Tuesday, February 13, 2007

VERSUS Part 2: Los Angeles Angels of Anaheim vs. Oakland Athletics

Editor's Note: Most of my baseball-related posts here at "Solvent is So Yesterday" have thus far focused on comparing teams across specific dimensions of baseball talent (e.g. hitting, fielding, and pitching). This method of presenting the data provides a lot of utility in explicating exactly what a team does well or not so well, but it also limits our ability to view teams at a holistic level. For example, the numbers indicate that the Red Sox will be strong on offense and in the rotation, but weak on defense and in relief. How good a team does that make them?

Like any good analyst, we need an alternate window into the data, different ways to mix and match the data. With that in mind, I think it is interesting to compare two teams across all the dimensions we have thus far studied, to get an idea of relative strength. The compared teams have been linked to each other, either by related characteristics or because they have a direct competition with each other.

In this, the second of a planned three part series, I compare the fortunes of the Los Angeles Angels of Anaheim and the Oakland Athletics, the key competitors for the AL West title.

Since the collapse of the Seattle Mariners in 2004, the Angels and Athletics have been the class of the AL West, with each team finishing first or second every year. The Angels took the division title in 2004 and 2005, but faded last year, allowing Oakland to overtake them. My analysis indicates that both the Rangers and Mariners will be improved next year, but neither team figures to contend for the division title barring misfortune amongst the top 2.

Although the A's and Angels have played foe to each other on the field, the comparison between these two teams is just as interesting for their divergent team-building philosophies. The A's and their celebrity GM Billy Beane were famously profiled in Moneyball: The Art of Winning an Unfair Game, and are generally recognized as among the most numerical projection-reliant teams in MLB. The Angels represent the traditional scouting based approach to the game (cynics will correctly note that the supposed dichotomy here is largely media induced, and that the true team building philosophy -- building around cheap pre-free agent players -- of both teams is identical. But that's a point for another day).

Competitive Landscape
In order to reach the playoffs, what types of hurdles will these teams have to clear? The race for the AL playoffs is shaping up as an absolute dogfight. I count eight teams with legitimate claims on four playoff spots (in alphabetical order: Boston, Chicago, Cleveland, Detroit, Los Angeles, Minnesota, New York, Oakland). The numbers indicate that the Angels and A's may be a notch below the other six teams, but the difference is not overwhelming. More succinctly, the AL wildcard could certainly come out of the AL West. However, there will be up to five teams fighting for one wildcard spot, so the most likely path to the playoffs is a divisonal win.

Taking the projections from my AL Lineup Preview, the starting line-ups compare as follow:

Los Angeles Angels of Anaheim
: 5.12 RPG
Oakland Athletics
: 5.01 RPG

The Angels have a slight advantage on the offensive side of the ball (on the order of 2 wins per year), but neither of these offenses will scare many teams. By comparison, the offenses of their six main competitors are all expected to be substantially better (with the exception of the Chicago White Sox, and the ChiSox always seem to squeeze more runs out of their offense than I expect them to. See: Dye, Jermaine circa 2006, or Konerko, Paul circa 2005).

Comparing the benches, while recognizing that bench composition is still a relative unknown at this point:

Los Angeles Angels of Anaheim

Jose Molina
Shea Hillenbrand
Maicer Izturis
Kendry Morales

Oakland Athletics

Adam Melhuse
Dan Johnson
Marco Scutaro
Shannon Stewart

Both benches are average to slightly above average in comparison to the rest of the AL. A weighted average Runs Created (RC) gives the Angels bench 4.4 RC and the A's 4.3 RC, an almost negligible difference. However, I would give the Angels the bench advantage because they seem better suited to complement their starting lineup than the A's. Adam Melhuse is not really what you would hope for from a backup catcher -- he is neither proficient offensively nor defensively, and the only other potential backup catcher on the A's roster is Mike Piazza. Jose Molina can't hit either, but at least he is excellent defensively. Additionally, the quality of the Angels' three other bench players and the positional flexibility of the starting nine (specifically Chone Figgins and DH Juan Rivera) mean that almost any starting player save Vlad Guerrero can be rested without significantly impacting the offense. By contrast, the A's don't have quality backups (when considering defensive ability) for 3B, SS, 2B, or RF. Add it all up:

Edge: Los Angeles of Anaheim (2-3 wins)

Taking the projections from my AL Rotation Analysis, the rotations compare as follow:

Los Angeles Angels of Anaheim: 4.06 EqERA, 994 IP
Oakland Athletics
: 4.31 EqERA, 956 IP

The Angels have some of the best pitching in the AL (and indeed, all of MLB), with five above-average starters. However, Oakland comes out surprisingly well in this analysis as well, finishing with an above average rotation. Of course, the assumption implicit in these rankings is that the A's will get 34 starts out of Rich Harden, so....yeah.

Edge: Los Angeles of Anaheim (3 wins)

Taking the projections from my AL Bullpen Analysis, the bullpens compare as follow:

Los Angeles Angels of Anaheim: 3.56 EqERA
Oakland Athletics: 3.52 EqERA

Two very good bullpens, with a good closer and strong set-up men.

Edge: Push

Taking the projections from my Team Fielding Preview, the defenses compare as follow:

Los Angeles Angels of Anaheim: +1 Runs/150 Games
Oakland Athletics: +21 Runs/150 Games

If there is one saving grace for the A's, it is that their team fielding is expected to once again be excellent. Regular readers of this blog know that I believe defense may currently be underrated by sabermetric fielding metrics. Note that this 20 run difference in fielding ability also assumes that Los Angeles will make the intelligent choice and put Juan Rivera in LF and use Garrett Anderson at DH. A full season of GA in left field might represent another 5 run advantage for the A's.

Edge: Oakland (2-3 wins)

Putting it all Together

The consensus view going into the season seems to be that the Angels are the favorites to take the division (TradeSports.com says a 50% chance of the Angels winning the division, and a 25% chance for the A's), and the numbers here don't refute that. Before giving my final predictions, I do think there are three points to consider that are currently difficult to incorporate numerically:

  1. Injury risk - Will Carroll at Baseball Prospectus recently released his Player Health Reports (subscription required). Carroll gives both these teams a fairly high injury risk, especially the Oakland infield and the Anaheim outfield. As discussed in the offensive section above, I believe the Angels positional depth in the outfield is far superior to Oakland's infield depth. The loss of Chavez, Crosby, or Ellis (all of whom are viewed as moderate to serious injury risks) will significantly impact Oakland's chances of contending.
  2. Defense - not to beat a dead horse, but Oakland's defense is going to be good, and this is an underrated component of team ability (in my opinion).
  3. The Sabermetric Demerit - this is a name I came up with to describe the general underperformance that numerically oriented teams seem to experience when compared to preseason projections (if anyone has a better name, feel free to leave it in the comments section). The general concept is that in any competitive endeavor, anything that we don't properly value will be exploited by our competitors. Oakland, being a numerically based team, is likely using a set of analysis tools that are generally aligned with those I am using here (although they are no doubt more sophisticated). In contrast, the Angels use a player valuation methodology extremely different from that used here. As a result, the Angels are likely incorporating factors that we (and Oakland) are not, and there will be a systematic bias in our projections, such that we will always project the A's (and similarly Boston and Cleveland) to be better than they actually are, and we will always project the Angels (and similarly, the White Sox) to be worse than they actually are.

Summing up the points above, the strict numerical interpretation is that the Angels are ~3 wins better than Oakland. Defense works in the A's favor, but injury risk and the Sabermetric Demerit are on Anaheim's side. The resulting integration of all those factors indicates to me that the Angels are probably nearly five wins better than Oakland. Five wins is a large deficit to overcome, and as a result I would put Oakland's probability of winning the division as quite low (on the order of 20%). I am also going to make a prediction that I did not expect to make when I started this exercise, and say that I think the Angels have to be heavy favorites to win the division (i.e. 75% chance). There does not appear to be a team that will offer the Angels significant competition in 2007.

Update: 2/14/07
Happy Valentine's day! Soon after publishing this post, I headed over to one of my favorite blogs, Athletics Nation, and found that they had also posted a Angels vs. A's comparison. Although their conclusions were similar, I was also surprised to learn that Juan Rivera recently broke his league in Winter League ball, and will be out at least until midseason. In most cases, the loss of one player does not affect my projections very much, but the A's and Angels are in a sort of statistical "sweet spot" where a small change in talent can make a big change in playoff odds. Specifically, the loss of Rivera hurts the Angels in at least three ways:
  1. Defensively. Garrett Anderson will now be the starting left fielder, a big downgrade from Rivera
  2. DH inflexibility - I thought one of the Halos' big advantages this year was that they had a nice rotation of players available for 3B, LF, and CF, with players getting some rotating rest at DH. They have now lost some of that flexibility, although Chone Figgins can play a passable left field still.
  3. Outfield depth. The Angels starting outfielders all got red lights from Will Carroll, meaning there is a good likelihood that Anaheim's outfield depth will become an issue this year. Previously, the Angels' fourth and fifth outfielders would have been Rivera and Figgins, both of whom are good enough to start in a pinch. Now the fifth outfielder will be...?? Maybe Angels fans know more about this than me - will Kotchman be asked to play some outfield? Does he have that ability?
Overall, I think the loss of Rivera is quite big, such that the five win advantage I had for the Angels before, might be closer to 3 wins now (maybe 3.5). The playoff odds now go from 80% Angels/20% A's, to more like 70/30, a big change.


Anonymous said...

I concur, but I feel more graphics and visual representations are needed to fully grasp your point. More pictures!!! JLo

Anonymous said...

I'm not really understanding what you're talking about with the "sabermetric demerit". I understand that the evaluations done are different, but it seems to me that the only effect that would have would be to say that your evaluation of Oakland would be more accurate than that of the Angels. Why do you say that it means that the Angels will be more likely to be better than your projection?

EA said...

Responding to the above comment:

First, thanks so much for commenting. I appreciate the feedback about what makes sense and what does not. The idea behind the sabermetric demerit is that teams like the Angels likely consider other sources of information in making their decisions that are not accounted for in our model (for example, scouting information that tells them a certain pitcher is likely to be better than the EqERA projections give them credit for). The Angels presumably choose players that are better at those skills than not (i.e. the Angels are likely to avoid players with POOR scouting reports), so over time they will skew towards player who are "undervalued" in their valuation system, while the A's will similarly choose players that are undervalued in their valuation system. If we use the same (or similar) valuation method as the A's, we will always like the A's more than we should, and dislike the Angels more than we should.