Since the first Baseball Abstract was published some 30 years ago, Bill James has been labeled a lot of things: revolutionary, heretic, genius, fraud, etc. His analysis and research has been praised and trashed at the same time. His work has been studied and acutely used by some big league clubs, and laughed at by others.

He has been credited for starting the suddenly not-so-underground sabermetric style of baseball analysis, and accused of reducing players and games down to mere data, not taking into account a manager’s keen eye or a player’s “makeup.”

Some of his research has even been labeled as unfounded, unusable and as something with no data to back it up. One of these innovations is the defensive spectrum.

Quite simply, it is a spectrum drawn out from all defensive positions ranging in order of the least difficult to play to the most difficult. The idea is that a player can easily go from right to left on the spectrum as he gets older and some of his speed diminishes, but it is much harder to go from left to right, no matter what point of his career the player is in. It looks like this:

1B - LF - RF - 3B - CF - 2B - SS - C

(We will leave pitchers out of the equation since this will inevitably be related to offense, but they typically fall on the far right)

With no easy way to compare defensive difficulty across position, this spectrum looks pretty good at first glance. Obviously the skill set of a catcher is much more rare than that of a second baseman, which is much more rare than what a first baseman can provide.

Also associated with this spectrum, James said, is an expected offensive production. As you move more to the left on the chart, and you are not expected to provide as much defensively, you better be carrying your weight offensively. And that makes sense at first glance. Any baseball fan or Rotisserie player can tell you, your sluggers tend to be your first basemen, outfielders, and third basemen, while you see very few guys who play up the middle slug 40 homers or accumulate an OPS of 1.000.

What I am going to look at is have offensive expectations changed since this spectrum was first introduced in the 1980s? Or, more precisely, how accurate is the assumption that just as catchers offer more defensively than shortstops who offer more then second baseman, etc., first basemen offer more offensively than left fielders who offer more than right fielders, etc.

You see, while we can look at metrics such as zone rating and Fielding Runs Above Average, there is no way to compare specific, defined skill sets like blocking wild pitches or turning the double play or routes taken to flyballs across positions. But with offense, we can equally compare contributions. We know the sole objective of offense, which is to create runs for your team and do this at a better rate than the other team. That is the responsibility of every player whether they are a catcher, DH, right fielder, or whatever. And, yes, there may be a number of different ways to accomplish that goal of pushing runs across the plate, but the eventual desired outcome is always the same: score runs.

So, for 2007, let’s see how close James’ spectrum is to predicting its reciprocal: offense.

From what I can tell, a little bit of research has been done on this before a few years ago. This link takes you to a post where Mike Mehl used the OPS of the top players in the 2003 season to plot the offensive range by position. His conclusions were that the majority of your low OBP, low SLG batters fall on the right side of the spectrum, with only a few exceptions. And most of the batters on the opposite side had high OPS, although admittedly skewed by the Barry Bonds’ of the world.

For this post, I am going to use Runs Created and VORP as the comparative metrics for offense. Runs Created because it is an easy way to compare apples to apples. A run is a run is a run. Did you create a lot of runs or did you not? And VORP because it compares positions to themselves. There are stated, specific offensive factors when discussing VORP for each defensive position and its relative value of replacement level. All OF positions are weighed the same in VORP, but that will suffice for our exercise.

What I have done is taken the average number of Runs Created and VORP for the top 30 in each of the eight listed positions of the spectrum. I then charted it on a graph in the order of the positions on the spectrum. So, if the theory is correct (at least for this year), we can expect to see two lines that go from the upper left corner of the graph to the bottom right. Below is the graph and the numbers represented with each. Click on the graph link to show the full image.

rc-and-vorp-2007-positions.jpg

So what do we notice here with these numbers? Generally the line we would expect is there, but with a few interesting variations to point out:

1. Shortstops - While this position is still generally considered to be one where you have to have a decent glove to play, you can see that the average RC and VORP for SS in 2007 has now reached 3B and CF levels. Someday, I will do this exercise again for 1987, 1967, and 1947, for example, to see how the numbers compare. But the facts are that there are only 31 players in MLB with over 100 Runs Created, and four of them are shortstops (Ramirez, Reyes, Jeter, and Rollins). These four carry the group, but you still have Guillen, Cabrera, Young, and Renteria with more than 85. And another nine have 70 or more RC. The total of SS over 70 RC is 17, or more than half of MLB. Clearly, this position is becoming one where both defense and offense are valued. But I bet if you polled GMs across the league, they would pick great defense if they were forced to pick one quality their SS had.

Unfortunately, only one of these top eight SS (Reyes at #2) is in the top nine for Revised Zone Rating for 2007. Jeter, Ramirez, and Guillen are all in the bottom 10. So it’s still tough to find a SS who does both well.

2. Right field - RF in 2007 has been the Magglio Ordonez and Vladimir Guerrero show. Can you name a player who is clearly the third best RF this year? It’s not easy. In fact, there is a 26 point gap in VORP between second place (Guerrero - 62) and third place (Corey Hart - 36.9). The same goes for RC: there is a 25 point gap between Guerrero (123) and Abreu in third (98). In fact, there are only eight RF with more than 90 RC while there are seven SS!

But is this a growing trend? Looking back, there have not been more than eight RF with more than 90 RC in a season since at least 2003. Looking at our graph, RF VORP is right where it should be based on those numbers, while Magglio’s astounding 144 RC skews that plot line north a little bit.

3. Catchers - Obviously, you see how wide the gap is between your average catchers and the rest of the seven positions. There are only three catchers (Martinez, Posada, and Martin) with more than 70 RC so far this season. And the same three catchers are the only ones at their position with VORPs over 27 for the season. I actually count eight starting catchers with a negative VORP, meaning whomever the team could call up from Triple-A would probably be better offensively.

While the defensive spectrum is nonscientific and purely speculation based on perceived defensive attributes and responsibilities, it does seem to serve the purpose of evaluating where a player can move on the diamond as his skills diminish. It also seems to prove for 2007, except for the recent outlier of shortstops, that players are increasing their offensive skills as they move down the spectrum.

If you know of any examples of players moving in the opposite direction of the spectrum as their careers moved on, let me know. I would be interested to study them and see why that was. Of course, the most recent famous example of the spectrum at work is Craig Biggio who moved from catcher to second base to center field to left field in his career. Now back at second because of lack of options for the Astros, I hear he is going to play all of his old positions in his final game on September 30.

I can’t tell you how much talk oxygen is wasted in this town talking about Adam Everett. People just can’t make up their mind whether they love him or hate him, whether he should be here or on his way out. On a team where the biggest struggle this season has been offense, someone who just last week got their average above .210 is normally vilified and persecuted to no end.

So it’s time to put this discussion to an end. I listen to a couple of sports talk shows every day on my home from work, and one host routinely laments that Adam Everett is a “net loss” when it comes to his overall body of work; that there is no way Everett’s outstanding glove can make up for his weak bat.

Really? Well, that is what I aim to find out.

First, the easy part. Let’s see how bad Everett is offensively. Through June 8, citing The Hardball Times, Everett has created 21 runs so far in the 2007 season (22nd amongst shortstops). His RC/27 is a horridly low 3.7 (what a full team of Everetts would score per game).** At this pace, Everett would have 57 Runs Created by the end of the season. To compare, in 2006, here are some final shortstop RC numbers:

Derek Jeter - 138
Jose Reyes - 125
Miguel Tejada - 104
Orlando Cabrera - 82
Yuniesky Betancourt - 64
Jack Wilson - 59
Khalil Greene - 54
Adam Everett (20th amongst SS) - 52

Using these numbers, Everett’s Value Over Replacement Player is -6.0 (23rd lowest in the majors for 2007). That means that he has created six runs less for his team that a replacement player off the bench or from the minors would have created in the same amount of plate appearances. That puts his VORP per game (or VORPr) at -.120. So, if after 60 games, his VORP was -6, and if he keeps up a pace of -.120 per game for the season, after 102 games, his total VORP for 2007 would be -18.24. That’s 18 runs created less than some scrub on the bench; or, in other words, gawdawful. To put it in comparison, the worst player in terms of VORP in 2006 was Clint Barmes at -20.5. Brad Ausmus was almost as bad with -17.5, and Everett had -8.7.

So a -18.5 would put him as one of the worst, if not the worst hitter in the league - which, granted, is what he is.

Now we need to look at the defensive numbers. First, we will throw out all of the ones that don’t matter. Here are some of Everett’s basic defensive stats and his rank amongst all fielders in the NL:

Errors: 6
Fielding Percentage: .978 - 8th
Assists: 179 - 3rd
Double Plays Started: 13 - 13th
Double Plays Turned: 20 - 2nd
Innings: 478 - 9th
(Balls In Zone does not come out until the end of the season)

Why do I say these don’t matter? Because Everett can not control the circumstances that determine these stats: a scorer decides error or not, and thus fielding percentage; and he can not control if there is a man on first and if the batter hits a groundball, starting a DP, etc.

So we will use something that does matter. For the first time, Michael Lichtman (working with Tangotiger) has released his defensive Ultimate Zone Ratings in-season. These stats are widely regarded around baseball as the ultimate source in defensive runs saved because, on defensive plays and players, Lichtman takes into account what park the game is played in, pitcher’s GB/FB tendencies, the base/out situation, ball trajectory, and more. The spreadsheet for this year, released on May 31, can be found here.

Lichtman has Everett third in the majors so far, having saved nine runs through May 31. Nine runs saved through 53 games averages out to 18.5 more for the rest of the season for a total of 27.5 runs saved with his defense in 2007.

So, we now have the projected total value of Everett’s run contribution (-18.5) and his projected total in defensive runs saved (27.5). The difference here is a plus nine, not a net negative number.

But just because he has a net positive, does not automatically mean he belongs in the lineup. He must be compared to his peers to show true value. So, I have done the work for you. Below, in decreasing order, is the net value of all NL shortstops for 2007, using the same projections I used for Everett:

So with as bad as Everett has been, he still has a net run value higher than four shortstops just in the NL. Over the next couple of days, I will run this for the AL and see how many he bests in that league as well. And remember, these are just the starting shortstops; don’t even get me started on Ronny Cedeno, Wilson Valdez, or others.

If Everett were to have even a little bit of a turnaround and reach last year’s numbers of -8.7 VORP, he would pass Eckstein and would be in Khalil Greene territory.

Like has been said many times before, Everett is the perfect player for a team that already has a strong offense. The problem is, the Astros decidedly do not. So, Everett’s obvious individual shortcomings at the plate get magnified and his defense gets overlooked.

But, if you don’t want to take my word for it, that’s fine. Lisa Gray found an excerpt from Michael Lichtman responding to Adam Everett doubters. He essentially says the same thing I have found here, and he at least knows what he is talking about.

_______________

**On the other hand, the Runs Created for Biggio, Ensberg and Lamb so far this year are 19, 17, and 9 respectively, but we’re not talking about them right now.