license	cc-by-4.0
pretty_name	ATP 2015 Tour-Level Singles Results
language	en
task_categories	tabular-classification tabular-regression time-series-forecasting
tags	tennis sports atp sports-analytics match-results
size_categories	1K<n<10K
configs	config_name data_files default split path train data/atp_2015_results.parquet

ATP 2015 Tour-Level Singles Results

A clean, ready-to-use table of every 2015 ATP tour-level men's singles match — 4,479 matches across 66 tournaments (Grand Slams, Masters 1000, ATP 250/500 events, and the season-ending ATP Finals). One row per match, with tournament context, both players' identifiers, seeding and ranking, the final score, and the match outcome.

Who publishes this (vendor disclosure)

This dataset is published by Live Tennis API (https://livetennisapi.com), the commercial tennis data provider that runs the API of the same name. It is a first-party sample drawn from our historical archive (men's and women's professional results, 1968–2022). We are the vendor, so treat this card as vendor-authored and judge it on the data, which you can inspect directly in the files.

This is a sample

This release is deliberately one ATP season of match-level results — a taste of the archive, not the whole thing. The full 1968–2022 archive (both tours, all levels, plus per-match statistics, head-to-head, and point-by-point tape as separate products) is available through the Live Tennis API:

• Product & pricing: https://livetennisapi.com
• API documentation: https://docs.livetennisapi.com
• Free keyed tier (develop-and-test): https://livetennisapi.com/subscribe/free

The sample contains match results only — it does not include point-by-point data or in-play statistics, which are separate paid products.

Dataset at a glance

Rows (matches)	4,479
Columns	26
Tournaments	66
Tour / gender	ATP (men's)
Discipline	Singles
Season	2015 (2015-01-04 → 2015-11-15)
Levels	Grand Slam (956), Masters 1000 (783), ATP 250/500 (2,725), Tour Finals (15)
Surfaces	hard (2,487), clay (1,423), grass (569)
Outcomes	completed (4,329), retired (138), walkover (12)
Format	CSV + Parquet
License	CC-BY-4.0

Files

• data/atp_2015_results.csv — UTF-8, header row, one match per line.
• data/atp_2015_results.parquet — same data, typed columns (nullable integers for missing ranks/seeds/heights), for direct loading with pandas / datasets.

Column dictionary

Column	Type	Description
tourney_name	string	Tournament name (e.g. Brisbane, Australian Open).
tourney_level	string	Event tier: G = Grand Slam, M = Masters 1000, A = ATP Tour 250/500, F = ATP (Tour) Finals.
surface	string	Court surface: hard, clay, or grass.
draw_size	int	Number of players in the singles draw.
tourney_date	date	Tournament start date, YYYY-MM-DD (shared by all matches in that event).
round	string	Round: R128, R64, R32, R16, QF, SF, F, plus RR/BR for round-robin events.
best_of	int	Maximum sets (3 or 5).
minutes	int, nullable	Match duration in minutes (present for ~52% of 2015 matches).
winner_name	string	Match winner.
winner_hand	string	Winner's playing hand: R, L, or U (unknown).
winner_ht	int, nullable	Winner's height in cm.
winner_ioc	string	Winner's country (IOC 3-letter code, e.g. SUI).
winner_age	float, nullable	Winner's age in years at the tournament date.
winner_rank	int, nullable	Winner's ATP singles ranking at the tournament date.
winner_seed	int, nullable	Winner's tournament seed (null if unseeded).
winner_entry	string, nullable	Winner's entry type: WC wild card, Q qualifier, LL lucky loser, PR protected ranking, SE special exempt, ALT alternate; blank = direct acceptance.
loser_name	string	Match loser.
loser_hand	string	Loser's playing hand: R, L, or U.
loser_ht	int, nullable	Loser's height in cm.
loser_ioc	string	Loser's country (IOC 3-letter code).
loser_age	float, nullable	Loser's age in years at the tournament date.
loser_rank	int, nullable	Loser's ATP singles ranking at the tournament date.
loser_seed	int, nullable	Loser's tournament seed (null if unseeded).
loser_entry	string, nullable	Loser's entry type (see winner_entry).
score	string	Final score from the winner's perspective, e.g. 6-3 6-1; tiebreak points in parentheses, e.g. 7-6(4).
outcome	string	completed, retired (loser retired), or walkover.

Notes on missing values: winner_rank/loser_rank are null when a player had no ranking at that date (e.g. protected-ranking or wild-card entries); winner_seed/loser_seed are null for unseeded players; minutes and heights are null where the archive lacks the value. Missing values are empty in the CSV and null (nullable-integer) in the Parquet.

Usage

from datasets import load_dataset

ds = load_dataset("livetennisapi/atp-2015-results", split="train")
print(ds[0])

Or straight from the Parquet with pandas:

import pandas as pd

df = pd.read_parquet("data/atp_2015_results.parquet")
# Federer's 2015 matches
fed = df[(df.winner_name == "Roger Federer") | (df.loser_name == "Roger Federer")]
print(len(fed), "matches")

Source & methodology

Rows are exported directly from the Live Tennis API historical archive (the same store that backs the API's /archive endpoints). The scope filter is: ATP tour (men's), singles, event levels Grand Slam / Masters 1000 / ATP 250-500 / Tour Finals, calendar year 2015. No rows are synthetic or interpolated — every match is a real recorded result. Player ages, heights, ranks, and seeds are the archive's stored values as of each tournament date.

License — CC-BY-4.0

This dataset is licensed under the Creative Commons Attribution 4.0 International License. You are free to share and adapt the data, including for commercial use, provided you give appropriate credit.

Required attribution:

Data: Live Tennis API (https://livetennisapi.com), ATP 2015 Tour-Level Singles Results, licensed under CC-BY-4.0.

See the LICENSE file for the full license text.

Citation

@misc{livetennisapi_atp_2015,
  title  = {ATP 2015 Tour-Level Singles Results},
  author = {Live Tennis API},
  year   = {2015},
  note   = {First-party sample of the 1968--2022 tennis archive},
  url    = {https://livetennisapi.com}
}