Friday, February 8, 2008

GAP's Greatest Shortstop?

For all the hype about Tampa's Miguel Torres, is he really the best SS in GAP? Let's take a look at the numbers and try to figure it out.

First, who is in the discussion? We'll start with Torres, and add active SSs who have earned at least two all-star berths at SS. Although, this criteria wouldn't include Torres (his only selection came way back in Season 1), it seems like a fair starting point. We'll knock out Eugene Hackman who hasn't played since Season 5, and Al Cradle, who is retired. That leaves us with Todd Linebrink (Oak), Alberto Abreu (Mem), and Theodore Haney (KC).

Trolling through the league depth charts, I found quite a few other names that should be considered for this discussion. But first, we should note some guys who have a chance to be in this discussion a few years down the road, when they have some big league results to back up their ratings:

Rick Gordon (Bal)
Joaquin Ortiz (Cle)
Domingo Sosa (Hou)
Ugueth Herrera (Hou)
Chip Oliver (Phi)
Russ Anderson (Van)

There are a couple other guys who could play SS with some defensive deficiences (like Cincinnati's Andres Mieses and Tampa's Pedro Guerrero), but without ML SS results, I don't want to include them here. In fact, that same criteria diminishes Haney's case pretty strongly, which we will discuss further shortly. However, there are a few guys with results that warrant them being included in this discussion:

Roosevelt Kohlmeier (ChW)
Matt Moore (SD)
Barry Clayton (StL)

Missing my arbitrary cut but still worthy of note: Matt Kubenka (Cha).

So let's take a look at these seven. First, on the offensive side. The easiest available tool to measure hitting production is OPS. Here are our seven with their career OPS, and S6 OPS in parentheses:

Haney 947 (951)
Linebrink 882 (919)
Kohlmeier 879 (978)
Moore 874 (793)
Clayton 854 (769)
Abreu 775 (605)
Torres 687 (712)

Another way to slice offensive output is RC/27. That Runs Created per 27 outs. I had to make a couple minor tweaks due to the limited historical data (player card doesn't show extended batting info). I used the stolen base RC formula so we were including SB/CS, and eliminated GIDP, SH, and SF from the outs side because we don't have that. Here are our candidates, with their RC/27 career, and S6 in parentheses:

Haney 7.69 (8.08)
Linebrink 7.23 (7.95)
Kohlmeier 6.95 (8.62)
Moore 6.68 (5.20)
Clayton 6.51 (5.21)
Abreu 5.27 (3.07)
Torres 4.33 (4.78)

Our players fall in the same order as in the OPS rankings. So we can feel good about this order, and weight them on a sliding scale (100 as the top of the scale) as follows:

Haney 100
Linebrink 94
Kohlmeier 92
Moore 90
Clayton 87
Abreu 75
Torres 64

Now let's take a look at defense. First, we should note the number of games played at SS, versus total games played:

Linebrink 906/916
Torres 870/875
Abreu 860/860
Kohlmeier 509/571
Moore 476/477
Clayton 408/749
Haney 224/701

All but Haney and Clayton have played more than 89% of their games at SS. Clayton played most of S2 at 2B and 3B, and continues to move around to other positions, but has played 54% of this games at SS. On the other hand, he only played 17 games at SS last year, filling the 2B slot most of the way. Haney, however, has played just 32% of his at SS. As a 20-year-old in S2, he played most of his games in RF, and did the same in S3. He split time evenly between SS, CF, and 2B in S4. His S4 games at SS primarily came early in the season, and were enough for him to be selected as an all-star at SS that year. He played 38 games at SS in S5, working primarily in CF that year. Last year was his first full year at SS, playing 130 of his 139 games there.

Each reader will have to make their own judgment call on whether one or both of these guys should be excluded from this discussion based on this. Haney, now 25, and Clayton, now 26, both have nearly reached their rating projections and are being used at SS here at the start of S7.

As to the defense, the most basic defensive stat is fielding percentage. Let's start there. Here are our guys, with career SS FPct and S6 in parentheses:

Torres .989 (.998)
Abreu .985 (.985)
Linebrink .973 (.986)
Clayton .966 (.986)
Haney .961 (.969)
Kohlmeier .958 (.974)
Moore .956 (.980)

Bill James created Range Factor, believing the number of outs a player participates in is more relevant than the number cleanly handled. Here our the range factors for our group, career and S6:

Haney 6.75 (6.55)
Torres 6.29 (6.14)
Linebrink 6.27 (5.90)
Abreu 6.19 (6.15)
Kohlmeier 5.87 (5.73)
Clayton 5.73 (5.20)
Moore 5.71 (5.76)

Lastly, HBD gives us a quantified number of excellent plays. To account for the varied amount of playing time, I converted the + plays to a rate, dividing by 1458 innings played (162*9), giving us an approximation of Great Plays Per Season (GPPS). Here are career and S6 numbers:

Torres 9.6 (12.8)
Abreu 5.8 (3.5)
Kohlmeier 2.6 (5.4)
Linebrink 2.0 (1.2)
Moore 1.5 (3.3)
Haney 1.5 (1.3)
Clayton 0.9 (0.0)

While FPct and RF reflect an entire season worth of play, GPPS only reflects a handful of plays. So by weighting these categories accordingly, and re-weighting again with 100 at the top of the scale, we get the following defensive scores:

Torres 100
Haney 98
Abreu 97

Linebrink 95
Kohlmeier 92
Moore 90
Clayton 90

Now, to go about the task of merging the offensive and defensive scores. It's clear we can't just average them, as fielding is only a portion of the defensive formula, along with pitching. Let's keep it simple and make fielding worth half as much. Using this method, here are the final rankings, again re-weighting to make 100 the top of the scale:

Haney 100
Linebrink 95
Kohlmeier 93
Moore 91
Clayton 89
Abreu 83
Torres 77

In the end, it's clear that while Torres and Abreu are both very strong defensively, their signifcant offensive shortcomings don't put them in the same category as a Haney or a Linebrink.

Clearly, there are a number of methodolgies that could be used that might vary these results slightly. Based on the statistics used and what they both mathematically represent and historically have proven, these results can't be far off. In fact, the only way to construct a case for Torres over the others would be to select only certain rudimentary statistics. However, even just using a basic stat like batting average, Torres comes up ahead of just two of the other six. There really is no argument here whatsoever. Torres is simply not a contender for GAP's Greatest Shortstop.