Scottfree Sports Historical Odds Data — Schema #

Document version: 2026-05-08 Supersedes: schema_20260507.mdCoverage snapshot: 2026-05-08 (the field-coverage table below is a snapshot — see "Live coverage" at the bottom for current per-CSV numbers)

This document describes the columns of the per-sport {sport}_game_scores_1g_*.csv files. Six sports are covered: MLB, NBA, NFL, NHL, NCAAB, NCAAF. Each row is one game.

Changes since schema_20240930 #

• Removed: index (was column 1) — no replacement.
• Retained: time_est from the raw game-score input. Product rows are uniquely identified by (date, time_est, away_team, home_team). This is required for MLB doubleheaders and for reconciling nightly product outputs back to raw game-score inputs.
• Added (all sports), at columns 16–20: open_home_money_line, open_away_money_line, open_home_point_spread, open_away_point_spread, open_over_under — opening lines captured before any in-day movement. See coverage matrix below for sport-specific availability.
• Removed: game_type — the MLB-only field was removed so every sport ships the same customer-facing schema.
• Added: season-to-date ATS record fields for the home and away teams: home_wins_ats, away_wins_ats, home_losses_ats, away_losses_ats, home_ties_ats, away_ties_ats.
• Added: home-team-keyed opponent ATS history fields: opponent_wins_ats, opponent_losses_ats, opponent_ties_ats.

Column count: 146 for every sport.

Field Reference #

Identifiers, scores, and lines (1–16) #

1. season — The league season the game took place in. Format depends on the sport: 2007-08 for cross-year seasons (NBA, NHL, NCAAB), single year for within-year seasons (MLB, NFL, NCAAF).
2. date — Game date in YYYY-MM-DD.
3. time_est — Scheduled game time in Eastern time. Together with date, away_team, and home_team, this uniquely identifies a game.
4. away_team — Team name as reported by the odds feed.
5. away_score — Final points scored by the away team.
6. away_point_spread — Closing point spread offered for the away team (positive = underdog, negative = favorite).
7. away_point_spread_line — Closing odds (American) attached to the away spread bet.
8. away_money_line — Closing moneyline (American) for the away team to win outright.
9. home_team — Team name as reported by the odds feed.
10. home_score — Final points scored by the home team.
11. home_point_spread — Closing point spread offered for the home team.
12. home_point_spread_line — Closing odds (American) attached to the home spread bet.
13. home_money_line — Closing moneyline (American) for the home team to win outright.
14. over_under — Closing over/under (total) line.
15. over_line — Closing odds (American) for the over.
16. under_line — Closing odds (American) for the under.

Opening lines (17–21) — added 2026 #

These columns capture the first observed market price of the day, before any in-day line movement. They support line-movement research. Coverage is non-uniform by sport × market — see the coverage matrix below before relying on these for any specific sport.

17. open_home_money_line — Opening moneyline for the home team (American odds).
18. open_away_money_line — Opening moneyline for the away team (American odds).
19. open_home_point_spread — Opening point spread for the home team.
20. open_away_point_spread — Opening point spread for the away team.
21. open_over_under — Opening over/under (total) line.

NaN convention: when the opener for a given market was not captured for a given game, the value is empty (NaN in pandas, null in JSON). Coverage percentages in the matrix below show the share of rows that have a non-null value.

Date breakdown (22–25) #

22. year — Calendar year extracted from date.
23. month — Calendar month extracted from date.
24. day — Day of month extracted from date.
25. dayofweek — Day of week (e.g. Monday, Tuesday) extracted from date.

Game-level outcomes (26–34) #

26. total_points — Sum of home_score + away_score.
27. point_margin_game — home_score - away_score (positive = home win).
28. won_on_points — 1 if home team won outright, else 0.
29. lost_on_points — 1 if home team lost outright, else 0.
30. cover_margin_game — point_margin_game + home_point_spread. Positive = home covered the spread.
31. won_on_spread — 1 if home team covered the spread.
32. lost_on_spread — 1 if home team failed to cover the spread.
33. overunder_margin — total_points - over_under.
34. over — 1 if total_points > over_under, else 0.
35. under — 1 if total_points < over_under, else 0.

Team season state and streaks (36–101) #

Home and away feature pairs are adjacent in the CSV. Season, streak, and rolling-window features are computed using only games prior to the current row.

36. home_wins — Home team's outright wins in the current season before this game.
37. away_wins — Away team's outright wins in the current season before this game.
38. home_losses — Home team's outright losses in the current season before this game.
39. away_losses — Away team's outright losses in the current season before this game.
40. home_ties — Home team's ties in the current season before this game.
41. away_ties — Away team's ties in the current season before this game.
42. home_days_since_first_game — Days since the home team's first game of the season.
43. away_days_since_first_game — Days since the away team's first game of the season.
44. home_days_since_previous_game — Days since the home team's previous game.
45. away_days_since_previous_game — Days since the away team's previous game.
46. home_point_win_streak — Current outright win streak length for the home team.
47. away_point_win_streak — Current outright win streak length for the away team.
48. home_point_loss_streak — Current outright loss streak length for the home team.
49. away_point_loss_streak — Current outright loss streak length for the away team.
50. home_point_margin_game — Current game point margin from the home team's perspective.
51. away_point_margin_game — Current game point margin from the away team's perspective.
52. home_point_margin_season — Home team's season-to-date point margin before this game.
53. away_point_margin_season — Away team's season-to-date point margin before this game.
54. home_point_margin_season_avg — Home team's average season-to-date point margin.
55. away_point_margin_season_avg — Away team's average season-to-date point margin.
56. home_point_margin_streak — Home team's point margin over its current streak.
57. away_point_margin_streak — Away team's point margin over its current streak.
58. home_point_margin_streak_avg — Home team's average point margin over its current streak.
59. away_point_margin_streak_avg — Away team's average point margin over its current streak.
60. home_point_margin_ngames — Home team's point margin over the rolling N-game window.
61. away_point_margin_ngames — Away team's point margin over the rolling N-game window.
62. home_point_margin_ngames_avg — Home team's average point margin over the rolling N-game window.
63. away_point_margin_ngames_avg — Away team's average point margin over the rolling N-game window.
64. home_cover_win_streak — Current ATS cover streak length for the home team.
65. away_cover_win_streak — Current ATS cover streak length for the away team.
66. home_cover_loss_streak — Current ATS non-cover streak length for the home team.
67. away_cover_loss_streak — Current ATS non-cover streak length for the away team.
68. home_cover_margin_game — Current game cover margin from the home team's perspective.
69. away_cover_margin_game — Current game cover margin from the away team's perspective.
70. home_cover_margin_season — Home team's season-to-date cover margin before this game.
71. away_cover_margin_season — Away team's season-to-date cover margin before this game.
72. home_cover_margin_season_avg — Home team's average season-to-date cover margin.
73. away_cover_margin_season_avg — Away team's average season-to-date cover margin.
74. home_cover_margin_streak — Home team's cover margin over its current ATS streak.
75. away_cover_margin_streak — Away team's cover margin over its current ATS streak.
76. home_cover_margin_streak_avg — Home team's average cover margin over its current ATS streak.
77. away_cover_margin_streak_avg — Away team's average cover margin over its current ATS streak.
78. home_cover_margin_ngames — Home team's cover margin over the rolling N-game window.
79. away_cover_margin_ngames — Away team's cover margin over the rolling N-game window.
80. home_cover_margin_ngames_avg — Home team's average cover margin over the rolling N-game window.
81. away_cover_margin_ngames_avg — Away team's average cover margin over the rolling N-game window.
82. home_total_points — Points scored by the home team in this game.
83. away_total_points — Points scored by the away team in this game.
84. home_overunder_margin — Current game total margin from the home team's perspective.
85. away_overunder_margin — Current game total margin from the away team's perspective.
86. home_over_streak — Consecutive games where the home team's games went over.
87. away_over_streak — Consecutive games where the away team's games went over.
88. home_under_streak — Consecutive games where the home team's games went under.
89. away_under_streak — Consecutive games where the away team's games went under.
90. home_overunder_season — Home team's season-to-date over/under margin.
91. away_overunder_season — Away team's season-to-date over/under margin.
92. home_overunder_season_avg — Home team's average season-to-date over/under margin.
93. away_overunder_season_avg — Away team's average season-to-date over/under margin.
94. home_overunder_streak — Home team's over/under margin over its current total streak.
95. away_overunder_streak — Away team's over/under margin over its current total streak.
96. home_overunder_streak_avg — Home team's average over/under margin over its current total streak.
97. away_overunder_streak_avg — Away team's average over/under margin over its current total streak.
98. home_overunder_ngames — Home team's over/under margin over the rolling N-game window.
99. away_overunder_ngames — Away team's over/under margin over the rolling N-game window.
100. home_overunder_ngames_avg — Home team's average over/under margin over the rolling N-game window.
101. away_overunder_ngames_avg — Away team's average over/under margin over the rolling N-game window.

ATS record features (102–107) #

These are season-to-date Against The Spread records for each team prior to the current game.

102. home_wins_ats — Home team's ATS wins in the current season before this game.
103. away_wins_ats — Away team's ATS wins in the current season before this game.
104. home_losses_ats — Home team's ATS losses in the current season before this game.
105. away_losses_ats — Away team's ATS losses in the current season before this game.
106. home_ties_ats — Home team's ATS pushes in the current season before this game.
107. away_ties_ats — Away team's ATS pushes in the current season before this game.

Home-vs-away deltas (108–140) #

Each delta_* field is home_* - away_* for the corresponding home/away pair, useful as a single matchup-strength feature.

108. delta_wins
109. delta_losses
110. delta_ties
111. delta_days_since_first_game
112. delta_days_since_previous_game
113. delta_point_win_streak
114. delta_point_loss_streak
115. delta_point_margin_game
116. delta_point_margin_season
117. delta_point_margin_season_avg
118. delta_point_margin_streak
119. delta_point_margin_streak_avg
120. delta_point_margin_ngames
121. delta_point_margin_ngames_avg
122. delta_cover_win_streak
123. delta_cover_loss_streak
124. delta_cover_margin_game
125. delta_cover_margin_season
126. delta_cover_margin_season_avg
127. delta_cover_margin_streak
128. delta_cover_margin_streak_avg
129. delta_cover_margin_ngames
130. delta_cover_margin_ngames_avg
131. delta_total_points
132. delta_overunder_margin
133. delta_over_streak
134. delta_under_streak
135. delta_overunder_season
136. delta_overunder_season_avg
137. delta_overunder_streak
138. delta_overunder_streak_avg
139. delta_overunder_ngames
140. delta_overunder_ngames_avg

Opponent ATS history (141–143) #

These fields are keyed off the home team and the current opponent across the whole dataset, prior to the current row.

141. opponent_wins_ats — Home team's historical ATS wins against this away opponent before this game.
142. opponent_losses_ats — Home team's historical ATS losses against this away opponent before this game.
143. opponent_ties_ats — Home team's historical ATS pushes against this away opponent before this game.

Daily-mean lag features (144–146) #

These columns are intentionally lagged by one day to prevent target leakage when training models on a per-row basis.

144. won_on_points_daily_mean_lag1 — Mean of won_on_points across all games on the previous calendar day.
145. won_on_spread_daily_mean_lag1 — Mean of won_on_spread across all games on the previous calendar day.
146. over_daily_mean_lag1 — Mean of over across all games on the previous calendar day.

Opener-Coverage Matrix #

The five opening-line columns are sourced from a one-time historical archive (where available per sport × market) plus forward capture starting 2026-04-18. Coverage is uneven across sports because no single free or commercial archive covers all six leagues uniformly. The table below is a snapshot from 2026-05-08 — see "Live coverage" below for current per-CSV numbers.

How to read this:

• High coverage (≥85%): usable as a primary feature.
• Medium coverage (50–85%): usable, but the model must be tolerant of NaN. NCAAB / NCAAF point-spread + over/under fall here — the archive covers most pre-2026 rows but not all conferences/games.
• Low coverage (<5%): forward-only, captured starting 2026-04-18. NBA / NCAAB / NCAAF moneyline openers fall here. Treat as a forward-going-feature only; do not use for training on pre-2026 history.

Why coverage varies:

• MLB / NHL / NFL: a long-running historical odds archive covers moneyline + total back to ~2008.
• NBA: archive covers spread + total back to 2006; moneyline was never archived publicly, so we capture it forward only.
• NCAAB / NCAAF: no public archive of money lines for college sports. Spread + over/under were back-filled from a historical-archive vendor; moneyline starts at 2026-04-18.

Live coverage #

Each refresh of the product writes a coverage sidecar JSON next to the CSV. Customers can use that sidecar to verify current numbers before each refresh.

{sport}_data_coverage_latest.json

Schema:

{
  "league": "MLB",
  "row_count": 28566,
  "first_game_date": "2014-03-22",
  "last_game_date": "2026-05-08",
  "openers": {
    "open_home_money_line": {
      "non_null": 27879,
      "coverage_pct": 97.59,
      "first_non_null_date": "2014-03-22"
    },
    ...
  },
  "generated_at": "2026-05-08T22:14:38+00:00"
}

Use the generated_at timestamp to confirm the coverage matches the CSV you downloaded — the two are written atomically per refresh.

Refresh policy #

• Subscribers to Scottfree Sports Data ($19/mo) can refresh up to 8 times per UTC calendar month via the /api/v1/scottfree-sports-data/refresh endpoint, returning a short-lived signed download URL.
• The underlying CSVs are regenerated daily at 12:15 PM ET by the pipeline-product-daily job, after the morning odds refresh and repair window. A refresh fetched after that job completes reflects all finalized games through the previous day.

Contact #

Questions: scottfree.analytics@scottfreellc.com. Documentation: https://scottfreellc.github.io/alphapy-sports