About Me

My photo
Bethesda, MD
A staunch believer in recovery, this newsletter provides consumers with the opportunity to express themselves both creatively and intellectually. We welcome your feedback! Please leave us comments :)

Friday, August 24, 2012

When Statistics Lie


Eleven statistics that attempt to measure a starting pitcher’s pitching performance throughout a game, season, or career include: 
A)      ERA
B)     Run Allowed/Nine innings
C)     Wins, Losses, No Decisions
D)     Home Runs Allowed/Nine innings
E)      Batting Average Against
F)      Strikeouts/Nine innings
G)     Walks/Nine innings
H)     Walk to Strikeout Ratio
I)        Innings Pitched
J)       Number of Starts
K)      Average  Pitches/Nine innings

Which statistics are more accurate indicators of the performance of a starting pitcher throughout a specific period of time?  Most people believe that E.R.A. is the most obvious indicator of a pitcher’s performance, whether or not the E.R.A. is compiled over the short, medium, or long run.  E.R.A. is a more reliable statistic than runs allowed/nine innings, because E.R.A. attempts to consider the defensive ability of the pitcher’s team.  However, it is not possible to accurately measure a team’s defense simply by the number of errors that a team incurs.  Furthermore, pitchers pitch in different home ballparks where most pitchers pitch in approximately half of their games.  There are no set dimensions to Major League Stadiums.  Pitchers who pitch in hitter’s stadiums where runs are plentiful, generally have higher E.R.A.’s than pitchers who pitch in pitcher’s stadiums where runs are scarcely distributed; Ceteris Paribus. 
What other biases exist to E.R.A. as an exact indicator of a pitcher’s performance.  All teams do not play the exact same schedule as all of the other teams.  Teams play more games against division foes than against League foes; furthermore, they play against different interleague teams than other teams.  Therefore, some pitchers will face different teams with different offensive run potential.  Pitchers who pitch in the American League relative to the National League, have higher ERA’s on average, because of the insertion of the designated hitter in the lineup in games played with American League rules.  Since in Interleague games the rules of the home team’s League apply, American League teams play considerably more games with American League rules than National League teams.
A second good indicator of a starting pitcher’s is his win/loss record.  This eliminates the biases of whether or not the pitcher pitches in a hitter’s or a pitcher’s ballpark, and any bias due to the insertion of the designated hitter in the lineup.  However, it probably has the largest bias that occurs, because some pitchers play on better teams than other pitchers (Won/Loss record).  It would be more accurate to divide the pitcher’s record by the team’s record to retrieve the pitcher’s performance level.   If the pitcher’s record is significantly better than the team’s record then the pitcher is probably performing well; even if he himself has a losing record.  The one exception where a starting pitcher would not be judged accurately by this method is if he plays on a team with four other exceptional or disastrous starting pitchers.
Furthermore, road statistics are more accurate than home statistics, because most teams play approximately the same number of road games in each stadium.  This is truer of teams within the same division.  This reduces the effect of the hitter’s or pitcher’s ballpark on E.R.A.  Still, a team would never play a road game at their own home stadium, leaving a slight bias.  All of the other statistics such as home runs allowed, and batting average against would also be considerately less biased if they were taken from a team’s road games for previously mentioned reasons. 
The same type of analysis could be made for offensive statistics as well as pitching statistics.  Eleven important offensive statistics include:
A)     Runs Scored/At Bat
B)     RBI/At Bat
C)     Batting Average
D)     Hits/At Bat
E)      Home Runs/At Bat
F)      Extra Base Hits/At Bat
G)     Walks/At Bat
H)     Strikeouts/At Bat
I)        Slugging Percentage
J)       On Base Percentage
K)      Stolen Bases
However, categories such as runs scored and RBI depend largely on the team’s run production, as well as the batter’s performance.  Where the batter is positioned in the batting order also has a large effect on runs scored and his RBI total.  The dimensions of the home stadium of a player have a large effect on a player’s number of Home Runs, because there are no standard dimensions to Major League Stadiums.  Offensive statistics do not have a similar statistic to Wins and Losses, because the batter is only one of eight position players in the lineup.

No comments:

Post a Comment