“Now that’s a well battled tail.”

Left Field

1. Twins: Delmon Young.  Delmon finally made “The Leap” last year, his fourth full MLB season, slugging over .425 for the first time (.493, to be exact).  The biggest question: was 2010 the norm, or the anomaly?  The stats suggest it will be the norm, as Delmon finally started hitting fly balls (.81 GB/FB%, as opposed to 1.22 and 1.00 in 2008 & 2009) and improved his K/BB ratio (2.89 in 2010, 7.67 in 2009).  His defense remains poor-to-comical.

1. Tigers: Ryan Raburn/Brennan Boesch.  Raburn is the favorite, and has put up good numbers in more of a part-time role the past 2 seasons (2009: .892 OPS in 261 ABs; 2010: .814 OPS in 371 ABs).  He looks like he will have a nice year if given the full-time job.  Boesch is more of a fourth OF, as he dropped off a cliff last year after a blistering first half.

1. White Sox: Juan Pierre.  Pierre can steal bases and hit for an extremely empty average.  He’s also jealous of Shannon Stewart’s arm.

Major Advantage: Twins, and potentially Tigers.  Production compared to last year: same or better for Twins, same or better for Tigers, same for White Sox.

Center Field

1. White Sox: Alex Rios.  Rios seems like the kind of guy that looks like a much better player than he actually is, and is generally scorned by the hometown fans because they know he’s not playing up to his potential, and makes many boneheaded plays (basically the White Sox version of 2007-2009 Delmon Young).  Still, he’s the only center fielder in the division that has a power-speed combo.  He’s #1, but not by much (generally regarded as a below-average defender, as well).

1. Twins: Denard Span.  In over 1,000 at-bats in ’08-’09, Span had an OPS over .800 and a batting average over .300.  In 2010, those numbers dropped significantly, to .679 and .264.  Not surprisingly, his BABIP in 2010 was .294, after he had a BABIP of .358 and .342 in ’08 & ’09.  Playing on grass might explain some of that (grounders didn’t scoot through the infield like they did at the Dome), but some correction can be expected in 2011.  His defense remains below-average.

1. Tigers: Austin Jackson.  Jackson is the poster child for “regression candidate” this season, as he had a league-high BABIP of .396 last season and a K/BB ratio of 3.62 (compared to Span’s .294 BABIP and K/BB ratio of 1.23).  Love his defense and speed, but he’s due for a big drop-off offensively.

Slight advantage: White Sox.  Production compared to last year: same for White Sox, better for Twins, worse for Tigers.

Right Field

1. Twins: Michael Cuddyer/Jason Kubel.  The Cuddyer/Kubel combo is the best based mostly on durability.  Both Cuddyer and Kubel had down years in 2010, and should improve in 2011, forming a nice RH-LH combo that hits 30 homers with an OPS over .800.  The defense isn’t good.

1. White Sox: Carlos Quentin.  Quentin certainly has the power potential to be a difference-maker, but his health and low batting average have prevented that from happening.  He has yet to record a 500 AB season, and he’s an extreme fly ball hitter (.66 GB/FB ratio for his career), so his BABIP is always low (.251 career).  Still, if he can stay healthy (a huge if), he’s a big factor in the division race.  Defense is not good.

1. Tigers: Magglio Ordonez.  Ordonez is 37 and coming off a broken ankle.  He’s what Bert Blyleven would call a “professional hitter.”  He’s never batted less than .292 or had an OPS less than .795 since 1998.  He’s still a valuable hitter, but his age, health, and terrible range in the OF put him third on the list.

Slight advantage: Twins.  Production compared to last year: improvement for Twins, potential improvement for Tigers & White Sox (depending on health).

Designated Hitter

1. White Sox: Adam Dunn.  The biggest free agent acquisition in the division goes to the White Sox.  Dunn is commonly referred to as your “three true outcomes” player: he either walks, strikes out, or hits a homerun.  You can basically write him down for 40 homers, a batting average below .260, and an OBP over .360.  It seems unlikely that a league switch would have much of an effect on him, and despite seemingly being around forever, he’s only 31.

1. Tigers: Victor Martinez.  Martinez is another free agent pickup slotted in the DH role.  He provides a better average but less power than Dunn or Thome/Kubel: you can assume an OPS around .850 with 20 homers.  Very good switch-hitting DH that loses some value because he’s not playing catcher full-time.

1. Twins: Jim Thome/Jason Kubel.  Thome had a great year last year, and Kubel had a poor year.  If they both revert a little closer to the mean, they should combine to provide similar production from the DH slot that the Twins got last year.  Which was very good.

Slight advantage: White Sox.  Production compared to last year: massive upgrade for White Sox (Andruw Jones/Mark Kotsay/Manny Ramirez’s corpses last year), large upgrade for Tigers (basically DH’d their fourth OF most of the year – Carlos Guillen, Johnny Damon, etc.), same for Twins.

Pitchers coming later . . .

The nice thing about defensive stats is that, no matter how trustworthy they are about giving us the actual defensive value of a particular player, they have done a bang-up job of giving us a good idea of how much defense is worth.

Before we talk more about stats, we need to lay a little groundwork.  In general, wins correlate to total runs scored and total runs allowed, via what’s called the “Pythagorean Theorem.”  This isn’t a perfect system, but it gives us a quick-and-dirty way to translate offensive, defensive, and pitching stats into win totals.  By the Pythagorean math, every ten runs scored or prevented is approximately worth one win.

The problem with talking about defense is that it’s actually two abilities: One is the ability to play a position at all, the other is ability relative to other players at the position in question.  We can think of the first skill basically as “can a straight-faced manager write down this player’s name in the lineup card at a particular position?”

I. Playing the position in the first place

So with our statistical groundwork in mind, let’s take a look at the first aspect of defense: ability to play a position at all.  This is very important because it makes sense that a good-hitting catcher is worth more than a good-hitting first baseman, simply because there are fewer good hitters who play catcher.

The reason that fewer good hitters play certain positions like catcher is because of factors like extra skills, athleticism and speed requirements, and handedness requirements, none of which correspond all that well to hitting skill.  For example, if you were forming a baseball team from a pool of a hundred random players, you’d have to put a coordinated, righthanded-throwing, strong-armed, athletic player at shortstop.  You might only have five guys who fit those criteria, and the chances that one of them is also one of your nine best hitters are not good.

It turns out that your best hitter can play DH no matter what, and can probably play first base if he’s not Mo Vaughan.  A little foot speed is required for LF and RF, but even marginal athleticism plays there if the player has enough hitting skill.  Once you get to 2B, 3B, and CF, you’re in need of more skill and athleticism (2B and 3B also come with handedness requirements).  SS and C are even more demanding.

All of this means that positions have different intrinsic value compared to each other.  Just playing a position at all has some effect on a player’s value, because the quality of hitters that can play each position varies drastically.  Based on some historical research, Tom Tango has come up with the following values for positions:

C: 12.5 runs

SS: 7.5 runs

2B, 3B, CF: 2.5 runs

LF, RF: -7.5 runs

1B: -12.5 runs

DH: -17.5 runs

(Interestingly, the DH value is more like 22.5 runs when you consider that any player can play DH.  But DHs get a bonus of five runs because players statistically perform much worse when hitting off of the bench than when playing in the field.  So Rondell White’s claims about hitting well when playing the field actually have some statistical backing.)

What do these numbers mean?  Over the course of a season, a catcher gets 12.5 runs worth of extra credit over the league average.  So if you take a catcher and move him to third base (as some have suggested with Mauer), you’re knocking about a win of value off of that player over the course of a full season.  That doesn’t account for how good the replacement player is at the position; that’s an entirely different issue.

What does this tell us about how the ability to suit up at a particular position affects player value?  It’s really important.  There are three wins between the most valuable and least valuable positions.

II. Skill compared to others who play the same position

Now let’s move on to ability relative to other players who play a position.  It’s debatable if we can use UZR to learn anything useful about a player’s defense in a particular season.  But the nice thing about UZR is that it gives us a range of reasonable fielding values.

Broadly speaking, UZR tells us that players usually contribute between plus-ten and minus-ten runs relative to position average.  In extreme seasons, UZR has players contributing up to and beyond plus- and minus-thirty.

Let’s do a little basic stats work here.  For 2010, the standard deviation of UZR was roughly 8.5.  95% of players will have a UZR two standard deviations away from the mean (plus or minus).  So in other words, 95% of UZRs will be between 17 and -17.  If the 5% (six of the ~150 qualified MLBers) were consistently the same players, then perhaps having a very good or very bad UZR is a skill.  But the list of top performers varies drastically from year to year (though especially awful fielders like Adam Dunn the Outfielder show up multiple times on the bad list).

UZR doesn’t seem to consistently identify the best fielders, but the best fielders it does identify all have about the same value.  This should make us comfortable with the idea that the almost every fielder is worth somewhere between 17 and -17 runs in any given season.  For familiarity’s sake, let’s analogize fielding prowess to batting average.  A plus-ten defender might be like a .300 hitter; plus seventeen is something like a .330 hitter.  A .300 career hitter is likely going to have some peak years at the .330 level, or maybe even above it.  Indeed, several players each season will have an average in excess of .330.  But predicting any one player to do it is probably a dumb thing to do.  And when the stat in question is less clear-cut than batting average (imagine a world in which .330 average might just be a .280 average with some weird fluke data collection error), trumpeting any given player as a .330 hitter is probably dumb too.

My basic thesis is that the range of values that UZR provides is useful and reasonable for quantifying how good defenders are when compared to others who play the same position.  The very best defenders (the .300+ hitters of defense) are worth somewhere in excess of ten runs over their positional peers.  The very worst are worth somewhere below ten runs under the positional average.

III. Putting it together

So let’s put this all together.  I calculated the total “defensive value” for the 150 qualified MLB regulars in 2010 by adding the UZR and the position adjustment.  The average is -1.12 runs (which makes sense, because most of the non-qualifiers in the MLB make their living by being “defensive replacements,” so the pool of qualified hitters is likely below average on the whole).  I then ran two correlations, one between UZR (approximating skill within a position) and total defensive value, and the other between position adjustment (approximating ability to play a particular position) and total defensive value.  I found that position adjustment correlates better by about 10% (position adjustment accounts for about 50% of defensive value; UZR for about 40%).  My guess is that, if we went by career averages and not just one season worth of UZR, position adjustment would account for an even greater percentage of total defensive value (UZR outliers would be diminished, which would get rid of the cases in which UZR contribution dwarfs that of position adjustment).*

This makes sense to me; position adjustment applies to every player, not just those who are good or bad at defense.  Most player aren’t either good or bad defenders, they’re just in the normal, acceptable range for a particular position.  To make up the position adjustment between a LF and a CF, the LF would have to be a consistent plus-ten defender.  No LF has had a plus-ten UZR in each of the past three seasons, so it’s probably more valuable to be just a run-of-the-mill CF than it is to be a really good LF.

In other words, even though UZR is trumpeted as the magic fielding number, what position you play is still more important than how well you play it.

*I didn’t do that particular study because UZR data isn’t available for the entire careers of players who aren’t young (and that would skew the sample since younger players are generally more athletic, better defenders).