NAME:  playbyplay2008.dat
TYPE:  Population  
SIZE:  193492 observations on  36 variables

DESCRIPTIVE ABSTRACT:
Play by play data for all plays for all games during the 2008 Major
League Baseball season.  

SOURCE:
The Regular Season Event Files from the Retrosheet organization at
www.retrosheet.org/game.htm.  The program bevent was used to
extract variables from thirty files corresponding to play by play
information at each of the thirty baseball stadiums.  All of the 
data was combined into a single file.

FORMATS:
The play by play dataset is available both as a data file "playbyplay2008.dat" 
and a R workspace "playbyplay2008.Rdata".  Tab characters are used to separate variables in 
the data file.

READING INTO R:
The play-by-play data file can be read into R by the command
retrosheet=read.delim("http://bayes.bgsu.edu/baseball/playbyplay2008.dat")

VARIABLE DESCRIPTIONS:
Each row represents information about a particular baseball play (either a batting event 
or a base running event) during a particular game.

game.id		game id  
v_team		code for visiting team    
inning		inning of game  
team_at_bat	id of team at bat    
outs			number of current outs   
balls		number of balls    
strikes		number of strikes  
v_score		current visitor score   
h_score		current home score    
batter		id code of batter    
batter_hand	batter side (L or R)
pitcher		id code of pitcher    
pitcher_hand	pitcher side (L or R)    
b1_runner	code of runner on first    
b2_runner	code of runner on second    
b3_runner	code of runner on third   
event		code description of event    
leadoff_flag	leadoff hitter? (TRUE or FALSE)   
pitchhit_flag	leadoff hitter? (TRUE or FALSE)
def_pos		defensive position of batter
  (1 through 9 correspond to numerical fielding locations,
  10 corresponds to DH, and 11 corresponds to pinch-hitter)
batting_pos	position in batting order 
event_code	event code
          0    Unknown event
          1    No event
          2    Generic out
          3    Strikeout
          4    Stolen base
          5    Defensive indifference
          6    Caught stealing
          7    Pickoff error
          8    Pickoff
          9    Wild pitch
          10   Passed ball
          11   Balk
          12   Other advance
          13   Foul error
          14   Walk
          15   Intentional walk
          16   Hit by pitch
          17   Interference
          18   Error
          19   Fielder's choice
          20   Single
          21   Double
          22   Triple
          23   Home run
          24   Missing play
bevent_flag	end of batting appearance (TRUE or FALSE)
ab_flag		indicator of at-bat (TRUE or FALSE)
hit_value		value of hit  (0, 1, 2, 3, 4)
sh_flag		sacrifice hit?  (TRUE or FALSE)
sf_flag		sacrifice fly?  (TRUE or FALSE)
outs_play		number of outs recorded
rbi_play		number of rbi's credited
wp_play		wild pitch? (TRUE or FALSE)
pb_flag		passed ball? (TRUE or FALSE)
nerrors		number of errors on play
batter_dest	base reached by batter (0, 1, 2, 3, or 4)
b1_runner_d	code of new runner on first
b2_runner_d	code of new runner on second
b3_runner_d	code of new runner on third
date			date of game
h_team		code for home team

Missing values are denoted with NA.

PEDAGOGICAL NOTES:
By use of the Retrosheet datasheet, one can investigate how batters and pitchers perform in 
different situations. For example, one can explore a particular batter's performance at home 
and away games, during different innings and bases situations during a game, and
against different pitchers. To understand so-called situational effects, it is helpful to 
look at a particular effect, say home versus away, for all hitters. Then one can understand 
the general situational bias (players typically perform better during home games) and detect 
the players who deviate from the general pattern.

REFERENCES:
Albert, Jim, Baseball Data at Season, Play-by-Play, and Pitch-by-Pitch Levels

SUBMITTED BY:
Jim Albert
Department of Mathematics and Statistics
Bowling Green State University
Bowling Green, OH  43403
albert@bgnet.bgsu.edu