Crooked Numbers: Do Not Pass Go, Do Not Collect $200

Three months ago, I took a look at the struggles of the A’s, Indians and Pirates with regards to their complete and utter inability to get runners home once they’d put them on base, a light stat called runner scoring percentage. While the Indians’ offensive struggles were more a result of overall struggles (they were batting .226/.296/.379 at the time and were plating a more respectable 35.4% of their baserunners), the A’s and Pirates were plating a lower percentage of their runners on base than any team since 1990, and it wasn’t particularly close.

Games in the last few days have highlighted the dramatic offensive turnarounds in both Oakland and Cleveland. After dropping 27 runs on the Royals over the weekend, the A’s were fourth in the AL in runs scored with 551 through Tuesday, a far cry from the 95 runs they totaled through the first month of the season. On Tuesday, eleven Indians crossed the plate in the ninth inning alone, vaulting the Tribe past the Royals and into the top half of teams in the AL in runs scored, fifth in Adjusted Equivalent Runs (AEQR).

For the Bucs, things have been better, but not to the degree of their junior circuit brethren. With only 471 runs on the season, the Pirates have the second-worst total in NL (and the major leagues, but comparing them to a league with a DH doesn’t tell us that much). Only the Cindarella-circa-11:58pm Washington Nationals have notched a lower total owing to their .252/.318/.385 team batting line, a performance only slightly behind the Pirates’ .256/.319/.396. The Pirates would have to be transported back to 1992–back when Jose Lind (.235/.270/.269) and Jeff King (.231/.268/.371) held regular jobs in the Pirate infield–for their performance this year to be even league average. Still, considering the Pittsburgh nine had only mustered 79 runs through the season’s first month, the fact that they’re not stuck under 400 runs should be seen as some slight improvement.

Let’s take a look at how things have changed since we last visited MLB’s more inept bats. This time, rather than looking at simply runs scored as a percentage of runs plus runners left on base, we’ll add home runs into the equation, subtracted from both the numerator–now R-HR–and the denominator–LOB+R-HR. This update doesn’t change the historical ineptitude of our three sample offenses, but it does give us a slightly better idea of which offenses are failing at driving runners home as opposed to simply failing to get them on base in the first place. Here’s what we get:

TEAM  LOB    R     HR     R%     MayR%    Change
----  ---   ---   ---   -----    -----    -----
ANA   729   527    99   37.0%    37.6%    -0.6%
CHA   683   540   143   36.8%    35.4%     1.4%
TOR   786   554   103   36.5%    35.6%     0.9%
SLN   798   582   127   36.3%    32.7%     3.6%
BOS   885   628   133   35.9%    36.7%    -0.8%
TBA   760   526   105   35.6%    35.6%     0.0%
TEX   787   621   189   35.4%    34.5%     0.9%
OAK   828   553    99   35.4%    27.2%     8.2%
KCA   728   485    91   35.1%    29.2%     5.9%
DET   773   516   101   34.9%    40.1%    -5.2%
SEA   734   485    92   34.9%    35.2%    -0.3%
CIN   798   577   157   34.5%    31.0%     3.5%
ATL   778   535   126   34.5%    33.1%     1.4%
NYA   867   601   157   33.9%    33.4%     0.5%
FLO   815   511    94   33.8%    38.9%    -5.1%
NYN   775   512   116   33.8%    32.9%     0.9%
SFN   764   473    84   33.7%    35.6%    -1.9%
MIN   784   495    99   33.6%    34.2%    -0.6%
CLE   799   529   129   33.4%    27.9%     5.5%
MIL   793   518   128   33.0%    34.8%    -1.8%
HOU   767   481   109   32.7%    32.9%    -0.2%
BAL   777   524   148   32.6%    36.5%    -3.9%
COL   824   503   106   32.5%    34.5%    -2.0%
PHI   865   525   111   32.4%    30.4%     2.0%
PIT   803   471    92   32.1%    25.8%     6.3%
LAN   791   485   112   32.0%    34.6%    -2.6%
SDN   846   495    99   31.9%    29.3%     2.6%
CHN   786   501   138   31.6%    33.1%    -1.5%
WAS   792   431    81   30.6%    31.6%    -1.0%
ARI   888   498   132   29.2%    31.3%    -2.1%

As we would expect going from a small sample size (games through May 3) to a large one (through Tuesday), the range of values shrinks dramatically, from 14.3% to 7.8%. Those outliers we saw in May–Oakland, Pittsburgh, and Cleveland as well as San Diego and Kansas City on the low end; Detroit and Florida on the high end–have all moved towards the middle of the pack, reversing their earlier trends.

While regression to the mean is a nice explanation, we can look at a few example teams to discern other possible reasons behind the change. The A’s, for example, spent the first two months of the season without shortstop Bobby Crosby or new first baseman Dan Johnson who are hitting .293/.362/.459 and .322/.407/.543 on the season. Replacing Erubiel Durazo (.237/.305/.368) and taking PAs away from Marco Scutaro (.244/.312/.363) and Mark Ellis (.303/.356/.418), Crosby and Johnson possess the slugging percentage that the A’s sorely lacked in the season’s first month. With the new power bats–as well as newly acquired Jay Payton who’s slugging .536 since coming over from Boston–the A’s went from hitting .239/.312/.340 as a team to .276/.345/.430. Their walk rate remained almost exactly the same, but their noticeable increase in batting average and slugging was the key to their turnaround.

Likewise, the Indians have batted .279/.340/.453 since May 3 as opposed to .227/.297/.380 before. Again, their walk rate is almost identical while most of their gains have come in the batting average department with an additional boost in isolated power (ISO). As opposed to the A’s shuffling their roster, the Indians’ success was more a result of many players underperforming their established levels of production turning things around. Victor Martinez (.198/.275/.308 before, .302/.381/.493 after), Grady Sizemore (.238/.274/.388, .294/.356/.486), and Coco Crisp (.252/.304/.346, .310/.357/.484) were the biggest culprits.

In Pittsburgh? Same story, .232/.301/.362 before, .263/.327/.406 after; note that the Pirates’ gains are significantly smaller than those of the Indians or Athletics, as is their improvement in their percentage of base runners plated. The Pirates have had to battle through the absence of Craig Wilson (.292/.427/.375 before) and Benito Santiago (though perhaps “battle through” isn’t quite appropriate to a player hitting .261/.261/.391). Jack Wilson (.165/.191/.231, .252/.302/.377) has rebounded from a terrible April to compliment the increase in power from Jason Bay, Rob Mackowiak, and the recently departed Matt Lawton to help the Pirates pull out of the cellar with regards to their runner scoring percentage. The division, however, is another matter.

So what can we expect as the season moves into the final two months? Let’s look at the same list but with teams’ batting stats and the projected runner scoring percentage (based on a regression on those stats for all teams from 1972-2005).

TEAM    R%      AVG      OBP     SLG    Proj    Trend
----    -----   ---      ---     ---    -----   -----
ANA     37.0%   .268    .321    .408    34.0%   -3.0%
CHA     36.8%   .262    .322    .425    33.6%   -3.2%
TOR     36.5%   .272    .336    .421    34.6%   -1.9%
SLN     36.3%   .271    .338    .433    34.7%   -1.6%
BOS     35.9%   .282    .360    .453    36.3%    0.4%
TBA     35.6%   .276    .330    .424    35.1%   -0.5%
TEX     35.4%   .272    .335    .477    35.5%    0.1%
OAK     35.4%   .267    .336    .409    33.9%   -1.5%
KCA     35.1%   .261    .316    .400    33.1%   -2.0%
DET     34.9%   .274    .325    .423    34.9%    0.0%
SEA     34.9%   .257    .313    .393    32.5%   -2.4%
CIN     34.5%   .263    .335    .454    34.2%   -0.3%
ATL     34.5%   .266    .328    .436    34.2%   -0.3%
NYA     33.9%   .274    .352    .450    35.4%    1.5%
FLO     33.8%   .276    .335    .418    35.0%    1.2%
NYN     33.8%   .262    .321    .417    33.5%   -0.3%
SFN     33.7%   .266    .321    .397    33.6%   -0.1%
MIN     33.6%   .260    .324    .398    33.0%   -0.6%
CLE     33.4%   .268    .329    .437    34.4%    1.0%
MIL     33.0%   .259    .328    .424    33.3%    0.3%
HOU     32.7%   .256    .317    .410    32.7%    0.0%
BAL     32.6%   .275    .330    .456    35.5%    2.9%
COL     32.5%   .266    .327    .413    33.9%    1.4%
PHI     32.4%   .267    .341    .412    33.9%    1.5%
PIT     32.1%   .256    .319    .396    32.5%    0.4%
LAN     32.0%   .258    .324    .404    32.9%    0.9%
SDN     31.9%   .262    .333    .402    33.2%    1.3%
CHN     31.6%   .271    .322    .444    34.8%    3.2%
WAS     30.6%   .252    .318    .385    31.9%    1.3%
ARI     29.2%   .259    .330    .424    33.3%    4.1%

The A’s have turned things around so much that they’re actually outperforming projected runner scoring percentage based on their composite hitting stats. We should expect them to plate about 33.9% of their runners for the rest of the season, as opposed to their current rate of 35.4%, if their hitting continues at its current pace. On the other hand, the Indians and Pirates are still short of their projected rates and should be expected to continue to improve. Interestingly, the Diamondbacks are underperforming their expected runner scoring percentage by a higher percentage than any team since 1972. If anybody should be expected to turn things around in the season’s final two months, it’s the Arizona offense.

As mentioned, it would be easy to say that the turnaround of the A’s, Pirates, and Indians in runner scoring percentage was inevitable and while some turnaround was very likely, the impressive turnaround of the A’s and Indians in particular is significantly more than could have been expected back in early May. While most of their offensive turnaround can be attributed to the individual performances of the players on the field, the turnaround in their ability to plate those runners they’re putting on the bases certainly didn’t hurt things.