Bill--To what extent do you suppose it's purely my imagination that tells me current MLB players are not as versatile as they used to be? When I was growing up, watching mostly Stengel-managed teams, it seemed that players would be shifted from position to position three or four times a game, and start at several different positions a month, and move up and down the lineup on a daily basis? (I remember Casey using Ed Kranepool as a leadoff batter--yes, I do.) Is there a quick-n-dirty way we could quantify the trend towards stability in lineups, position, and in-game position switches--or refute that trend? Do any of these three perceived changes seem more real than any other?

--SGoldleaf

My QND (quick-n-dirty) study of this suggests that your perception may not be consistent with the facts.  I took all players who

a)      played 100 or more games (in a season),

b)      had 300 or more PA, and

c)      played in either the 1960s or the 2000s. 

These players I assigned random numbers (to decide who to check first), and began checking to what extent the players were used in multiple roles.

The method I outlined in the “Hey, Bill” section didn’t exactly work optimally, so I switched to a different method. . ..let’s forget the one outlined before, doesn’t matter.   The better method was to record, for each player,

1)      Games played at his number one defensive position,

2)      Games played at his second defensive position,

3)      Games played at his third defensive position,

4)      Games played at his primary lineup position,

5)      Games played at his second lineup position,

6)      Games played at his third most-common lineup position.

And all of this data, I should remember to say, was from Retrosheet, the brainchild of the immortal Dave Smith. 

The top three points were the “fielding versatility” elements, which were made into a “Versatility Score” by dividing the square of the sum by the sum of the squares.  In other words, if a player played 10 games, of which 4 were at one position, 3 at a second, and 3 at a third, then his Fielding Versatility Scores would be 2.94:

10² 

-------                        &nb​sp;              =   2.94

4²  + 3²  + 3²

If a player played 10 games, 9 of them at one position and 1 at a second position, his Fielding Versatility Score would be 1.22—100 divided by 82.  

The “Lineup Versatility Score” was the same, except that there was a fourth element.   Since a player can’t occupy multiple lineup positions in the same game and must occupy some lineup position in every game, we can easily calculate how many games the player has played at other lineup positions by simply subtracting the total played at his three top lineup positions from his games played. 

The player’s “Versatility Score” is the average of his Fielding Versatility Score and his Lineup Versatility Score.  

I figured the Versatility Scores for 60 players from the 1960s, and for 68 players from the years 2000-2007.   The most versatile players in the study were John Briggs in 1968 and Jolbert Cabrera in 2004:

Briggs in the field:   1B—36, CF—34, RF—22, LF—9

    Games by lineup position    5-6-23     1-21-11    29-2-12

Cabrera in the field:   3B—36, 1B—23, LF—21, 2B—18, SS—14, DH—5,

                                    CF—1, RF—1

    Games by lineup position     1-10-12    1-16-35   20-10-8

Briggs’ Versatility Score is 3.34; Cabrera’s is 3.18.

On the other end of the scale, the lowest possible Versatility Score is 1.00, and there was one player in the study who achieved that score.   Rocky Colavito in 1965 just missed.   Rocky played 162 games, all of them in right field, but also played two innings in left field.  In the lineup he was in the cleanup spot in every game (leading the league in RBI), giving him a Versatility Score of 1.0062.  The only player to hit 1.0000 was A-Rod in 2000—148 games, every inning at shortstop, every at bat hitting third. 

A study of 128 players is not entirely convincing, but the Versatility Scores for modern players were actually higher than for the 1960s players:

2000-2007       Fielding:    1.38           Lineup:   2.69       Overall:   2.04

1960-1969       Fielding:    1.29           Lineup:   2.43       Overall:   1.86

The standard deviations are 0.57 for the 21^st century players, 0.59 for the 1960s. 

As I say, the study is not large enough to be entirely convincing, but. . ..pitching staffs have expanded since the 1960s.   In the 1960s teams would carry nine or ten pitchers, leaving 15 or 16 position players.   Now teams carry about twelve pitchers, meaning thirteen position players.   One would assume that this would mean that modern players would have to play more positions.   

Goldleaf posted a later question, in which he suggested that his initial suggestion was too broad, and that what we was really remembering was something more specific, like catchers doubling as first basemen and slow, mediocre first basemen being used as leadoff men (who is the modern Ed Kranepool, anyway?).  

            It is possible, consistent with this study, that modern players play more positions, but with less “eccentric” combinations.   There are other things that could be studied here, like players switching defensive positions in the game.   I’m just cracking open a little window on this subject.   But the view from the little window looks like modern players may be used in more different roles, rather than less.