Skip to content

Latest commit

 

History

History
44 lines (38 loc) · 3.49 KB

File metadata and controls

44 lines (38 loc) · 3.49 KB

Data dictionary — churn_dataset.csv (Live betting)

Churn label definition

Term Value
observation_start_date 2025-10-03
observation_end_date 2025-12-31 (model snapshot date)
Outcome window the next 30 calendar days after observation_end
churn = 1 player has no live bets in the outcome window
churn = 0 player has at least one live bet in the outcome window

Columns

Column Type Description
user_id string Unique player ID
registration_date date Account registration date (always ≤ first_live_bet_date if player has bets)
observation_start_date date Start of the observation period
observation_end_date date End of the observation period
last_live_bet_date date Date of the last live bet in the observation period
first_live_bet_date date Date of the first live bet in the observation period
tenure_days int Days from registration to observation_end
live_bets_count int Number of live bets in the observation period
avg_bet_amount float Average stake per bet (EUR); equals total_turnover / live_bets_count
total_turnover float Total live turnover in the observation period
total_payout float Total payout in the observation period
ggr float turnover − payout
ggr_margin float ggr / turnover
deposit_count int Number of deposits in the observation period
total_deposit_amount float Total deposit amount in the observation period
deposit_to_turnover_ratio float total_deposit_amount / total_turnover
days_active_in_observation int Number of days with at least one live bet in the observation period
bet_day_rate float days_active_in_observation / 90
days_since_last_bet int Days from last_live_bet_date to observation_end
live_bets_count_outcome_window int Number of live bets in the outcome window (after observation_end)
churn int Target variable (0/1) — do not use as a feature

Notes

  • Data is aggregated at player level (no individual tickets).
  • Currency: EUR. Product: live betting.
  • Some columns contain missing values (avg_bet_amount, deposits); handle them in your code.
  • Review the date definitions and decide which columns are appropriate for model training.