The Hardball Times

Best Outfield Arms of 2006

by John Walsh
February 15, 2007

It was the last day of 2006 and I was getting ready for a little New Year's Eve get-together. While I was blowing up some balloons, I saw a new message in my inbox, and after a click I realized that I had just received a late Christmas gift—the 2006 play-by-play data had just been released by Retrosheet. Oh, joy! I know, it's pretty nerdy to get worked up over a big bunch of numbers, but I usually can't wait for the new release to come out, and this year the crew at Retrosheet.org outdid themselves by bringing home the data two or three months ahead of schedule. Nice work, boys.

So, the first order of business was to determine the best outfield arms of 2006. Last year I developed a method for evaluating outfield arms, which you can read about here and here. The first article has all the gory details of the methodology, here is the short, painless version: using play-by-play data, I consider five different situations when an outfielder's throwing ability comes into play:

  1. Single with runner on first base (second base unoccupied).
  2. Double with runner on first base.
  3. Single with runner on second base.
  4. Fly out with runner on third base, fewer than two outs.
  5. Fly out with runner on second base, fewer than two outs (third base unoccupied).
For those plays, I add up how often the runner is thrown out or how often the runner is "held," i.e. prevented from taking the extra base. A comparison with league average allows me to rate the outfielder's arm.

By the way, I used this same method to rank the arms of outfielders from the full Retrosheet era, which roughly covers the years from Roberto Clemente through Juan Pierre. Actually, the rankings pretty much go from Clemente to Pierre, as well. Anyway, if you enjoy this kind of thing, the article appears in The Hardball Times Annual 2007, so be sure to check it out. Now, let's get to the matter at hand.

Right Fielders

I always start with right field, I suppose because that's where we expect to find the strongest outfield arms, and I want to get right to the best stuff. In my view, the best right field arm in 2006 belongs to Alexis Rios, the Blue Jays right fielder. Here's a table showing Rios' performance:

Alexis Rios - RF
   Situation  |   Opps  |    H    K  |   Hexp   Kexp 
        S-1B  |     44  |   31    2  |   25.4    0.6 
        S-2B  |     30  |   20    1  |   12.7    1.6 
        D-1B  |     16  |   10    1  |    9.3    0.8 
       OF-3B  |     11  |    5    0  |    2.6    0.4 
       OF-2B  |     22  |   14    1  |   12.5    0.5 
     Overall  |    123  |   80    5  |   62.5    3.8 
Notation:
Opps - number of opportunities
H - Hold: runner did not take extra base.
K - Kill: runner (or batter) thrown out on bases. 
Hexp - Expected Holds based on league average play.
Kexp - Expected Kills based on league average play.
Each of the first five rows shows how Rios did in the five situations considered. You can see that in every category, runners tended to hold more than they did on average. Overall, Rios "saved" around 18 extra bases by holding runners. For good measure, he also had one more "kill" than expected given his opportunities. Rios ranked pretty high last year (5th among right fielders), but really aired it out this year.

The worst performance in right field—in any field, actually—was turned in by Shawn Green:

Shawn Green - RF 
 Situatation  |   Opps  |    H    K  |   Hexp   Kexp 
        S-1B  |     57  |   22    0  |   32.6    0.7 
        S-2B  |     35  |   11    0  |   14.0    2.0 
        D-1B  |     21  |   11    0  |   13.1    0.9 
       OF-3B  |     12  |    1    1  |    2.8    0.5 
       OF-2B  |     15  |    8    0  |    8.5    0.3 
     Overall  |    140  |   53    1  |   71.0    4.3 
Ouch! That is bad. Green can't throw anybody out (only one kill, when four were expected) and the runners know it. They ran hog-wild on the Diamondbacks/Mets right fielder, taking the extra base fully 18 times more than they would have against an average right fielder. Green also ranked dead last among right field arms last year—they need to get this guy closer to the runners he's trying to throw out. Maybe he can swap positions with Carlos Delgado.

I have converted these hold and kill numbers into runs saved above average, and ranked all right fielders according to the number of runs saved per 200 opportunities (generally, around one full season's worth). Here's the list of all right fielders who had at least 100 opportunities in 2006:

                      Inn/9   Opps   Kill+  Hold+  Runs  Runs/200
Rios_Alexis             106    123    133    129    7.1   11.6
Hawpe_Brad              134    187    202    110    9.6   10.2
Ordonez_Magglio         141    150    164    106    5.4    7.2
Cuddyer_Mike            136    149    143    115    4.9    6.5
Suzuki_Ichiro           118    154    125    107    3.9    5.1
Francoeur_Jeff          158    223    127    110    4.6    4.1
Blake_Casey              90    111    155     88    0.9    1.6
Markakis_Nick           102    143     86    100    0.9    1.3
Nixon_Trot               99    102    117     93    0.2    0.4
Kearns_Austin           137    196     73    110    0.1    0.1
Guerrero_Vladimir       121    146     61    110    0.0    0.1
Abreu_Bobby             144    192    124     96   -0.4   -0.4
Bradley_Milton           89    112     75    102   -1.0   -1.7
Giles_Brian             155    161    108     94   -2.4   -3.0
Dye_Jermaine            138    172     34     99   -2.7   -3.2
Jones_Jacque            133    158     54    103   -2.8   -3.5
Encarnacion_Juan        109    134     68     95   -2.6   -3.8
Drew_J.D.               124    147     60    101   -2.9   -4.0
Nady_Xavier              95    119    137     79   -2.8   -4.7
Jenkins_Geoff           122    138     46     96   -3.7   -5.4
Green_Shawn             125    140     23     73   -9.1  -13.0
Notation:
Inn/9 - Defensive games
Opps - Opportunities
Kill+ - Normalized kill rate, e.g. 110 means 10% more kills than average
Hold+ - Normalized hold rate
Runs - Runs saved above average
Runs/200 - Runs saved per 200 opps
You can see that Brad Hawpe was excellent as well—in fact, since he played more, he actually saved his team more runs than Rios. Seeing J.D. Drew so low surprised me, since he is considered to have a strong arm. It may be that his shoulder problems in 2006 were hampering his throwing.

Center Fielders

You know what? Willy Taveras has a great throwing arm. Last year I was surprised to see him rank quite highly (sixth out of about 30 center fielders with 75 opps). Here's what I wrote about him then:

Taveras eliminated more than twice as many base runners as expected. He did not hold runners well, but I think it's fairly common for rookies to have poor hold rates, since opposing teams are going to test their arms. Especially, when they are six feet tall and weigh 160 lbs. Six of Taveras' kills came in the critical situations 2-4, where the key play occurs at the plate. This results in Taveras having a large number of runs saved.
In 2006, word had definitely got around about Taveras' arm: his hold rate of 108 was well above average. But, he still managed to gun down more than twice as many runners as the average center fielder. Very impressive work, by the now-Rockies center fielder. Here's the complete list of center fielders with 100 opportunities in 2006:
                      Inn/9   Opps   Kill+  Hold+  Runs  Runs/200
Taveras_Willy           124    119    211    108    3.5    5.9
Sullivan_Cory            94    122    106    102    1.6    2.6
Figgins_Chone            92    101    101     97    1.1    2.2
Beltran_Carlos          132    173    142    101    1.9    2.2
Byrnes_Eric             117    136    128    107    1.0    1.5
Rowand_Aaron            100    127     82    115    0.8    1.2
Edmonds_Jim              88    118     92    104    0.4    0.6
Sizemore_Grady          153    227    113    101   -0.3   -0.3
Kotsay_Mark             116    144    144     95   -0.7   -1.0
Wells_Vernon            143    176     60    109   -1.0   -1.1
Matthews_Gary           136    201    106    101   -1.2   -1.2
Jones_Andruw            146    190     68     99   -1.7   -1.8
Baldelli_Rocco           83    130    114     94   -1.2   -1.8
Hunter_Torii            137    164     84     94   -1.6   -1.9
Pierre_Juan             158    173    107     89   -1.8   -2.1
Griffey_Ken              97    129     27    105   -1.5   -2.4
Clark_Brady             101    128     28    100   -1.7   -2.7
Patterson_Corey         120    161     94     89   -2.3   -2.9
Lofton_Kenny            107    145     72     86   -2.7   -3.7
Anderson_Brian_N.       107    151     24    107   -3.0   -3.9
Granderson_Curtis       146    163     47     93   -3.4   -4.2
Crisp_Coco              100    133     53     92   -2.9   -4.4
Finley_Steve            108    148     72     90   -3.4   -4.6
Cameron_Mike            138    151     90     88   -3.5   -4.6
Gathright_Joey          113    175     77     80   -5.8   -6.6
Damon_Johnny            121    146     49     87   -5.0   -6.9
One thing to note right away: aside from Taveras, nobody really stood out in center field in 2006. In fact, there isn't really much difference between number three, Chone Figgins, and number 15, Juan Pierre. The bottom two, though, Joey Gathright and Johnny Damon are genuinely atrocious throwers, at least they were in 2006. Damon ranked very poorly last year—well, nobody is surprised by this. I think most of us when seeing Damon throw conjure up an image something like this.

Left Fielders

The Rodney Dangerfields of the outfield throwers, left fielders never get no respect. Nevertheless, many left fielders save (or cost) their team significant amounts of runs each season. The cream of the crop in 2006 was the Dodgers' Andre Ethier, who doubled his expected kill total and was also excellent at holding runners.

                      Inn/9   Opps   Kill+  Hold+  Runs  Runs/200
Ethier_Andre            100    123    206    108    5.6    9.1
DeJesus_David            61    110     90    118    4.5    8.1
Ramirez_Manny           115    135     94    109    4.2    6.2
Monroe_Craig            103    114    187    101    2.8    4.9
Cabrera_Melky           111    144    132    101    2.9    4.0
Soriano_Alfonso         153    202    135     96    1.9    1.9
Ibanez_Raul             155    213    118    105    1.9    1.8
Willingham_Josh         119    134     64    105    1.1    1.6
Murton_Matt             116    135     42    101   -0.6   -0.9
Brown_Emil               80    131    118     96   -1.2   -1.9
Swisher_Nick             73    102    114    100   -1.0   -2.0
Crawford_Carl           139    200     68     99   -2.0   -2.0
Michaels_Jason          112    150     93     98   -1.6   -2.1
Gonzalez_Luis           146    158     36    101   -2.6   -3.3
Lee_Carlos              140    171     68     97   -2.9   -3.4
Bay_Jason               153    224     88     95   -4.3   -3.8
Holliday_Matt           148    176    106     90   -3.4   -3.9
Wilkerson_Brad           74    104     84     97   -2.1   -4.0
Dunn_Adam               147    177     56     97   -4.6   -5.2
Burrell_Pat             110    138    130     91   -3.6   -5.2
Podsednik_Scott         121    130     47     91   -4.9   -7.6
David DeJesus had a curious performance: a low Kill+ with a very high Hold+. Typically, you see this kind of pattern on established players who are living more on reputation than present value. DeJesus, of course, doesn't fit that description, so it'll be interesting to see how he rates going forward. DeJesus also was great in center field in 2006. He did not meet the 100 opp minimum, but his Runs/200 figure was off the charts at 15.0.

A Picture is Worth 1000 Words

image It's not really, otherwise I could submit a couple of plots each week and my THT editors would be satisfied. Oh, were life that easy. Anyway, as I wrote in a previous article, I like to make graphs with colored dots and stuff so here you see a plot of outfield arms for 2006.

I have have a suspicion that some readers are put off by these scatterplots. I once came across a discussion of one of my articles, an article that happened not to have any plots, on a baseball forum. One poster commented that it was so nice to be able to read a Hardball Times article without any of "those meaningless scatterplots." Well, I'd like to convince you that this plot is not meaningless. Here goes...

Each point on the plot represents an outfielder for the 2006 season. The position along the x-axis (that's horizontal) gives Hold+, while the y-axis (vertical) shows Kill+. Left, right and center fielders are color-coded as noted on the plot. The dotted grey lines show the average for each quantity: players above the horizontal grey line were better than average at killing baserunners and those to the right of the vertical line were better than expected at preventing runners from taking the extra base.

Players appearing in the northeast quadrant are better than average in both measures and have the best outfield arms. And conversely, players in the southwest quadrant have the weakest arms. Geez, look at Shawn Green there, he looks like Easter Island on a map of the South Pacific Ocean.

Better Than Their Stats

Although we are capturing quite a bit of an outfielder's throwing duties with our five situations, we don't get all of it. For example, an outfielder guns down a runner trying to stretch a single into a double. It's an important play and not uncommon, but unfortunately, it's hard to use that kind of play in this analysis. That's because, although we know how many kills of this type an outfielder had, we don't know how many opportunities he was given. In fact, it's not even easy to define a "single stretched into a double" and in any case, such plays (when the runner is safe) just appear as a double in the play-by-play data.

So, while we can't somehow work these plays into the system, we can highlight the outfielders who had a large number of assists that were not captured by our metric. Head and shoulders above everybody else in this regard stands Alfonso Soriano. The first-time left fielder led the majors with 22 outfield assists last season, but only 10 of them came in one of our five situations. Twelve "unaccounted for" assists is an unusually high number. In Soriano's case most of the missing assists occurred when the batter tried to stretch a hit, but Soriano also doubled off the runner after catching a fly ball five times. A perfect ranking system would probably rank Soriano as the top left field thrower in 2006.

Here is a list of outfielders with at least five "lost assists":

Soriano     (LF)  12
Beltran     (CF)   6
Griffey     (CF)   5
Cabrera, Me (LF)   5
Along with Soriano, each of the guys probably deserve a boost in their rankings shown above.

Coming Soon

Next time I'll have a look at the most valuable outfield throws of 2006 from the perspective of Win Probability Added. I'll also have a rankings by team and perhaps a couple of other things on outfield arms.



John Walsh dabbles in baseball analysis in his spare time. He welcomes questions and comments via e-mail.