Errata

On line 20 of the rosters.pl code (which also appears in the hack_22_rosters.pl code
that can be downloaded from the web site), the regular expression for verifying the
Retrosheet ID is incorrect. The code reads:

if (/[a-z]{5}d{3}/) {

Retrosheet ID's can include a dash when a player's last name is shorter than four
characters (like Derrek Lee, lee-d002). The line should read:

if (/[a-z-]{5}d{3}/) {

The dash in the first character class will match players like Lee or Jason Bay, etc.

In line 5 of the code,

print "retroID,lastName,firstName,bats,throws,team,pos
";

should read:

print "year,retroID,lastName,firstName,bats,throws,team,pos
";

Otherwise, all of the columns will be shifted off by one (and position information
will not make it into the database).

The changes to the code in translate.pl for Windows (listed in the confirmed errata
list) are not enough to get this script working on Windows XP. I had to make the
following changes to get this code to work on Windows. (Note this fix uses the free
7-zip archive tool):

Change from this:
print `cat all_hdr.txt > $outfile`;
print `cat all_hdr.txt > $outfile2k`;

to this:
print `type all_hdr.txt > $outfile`;
print `type all_hdr.txt > $outfile2k`;

Change from this:
print `unzip -qq -o $archive`;

to this:
print `7z x -y $archive`;

Change from this:
print `./BEVENT.EXE -y $century$year -f 0-96 $file >> $outfile`;

to this:
print `BEVENT.EXE -y $century$year -f 0-96 $file >> $outfile`;

Change from this:
print `./BEVENT.EXE -y $century$year -f 0-96 $file >> $outfile2k`;

to this:
print `BEVENT.EXE -y $century$year -f 0-96 $file >> $outfile2k`;

the line
my $fday = length($mday) == 1 ? '0' . $mon : $mday;
should read
my $fday = length($mday) == 1 ? '0' . $mday : $mday;

to format the 'day of the month' portion of the URL that will fetch data from mlb.com

When running this on Windows, using ActivePerl, the DBD-mysql module package was not
yet installed.

In order to run this you will need to load the Perl Package Manager (as in Hack #12).
Then type "install DBD-mysql". This will download the DBI and DBD-mysql packages,
which will allow the perl to run and connect to the database.

(The DBD-mysql package should probably be included in Hack #12 on page 66 in the list
of hidden pacakges. I do not know if users of other operating systems will need to
separately download this as well.)

This is not a mistake, but a technical update.

For the 2006 season, GameDay has replaced the players.txt file with the players.xml
file. Both files were available from Opening Day through 4/25/06, but as of 4/26/06
only the XML file exists.

Contents of pbp can no longer be fetched. Pitch-by-pitch data has not been available
as of the 2006 All-Star game, 7/11/06. Cannot find pitch location information stored
anywhere, except for final batter of a game.

All references to a field named playerID should instead be to idxLahman.

In R, the last line of the following code sample causes an error message stating "Error: object 'H99' not found". I triple-checked and it seems i am following the code example :



batting1999 <- subset(batting, batting$yearID == 1999 & batting$AB > 249, select = c("H", "AB", "playerID"))
batting2000 <- subset(batting, batting$yearID == 2000 & batting$AB > 249, select = c("H", "AB", "playerID"))
batting2001 <- subset(batting, batting$yearID == 2001 & batting$AB > 249, select = c("H", "AB", "playerID"))
batting2002 <- subset(batting, batting$yearID == 2002 & batting$AB > 249, select = c("H", "AB", "playerID"))

names(batting1999) <- c("H99", "AB99", "playerID")
names(batting2000) <- c("H00", "AB00", "playerID")
names(batting2001) <- c("H01", "AB01", "playerID")
names(batting2002) <- c("H02", "AB02", "playerID")

batting.test <- merge(merge(merge(batting1999, batting2000, by="playerID"), batting2001, by="playerID"), batting2002, by="playerID")

batting.test$sample.AVG <- (H99 + H00 + H01)/(AB99 + AB00 + AB01)