Sportsbook Player Features

Overview

Sportsbook Player Features are a set of seven Player Features that describe sports betting behaviour across stake size, activity, recency, bonus usage, betting preferences, accumulator complexity, and betting timing.
The features are recalculated daily at 07:00 UTC and can be used for segmentation, movement triggers, reporting, and campaign personalisation in the same way as any other Player Feature.
These classifications help operators better understand Sportsbook customers and create more relevant experiences throughout the player lifecycle.

Shared Concepts

The following rules apply across all Sportsbook Player Features unless otherwise stated.
  1. Time Window - All features analyse Sportsbook activity over a rolling 365-day period.
  2. Eligibility Threshold - Most features require a player to have at least 10 approved bet slips within the rolling time window. Players below this threshold are assigned the class: Insufficient. This prevents classifications from being based on too little activity. Exception: Sportsbook Bet Timing uses a separate requirement of 20 classifiable bet legs.
  3. Data Scope - Only Sortsbook slips meeting all of the following conditions are included:
  1. Status = Approved
  2. Type = Bet
  3. Stake greater than zero
Settlements and non-betting transactions are excluded.
  1. Currency Standardisation - All monetary values are converted into the base currency before calculations are performed. This ensures a fair comparison between players betting in different currencies.

1. Sportsbook Stake Tier

Usage & Value

Sportsbook Stake Tier identifies how much a player typically wagers on an individual bet compared to other players within the same brand.
This feature is particularly valuable for:
  1. Personalising bonus and free-bet values
  2. Identifying VIP or high-value sportsbook players
  3. Excluding very high-stake bettors from mass-market promotions
  4. Creating stake-based segmentation strategies
  5. Understanding player value without relying solely on turnover
Because the classification is relative to the brand's player base, it automatically adapts to different market profiles and betting behaviours.

Product Definition

The feature measures a player's typical bet size using their median sportsbook stake over the last 365 days.
Players are grouped into brand-relative tiers ranging from low-stake recreational bettors to the highest-staking players.
A player classified as a high-stake bettor on one brand may still wager less in absolute terms than a regular bettor on another brand. The classification reflects their position within the brand's ecosystem rather than a fixed monetary amount.
The highest tier, Whale Stake, represents approximately the top 1% of eligible players by typical stake size.
Classification
Condition (Player's median slip amount vs brand percentiles)
Micro Stake
<= P50
Casual Stake
<= P80
Regular Stake
<= P95
High Stake
<= P99
Whale Stake
> P99
Insufficient
fewer than 10 slips in the window
Tier boundaries are calculated using percentile distributions within each brand rather than fixed stake values.
Median bet stake is used instead of average stake to reduce the impact of occasional unusually large bets.

2. Sportsbook Frequency Tier

Usage & Value

Sportsbook Frequency Tier identifies how often a player places bets relative to other players on the same brand.
Typical use cases include:
  1. Communication cadence optimisation
  2. Identifying highly engaged sportsbook customers
  3. Re-engagement targeting for infrequent bettors
  4. Loyalty and retention programmes
  5. Activity-based customer journeys
The feature helps distinguish players who bet regularly from those who participate only occasionally.

Product Definition

The feature measures Sportsbook activity volume by analysing how many approved bet slips a player has placed during the last 365 days.
Players are assigned a relative activity tier based on their position within the brand.
High Frequency players represent the most active segment of Sportsbook users, while Low Frequency players participate less often.
Classification
Condition (slip count vs brand percentiles)
Low Frequency
<= P50
Mid Frequency
<= P90
High Frequency
> P90
Insufficient
fewer than 10 slips in the window
Tier boundaries are calculated using percentile distributions within each brand.

3. Sportsbook Recency Tier

Usage & Value

Sportsbook Recency Tier measures how recently a player engaged with Sportsbook betting.
This feature is useful for:
  1. Lifecycle marketing
  2. Churn prevention
  3. Win-back campaigns
  4. Reactivation journeys
  5. Engagement monitoring
Recency is often one of the strongest indicators of future Sportsbook activity.

Product Definition

The feature classifies players according to the number of days since their most recent approved sportsbook bet.
Players are grouped into lifecycle-oriented states:
  1. Active
  2. Cooling
  3. Dormant
These classifications provide an immediate view of sportsbook engagement status.
Classification
Condition
Active
last bet within 30 days
Cooling
last bet 31-90 days ago
Dormant
last bet more than 90 days ago
Insufficient
fewer than 10 slips in the window
Classification is based on:
Days since last approved sportsbook bet
Unlike stake and frequency tiers, recency uses fixed day thresholds rather than percentile calculations.

4. Sportsbook Bonus Reliance Tier

Usage & Value

Sportsbook Bonus Reliance Tier indicates how dependent a player's betting activity is on promotional funds.
This feature can be used to:
  1. Evaluate promotional efficiency
  2. Identify bonus-driven behaviour
  3. Improve offer allocation
  4. Encourage sustainable real-money activity
  5. Reduce over-incentivisation
It provides visibility into whether sportsbook engagement is primarily funded by bonuses or by player deposits.

Product Definition

The feature measures the proportion of a player's total sportsbook wagering that was funded using bonus money.
Players who mainly use their own deposited funds will fall into lower-reliance categories.
Players who wager heavily using bonus balances are classified into higher-reliance categories.
Classification
Condition (bonus share of total stake)
None
0% (no bonus wagering)
Low
up to 10%
Mid
10-30%
High
more than 30%
Insufficient
fewer than 10 slips in the window
Classification is based on:
SUM(bonus_wager_amount) / SUM(amount)
Fixed thresholds are applied globally rather than using percentile distributions.

5. Sportsbook Betting Style

Usage & Value

Sportsbook Betting Style identifies the type of bets a player prefers.
The feature supports:
  1. Content personalisation
  2. Bet-type-specific promotions
  3. Educational campaigns
  4. Product discovery initiatives
  5. Player preference analysis
Understanding betting preference allows operators to present more relevant offers and experiences.

Product Definition

Players are classified according to the dominant sportsbook bet type they use.
Possible classifications include:
  1. Single
  2. Multi
  3. System
  4. Mixed
Single players typically place straightforward bets on individual outcomes.
Multi players prefer accumulator-style bets with multiple selections.
System players regularly use more advanced combination betting structures.
Mixed players do not exhibit a strong preference for any classified betting style.
Classification
Condition
System
system-bet share is dominant (threshold 20%)
Multi
multi share >= 60%
Single
single share >= 60%
Mixed
no slip type dominates
Insufficient
fewer than 10 slips in the window
The lower required threshold for System betting reflects its naturally lower prevalence.

6. Sportsbook Combo Complexity

Usage & Value

Sportsbook Combo Complexity provides additional insight into players who prefer accumulator betting.
The feature can be used to:
  1. Tailor accumulator promotions
  2. Match offers to risk appetite
  3. Create advanced sportsbook segments
  4. Personalise accumulator-related content
  5. Identify players chasing larger potential payouts
It helps distinguish casual accumulator users from players who consistently build large combinations.

Product Definition

This feature applies only to players classified as Multi bettors.
It measures how many selections are typically included in a player's accumulator bets.
Classifications range from smaller combinations to long accumulators with many selections.
Classification
Condition (median legs per multi slip)
Light Combo
up to 3 legs
Mid Combo
4-6 legs
Long Accumulator
more than 6 legs
Not Applicable
player is not Multi-dominant or has fewer than 10 slips
Classification is based on:
Median number of legs per multi slip
Leg counts are extracted from sportsbook slip payloads.
Eligibility:
  1. Betting Style must be classified as Multi
Thresholds are fixed and based on median accumulator size.
Players who are not Multi bettors are assigned:
Not Applicable

7. Sportsbook Bet Timing

Usage & Value

Sportsbook Bet Timing identifies whether players prefer to bet before events begin or while events are in progress.
This enables:
  1. Better campaign timing
  2. Live betting engagement strategies
  3. In-play promotion targeting
  4. Event-day communications
  5. Tournament-specific activation campaigns
Bet timing preferences can significantly improve the relevance and effectiveness of sportsbook messaging.

Product Definition

The feature analyses whether a player predominantly places bets:
  1. Before matches start (Prematch)
  2. During live events (Live)
  3. Across both contexts (Mixed)
The classification reflects the player's overall betting behaviour rather than individual bets.
Classification
Condition (live share of legs)
Prematch
less than 30% live
Mixed
30-70% live
Live
70% or more live
Insufficient
fewer than 20 classifiable legs in the window
Live Share = Live Legs / (Live Legs + Prematch Legs)
Data is sourced from sportsbook activity leg counts rather than slip-level betting records.
Eligibility:
  1. Minimum 20 classifiable legs
This threshold replaces the standard 10-slip requirement used by the other sportsbook features.
The feature is not percentile-based and uses fixed classification thresholds.