February 20, 2004

The A-Rod/Jeter Defense Dilemma

Shortstops' Moves to Third Base

by James Click

Printer-friendly

Alex Rodriguez's trade to the Yankees has elicited plenty of spirited debate on several related topics, notably what to do with Derek Jeter and his matador defense at short. Reader Mark Shirk had this to say:

With an nearly imminent A-Rod to the Yankees trade, I got to thinking about how a move to 3B would affect the value of Derek Jeter. I figured out, using Clay Davenport's equations, that a move to 3B would mean that Jeter's RARP would drop about 4 runs over the course of a full season or roughly 154 games. However since Jeter is such a bad defensive player (-22.5 FRAA per 154 games from 2001-2003) the move might actually benefit him.

Is it unreasonable to think that Jeter would be 15 runs below average as a 3B? I don't think it is even out of the realm of potability for him to be only 10 runs below average. All told that is an 8-run gain in value for Jeter, a pretty significant sum. Am I wrong in thinking this?

Thanks for the note, Mark. Before delving into the question at hand, it's worth noting, as BP's Mark Armour pointed out, the sudden and surprising reversal of the opinion of Jeter's defensive abilities. Jeter has quickly gone from a heralded and graceful fielder to the obvious stepchild next to Rodriguez, the same player who was chided for tripping over his own feet his first day as a Ranger. There seems little public opposition to the conclusion that it is Jeter that should move to the hot corner.

The basis for these conclusions are largely based on empirical evidence, backed lightly with what few defensive statistics are available. With the shortcomings inherent to current defensive statistics and our own visual impressions of players in the field, there is folly to be found by basing any decisions about the optimal defensive alignment of the Yankee infield on this data. Instead, by combining Clay Davenport's defensive statistics and some similar situations in the past, a better idea of what awaits the Bronx Basemen can be gathered.

Back to the question at hand, looking for examples in the past couple decades, there are a few examples of players who moved from the role of everyday shortstop to everyday third baseman. Restricting the search to all seasons 1980-2003 and only considering players who logged at least 500 innings in the field, there are 14 players who played every day at short one season and every day at third the next. However, only five continued to play everyday third the next season. The list, followed by the years of the switch:

Vance Law, 1982-83
Travis Fryman, 1992-93
Cal Ripken Jr., 1996-97
John Valentin, 1996-97
Tony Batista, 1999-2000

The rest of the list--the guys who only played one season at third--does have some qualifying examples, including a young Gary Sheffield, but for the most part, they are players who switched between the positions routinely. Howard Johnson and Jose Hernandez are two of the better-known examples.

Looking at the players above and throwing in a few other examples that fit the profile, we can look at a variety of defensive statistics to get an idea of how well these players adapted to the hot corner. First, look at how their last few seasons before the switch went:

Player          Year    Team    Pos  RAR/AdjG  BO/AdjG  Rate
CAL RIPKEN      1995    BAL-A   SS      21.8    2.93    106
CAL RIPKEN      1996    BAL-A   SS      11.8    2.95    96
GARY SHEFFIELD  1988    MIL-A   SS      -4.2    2.71    75
GARY SHEFFIELD  1989    MIL-A   SS      7.4     3.16    89
TRAVIS FRYMAN   1992    DET-A   SS      18.4    3.10    101
TRAVIS FRYMAN   1993    DET-A   SS      12.3    2.92    96
VANCE LAW       1982    CHI-A   SS      9.5     2.65    91
TONY BATISTA    1998    ARI-N   SS      17.9    2.87    104
TONY BATISTA    1999    ARI-N   SS      26.1    2.82    111
TONY BATISTA    1999    TOR-A   SS      23.7    2.85    108
JOHN VALENTIN   1995    BOS-A   SS      15.0    2.86    98
JOHN VALENTIN   1996    BOS-A   SS      22.4    3.05    107
TOTAL           ---     ---     SS      15.2    2.9     98.5

Admittedly, this is rather confusing. What's been done here is an adjustment to several of Clay Davenport's defensive statistics. In order to compare between seasons and eliminate the effect of playing time discrepancies, several of Clay's stats have been divided by the estimated number of full nine-inning games these players played (AdjG). (Note: In the case of RAR/AdjG, the result has been multiplied by 100 to make the result more easily readable.) The first column is Runs Above Replacement (RAR), the next Batting Outs (BO), and the final column is "Rate", a metric centered around 100 that calculates the approximate number of runs a fielder is worth per 100 games. A fielder worth 10 runs more than average will be 110; similarly, 87 is 13 runs worse than average.

From the numbers, it's clear that the group was, at least immediately before switching to the hot corner, slightly below average for a shortstop.

The numbers for the players' first two years after switching to third:

Player          Year    Team    Pos  RAR/AdjG  BO/AdjG  Rate
CAL RIPKEN      1997    BAL-A   3B      8.4     2.94    98
CAL RIPKEN      1998    BAL-A   3B      5.8     2.87    96
GARY SHEFFIELD  1990    MIL-A   3B      3.3     3.06    92
GARY SHEFFIELD  1991    MIL-A   3B      -2.5    3.19    86
TRAVIS FRYMAN   1994    DET-A   3B      3.5     3.14    94
TRAVIS FRYMAN   1995    DET-A   3B      28.5    2.92    120
VANCE LAW       1983    CHI-A   3B      14.9    2.48    101
VANCE LAW       1984    CHI-A   3B      0.0     2.69    88
TONY BATISTA    2000    TOR-A   3B      20.1    3.01    111
TONY BATISTA    2001    BAL-A   3B      17.9    2.79    108
TONY BATISTA    2001    TOR-A   3B      16.1    3.20    106
JOHN VALENTIN   1997    BOS-A   3B      13.0    2.90    102
JOHN VALENTIN   1998    BOS-A   3B      28.8    3.04    119
TOTAL           ---     ---     3B      12.4    2.9     101.9

The numbers are both remarkably similar in total and when looking at individual players. Note that all of these metrics are adjusted for position, meaning that the results point to the fact that average shortstops become average third basemen. Similar conclusions can be drawn from looking at more traditional mainstream defensive statistics such as range factor and fielding percentage. The sample group posted a range factor of 98.8% against average while at short, 99.6% at third; fielding percentage was virtually league average in both cases.

In the end, the bottom line may lie in the fact that shortstops handle, on average, two more balls per game than third basemen. While there's no reason to think that the fielding quality of Jeter or Rodriguez will be substantially improved or worsened by the move, putting the worse fielder in front of 300 more balls a year is akin to batting Barry Bonds eighth so that Edgardo Alfonzo can bat cleanup; Bonds will still mash the ball and Alfonso will still clutch at his back between swings, but it's a drastic misallocation of resources.