Player Analytics and Metrics for Interactive Fiction¶

Craft guidance for tracking and analyzing player behavior in IF—choice analytics, engagement metrics, and data-driven narrative improvement.

Why Analytics for IF¶

Unique Opportunities¶

Interactive fiction generates rich behavioral data:

Every choice recorded
Reading pace measurable
Path selection trackable
Replay patterns visible
Abandonment points identifiable

What Analytics Can Answer¶

Question	Data Needed
Which choices are hard?	Decision time, hover patterns
Where do players quit?	Abandonment by passage
Which content is skipped?	Reading time, scroll depth
Do choices feel meaningful?	Replay rates, path diversity
What paths are popular?	Choice distribution
Is pacing working?	Session duration, completion rate

Key Metrics Categories¶

Engagement Metrics¶

Session metrics:

Session duration
Sessions per player
Time between sessions
Completion rate

Progress metrics:

Passages read per session
Branches explored
Replay frequency
Drop-off points

Choice Metrics¶

Distribution:

% selecting each option
Variation across player segments
Comparison to intended distribution

Decision process:

Time spent on choice
Hover/focus patterns
Back-button usage before choice

Outcome perception:

Replay to try other options
Player feedback/reactions
"Would choose differently" surveys

Narrative Quality Indicators¶

Research has identified user-log indicators for automatic IF evaluation:

Indicator	What It Measures
Length	Total content consumed
Duration	Time spent in experience
Diversity	Variety of paths explored
Renewal	Replay behavior
Choice range	Options explored across sessions
Choice frequency	Decisions per time unit
Choice variety	Different choices across replays

Data Collection Approaches¶

Event Logging¶

Track discrete events:

{passage_viewed: "forest_path", timestamp: "2024-01-15T14:32:00"}
{choice_made: "help_stranger", from: "forest_path", timestamp: "2024-01-15T14:33:15"}
{session_end: "natural", last_passage: "village_gate", duration: 847}

State Snapshots¶

Periodic captures of game state:

{checkpoint: "chapter_2_start",
 relationship_sarah: 3,
 inventory: ["key", "letter"],
 flags: ["met_stranger", "helped_merchant"]}

Aggregation Strategies¶

Real-time: Dashboard updates immediately Batch: Process daily/weekly for trends Cohort: Compare player groups over time

Privacy Considerations¶

Anonymization: No personally identifiable data
Consent: Clear data collection disclosure
Minimization: Collect only what's needed
Security: Protect collected data
Deletion: Honor removal requests

Implementation Options¶

For Twine/Web-Based IF¶

Simple approach: Google Analytics events

gtag('event', 'choice_made', {
  'passage': 'forest_path',
  'choice': 'help_stranger'
});

Dedicated tools:

PlayFab-Twine — Free analytics integration
Custom backends — Full control, more work

For Ink/Game Engine IF¶

Unity Analytics:

Built-in event tracking
Dashboard visualizations
Cohort analysis

Third-party:

GameAnalytics
Amplitude
Mixpanel

For ChoiceScript¶

Limited built-in analytics; Choice of Games may share aggregate data with published authors.

Telltale-Style Statistics¶

Showing Choices to Players¶

A notable approach involves showing "Telltale style metrics" to users, where after completing the game, players can see "for this major choice, X% of players made the same choice as you."

Benefits:

Validates player decisions
Creates social experience
Encourages replay
Generates discussion

Implementation:

Aggregate on server
Display at episode/game end
Highlight meaningful choices only

Design Considerations¶

What to show:

Major story choices
Character-defining moments
Surprising distributions

What to hide:

Trivial choices
Spoiler-heavy statistics
Embarrassingly lopsided choices

Using Data to Improve IF¶

Identifying Problems¶

Data Pattern	Possible Problem
High abandonment at X	Pacing, confusion, difficulty
Choice split 99/1	One option seems wrong
Long decision time	Unclear choices
No replays	Choices feel meaningless
Skipped passages	Content too slow

Common Findings¶

From research on narrative games:

Understanding player behaviors helps craft narratives that resonate more deeply, including tracking choice patterns across branching storylines to refine underutilized quests or dialogue branches.

Typical discoveries:

Players skip long exposition
Moral choices split evenly; optimal choices don't
Replays drop sharply after first playthrough
Abandonment clusters at specific points

Iterative Improvement¶

Baseline: Measure before changes
Hypothesize: What might improve metrics?
Change: Implement modification
Measure: Compare to baseline
Learn: Adopt or revert based on data

Player Modeling¶

Beyond Aggregate Data¶

Individual player modeling enables:

Personalized content based on play style
Difficulty adjustment for struggle detection
Recommendation of paths/content
Prediction of likely choices

Research Approaches¶

Academic research has explored:

Drama managers can learn models of player storytelling preferences and automatically recommend narrative experiences predicted to optimize the player's experience.

Techniques:

Collaborative filtering
Personality modeling
Behavior clustering
Preference learning

Practical Applications¶

Model	Application
Pacing preference	Adjust text length dynamically
Risk tolerance	Offer appropriate challenges
Exploration style	Surface or hide optional content
Narrative preference	Emphasize character or plot

Limitations and Caveats¶

What Analytics Can't Tell You¶

Why players made choices
Emotional response to content
Quality of writing
Meaning derived from experience

Metric Pitfalls¶

Goodhart's Law: When a measure becomes a target, it ceases to be a good measure.

Optimizing for completion rate might make IF shorter, not better.

Survivorship bias: Only see data from players who stayed.

High engagement among completers doesn't show why others left.

Correlation vs causation: A precedes B doesn't mean A caused B.

Long passages correlate with abandonment, but cutting them might not help.

Complementary Methods¶

Analytics work best alongside:

Playtesting: Qualitative observation
Surveys: Direct player feedback
Interviews: Deep understanding of experience

Quick Reference¶

Goal	Metric	Tool
Engagement	Session duration, completion	Event logging
Choice balance	Distribution %	Choice tracking
Pacing	Reading time per passage	Timestamp analysis
Problems	Abandonment clusters	Funnel analysis
Replay value	Return sessions, path diversity	Cohort tracking
Player satisfaction	Survey responses	Post-play feedback

Research Basis¶

Key sources on game analytics and IF metrics:

Concept	Source
Game analytics foundations	Magy Seif El-Nasr et al., Game Analytics (2013)
IF user-log indicators	Sali et al., "Measuring the User Experience in Narrative-Rich Games"
Player modeling	Yannakakis & Togelius, Artificial Intelligence and Games (2018)
Drama manager recommendation	Mark Riedl et al., research on automated storytelling

Professor Magy Seif El-Nasr's work establishing game analytics as a field includes "developing evidence-based methodologies to measure game environment effectiveness through novel behavior mining and visual analytics tools."