added tour-level doubles 2000-20

JeffSackmann · Mar 13, 2020 · 980a85a · 980a85a
1 parent ca3ce9a
commit 980a85a
Show file tree

Hide file tree

Showing 23 changed files with 26,438 additions and 4 deletions.
diff --git a/README.md b/README.md
@@ -18,11 +18,15 @@ MatchStats are included where I have them. In general, that means 1991-present f
 
 There are some tour-level matches with missing stats. Some are missing because ATP doesn't have them. Others I've deleted because they didn't pass some sanity check (loser won 60% of points, or match time was under 20 minutes, etc). Also, Davis Cup matches are included in the tour-level files, but there are no stats for Davis Cup matches until the last few seasons.
 
----
+# Doubles
 
-Please read, understand, and abide by the license below. It seems like a reasonable thing to ask, given the hundreds of hours I've put into amassing and maintaining this dataset. Unfortunately, a few bad apples have violated the license, and when people do that, it makes me considerably less motivated to continue updating. That's one reason why I no longer keep these files completely up-to-date.
+I've added tour-level doubles back to 2000. Filenames follow the convention atp_matches_doubles_yyyy.csv. I may eventually be able to add tour-level doubles from before 2000, as well as lower-level doubles for some years. Most of the columns are the same, though in a different order.
 
----
+# Attention
+
+Please read, understand, and abide by the license below. It seems like a reasonable thing to ask, given the hundreds of hours I've put into amassing and maintaining this dataset. Unfortunately, a few bad apples have violated the license, and when people do that, it makes me considerably less motivated to continue updating.
+
+Also, if you're using this for academic/research purposes (great!), take a minute and cite it properly. It's not that hard, it helps others find a useful resource, and let's face it, you should be doing it anyway.
 
 # License
 

diff --git a/atp_matches_doubles_2000.csv b/atp_matches_doubles_2000.csv
diff --git a/atp_matches_doubles_2001.csv b/atp_matches_doubles_2001.csv
diff --git a/atp_matches_doubles_2002.csv b/atp_matches_doubles_2002.csv
diff --git a/atp_matches_doubles_2003.csv b/atp_matches_doubles_2003.csv
diff --git a/atp_matches_doubles_2004.csv b/atp_matches_doubles_2004.csv
diff --git a/atp_matches_doubles_2005.csv b/atp_matches_doubles_2005.csv
diff --git a/atp_matches_doubles_2006.csv b/atp_matches_doubles_2006.csv
diff --git a/atp_matches_doubles_2007.csv b/atp_matches_doubles_2007.csv
diff --git a/atp_matches_doubles_2008.csv b/atp_matches_doubles_2008.csv
diff --git a/atp_matches_doubles_2009.csv b/atp_matches_doubles_2009.csv
diff --git a/atp_matches_doubles_2010.csv b/atp_matches_doubles_2010.csv
diff --git a/atp_matches_doubles_2011.csv b/atp_matches_doubles_2011.csv
diff --git a/atp_matches_doubles_2012.csv b/atp_matches_doubles_2012.csv
diff --git a/atp_matches_doubles_2013.csv b/atp_matches_doubles_2013.csv
diff --git a/atp_matches_doubles_2014.csv b/atp_matches_doubles_2014.csv
diff --git a/atp_matches_doubles_2015.csv b/atp_matches_doubles_2015.csv
diff --git a/atp_matches_doubles_2016.csv b/atp_matches_doubles_2016.csv
diff --git a/atp_matches_doubles_2017.csv b/atp_matches_doubles_2017.csv
diff --git a/atp_matches_doubles_2018.csv b/atp_matches_doubles_2018.csv
diff --git a/atp_matches_doubles_2019.csv b/atp_matches_doubles_2019.csv
diff --git a/atp_matches_doubles_2020.csv b/atp_matches_doubles_2020.csv
diff --git a/matches_data_dictionary.txt b/matches_data_dictionary.txt
@@ -86,4 +86,14 @@ winner_rank
 winner_rank_points
 - number of ranking points, where available
 loser_rank
-loser_rank_points
+loser_rank_points
+
+* _doubles_ files notes
+
+The matches_doubles files have similar columns, though not all in the same order.
+
+The identifying information for each player refers to 'winner1', 'winner2', 'loser1', and 'loser2'. The labels 1 and 2 are not assigned for any particular reason.
+
+In general, the tournament IDs for doubles results are the same as for singles results (so, for instance, you can see which players entered both draws at the same event), though this is not guaranteed for every single tournament, since some of the data came from different sources.
+
+The stats columns ('w_ace' etc) are per *team*, not per player. That's a function of how tennis stats are typically recorded, not a decision on my part.