BtB Power Rankings: Week 1

See the complete rankings below the jump!

All data through April 7, 2010.

On Paper Playoff Leaders

American League: E=Yankees C=Royals W=Athletics WC=Rays
National League: E=Phillies C=Cardinals W=Giants WC=Braves

This Week's Breakdown

Usually, I will use this space to talk about interesting teams in the rankings.  However, since this is the season's inaugural BtB Power Ranking post (and more of a trial run of the spreadsheets than anything else), I thought I'd instead use this space to describe what the power rankings are...and what they are not.

What They Are

The power rankings are an objective way of ranking all MLB teams based on their season-to-date statistics (i.e. performance).  The basic question the answer is this: if we put teams with these statistics into a large, fictitious league and let them fight it out for 10,000 games, how would we expect them to rank at the end of those games?  

The statistics we use are the same sophisticated statistics that we to evaluate players: linear weights for offense, advanced baserunning, quality fielding statistics, and defense-independent pitching statistics.  We make an effort to make them as context-neutral as possible.  They are park adjusted, and they do include league adjustments as well: the American League is a vastly superior league to the National League (see here, for example), and we give them credit for that in these rankings.  Sometimes, this adjustment can be jarring: while the National League is starting strong, you can expect to see a disproportionate number of AL teams in the top-15 by mid-season.  As a fan of an NL team, I don't like this...but the data are clear: AL teams are better, on average, than NL teams.

What They Are NOT

1. They are not based on actual team wins or actual team runs.  Instead, we use statistics to predict team run scored and run allowed totals, and from that infer team winning percentage.  Why do this?  For one thing, most other power rankings we see around the internet are based on team wins, so this provides a different look at team performances than you will find elsewhere.  For another, component statistics tend to be less volatile than context-specific totals like runs or wins, because they are less influenced by timing and other essentially random events.

2. They are not projections.  They do not include prior seasons' data, aging, regression, etc, and therefore are not necessarily designed to be predictive.  Furthermore, they do not account for team personnel changes or injuries. In fact, they are primarily historical, giving us a look at what has happened in a way that gets away from actual win (or actual run) totals.  That said, they are based on statistics that are known to be relatively quick to stabilize, and so they may be expected to be  more predictive of future team performance than other rankings based on win/loss records.

3. They are not the be all and end all.  There are absolutely important pieces of information, some of which represent repeatable skill or tendencies, that are not factored into these rankings.  At this time, for example, they do not look at how bullpens are leveraged.  Pitching is based this year on xFIP, which works well for most, but not all pitchers at the career level.  They also do not include strength of schedule adjustments.  But for all their shortcomings, I think they present one of the best overall rankings of team performance that you will find.

What follows is mostly for those who want to peer under the hood.  First is a table that summarizes actual vs. expected run totals, and how those convert into expected winning percentages.  The second table breaks down the statistics that go into the estimated runs scored and estimated runs allowed totals that you see in the first table.  Tables are sortable by clicking in the header.

Converting Runs to Wins

G=Games
RS = Actual Runs Scored, after a park adjustment
eRS = Estimated Runs Scored, after park adjustment (see "Offense" table below)
RA = Actual Runs Allowed, after a park adjustment
eRA = Estimated Runs Allowed, after park adjustments (see "Defense" table below)
W% = Actual Winning Percentage
pW% = PythagenPat Winning Percentage, based on actual runs scored and run allowed totals
cW% = Component Winning Percentage, using estimated runs scored and estimated runs allowed totals
LgAdj = League adjustment, based on differences in league quality (justification here and here).
TPI = Team Quality Index, a hypothetical winning % based on component estimates of runs scored and runs allowed after the league adjustment.

 

Team Offenses and Defenses

RS = Actual Runs Scored
eRS = Estimated Runs Scored: wRC + EqBRR
wOBA = The Book's statistic, but park adjusted, and using data from both wRC and EqBRR
OBP = On Base Percentage (Times on Base / Plate Appearances)
SLG = Slugging Percentage (Total Bases / At Bats)
wRC = From FanGraphs, with baserunning removed, after park adjustments
EqBRR = Dan Fox's composite baserunning statistics from Baseball Prospectus
Clutch = "Clutchiness" measure from fangraphs; difference between actual WPA and expected WPA based on component statistics.  We report this in runs.

RA = Actual Runs Allowed, after park adjustment
eRA = Estimated Runs Allowed: tRuns - Field - Catch
ERA = Straight-up Earned Run Average
FIP = Fielding-Independent Runs, based strictly on K-, BB-, and HR-rates.
xFIP = Experimental Fielding-Independent Runs from FanGraphs.  Like FIP, but with HR/Outfield Fly Ball rates regressed completely to league average.  xFIP is as predictive as any other DIPS-like stat.
xFIPrns = Pitching Runs Allowed, based on xFIP
Field = Team UZR.  I typically prefer to have multiple fielding statistics, but with THT's team pages no longer updated, I don't currently have a second option on par with UZR.
Catch = Catcher Fielding Runs, based on SB's, CS's, WP's, PB's, and E's.  The methods are essentially those described here.  But I'm using B-Ref data this year, and so there are slight tweaks to the methodology, generally in ways that should lead to greater precision.
BABIP = Batting Average on Balls In Play.  Fluctuates at the team level with fielding, although chance events can have effects as well.