimidating, this is high praise, or that a linebacker is intimidating or a shot-blocking center "intimidates" the shooters, that is always meant as a form of high praise, that his presence creates fear in the opposition in a way that goes well beyond the norm.
But in the rest of the world, what is the difference between "bullying" and "intimidation"? There is none; they’re the exact same things. Intimidation is what Tony Soprano does for a living. Do athletes bully their wives, sometimes, some athletes? Well. . .I’ve seen news stories.
We should not talk about athletes having "character", in my view, because to say that implies things which are simply not true. It implies some things which are true, yes, but also many things that are not.
Another phrase on this continuum is "team chemistry". "Team chemistry" means the interaction of members of a team to bring out the best in one another. Team chemistry is still a fairly broad and difficult concept, although nowhere near so broad as "character". MarisFan assumes that team chemistry is a real thing, whereas the Other Poster derides it as Mystical Powers.
There is nothing in the statistical record of baseball which is in any way inconsistent with the concept of team chemistry. The concept of team chemistry is still overbroad, and because it is overbroad it is vulnerable to being used in deceptive ways. The announcers for every team which has ever hired an announcer have all proclaimed that their team had great chemistry. The players all get along great, they all hang out together after the game and talk baseball. Only when the season is over, only in discussing last season or the season before, do we ever see bad team chemistry.
But as to the existence of team chemistry, there is a very obvious place, in the statistical record of the game, where team chemistry could reside. Team chemistry could be found, at least in theory and probably in fact, in the vast gap between a team’s potential and their actual accomplishment.
There are, of course, many reasons why the players on a team do not meet 100% of their potential. Each player has a potential of what he could accomplish given 650 plate appearances, but not every player can be given 650 plate appearances. There is luck, there is limited opportunity, there are injuries, and there are personal issues such as a lack of motivation and a lack of self-confidence which cause individuals to fall short of what they might accomplish. All of these contribute to what we might call shortfall, shortfall being the gap between potential and production.
But shortfall, on a team level, is enormous, and because it is enormous, there is plenty of space there for team chemistry. If a team accomplishes 52, 53, 54% of what they have the individuals on the team have the potential to accomplish, that team will have a successful season. That can reasonably be attributed to good team chemistry. If they’re at 46, 47, 48%, that can be poor team chemistry.
The concept of team chemistry, far from being inconsistent with the statistical record of the game, actually helps to explain the statistical record of the game. If you look at the Boys of Summer, what you find is a large number of players who, year in and year out, achieved at or near the level of their potential. I will document this later in the article, will try to, but Pee Wee Reese, Gil Hodges, Jackie Robinson, Duke Snider, Junior Gilliam, Carl Furillo and others on that team achieved year-in and year-out at a very high percentage of their potential. I have Pee Wee Reese achieving 90+% of his potential seven times in his career—very unusual. If you check all of the other top shortstops of that era (Rizzuto, Marion, Boudreau, Vern Stephens, Granny Hamner, Al Dark) you don’t find anyone else who did that. If you compare Gil Hodges to Wertz, to Kluszewski, to Sievers or to Adcock, it is obvious that Hodges was achieving 80, 90% of his potential year-in and year-out, while the other players bounced from 30% to 70%.
Why is it unreasonable to believe that Hodges, Snider, Reese and Jackie helped one another to achieve, as a team, more than they would have achieved had they not been a team?
Well, it is not unreasonable at all, and in fact it seems totally unreasonable, to me, to argue that this sort of thing does not happen. What I would ask is, in your line of work, can those you work with make you more or less productive?
For maybe 1% of you, the answer might be "no", but for 99% of you, it is "yes". If you work in an office, and you have a co-worker who is always pestering you for some report that should be your 12th priority but which he wants to move up to #1, that interferes with your productivity. If a co-worker is annoying, it interferes with the operation of the office. If a co-worker is greedy, always complaining about her salary and her hours, it creates a negative atmosphere that discourages others from working at peak efficiency. If you work in a restaurant, and you have a co-worker who is lazy, it creates more work for everybody else. If a co-worker has anger issues, it creates distractions, and makes an unpleasant workspace.
There are a million ways in which a co-worker can make you more or less productive, and there are a million ways in which a baseball player can make a teammate more or less productive; surely I don’t have to suggest these to you? It is the common experience of the human race that we benefit or suffer by the actions of those around us. It seems quite remarkable that, in the context of a baseball team, anyone would deny there this is the common interaction.
A baseball team, far from being an exception to the common rule of humanity in this respect, is actually a closed chamber in which these interactions are exaggerated. The baseball team experience is almost claustrophobic, in a sense. You’re around the same guys seven days a week from mid-February until at least the end of September, and you’re under immense and unusual pressure to succeed as a group. Teammates depend on one another to do things that 99.999% of the population would be unable to do—and if YOU don’t succeed, then I, to an extent, will be perceived as a failure. If you don’t get on base, I don’t get an RBI. If you strike out, I get charged with a loss. If you don’t make the play in the field, I don’t get out of the inning—and it will be on television and in the newspapers, so that tomorrow morning tens of millions of people will know that I have failed, without knowing that I failed because you didn’t do your job.
Baseball players fly to California together three, four, five times every year. Do you fly to California with 30 or 40 co-workers several times every year, and do you go together to Cincinnati, and Pittsburgh, and Arizona, and New York and Atlanta? Don’t you think that, seven days a week without a break, the failings and foibles of your co-workers would get to be greatly exaggerated in your mind?
Because this is true—and this is perhaps the one thing that the public least understands about professional baseball—because this is true, the standards of conduct for major league baseball players are very, very high—ridiculously high, so high that someone like me could basically never fit in. I don’t mean all standards of conduct, of course; professional athletes are allowed to exempt themselves from certain standards of conduct that almost everybody else must meet. You’ve all heard stories about athletes who routinely park in Disabled/ Reserved parking spaces, because they figure the rules don’t apply to them. It’s true; there are certain areas of conduct in which athletes see themselves as exempt from the rules.
But what people don’t understand, generally: when baseball men talk about team-responsible behavior, it’s not just talk. We mean it. The expectations for good conduct in terms of team-oriented behavior, for major league athletes, are absurdly high. Team-oriented behavior means punctuality, self-discipline, doing your own work and not getting in the way of others, respect for the coaches and manager, respect for your teammates, respect for the rules, written and unwritten. It involves cleanliness and good personal habits. It means maintaining a positive attitude, not grousing, not complaining, not blaming others for your failures. Good humor.
You probably believe that major league athletes are like anybody else in these ways—but they’re not. If you created a scale to measure the people you know in these ways and the average person is a "5", basically all major league athletes are 7s, 8s, 9s, and 10s—when it comes to dealing with their teammates. They may be boorish in dealing with the press or with the public, and you do sometimes get a superstar who is a jackass but who has to be tolerated because of his ability, and you do occasionally get a rogue athlete who is unhappy in his situation and who makes himself a pain in the ass because he is trying to get out of town, but basically. . ..if you don’t behave yourself, you’re not going to be a part of any good major league team. If you’re a fringe guy, a utility infielder or a backup catcher, it is very unlikely that you’re going to be on a major league roster unless you’re a "10". We don’t have any place for fringe players who are negative people. There is no such spot on our roster—or anybody’s roster.
This is true generally in all sports: You have to behave yourself, on the team, if you want to be a part of the team. If you’re a student and you’re five minutes late to class, no big deal. If you’re an athlete and you’re five minutes late to practice, it’s a BIG deal. If you cut a class, if you cut a third of your classes, nobody will say anything. You miss a practice, you had damned well better have a good reason. One of the best sports scenes I have ever seen in a movie is in The Prince of Tides, when Nick Nolte is hired to "coach" Barbra Streisand’s son, who wants to be an athlete but can’t make the team in anything. The kid is angry and disrespectful. What Nolte has to teach him is, you can’t pull that crap in sports. You can get by with that in the classroom; you can get by with that in the family, dealing with your mother. You can get by with it among your friends. If you want to be part of the team, lose the attitude.
The misbehaviors you hear about on a major league team, the conflicts and loser behavior; most of the time that’s really just normal stuff. It’s normal stuff in the rest of the world; it sticks out like a sore thumb on baseball team.
Team Chemistry is a relatively broad term. "Leadership" is a much narrower and more focused term. To be a leader means
a) Being a member of a group,
b) Having a position of prominence within that group,
c) Acting in that position to reinforce the shared values of the group, and
d) Taking actions designed to help the group be successful.
Leadership in sports means taking an active role to reinforce the codes of behavior that are important on a team. That is a great deal narrower and more workable definition than "Character". If the definition of "Leadership" covers the city of Denver, then the definition of "Team Chemistry" would cover the state of Colorado, and the definition of "Character" would cover everything west of the Mississippi.
If this stuff is normal is all of the world and of exaggerated importance in sports, why then would anyone deny that it exists, or that it is relevant?
Linear thinking. Linear logic.
Linear logic assumes that there is a single starting point to the analysis, and that that starting point represents an absolute truth. Statistical analysis is peculiarly prone to linear logic, because each statistic generally represents a single point, which becomes a point of departure. Statistical analysis tends to move from Point A to Point B, with Point A being the statistics that a player has compiled.
But understanding the truth requires remembering that that which has happened is not the only thing which could have happened.
Let us go back now to the Gil Hodges debate that I quoted earlier. The Other Poster wrote that "what you're saying is that Gil's leadership is why Jackie was such a good hitter, such a good defender, such a good base runner, and such a fiery competitor. That these weren't things he flashed at John Muir, Pasadena City College, and UCLA. Nor in the Negro League, nor in the minors before coming up. Gil Hodges is the reason that Jackie was X% better than he'd shown growing up." Well, no, that is not what he is saying. What he is saying is that on a good team. . ..well, this was a great team. What he is saying is that on a great team, players help one another to succeed.
The other poster misrepresents MarisFan’s position in a systematic way, replacing non-linear assertions with linear substitutes. Maris asserts—reasonably and correctly, I think—that we may generally assume that a player who becomes a manager has positive leadership qualities that may have helped his teammates be successful. The Other Poster responds as if Maris had said that this meant that the other players on the team, who did not become managers, did not have these skills:
But beyond that, look at the Boys of Summer and the players who *didn't* become managers. You don't think they didn't also have some similar "valuable qualities"?
Duke
Jackie
Pee Wee
Campy
Amoros
Gilliam
Furillo
Erskine
Podres
Newcombe
But Maris had never in any way suggested that these players did not also have leadership ability. Some of them may have had similar skills; some of them may not have had. Maris is silent on the issue. Suppose that there were eight players on a team (Joe, Jack, John, Jerry, Jess, Junior, Jed and Jason), and suppose that someone said that Jerry was a Methodist. Would any reasonable person read that as saying that Junior was a sinner, or even that any of the others were not also Methodists? Suppose that he said that Jason was rapist; would anyone take that to be a representation that Joe could not also be a rapist?
Of course they would not; it is an entirely arbitrary and capricious misreading of Maris’ statement. Why, then, does the Other Poster rush to this misrepresentation?
I believe that he does so because he is discomfited by the belief that players interact with one another to produce statistical outcomes, rather than that they produce statistical outcomes individually. We are in the habit of seeing individual statistical outcomes as representing the unique skills of those players, without interaction with other forces—in other words, of seeing them as hard, fixed dots which may be used as the foundations of research.
But this is not what they are, in fact; they are happenstance outcomes selected by fate from a wide array of opportunities. The player who hit 23 home runs could just as well have hit 30; he could just as well have hit 5.
It shakes the universe of the Other Poster to see the numbers in this way, but this is unnecessary. The way to avoid that problem is: Potential. If we have statistics that represent the potential of each player, numbers that represent what a player with this background might realistically have done in that season, then we are no longer trapped on that small island that represents the chance outcome of the season, therefore no longer threatened or discomfited to see the actual outcome as coming from an array of options.
By adding Potential to the set of measurements that describe a season, we create a space in which Leadership and Team Chemistry can exist. It is vital to do this because Leadership and Team Chemistry do in fact exist; therefore, if we have a conceptual structure in which they cannot exist, we have a bad conceptual construct.
The world is not linear. When you pretend that is linear to create an understanding of it, what you have is not understanding, it is pretense.
Another problem with the Other Poster’s argument is reflected in this passage:
The problem is that there were *many* players on that team who have similar "leadership" bullshit tossed around. And that's why I was asking you to quantify just how much of it comes down to Gil's Mystical Inspirational Leadership Qualites and how much of it comes down to those same qualities of everyone else... and then to the fact that the all were also really good Baseball Players.
To which you can't. Worse, you give onto Gil special credit while refusing to acknowledge that it because an extremely small "value" if you're forced to give similar "value" to everyone else on the club.
Well, yes, it is true that we can’t measure how large these contributions of Hodges’ were. But the fact that we can’t measure them is no evidence whatsoever that they do not exist. After all, the world can now measure tens of thousands of things that we did not know how to quantify a hundred years ago.
But it’s not a small value. Shortfall—the gap between what players do and what they might potentially have done—is essentially equal to production. Even if there are ten leaders on a team. ..there is plenty of room in the statistical universe to accommodate that.
Look, we do not understand the world, right? We don’t know everything. Because we don’t know everything, we do not always know how much credit might be given where:
The 1955 Dodgers won 98 games, then 4 in the World Series. So how would you split up the credit for the Wins and Losses between:
Offense
Pitching
Defense
Manager
Coaching
General Management
Ownership
Mystical Player Powers
Fan Support
Dumb Ass Luck
Then for each of those things, how would you parcel them out for each player / manager / coach / front office person?
It's fine to talk about Mystical Player Powers, but how many wins did it add? Was Gil worth 2 wins a year just on Mystical Player Powers?
It is true that we can’t measure these things. But we absolutely cannot say that, because these things cannot be measured (yet), therefore they should be ridiculed, dismissed, or ignored. If we can’t measure it yet, we should turn out attention to the question of how it might be measured.
Measuring Potential
I’ve made two separate approaches to measuring potential, one based on Win Shares, and one based on Runs Scored and RBI. The two methods work about the same and reach about the same conclusions, but let’s start with the Runs Scored/RBI method.
Take Mickey Mantle in 1963. Mickey Mantle in 1963 played only 65 games, scored 40 runs and drove in 35. Is that the best season that Mantle might reasonably have had?
Of course it is not; no one would say that it is, short of making the argument that that which happened is the only thing that could possibly have happened. But what is the best season that Mantle might reasonably have had in 1963?
The number of runs that Mantle might potentially have driven and scored in 1963 can be estimated as the number of plate appearances that he might potentially have had, times the number of runs per plate appearance that he might potentially have produced.
OK, how many plate appearances might Mantle potentially have had in 1963?
We assume that a player might potentially have as many as 4.75 plate appearances per scheduled game. In a 162-game season, that’s 770 plate appearances.
However, 770 plate appearances is not a realistic best-case scenario for, let us say, a player who has 300, 250, and 275 plate appearances over the last three seasons. We have to modify the 770 based on the player’s history.
Mickey Mantle in the three previous seasons had had 644, 646 and 502 plate appearances (1960-1962). The highest figure there is 646, but what a player might do next year is not absolutely limited by what he did last year or the year before.
We have these two poles, 646 plate appearances and 770, and we need to pick a point between them. We pick a point between them based on the player’s age by using the formula:
(52 minus Age) / 40
So that if a player is 20 years old, we assume that he might travel 80% of the distance between his "history" (646) and the maximum (770), whereas if he is 31 years old, then we assume that he might traverse 52.5% of the distance. Mantle in 1963 was 31 years old, so we will assume that he might travel 52.5% of the distance. "The distance" is 124 plate appearances; 52.5% of that is 65. 646 + 65 is 711, so we will assume that Mantle in 1963 might potentially have had 711 plate appearances.
711 plate appearances, at what rate of production?
Mantle in his career before 1963 had 1,341 runs scored, 1,152 RBI in 7,201 plate appearances. This is a rate of .346 runs scored/RBI per plate appearance. However, a player can play better this year than his career norms; it is always possible. How much better?
Similar process. We assume that a player’s productivity can go as high as .400 runs scored/RBI per plate appearance. That’s Babe Ruth/Ted Williams territory; that’s about as good as you can do. So then we have two poles: .346 runs/PA, and .400. The player might travel some portion of that distance?
What portion?
Depends on his experience. If the player has 300 career plate appearances, we’ll assume that his potential for improvement relative to previous performance is pretty high. Mantle, however, has 7201 career plate appearances. His potential to out-perform his previous numbers, that late in his career, is more limited.
This is the formula I came up with to estimate the "gap coverage" for the player’s upcoming season; CPA is "career plate appearances":
500 / (1000 + CPA)
For Mantle, with 7,201 plate appearances, this is .061. We thus assume that Mantle’s potential improvement is .061 of the .054 gap between what Mantle has done (.346) and what he might potentially do (.400). .346 + (.054 * .061) = .349; I think that actually would be .350, but if you save all of the decimals it works out to .34944.
If Mantle had had 711 plate appearances in 1963 and produced .34944 runs per plate appearances, that would have been 248 Runs Scored + RBI. Since Mantle had in fact driven in and scored 260 Runs in 1961, just two years earlier, it’s not crazy optimistic to suggest that he might have gotten to 248 in 1963. It’s a reasonable, moderately optimistic estimate of what he might have done.
But, of course, he didn’t do it. On June 5, 1963, chasing a home run hit by Brooks Robinson, Mantle caught his cleats in the chain-link outfield fence, broke his foot and tore his knee all to hell (not a technical description.) He missed most of the rest of the year. He drove in and scored 75 runs, which is 30% of what we estimated that he might potentially have done. He achieved 30% of his potential in that season.
Pee Wee Reese in 1941, his rookie season. It’s a 154-game season, so the upper boundary of his potential plate appearances is 732 (154 * 4.75). His high number of PA for the previous three seasons is zero, so our two poles are zero and 732.
Reese was 21 years old, so we’ll estimate that he could pick up 77.5% of the gap between previous and upper-boundary at bats. 77.5% of 732 is 567, so Reese in 1941 potentially has 567 plate appearances.
Reese has no history as a hitter, but we know that he can’t be worse than about .125 runs scored/RBI per plate appearance. We’ll assume that’s what he is, since we have no other information. The poles are .125 and .400—the upper and lower boundaries of what hitters realistically might do.
However, since he has no experience, we will assume that he has the potential to travel half the distance between these poles:
500 / (1000 + 0 ) = .500
The distance between the poles is .275; if he travels half of that that would be .1375, plus the .125 lower boundary; he has the potential to produce .2625 runs per plate appearance.
If he has 567 plate appearances and produces .2625 runs per plate appearance, that would be 149 runs scored + RBI. Reese in 1941 actually did score and drive in 86 runs (58 scored, 28 driven in.) 86 is 58% of 149, so Pee Wee Reese in 1941 is believed to have achieved about 58% of his potential. Of course, the first-year number is just a wild-ass guess, frankly; the only thing we’re really operating with is his age. After the first year he has a history, and after he has a history we can start to zero in on what the real expectations for him might be.
|
|
|
|
|
Potential
|
|
|
|
|
|
|
|
Previous
|
Production
|
Potential
|
Actual
|
Achievement
|
Player
|
Year
|
Age
|
Productivity
|
Rate
|
Runs
|
Runs
|
Percentage
|
Pee Wee
|
Reese
|
1940
|
21
|
.000
|
.263
|
149
|
86
|
58%
|
Pee Wee
|
Reese
|
1941
|
22
|
.238
|
.277
|
177
|
122
|
69%
|
Pee Wee
|
Reese
|
1942
|
23
|
.201
|
.250
|
179
|
140
|
78%
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Pee Wee
|
Reese
|
1946
|
27
|
.206
|
.242
|
172
|
139
|
81%
|
Pee Wee
|
Reese
|
1947
|
28
|
.209
|
.237
|
168
|
154
|
92%
|
Pee Wee
|
Reese
|
1948
|
29
|
.219
|
.242
|
169
|
171
|
101%
|
Pee Wee
|
Reese
|
1949
|
30
|
.227
|
.246
|
171
|
205
|
120%
|
Pee Wee
|
Reese
|
1950
|
31
|
.235
|
.251
|
185
|
149
|
81%
|
Pee Wee
|
Reese
|
1951
|
32
|
.235
|
.249
|
184
|
178
|
97%
|
Pee Wee
|
Reese
|
1952
|
33
|
.237
|
.249
|
184
|
152
|
83%
|
Pee Wee
|
Reese
|
1953
|
34
|
.236
|
.248
|
178
|
169
|
95%
|
Pee Wee
|
Reese
|
1954
|
35
|
.240
|
.250
|
180
|
167
|
93%
|
Pee Wee
|
Reese
|
1955
|
36
|
.241
|
.250
|
172
|
160
|
93%
|
Pee Wee
|
Reese
|
1956
|
37
|
.241
|
.250
|
172
|
131
|
76%
|
Pee Wee
|
Reese
|
1957
|
38
|
.238
|
.246
|
169
|
62
|
37%
|
Pee Wee
|
Reese
|
1958
|
39
|
.235
|
.243
|
165
|
38
|
23%
|
In 1948 and 1949 Reese scored and drove in more runs than we had projected for him. That’s what I call a Dream Season, when a player exceeds what we perceive as his potential. It happens sometimes, I would guess about one player per team per season has a Dream Season.
When that happens, I just adjust the "potential" number so that it matches the "actual" number, and assume that the player has achieved 100% of his potential in that season. I think that is the obvious thing to do. All we are saying is that our preseason estimate of the player’s upper boundary turned out to be wrong, and his potential was actually higher than we believed it to be. That’s no big deal; that’s just real life. Sometimes players are better than you figure they might be.
Anyway, you can see that there are seven seasons in which Pee Wee Reese achieved 90% of his potential—all of them after Gil Hodges (and Jackie Robinson) joined the Dodgers in 1947, for whatever that’s worth. When you compare that to the other top shortstops of Reese’s era, you can see that none of them had as many high-achievement seasons as did Reese:
Pee Wee Reese
|
Phil Rizzuto
|
Lou Boudreau
|
Al Dark
|
YEAR
|
AGE
|
Achi %
|
Year
|
Age
|
Achi %
|
YEAR
|
AGE
|
Achi %
|
YEAR
|
AGE
|
Achi %
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1940
|
21
|
58%
|
|
|
|
1939
|
21
|
41%
|
|
|
|
1941
|
22
|
69%
|
|
|
|
1940
|
22
|
100%
|
|
|
|
1942
|
23
|
78%
|
1941
|
23
|
80%
|
1941
|
23
|
69%
|
|
|
|
|
|
|
1942
|
24
|
82%
|
1942
|
24
|
57%
|
|
|
|
|
|
|
|
|
|
1943
|
25
|
73%
|
|
|
|
|
|
|
|
|
|
1944
|
26
|
88%
|
1948
|
26
|
100%
|
1946
|
27
|
81%
|
|
|
|
1945
|
27
|
56%
|
1949
|
27
|
67%
|
1947
|
28
|
92%
|
1946
|
28
|
51%
|
1946
|
28
|
64%
|
1950
|
28
|
82%
|
1948
|
29
|
100%
|
1947
|
29
|
83%
|
1947
|
29
|
85%
|
1951
|
29
|
100%
|
1949
|
30
|
100%
|
1948
|
30
|
70%
|
1948
|
30
|
100%
|
1952
|
30
|
89%
|
1950
|
31
|
81%
|
1949
|
31
|
100%
|
1949
|
31
|
64%
|
1953
|
31
|
100%
|
1951
|
32
|
97%
|
1950
|
32
|
100%
|
1950
|
32
|
30%
|
1954
|
32
|
88%
|
1952
|
33
|
83%
|
1951
|
33
|
73%
|
1951
|
33
|
49%
|
1955
|
33
|
65%
|
1953
|
34
|
95%
|
1952
|
34
|
75%
|
|
|
|
1956
|
34
|
68%
|
1954
|
35
|
93%
|
1953
|
35
|
63%
|
|
|
|
1957
|
35
|
80%
|
1955
|
36
|
93%
|
1954
|
36
|
38%
|
|
|
|
1958
|
36
|
63%
|
1956
|
37
|
76%
|
1955
|
37
|
18%
|
|
|
|
1959
|
37
|
61%
|
1957
|
38
|
37%
|
1956
|
38
|
9%
|
|
|
|
1960
|
38
|
48%
|
1958
|
39
|
23%
|
|
|
|
|
|
|
|
|
|
Rizzuto and Boudreau had two seasons each in which they achieved 90% of their potential; Al Dark had three. Reese had as many as all three combined.
Pee Wee Reese
|
Granny Hamner
|
Marty Marion
|
Vern Stephens
|
YEAR
|
AGE
|
Achi %
|
YEAR
|
AGE
|
Achi %
|
YEAR
|
AGE
|
Achi %
|
YEAR
|
AGE
|
Achi %
|
|
|
|
1944
|
17
|
7%
|
|
|
|
|
|
|
|
|
|
1945
|
18
|
6%
|
|
|
|
|
|
|
1940
|
21
|
58%
|
1948
|
21
|
62%
|
|
|
|
1942
|
21
|
100%
|
1941
|
22
|
69%
|
1949
|
22
|
82%
|
1940
|
22
|
63%
|
1943
|
22
|
74%
|
1942
|
23
|
78%
|
1950
|
23
|
95%
|
1941
|
23
|
63%
|
1944
|
23
|
91%
|
|
|
|
1951
|
24
|
78%
|
1942
|
24
|
73%
|
1945
|
24
|
81%
|
|
|
|
1952
|
25
|
96%
|
1943
|
25
|
56%
|
1946
|
25
|
61%
|
|
|
|
1953
|
26
|
100%
|
1944
|
26
|
72%
|
1947
|
26
|
74%
|
1946
|
27
|
81%
|
1954
|
27
|
100%
|
1945
|
27
|
82%
|
1948
|
27
|
100%
|
1947
|
28
|
92%
|
1955
|
28
|
58%
|
1946
|
28
|
64%
|
1949
|
28
|
100%
|
1948
|
29
|
100%
|
1956
|
29
|
49%
|
1947
|
29
|
90%
|
1950
|
29
|
100%
|
1949
|
30
|
100%
|
1957
|
30
|
72%
|
1948
|
30
|
76%
|
1951
|
30
|
60%
|
1950
|
31
|
81%
|
1958
|
31
|
23%
|
1949
|
31
|
90%
|
1952
|
31
|
34%
|
1951
|
32
|
97%
|
1959
|
32
|
15%
|
1950
|
32
|
52%
|
1953
|
32
|
27%
|
1952
|
33
|
83%
|
|
|
|
|
|
|
1954
|
33
|
44%
|
1953
|
34
|
95%
|
|
|
|
1952
|
34
|
24%
|
1955
|
34
|
10%
|
1954
|
35
|
93%
|
|
|
|
|
|
|
|
|
|
1955
|
36
|
93%
|
|
|
|
|
|
|
|
|
|
1956
|
37
|
76%
|
|
|
|
|
|
|
|
|
|
1957
|
38
|
37%
|
|
|
|
|
|
|
|
|
|
1958
|
39
|
23%
|
|
|
|
|
|
|
|
|
|
Granny Hamner had four seasons at 90% of his potential, Marion had two, Vern Stephens five. But these are the best shortstops of the era. Rizzuto and Boudreau are in the Hall of Fame; the other guys were All Stars and perennial All Stars. What may be more instructive is to compare Reese to the shortstops who weren’t perennial All Stars, like Sam Dente, Johnny Lipon and Solly Hemus:
Pee Wee Reese
|
Sam Dente
|
Johnny Lipon
|
Solly Hemus
|
YEAR
|
AGE
|
Achi %
|
Year
|
Age
|
Achi%
|
YEAR
|
AGE
|
Achi %
|
YEAR
|
AGE
|
Achi %
|
|
|
|
|
|
|
1942
|
19
|
9%
|
|
|
|
1940
|
21
|
58%
|
|
|
|
|
|
|
|
|
|
1941
|
22
|
69%
|
|
|
|
|
|
|
|
|
|
1942
|
23
|
78%
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1947
|
25
|
19%
|
1948
|
25
|
90%
|
|
|
|
|
|
|
1948
|
26
|
37%
|
1949
|
26
|
69%
|
1949
|
26
|
8%
|
1946
|
27
|
81%
|
1949
|
27
|
76%
|
1950
|
27
|
100%
|
|
|
|
1947
|
28
|
92%
|
1950
|
28
|
78%
|
1951
|
28
|
53%
|
1951
|
28
|
85%
|
1948
|
29
|
100%
|
1951
|
29
|
35%
|
1952
|
29
|
43%
|
1952
|
29
|
95%
|
1949
|
30
|
100%
|
1952
|
30
|
16%
|
1953
|
30
|
20%
|
1953
|
30
|
94%
|
1950
|
31
|
81%
|
|
|
|
|
|
|
1954
|
31
|
38%
|
1951
|
32
|
97%
|
1954
|
32
|
27%
|
|
|
|
1955
|
32
|
31%
|
1952
|
33
|
83%
|
1955
|
33
|
20%
|
|
|
|
1956
|
33
|
28%
|
1953
|
34
|
95%
|
|
|
|
|
|
|
1957
|
34
|
11%
|
1954
|
35
|
93%
|
|
|
|
|
|
|
1958
|
35
|
80%
|
1955
|
36
|
93%
|
|
|
|
|
|
|
|
|
|
1956
|
37
|
76%
|
|
|
|
|
|
|
|
|
|
1957
|
38
|
37%
|
|
|
|
|
|
|
|
|
|
1958
|
39
|
23%
|
|
|
|
|
|
|
|
|
|
These were not players who had no ability. Johnny Lipon had enough ability that he was in the majors at the age of 19, before the War had really begun to thin the talent; he had enough ability to hit .290 as a first-time regular in 1948, enough ability to hit .293 and score 104 runs in 1950. Solly Hemus’ best seasons are comparable in value to Pee Wee Reese’s best seasons. But Johnny Lipon had two seasons of performing near the level of his potential; Solly Hemus had three. That is the difference between them and Pee Wee Reese.
Three other shortstops of that era were Stan Rojek, Virgil Stallcup and Bill Rigney:
Pee Wee Reese
|
Stan Rojek
|
Virgil Stallcup
|
Bill Rigney
|
YEAR
|
AGE
|
Achi %
|
Year
|
Age
|
Achi%
|
YEAR
|
AGE
|
Achi %
|
YEAR
|
AGE
|
Achi %
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1940
|
21
|
58%
|
|
|
|
|
|
|
|
|
|
1941
|
22
|
69%
|
|
|
|
|
|
|
|
|
|
1942
|
23
|
78%
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1948
|
26
|
84%
|
|
|
|
1946
|
27
|
81%
|
1946
|
27
|
11%
|
1949
|
27
|
55%
|
|
|
|
1947
|
28
|
92%
|
1947
|
28
|
12%
|
1950
|
28
|
64%
|
1946
|
28
|
60%
|
1948
|
29
|
100%
|
1948
|
29
|
100%
|
1951
|
29
|
55%
|
1947
|
29
|
98%
|
1949
|
30
|
100%
|
1949
|
30
|
57%
|
|
|
|
1948
|
30
|
66%
|
1950
|
31
|
81%
|
1950
|
31
|
28%
|
|
|
|
1949
|
31
|
58%
|
1951
|
32
|
97%
|
1951
|
32
|
22%
|
|
|
|
1950
|
32
|
10%
|
1952
|
33
|
83%
|
|
|
|
|
|
|
1951
|
33
|
12%
|
1953
|
34
|
95%
|
|
|
|
|
|
|
1952
|
34
|
20%
|
1954
|
35
|
93%
|
|
|
|
|
|
|
|
|
|
1955
|
36
|
93%
|
|
|
|
|
|
|
|
|
|
1956
|
37
|
76%
|
|
|
|
|
|
|
|
|
|
1957
|
38
|
37%
|
|
|
|
|
|
|
|
|
|
1958
|
39
|
23%
|
|
|
|
|
|
|
|
|
|
Dente, Lipon, Hemus, Rojek, Stallcup and Rigney are typical of the great mass of players, who normally achieve at about 50% of what we estimate to be their potential. I figured potential in two different ways; with both approaches, I came out with the estimate that normal achievement percentages are about 50%.
What is true of Reese is also true of Gil Hodges. Compare Hodges to Vic Wertz, Ted Kluszewski and Joe Adcock:
Gil Hodges
|
Ted Kluszewski
|
Vic Wertz
|
Joe Adcock
|
YEAR
|
AGE
|
Achi %
|
Year
|
Age
|
Achi%
|
YEAR
|
AGE
|
Achi %
|
YEAR
|
AGE
|
Achi %
|
|
|
|
|
|
|
1947
|
22
|
72%
|
1950
|
22
|
70%
|
1947
|
23
|
12%
|
1948
|
23
|
76%
|
1948
|
23
|
62%
|
1951
|
23
|
47%
|
1948
|
24
|
86%
|
1949
|
24
|
70%
|
1949
|
24
|
100%
|
1952
|
24
|
54%
|
1949
|
25
|
100%
|
1950
|
25
|
96%
|
1950
|
25
|
98%
|
1953
|
25
|
89%
|
1950
|
26
|
100%
|
1951
|
26
|
74%
|
1951
|
26
|
78%
|
1954
|
26
|
87%
|
1951
|
27
|
100%
|
1952
|
27
|
74%
|
1952
|
27
|
60%
|
1955
|
27
|
45%
|
1952
|
28
|
86%
|
1953
|
28
|
100%
|
1953
|
28
|
59%
|
1956
|
28
|
96%
|
1953
|
29
|
100%
|
1954
|
29
|
100%
|
1954
|
29
|
48%
|
1957
|
29
|
37%
|
1954
|
30
|
100%
|
1955
|
30
|
100%
|
1955
|
30
|
45%
|
1958
|
30
|
53%
|
1955
|
31
|
78%
|
1956
|
31
|
88%
|
1956
|
31
|
90%
|
1959
|
31
|
74%
|
1956
|
32
|
77%
|
1957
|
32
|
15%
|
1957
|
32
|
97%
|
1960
|
32
|
88%
|
1957
|
33
|
87%
|
1958
|
33
|
30%
|
1958
|
33
|
9%
|
1961
|
33
|
100%
|
1958
|
34
|
62%
|
1959
|
34
|
25%
|
1959
|
34
|
44%
|
1962
|
34
|
65%
|
1959
|
35
|
65%
|
1960
|
35
|
39%
|
1960
|
35
|
74%
|
1963
|
35
|
40%
|
1960
|
36
|
25%
|
1961
|
36
|
47%
|
1961
|
36
|
52%
|
1964
|
36
|
54%
|
1961
|
37
|
30%
|
|
|
|
1962
|
37
|
14%
|
1965
|
37
|
49%
|
1962
|
38
|
18%
|
|
|
|
1963
|
38
|
6%
|
1966
|
38
|
55%
|
1958
|
39
|
23%
|
|
|
|
|
|
|
|
|
|
Hodges was not a better player than Kluszewski, Wertz or Adcock; in fact, he was probably not as good. All three had Career Winning Percentages (Win Shares and Loss Shares) of .639 to .667, whereas Hodges was at .603. But Gil Hodges had five seasons playing at 90% of his potential, and eight seasons over 85%. Kluszewski had four over 90, five over 85, Wertz had four and four, Adcock had two and five. And these, of course, are among the best first basemen of that era; I’m not going to show you the Virgil Stallcups and Stan Rojeks among the first basemen of that era; I’m sure you’ve got the point.
Gil Hodges does not belong in the Hall of Fame, in my view, because he fails to reach the Gray Area. There are two practical standards in the Hall of Fame debate: a minimum standard, above which some players are Hall of Famers and some are not, and a "hard line" standard, above which essentially all players have historically been elected. If Hodges (as a player) was in the Gray Area, then, in my view, Maris’ argument for him to selected based on his intangible contributions might be persuasive. But since Hodges failed to reach the minimum standards, to me he isn’t in the discussion. To put him in the Hall of Fame, you would have to put him ahead of a large number of men who were simply better players.
Let’s compare Carl Furillo’s Achivement Percentage to Hank Bauer, the right fielder of the Yankees, or Don Mueller, the right fielder in that era of the Dodgers’ rival, the Giants, or Wally Post of the Reds:
Carl Furillo
|
Hank Bauer
|
Don Mueller
|
Wally Post
|
YEAR
|
AGE
|
Achi %
|
Year
|
Age
|
Achi%
|
YEAR
|
AGE
|
Achi %
|
YEAR
|
AGE
|
Achi %
|
|
|
|
|
|
|
1948
|
21
|
14%
|
1951
|
21
|
9%
|
|
|
|
|
|
|
1949
|
22
|
4%
|
1952
|
22
|
8%
|
|
|
|
|
|
|
1950
|
23
|
100%
|
|
|
|
1946
|
24
|
47%
|
|
|
|
1951
|
24
|
64%
|
1954
|
24
|
93%
|
1947
|
25
|
98%
|
1948
|
25
|
12%
|
1952
|
25
|
57%
|
1955
|
25
|
100%
|
1948
|
26
|
53%
|
1949
|
26
|
78%
|
1953
|
26
|
64%
|
1956
|
26
|
78%
|
1949
|
27
|
100%
|
1950
|
27
|
79%
|
1954
|
27
|
94%
|
1957
|
27
|
63%
|
1950
|
28
|
100%
|
1951
|
28
|
52%
|
1955
|
28
|
82%
|
1958
|
28
|
52%
|
1951
|
29
|
87%
|
1952
|
29
|
82%
|
1956
|
29
|
44%
|
1959
|
29
|
76%
|
1952
|
30
|
52%
|
1953
|
30
|
65%
|
1957
|
30
|
47%
|
1960
|
30
|
50%
|
1953
|
31
|
84%
|
1954
|
31
|
64%
|
1958
|
31
|
14%
|
1961
|
31
|
52%
|
1954
|
32
|
72%
|
1955
|
32
|
75%
|
|
|
|
1962
|
32
|
54%
|
1955
|
33
|
93%
|
1956
|
33
|
96%
|
|
|
|
1963
|
33
|
8%
|
1956
|
34
|
78%
|
1957
|
34
|
69%
|
|
|
|
|
|
|
1957
|
35
|
67%
|
1958
|
35
|
58%
|
|
|
|
|
|
|
1958
|
36
|
74%
|
1959
|
36
|
44%
|
|
|
|
|
|
|
1959
|
37
|
11%
|
1960
|
37
|
36%
|
|
|
|
|
|
|
|
|
|
1961
|
38
|
18%
|
|
|
|
|
|
|
Furillo has five seasons at 85% of his potential. Bauer has one, Mueller has two, Post has two.
You can compare Jackie Robinson to Red Schoendienst and Nellie Fox, the other Hall of Fame second basemen of his era; they have more seasons at 85 or 90% of their potential than Jackie does:
Jackie Robinson
|
Red Schoendienst
|
Nellie Fox
|
YEAR
|
AGE
|
Achi %
|
Year
|
Age
|
Achi%
|
YEAR
|
AGE
|
Achi %
|
|
|
|
|
|
|
1949
|
21
|
43%
|
|
|
|
1945
|
22
|
94%
|
1950
|
22
|
46%
|
|
|
|
1946
|
23
|
65%
|
1951
|
23
|
94%
|
|
|
|
1947
|
24
|
78%
|
1952
|
24
|
69%
|
|
|
|
1948
|
25
|
58%
|
1953
|
25
|
100%
|
|
|
|
1949
|
26
|
91%
|
1954
|
26
|
99%
|
|
|
|
1950
|
27
|
85%
|
1955
|
27
|
99%
|
1947
|
28
|
100%
|
1951
|
28
|
85%
|
1956
|
28
|
99%
|
1948
|
29
|
92%
|
1952
|
29
|
95%
|
1957
|
29
|
100%
|
1949
|
30
|
100%
|
1953
|
30
|
100%
|
1958
|
30
|
80%
|
1950
|
31
|
80%
|
1954
|
31
|
100%
|
1959
|
31
|
96%
|
1951
|
32
|
87%
|
1955
|
32
|
70%
|
1960
|
32
|
90%
|
1952
|
33
|
81%
|
1956
|
33
|
54%
|
1961
|
33
|
72%
|
1953
|
34
|
98%
|
1957
|
34
|
95%
|
1962
|
34
|
83%
|
1954
|
35
|
57%
|
1958
|
35
|
43%
|
1963
|
35
|
62%
|
1955
|
36
|
42%
|
1960
|
37
|
25%
|
1964
|
36
|
47%
|
1956
|
37
|
54%
|
1961
|
38
|
13%
|
1965
|
37
|
3%
|
|
|
|
1962
|
39
|
26%
|
|
|
|
I’m not suggesting that ONLY the Dodgers had players who achieved at the level of their potential. Other teams had some of these players as well. But Jackie, starting his career very late, still has four seasons achieving at 90% of his potential, and seven seasons over 80%. Those are still very, very good achievement percentages. Like Furillo, Reese and Hodges, Jackie did what he was capable of doing, year in and year out. As did Junior Gilliam; as did Duke Snider, although not as much so as Reese or Hodges:
Jim Gilliam
|
Duke Snider
|
YEAR
|
AGE
|
Achi %
|
Year
|
Age
|
Achi%
|
|
|
|
1947
|
20
|
7%
|
|
|
|
1948
|
21
|
29%
|
|
|
|
1949
|
22
|
100%
|
|
|
|
1950
|
23
|
99%
|
1953
|
24
|
100%
|
1951
|
24
|
87%
|
1954
|
25
|
72%
|
1952
|
25
|
77%
|
1955
|
26
|
74%
|
1953
|
26
|
100%
|
1956
|
27
|
74%
|
1954
|
27
|
100%
|
1957
|
28
|
68%
|
1955
|
28
|
100%
|
1958
|
29
|
71%
|
1956
|
29
|
89%
|
1959
|
30
|
74%
|
1957
|
30
|
77%
|
1960
|
31
|
82%
|
1958
|
31
|
45%
|
1961
|
32
|
65%
|
1959
|
32
|
64%
|
1962
|
33
|
78%
|
1960
|
33
|
34%
|
1963
|
34
|
78%
|
1961
|
34
|
47%
|
1964
|
35
|
44%
|
1962
|
35
|
31%
|
1965
|
36
|
58%
|
1963
|
36
|
57%
|
1966
|
37
|
32%
|
1964
|
37
|
19%
|
Even Campanella. . ..of course we all know that Campanella had injuries that shortened his career and kept him from performing at his best in several of his later seasons. But if you compare Campanella to other talented catchers of his era, to Ed Bailey and Smoky Burgess and Del Rice, you can see that Campanella, while he was not Yogi Berra in this respect, was still achieving at a high level relative to his potential:
Roy Campanella
|
Ed Bailey
|
Smoky Burgess
|
Del Rice
|
YEAR
|
AGE
|
Achi %
|
Year
|
Age
|
Achi%
|
YEAR
|
AGE
|
Achi %
|
YEAR
|
AGE
|
Achi %
|
|
|
|
|
|
|
1949
|
22
|
11%
|
1945
|
22
|
38%
|
|
|
|
|
|
|
1951
|
24
|
29%
|
1946
|
23
|
14%
|
|
|
|
1955
|
24
|
5%
|
1952
|
25
|
72%
|
1947
|
24
|
48%
|
|
|
|
1956
|
25
|
94%
|
1953
|
26
|
39%
|
1948
|
25
|
38%
|
1948
|
26
|
62%
|
1957
|
26
|
55%
|
1954
|
27
|
55%
|
1949
|
26
|
37%
|
1949
|
27
|
92%
|
1958
|
27
|
56%
|
1955
|
28
|
97%
|
1950
|
27
|
68%
|
1950
|
28
|
81%
|
1959
|
28
|
49%
|
1956
|
29
|
40%
|
1951
|
28
|
57%
|
1951
|
29
|
99%
|
1960
|
29
|
75%
|
1957
|
30
|
41%
|
1952
|
29
|
78%
|
1952
|
30
|
80%
|
1961
|
30
|
58%
|
1958
|
31
|
36%
|
1953
|
30
|
49%
|
1953
|
31
|
100%
|
1962
|
31
|
48%
|
1959
|
32
|
76%
|
1954
|
31
|
21%
|
1954
|
32
|
42%
|
1963
|
32
|
69%
|
1960
|
33
|
49%
|
1955
|
32
|
19%
|
1955
|
33
|
88%
|
1964
|
33
|
42%
|
1961
|
34
|
61%
|
1956
|
33
|
26%
|
1956
|
34
|
52%
|
1965
|
34
|
29%
|
1962
|
35
|
69%
|
1957
|
34
|
38%
|
1957
|
35
|
47%
|
|
|
|
1963
|
36
|
41%
|
1958
|
35
|
20%
|
|
|
|
|
|
|
1964
|
37
|
21%
|
1960
|
37
|
7%
|
|
|
|
|
|
|
1965
|
38
|
20%
|
1961
|
38
|
29%
|
|
|
|
|
|
|
1966
|
39
|
14%
|
|
|
|
|
|
|
|
|
|
1967
|
40
|
14%
|
|
|
|
Campanella had three seasons playing at 90% of his potential; Ed Bailey and Smoky Burgess had one each. Catchers get hurt.
The difference between a Hall of Famer and a pretty good player, in many cases, is simply that the Hall of Famer achieves consistently at or near the level of his potential. Bob Bailey was a third baseman of about the same age as Ron Santo, and he had potential as a hitter not that different from Santo’s. Bailey, signed as an amateur, received the largest signing bonus that any amateur had ever received—certainly attesting to the perception of his potential.
Ron Santo
|
Bob Bailey
|
YEAR
|
AGE
|
Achi %
|
Year
|
Age
|
Achi%
|
|
|
|
1962
|
19
|
7%
|
1960
|
20
|
57%
|
1963
|
20
|
64%
|
1961
|
21
|
89%
|
1964
|
21
|
70%
|
1962
|
22
|
60%
|
1965
|
22
|
78%
|
1963
|
23
|
94%
|
1966
|
23
|
57%
|
1964
|
24
|
100%
|
1967
|
24
|
29%
|
1965
|
25
|
95%
|
1968
|
25
|
39%
|
1966
|
26
|
94%
|
1969
|
26
|
72%
|
1967
|
27
|
100%
|
1970
|
27
|
100%
|
1968
|
28
|
91%
|
1971
|
28
|
100%
|
1969
|
29
|
100%
|
1972
|
29
|
69%
|
1970
|
30
|
97%
|
1973
|
30
|
99%
|
1971
|
31
|
81%
|
1974
|
31
|
86%
|
1972
|
32
|
72%
|
1975
|
32
|
33%
|
1973
|
33
|
72%
|
1976
|
33
|
25%
|
1974
|
34
|
36%
|
1977
|
34
|
13%
|
|
|
|
1978
|
35
|
19%
|
The two players are otherwise almost identical—right-handed hitting third basemen, National Leaguers of the same era, both reached the majors at a young age, both men right-handed power hitters, both quite slow afoot by the time they were 28, 29 years old. Ron Santo was fighting diabetes all of his career, all of his life. But Santo, for whatever reason, achieved near the level of his potential in eight or nine seasons; Bailey, in three seasons.
You all know, I assume, that these numbers are fairly imprecise, that this method is crude and that the approach is unproven. All true, but let me point this out: that almost all that I have said here is obviously true if you set aside the numbers. What is new here is that I have proposed a way to put numbers on these things. But set aside the numbers, what I am saying is that Pee Wee Reese played at the level of his ability in an unusually high number of seasons. I think that it obviously true.
I have created a crude statistical companion to an obviously true statement. But by doing this, we enable ourselves to quasi-document a further statement which is less obvious, but still has the ring of truth when you think about it: that the Dodgers of the 1950s had a great team in part because they had a core of very good players who played at the full level of their potential year in and year out.
And simply by doing that, we are one step further on the road toward being able to measure Team Chemistry—not that we are close to that goal, not that we are actively pursuing that goal, even. We are a little bit closer to it than we were before.
And, of course, there is more to winning pennants than that. The Philadelphia Phillies of the 1950s also had several core players who consistently achieved near the level of their potential (Richie Ashburn, Del Ennis, Willie Jones, Granny Hamner, Robin Roberts, Curt Simmons), but they won only the one pennant, 1950, because they just didn’t have enough talent. The Cubs of the 1960s had a group of players who played consistently near the level of their potential (Santo, Banks, Billy Williams, Beckert, Kessinger, Ferguson Jenkins.) Good team chemistry doesn’t win for you if you don’t have the raw materials. Team chemistry is not the only variable in which teams win—but it is one variable.
And back to Gil Hodges. ..let’s look at Tommie Agee. Hodges managed Agee from 1969 to 1971:
YEAR
|
AGE
|
Achi %
|
1966
|
23
|
100%
|
1967
|
24
|
56%
|
1968
|
25
|
23%
|
1969
|
26
|
95%
|
1970
|
27
|
100%
|
1971
|
28
|
57%
|
1972
|
29
|
54%
|
1973
|
30
|
33%
|
As far as I know, every person who was there will tell you that Gil Hodges helped Tommie Agee to achieve his potential. Agee was floundering before he came to Hodges; he floundered again after Hodges died. Under Hodges, he was the player he had the ability to be.
Again, not saying that Hodges uniquely had this ability, or that he alone was the reason the Dodgers of the 50s were consistently successful. It is very likely that Gil Hodges learned some of what he knew about helping other players achieve their potential from Pee Wee Reese, from Jackie Robinson, from Walter Alston and Leo Durocher.
But the statistical universe makes more sense if you assume that this "interactive ability" exists than if you assume that it doesn’t. Earlier I showed you that Nellie Fox had eight seasons in which he achieved at 90% of his potential—very unusual. When I talked to Joe Morgan a couple of months ago, Morgan spoke at length, and emotionally, about how much Nellie Fox helped him to achieve his potential. Why in the world would we not believe him?
OK, I told you I developed a second approach to measuring potential, achievement and shortfall. .. this research was actually done last year, so I’m not sure that I’ll remember how I did all of this.
(Author spends a half hour rummaging through old computer files, trying to find last year’s work on potential.)
Oh, wow; I can see why I dropped this; it leaves too many things out. OK, my other approach to estimating potential was to start with the player’s Win Shares over the last three seasons, and use that to calculate his Established Win Share Level. The Established Win Share Level for a player entering 2014 is
1 times his Win Shares in 2011, plus
2 times his Win Shares in 2012, plus
3 times his Win Shares in 2013,
Divided by six,
But not less than 75% of his Win Shares in 2013.
Let’s go back to Mickey Mantle in 1963. Mantle’s Win Shares in 1960-61-62 are 36-48-33, which makes an Established Level of 38.50.
Mantle was 31 years old in 1963; we assume that he has the potential to do better than he has in recent years based on (42 minus age); in other words, we assume that a 31-year-old has the potential to improve by 11 Win Shares from his previous established level. That puts Mantle at 49.50 potential Win Shares for 1963.
However, that’s frankly a crazy level of production, 50 Win Shares, and we don’t want to say that anybody has THAT kind of potential, so we moderate the previous estimate by taking two times this number, adding 25, and dividing by three. That puts Mantle at 41.33 Potential Win Shares for 1963—still an MVP number. Mantle had in fact won the MVP Award in 1962, with 33 Win Shares.
The "plus 25/divide by 3" stage has the effect of creating larger increases for young, unproven players, but smaller increases for established stars. We’ll call this adjustment the governor; it acts like a governor on an engine. Mantle actually earned 14 Win Shares in 1963, so that’s a little more than 30% (14 of 41.3). . .basically the same number we got by using the other method.
Let’s do Pee Wee Reese. For purposes of this analysis we will treat Pee Wee’s 1942 and 1946 seasons as consecutive seasons. These are Pee Wee’s Win Shares, by season and by age:
Year
|
Age
|
WS
|
1940
|
21
|
13
|
1941
|
22
|
15
|
1942
|
23
|
27
|
1946
|
27
|
26
|
1947
|
28
|
26
|
1948
|
29
|
23
|
1949
|
30
|
32
|
1950
|
31
|
20
|
1951
|
32
|
22
|
1952
|
33
|
23
|
1953
|
34
|
21
|
1954
|
35
|
26
|
1955
|
36
|
18
|
1956
|
37
|
14
|
1957
|
38
|
4
|
1958
|
39
|
4
|
Based on that, we can figure his Established Win Share Level after each season:
Year
|
Age
|
WS
|
EWSL
|
1940
|
21
|
13
|
9.8
|
1941
|
22
|
15
|
13.5
|
1942
|
23
|
27
|
20.7
|
1946
|
27
|
26
|
24.5
|
1947
|
28
|
26
|
26.2
|
1948
|
29
|
23
|
24.5
|
1949
|
30
|
32
|
28.0
|
1950
|
31
|
20
|
24.5
|
1951
|
32
|
22
|
23.0
|
1952
|
33
|
23
|
22.2
|
1953
|
34
|
21
|
21.8
|
1954
|
35
|
26
|
23.8
|
1955
|
36
|
18
|
21.2
|
1956
|
37
|
14
|
17.3
|
1957
|
38
|
4
|
9.7
|
1958
|
39
|
4
|
5.7
|
But what we need, of course, is not his Established Win Share Level AFTER that season, but his level BEFORE that season:
Year
|
Age
|
EWSL
|
PEWSL
|
1940
|
21
|
9.8
|
0
|
1941
|
22
|
13.5
|
9.8
|
1942
|
23
|
20.7
|
13.5
|
1946
|
27
|
24.5
|
20.7
|
1947
|
28
|
26.2
|
24.5
|
1948
|
29
|
24.5
|
26.2
|
1949
|
30
|
28.0
|
24.5
|
1950
|
31
|
24.5
|
28.0
|
1951
|
32
|
23.0
|
24.5
|
1952
|
33
|
22.2
|
23.0
|
1953
|
34
|
21.8
|
22.2
|
1954
|
35
|
23.8
|
21.8
|
1955
|
36
|
21.2
|
23.8
|
1956
|
37
|
17.3
|
21.2
|
1957
|
38
|
9.7
|
17.3
|
1958
|
39
|
5.7
|
9.7
|
From the Previously Established Win Share Level we make an adjustment based on the player’s age (42 – Age), assuming that he might over-achieve until he is 42 years old:
Year
|
Age
|
PEWSL
|
Age Adjustment
|
1940
|
21
|
0
|
21.0
|
1941
|
22
|
9.8
|
29.8
|
1942
|
23
|
13.5
|
32.5
|
1946
|
27
|
20.7
|
35.7
|
1947
|
28
|
24.5
|
38.5
|
1948
|
29
|
26.2
|
39.2
|
1949
|
30
|
24.5
|
36.5
|
1950
|
31
|
28.0
|
39.0
|
1951
|
32
|
24.5
|
34.5
|
1952
|
33
|
23.0
|
32.0
|
1953
|
34
|
22.2
|
30.2
|
1954
|
35
|
21.8
|
28.8
|
1955
|
36
|
23.8
|
29.8
|
1956
|
37
|
21.2
|
26.2
|
1957
|
38
|
17.3
|
21.3
|
1958
|
39
|
9.7
|
12.7
|
And then we apply the governor, moving everything in the direction of 25.0:
Year
|
Age
|
Age Adjustment
|
Governor
|
Potential
|
1940
|
21
|
21.0
|
25
|
22.3
|
1941
|
22
|
29.8
|
25
|
28.2
|
1942
|
23
|
32.5
|
25
|
30.0
|
1946
|
27
|
35.7
|
25
|
32.1
|
1947
|
28
|
38.5
|
25
|
34.0
|
1948
|
29
|
39.2
|
25
|
34.4
|
1949
|
30
|
36.5
|
25
|
32.7
|
1950
|
31
|
39.0
|
25
|
34.3
|
1951
|
32
|
34.5
|
25
|
31.3
|
1952
|
33
|
32.0
|
25
|
29.7
|
1953
|
34
|
30.2
|
25
|
28.4
|
1954
|
35
|
28.8
|
25
|
27.6
|
1955
|
36
|
29.8
|
25
|
28.2
|
1956
|
37
|
26.2
|
25
|
25.8
|
1957
|
38
|
21.3
|
25
|
22.6
|
1958
|
39
|
12.7
|
25
|
16.8
|
On which basis, we can estimate Pee Wee Reese’s achievement percentage in each season of his career:
|
|
|
Achievement
|
Year
|
Age
|
Actual
|
Percentage
|
1940
|
21
|
13
|
58%
|
1941
|
22
|
15
|
53%
|
1942
|
23
|
27
|
90%
|
1946
|
27
|
26
|
81%
|
1947
|
28
|
26
|
76%
|
1948
|
29
|
23
|
67%
|
1949
|
30
|
32
|
100%
|
1950
|
31
|
20
|
58%
|
1951
|
32
|
22
|
70%
|
1952
|
33
|
23
|
78%
|
1953
|
34
|
21
|
74%
|
1954
|
35
|
26
|
94%
|
1955
|
36
|
18
|
64%
|
1956
|
37
|
14
|
54%
|
1957
|
38
|
4
|
18%
|
1958
|
39
|
4
|
24%
|
Some of these numbers are very similar to what we got by the other method, some of them not so much. The other method is better than this one, but. . ..it’s a new area, we want to examine all the options.
The way this should really be done, I think, is to adapt the methodology we have developed from doing projections; I don’t have the technical ability to do this, but this is how I think it should be done. Take a next-season projection system, and modify it so that it makes an optimistic projection for the player’s next season, rather than a central-tendency projection. Actually, in the Handbook, we DO make optimistic projections, for each player, but I mean more optimistic; if you think a player should hit .270, project him to hit .290; if you think he should play 115 games, project him to play in 135. Call those projections "potential", then compare the real-life numbers with the projections, figure the achievement percentage, figure the shortfall.
What some of you are going to want to do is do projections based on WAR. That won’t work. . .I mean, go ahead and try, but I am telling you absolutely that it isn’t going to work.
The reason it won’t work is that WAR is a "sensitive" measurement, because WAR is derived from a comparison of two calculations: the number of wins the player HAS contributed, and the number of wins a replacement player WOULD HAVE contributed. When you derive a value from a comparison in that way, it makes the measurement unstable from season to season. In other words, suppose that we said that a player’s value was not in his batting average, but in his ability to hit more than .200. .. .a true statement in general terms. But whereas a player’s batting average will go up or down by 10 to 15% each season, his batting average over .200 will go up or down by 30 to 50% each season. It is much less stable. When you then compare the ACTUAL batting average to the POTENTIAL batting average, that’s a second derivative, the same type of instability added into the system a second time. The result is going to be massive instability in the outcome—numbers that bounce around so wildly that nobody will take them seriously.
It’s like this: You can stand on a ladder to reach a high place, or you can stand on a sawhorse. But you can’t place a ladder on top of a sawhorse and stand on that, because if you do, you have the combined instability of the ladder and the sawhorse, and the result is that you absolutely are going to fall and injure yourself. Same thing; figuring achievement potential based on WAR would be placing a ladder on top of a sawhorse. That’s actually one of the big problems with using Win Shares for this method; Win Shares are more stable than WAR, but less stable than raw batting statistics. You need to start with a stable base.
Well. . ..long article and I don’t have a closing paragraph for you; hope you got something out of it. A great deal of my career has been based on pointing out things that are entirely obvious once you point them out. This is one of those things; what I am saying here is both (I believe) entirely new to sabermetrics, and mostly pretty obvious. I am merely pointing out that potential can be measured, and that there are tremendous analytical opportunities that will be open to us once we can agree on how to measure it.