This dataset requires a paid license. Please purchase access at the link below and provide your Hugging Face username. Once purchased, submit an access request on this page. Access is granted manually.

NFL Rest Advantage and Travel Spot Dataset for Spread Research

Question this dataset helps answer: Which NFL teams historically cover, miss, or play over or under expectations in short-week, rest-advantage, and travel-heavy schedule spots?

This dataset helps sports researchers, dashboard builders, fantasy analysts, and betting-market historians compare NFL schedule spots using public rest fields, approximate travel load, closing spread and total context, and rolling team trend features that are available before kickoff.

Use the public sample to check:

Which teams and seasons are covered, and how often 3-plus-day rest edges appear in the preview
Which columns are safe pre-event features versus post-event outcome fields
Whether the row structure supports ATS, total, travel-bucket, and team-trend analysis without extra joins

Preview vs full dataset:

Preview: 727 rows and 87 columns
Full: 14552 rows and 87 columns Full dataset: here

from datasets import load_dataset

dataset = load_dataset("Karmane/nfl-rest-advantage-travel-spot-research")
print(dataset)

What this product is

This is not a generic NFL scoreboard export. It is a schedule-spot research table designed around a repeated buyer problem: analysts want to ask whether teams behave differently when the schedule context changes, but the required joins are annoying enough that many workflows stall at the data-prep stage. Rest days, opponent rest, home-away splits, divisional context, weather, roof, surface, spread, total, moneyline, and rolling trend features often live in different notebooks, ad hoc spreadsheets, or half-maintained scripts. That makes repeatable historical research slower than it should be.

The dataset solves that by standardizing the grain at one team in one completed NFL game. That row design matters because the practical research questions are almost always team-side questions. Did this team cover with a rest advantage? Did this away team travel across time zones? Did a short-week underdog stay under the closing total? Did the same team enter on stronger recent ATS form than it had three weeks earlier? A team-side row answers those questions naturally without requiring the buyer to manually explode one game row into two observations first.

Coverage and why the row count matters

The current build includes 14,552 rows across 27 seasons of completed NFL games, with the latest completed game date in the source currently at 2026-02-08. That volume matters because schedule-spot research breaks the league into many smaller buckets. Once a buyer filters to away teams, then to short weeks, then to divisional games, then to favorites or underdogs, tiny datasets become useless very quickly. The full dataset is large enough for dashboard slicing, rolling-trend inspection, historical filtering, and exploratory modeling across many subgroups without collapsing to a handful of observations.

The build also includes 618 playoff team-game rows, 2,418 team-game rows with a rest advantage, and 581 short-week rows. The average away-team scheduled travel proxy is about 953.8 miles. Those numbers are not marketing filler. They show that the table is broad enough to support real screening work instead of just anecdotal spot checks.

What each row represents

Each row represents one team in one completed NFL game. Core identifiers include the source game id, season, week, game type, team, opponent, and side. Pregame schedule-context fields include official rest days, opponent rest days, the rest differential, short-week and long-rest flags, post-bye flags, divisional status, prime-time windows, roof, surface, weather, and an approximate travel proxy derived from team home-market coordinates and the listed venue. Market-context fields include the team-side spread, opponent-side spread, spread prices, total, over and under prices, moneylines when available, implied win probabilities, and implied team totals derived from the closing line.

The same row then includes postgame outcomes such as team score, opponent score, margin, ATS margin, cover result, total result, and score versus implied total. Rolling pregame features are shifted so they only use earlier games. That separation is critical. Buyers can keep pre-event fields on the modeling side and reserve outcomes for target variables or evaluation.

Three concrete research checks the sample supports

First, the sample lets you inspect which seasons, teams, and game types are covered and confirm that rest fields, line fields, and outcome fields are all present in one schema. Second, the sample lets you test the pre-event versus post-event split by reviewing rolling features such as prior cover rate or prior average margin against outcome columns such as ATS margin and total margin. Third, the sample lets you run immediate spot checks, such as filtering to away teams with 3-plus days of rest disadvantage or comparing travel buckets by spread status, without building custom joins.

Those are exactly the types of checks a buyer should run before purchasing a larger sports research product. The sample is not there to tell someone what to bet. It is there to prove that the data model is useful, coherent, and broad enough to support a real analysis workflow.

Why this category has commercial demand

Demand in this lane is materialized rather than hypothetical. There are active public tools and articles built around NFL rest-advantage ATS records, schedule-rest disparity, and weekly betting-trends research. That means buyers already spend attention and tooling effort on this exact family of questions. They may phrase the problem differently depending on the workflow, but the core task is the same: connect schedule context to market expectations and historical outcomes in a form that is quick to filter and compare.

This is especially useful for buyers who need repeatable research infrastructure rather than one-off blog content. A fantasy analyst may want to blend the schedule-spot context into player or team projections. A sports researcher may want to benchmark which teams repeatedly face ugly travel or rest sequences. A dashboard builder may want one stable table that can power season filters, bucket comparisons, and rolling trend cards. A betting-market historian may simply want a clean completed-game panel with both pregame and outcome fields in one place.

Responsible use note

This dataset is for research, analysis, dashboard building, educational use, and exploratory modeling. It does not provide betting advice, picks, guarantees, or predictions of future outcomes.

Modeling notes

Pregame features include rest days, rest differential, travel proxy, time-zone shift, home-away status, divisional context, spread, total, moneyline, implied totals, and rolling prior-game metrics. Outcome fields include score, margin, ATS margin, cover result, and total result. A clean train-test split can be done chronologically by season or by game date because each rolling feature is shifted and only uses earlier games for the same team. Neutral-site or special-travel situations are simplified through the schedule-proxy approach, so travel fields should be treated as approximate context rather than exact operational logistics. Older seasons also have less complete moneyline coverage than later seasons.

If your workflow needs a reusable NFL team-game table for schedule-spot research rather than another raw game list, this is the product.

Dataset Preview

record_id	game_id	season	game_type	week	gameday	weekday	is_playoff_game	team	opponent	team_side	is_home	team_conference	team_division	opponent_conference	opponent_division	div_game	rest_days	opponent_rest_days	has_rest_advantage	has_rest_disadvantage	is_short_week	is_long_rest	is_off_bye	rest_bucket	rest_edge_bucket	temp	wind	roof	surface	stadium	travel_miles	timezone_shift_hours	traveled_eastward	traveled_westward	travel_bucket	closing_total_line	team_spread_line	opponent_spread_line	team_implied_total	opponent_implied_total	team_favorite_flag	pickem_flag	prime_time_flag	team_score	opponent_score	margin	won_game	tied_game	ats_margin	covered_spread	pushed_spread	total_points	total_margin	went_over_total	pushed_total	score_vs_implied_total	source_url	source_domain	last_collected_at
1999_01_ARI_PHI_ARI	1999_01_ARI_PHI	1999	REG	1	1999-09-12 00:00:00	Sunday	False	ARI	PHI	away	False	NFC	West	NFC	East	True	7	7	False	False	False	False	False	standard_rest	even_rest	75.0	3.0	outdoors	astroturf	Veterans Stadium	2085.6	2	True	False	2000_plus_miles	37	-3	3	20	17	True	False	False	25	24	1	True	False	-2	False	False	49	12	True	False	5	https://github.com/nflverse/nflverse-data/releases/download/schedules/games.parquet	github.com, nflverse.com	2026-06-11 14:37:58
1999_01_ARI_PHI_PHI	1999_01_ARI_PHI	1999	REG	1	1999-09-12 00:00:00	Sunday	False	PHI	ARI	home	True	NFC	East	NFC	West	True	7	7	False	False	False	False	False	standard_rest	even_rest	75.0	3.0	outdoors	astroturf	Veterans Stadium	0	0	False	False	home	37	3	-3	17	20	False	False	False	24	25	-1	False	False	2	True	False	49	12	True	False	7	https://github.com/nflverse/nflverse-data/releases/download/schedules/games.parquet	github.com, nflverse.com	2026-06-11 14:37:58
1999_01_BAL_STL_BAL	1999_01_BAL_STL	1999	REG	1	1999-09-12 00:00:00	Sunday	False	BAL	STL	away	False	AFC	North	NFC	West	False	7	7	False	False	False	False	False	standard_rest	even_rest			dome	astroturf	TWA Dome	729.6	-1	False	True	250_to_999_miles	39	0	-0	19.5	19.5	False	True	False	10	27	-17	False	False	-17	False	False	37	-2	False	False	-9.5	https://github.com/nflverse/nflverse-data/releases/download/schedules/games.parquet	github.com, nflverse.com	2026-06-11 14:37:58
1999_01_BAL_STL_STL	1999_01_BAL_STL	1999	REG	1	1999-09-12 00:00:00	Sunday	False	STL	BAL	home	True	NFC	West	AFC	North	False	7	7	False	False	False	False	False	standard_rest	even_rest			dome	astroturf	TWA Dome	0	0	False	False	home	39	-0	0	19.5	19.5	False	True	False	27	10	17	True	False	17	True	False	37	-2	False	False	7.5	https://github.com/nflverse/nflverse-data/releases/download/schedules/games.parquet	github.com, nflverse.com	2026-06-11 14:37:58
1999_01_BUF_IND_BUF	1999_01_BUF_IND	1999	REG	1	1999-09-12 00:00:00	Sunday	False	BUF	IND	away	False	AFC	East	AFC	South	True	7	7	False	False	False	False	False	standard_rest	even_rest			dome	astroturf	RCA Dome	435.8	0	False	False	250_to_999_miles	45.5	-3	3	24.25	21.25	True	False	False	14	31	-17	False	False	-20	False	False	45	-0.5	False	False	-10.25	https://github.com/nflverse/nflverse-data/releases/download/schedules/games.parquet	github.com, nflverse.com	2026-06-11 14:37:58

Access Requirements (Paid Dataset)

This dataset is behind manual gated access.

To obtain access:

Purchase the dataset here:
https://thearticulated.gumroad.com/l/isqxw?utm_source=hf_paid_readme&utm_medium=referral&utm_campaign=nfl-rest-advantage-travel-spot-research
Provide your Hugging Face username at checkout.
Return to this Hugging Face page and click:
"Request Access"
Your access will be approved within 1-12 hours.

Once approved, you can use the Python snippet at the bottom of this README to load the dataset.

Dataset Structure

Total rows: 14,552

Total columns: 87

Splits

data: 14,552 rows

Data Files

data: data/data.parquet

Data Dictionary

The table below describes the columns included in this dataset.

column	pandas_dtype	dataset_type	description
record_id	object	string	Text column. Stable unique identifier for one team-side observation in one completed NFL game.
game_id	object	string	Text column. Original nflverse game identifier.
season	Int64	integer	Whole-number numeric column. NFL season year attached to the game record.
game_type	object	string	Text column. Game type code such as REG, WC, DIV, CON, or SB.
week	Int64	integer	Whole-number numeric column. Week number from the source schedule file.
gameday	datetime64[ns]	datetime	Date or timestamp column. Calendar date of the game.
weekday	object	string	Text column. Weekday label from the source schedule file.
gametime	object	string	Text column. Scheduled local kickoff time string from the source schedule file.
is_playoff_game	bool	boolean	True/false column. Boolean flag indicating whether the row comes from a playoff game.
team	object	string	Text column. Team abbreviation for the row side.
opponent	object	string	Text column. Opponent team abbreviation for the row side.
team_side	object	string	Text column. Home or away indicator stored as text.
is_home	bool	boolean	True/false column. Boolean flag indicating whether the team played at home.
team_conference	object	string	Text column. Conference of the row team.
team_division	object	string	Text column. Division of the row team.
opponent_conference	object	string	Text column. Conference of the opponent team.
opponent_division	object	string	Text column. Division of the opponent team.
div_game	bool	boolean	True/false column. Boolean flag showing whether the matchup was a divisional game.
rest_days	Int64	integer	Whole-number numeric column. Official days of rest for the row team from the source schedule data.
opponent_rest_days	Int64	integer	Whole-number numeric column. Official days of rest for the opponent from the source schedule data.
rest_advantage_days	Int64	integer	Whole-number numeric column. Row-team rest days minus opponent rest days.
has_rest_advantage	bool	boolean	True/false column. Boolean flag showing the team entered with more rest than the opponent.
has_rest_disadvantage	bool	boolean	True/false column. Boolean flag showing the team entered with less rest than the opponent.
is_short_week	bool	boolean	True/false column. Boolean flag for rest windows of three days or fewer.
is_long_rest	bool	boolean	True/false column. Boolean flag for rest windows of eight days or more.
is_off_bye	bool	boolean	True/false column. Boolean flag for likely post-bye rest windows of thirteen days or more.
rest_bucket	object	string	Text column. Bucketed rest label derived from rest_days.
rest_edge_bucket	object	string	Text column. Bucketed rest-differential label derived from rest_advantage_days.
temp	float64	float	Decimal numeric column. Temperature field from the source schedule data when available.
wind	float64	float	Decimal numeric column. Wind field from the source schedule data when available.
roof	object	string	Text column. Roof classification for the game venue.
surface	object	string	Text column. Playing surface classification for the game venue.
stadium	object	string	Text column. Venue name from the source schedule data.
travel_miles	float64	float	Decimal numeric column. Approximate scheduled travel distance from the team home market to the game venue, using home-stadium coordinates.
timezone_shift_hours	Int64	integer	Whole-number numeric column. Approximate time-zone shift in hours from the team home market to the game venue.
traveled_eastward	bool	boolean	True/false column. Boolean flag for positive time-zone travel toward the east.
traveled_westward	bool	boolean	True/false column. Boolean flag for negative time-zone travel toward the west.
travel_bucket	object	string	Text column. Bucketed scheduled-travel label for dashboard filtering.
team_moneyline	Int64	integer	Whole-number numeric column. Team-side closing moneyline when available.
opponent_moneyline	Int64	integer	Whole-number numeric column. Opponent-side closing moneyline when available.
team_spread_odds	Int64	integer	Whole-number numeric column. Team-side spread price when available.
opponent_spread_odds	Int64	integer	Whole-number numeric column. Opponent-side spread price when available.
closing_total_line	float64	float	Decimal numeric column. Closing game total line.
over_odds	Int64	integer	Whole-number numeric column. Over price for the closing total when available.
under_odds	Int64	integer	Whole-number numeric column. Under price for the closing total when available.
team_spread_line	float64	float	Decimal numeric column. Team-side closing spread line. Negative values indicate the team was favored.
opponent_spread_line	float64	float	Decimal numeric column. Opponent-side closing spread line from the row-team perspective.
team_implied_win_prob	float64	float	Decimal numeric column. Implied win probability derived from the team moneyline when available.
opponent_implied_win_prob	float64	float	Decimal numeric column. Implied win probability derived from the opponent moneyline when available.
team_implied_total	float64	float	Decimal numeric column. Approximate team implied total derived from the closing spread and total.
opponent_implied_total	float64	float	Decimal numeric column. Approximate opponent implied total derived from the closing spread and total.
team_favorite_flag	bool	boolean	True/false column. Boolean flag indicating the row team closed as the favorite.
pickem_flag	bool	boolean	True/false column. Boolean flag indicating the closing spread was effectively pick'em.
prime_time_flag	bool	boolean	True/false column. Boolean flag for Thursday, Monday, or Saturday standalone windows.
team_score	Int64	integer	Whole-number numeric column. Points scored by the row team.
opponent_score	Int64	integer	Whole-number numeric column. Points scored by the opponent team.
margin	Int64	integer	Whole-number numeric column. Point differential from the row-team perspective.
won_game	bool	boolean	True/false column. Boolean flag indicating the row team won the game outright.
tied_game	bool	boolean	True/false column. Boolean flag indicating the game finished tied.
ats_margin	Float64	float	Decimal numeric column. Against-the-spread margin from the row-team perspective.
covered_spread	bool	boolean	True/false column. Boolean flag indicating the team covered the closing spread.
pushed_spread	bool	boolean	True/false column. Boolean flag indicating the spread landed exactly on the closing number.
total_points	Int64	integer	Whole-number numeric column. Combined points scored by both teams.
total_margin	Float64	float	Decimal numeric column. Actual total points minus the closing total line.
went_over_total	bool	boolean	True/false column. Boolean flag indicating the game finished over the closing total.
pushed_total	bool	boolean	True/false column. Boolean flag indicating the total landed exactly on the closing number.
score_vs_implied_total	Float64	float	Decimal numeric column. Actual team score minus the team implied total derived from the market.
prior_games_played	Int64	integer	Whole-number numeric column. Count of prior games for the same team before this row.
season_to_date_games	Int64	integer	Whole-number numeric column. Count of prior same-season games for the same team before this row.
rolling_win_rate_last_5	float64	float	Decimal numeric column. Pregame rolling mean of outright wins over the previous five games.
rolling_cover_rate_last_5	float64	float	Decimal numeric column. Pregame rolling mean of ATS cover value over the previous five games.
rolling_over_rate_last_5	float64	float	Decimal numeric column. Pregame rolling mean of total-over value over the previous five games.
rolling_margin_last_3	float64	float	Decimal numeric column. Pregame rolling average margin over the previous three games.
rolling_margin_last_5	float64	float	Decimal numeric column. Pregame rolling average margin over the previous five games.
rolling_ats_margin_last_3	float64	float	Decimal numeric column. Pregame rolling average ATS margin over the previous three games.
rolling_ats_margin_last_5	float64	float	Decimal numeric column. Pregame rolling average ATS margin over the previous five games.
rolling_points_for_last_5	float64	float	Decimal numeric column. Pregame rolling average points scored over the previous five games.
rolling_points_allowed_last_5	float64	float	Decimal numeric column. Pregame rolling average points allowed over the previous five games.
rolling_total_margin_last_5	float64	float	Decimal numeric column. Pregame rolling average total-line margin over the previous five games.
rolling_travel_miles_last_3	float64	float	Decimal numeric column. Pregame rolling average scheduled travel miles over the previous three games.
season_to_date_win_pct	float64	float	Decimal numeric column. Pregame same-season outright win percentage before this game.
season_to_date_cover_pct	float64	float	Decimal numeric column. Pregame same-season ATS cover percentage before this game.
season_to_date_avg_margin	float64	float	Decimal numeric column. Pregame same-season average point differential before this game.
season_to_date_avg_ats_margin	float64	float	Decimal numeric column. Pregame same-season average ATS margin before this game.
source_url	object	string	Text column. Primary source file URL used for this build.
source_domain	object	string	Text column. Source domain summary for the row.
last_collected_at	datetime64[us]	datetime	Date or timestamp column. UTC timestamp when this dataset build collected the row.

Intended Use

This dataset is intended for research, experimentation, analysis, and model prototyping.

Loading the Dataset


import os
from datasets import load_dataset

HUGGINGFACE_API_KEY_KARMANE = os.environ.get("HUGGINGFACE_API_KEY_KARMANE")

dataset = load_dataset(
    "Karmane/nfl-rest-advantage-travel-spot-research",
    token=HUGGINGFACE_API_KEY_KARMANE,
)

print(dataset)
print(dataset[list(dataset.keys())[0]][0])

# getting the DataFrame itself
# df = dataset[list(dataset.keys())[0]].to_pandas()

Karmane. (2025). NFL Rest Advantage and Travel Spot Dataset for Spread Research. Hugging Face. /datasets/Karmane/nfl-rest-advantage-travel-spot-research

Downloads last month: 14

Total file size:

1.11 MB