Generative AI in Fantasy Sports
Generative AI in Fantasy Sports
Generative AI in Fantasy Sports
In a 5-person interdisciplinary team, we worked with FanDuel to conduct user research and determine how to integrate generative AI into their infrastructure and improve the fantasy sports user experience.
In a 5-person interdisciplinary team, we worked with FanDuel to conduct user research and determine how to integrate generative AI into their infrastructure and improve the fantasy sports user experience.
Background
Context
FanDuel, a prominent player in the fantasy sports arena, has a substantial user base of 12 million registered users. The growing adoption of AI in fantasy sports, leveraging techniques like pattern recognition and data analytics, has significantly enhanced the overall user experience. However, there are many existing issues. Users will often go to third-party sources to research, increasing their workload. There is also an absence of generative AI for drafting within existing apps and a lack of transparency on how Fantasy sports apps operate.
FanDuel, partnered with our team, was exploring the use of generative AI to enhance the Fantasy app user experience. Our team's task was to conduct user research, focusing on how fantasy players use AI to improve their experience and integrate it into the app's infrastructure.
Integrating generative AI into FanDuel's services would streamline the user experience, setting them apart from competitors. This AI implementation can enhance operational efficiency and user engagement, acting as virtual assistants or chatbots to assist users in team lineup creation, making FanDuel more user-friendly and technologically advanced.
Problem Statement
How might we utilize generative AI to improve the user experience and performance on the FanDuel Fantasy Sports app?
Who were our target users?
Fantasy Sports players
Novice Users
Professional Users
Our goals were to study interaction patterns among fantasy sports app users, identify prioritized features, and assess the potential of Generative AI in customizing the app for an improved drafting process.
My Role
Contributed to background research and planning
Conducted 3 user interviews
Analyzed results from respondents to the Qualtrics surveys
Created bar charts and the heat map based on open coding
Created some of the initial concept sketches
Assisted with information architecture and 2 user feedback sessions
Worked on creating the final design and prototype
Research Methods
Semi-structured Interviews
We interviewed 5 current fantasy sports players.
Goals
Understand app interactions, identify valued features, and uncover motivations, attitudes, and behaviors
Interview Structure
4 sections focused on different parts of the user experience: General Information, Recommendation System, Drafting, and Generative AI
ANALYSIS
Open Coding
Interviews conducted by a minimum of two researchers
Dedoose software for open coding interview transcripts
Heat Map
From bar charts tracking code occurrences
Clustered parent codes into groups and used color coding for visual accessibility
Background
Context
FanDuel, a prominent player in the fantasy sports arena, has a substantial user base of 12 million registered users. The growing adoption of AI in fantasy sports, leveraging techniques like pattern recognition and data analytics, has significantly enhanced the overall user experience. However, there are many existing issues. Users will often go to third-party sources to research, increasing their workload. There is also an absence of generative AI for drafting within existing apps and a lack of transparency on how Fantasy sports apps operate.
FanDuel, partnered with our team, was exploring the use of generative AI to enhance the Fantasy app user experience. Our team's task was to conduct user research, focusing on how fantasy players use AI to improve their experience and integrate it into the app's infrastructure. Why is this significant for FanDuel?
Integrating generative AI into FanDuel's services would streamline the user experience, setting them apart from competitors. This AI implementation can enhance operational efficiency and user engagement, acting as virtual assistants or chatbots to assist users in team lineup creation, making FanDuel more user-friendly and technologically advanced.
Problem Statement
How might we utilize generative AI to improve the user experience and performance on the FanDuel Fantasy Sports app?
Who were our target users?
Fantasy Sports players
Novice Users
Professional Users
Our goals were to study interaction patterns among fantasy sports app users, identify prioritized features, and assess the potential of Generative AI in customizing the app for an improved drafting process.
My Role
Contributed to background research and planning
Conducted 3 user interviews
Analyzed results from respondents to the Qualtrics surveys
Created bar charts and the heat map based on open coding
Created some of the initial concept sketches
Assisted with information architecture and 2 user feedback sessions
Worked on creating the final design and prototype
Research Methods
Semi-structured Interviews
We interviewed 5 current fantasy sports players.
Goals
Understand app interactions, identify valued features, and uncover motivations, attitudes, and behaviors
Interview Structure
4 sections focused on different parts of the user experience: General Information, Recommendation System, Drafting, and Generative AI
ANALYSIS
Open Coding
Interviews conducted by a minimum of two researchers
Dedoose software for open coding interview transcripts
Heat Map
From bar charts tracking code occurrences
Clustered parent codes into groups and used color coding for visual accessibility
Background
Context
FanDuel, a prominent player in the fantasy sports arena, has a substantial user base of 12 million registered users. The growing adoption of AI in fantasy sports, leveraging techniques like pattern recognition and data analytics, has significantly enhanced the overall user experience. However, there are many existing issues. Users will often go to third-party sources to research, increasing their workload. There is also an absence of generative AI for drafting within existing apps and a lack of transparency on how Fantasy sports apps operate.
FanDuel, partnered with our team, was exploring the use of generative AI to enhance the Fantasy app user experience. Our team's task was to conduct user research, focusing on how fantasy players use AI to improve their experience and integrate it into the app's infrastructure. Why is this significant for FanDuel?
Integrating generative AI into FanDuel's services would streamline the user experience, setting them apart from competitors. This AI implementation can enhance operational efficiency and user engagement, acting as virtual assistants or chatbots to assist users in team lineup creation, making FanDuel more user-friendly and technologically advanced.
Problem Statement
How might we utilize generative AI to improve the user experience and performance on the FanDuel Fantasy Sports app?
Who were our target users?
Fantasy Sports players
Novice Users
Professional Users
Our goals were to study interaction patterns among fantasy sports app users, identify prioritized features, and assess the potential of Generative AI in customizing the app for an improved drafting process.
My Role
Contributed to background research and planning
Conducted 3 user interviews
Analyzed results from respondents to the Qualtrics surveys
Created bar charts and the heat map based on open coding
Created some of the initial concept sketches
Assisted with information architecture and 2 user feedback sessions
Worked on creating the final design and prototype
Research Methods
Semi-structured Interviews
We interviewed 5 current fantasy sports players.
Goals
Understand app interactions, identify valued features, and uncover motivations, attitudes, and behaviors
Interview Structure
4 sections focused on different parts of the user experience: General Information, Recommendation System, Drafting, and Generative AI
ANALYSIS
Open Coding
Interviews conducted by a minimum of two researchers
Dedoose software for open coding interview transcripts
Heat Map
From bar charts tracking code occurrences
Clustered parent codes into groups and used color coding for visual accessibility
KEY FINDINGS FROM CODES
OPEN CODING HEAT MAP
Surveys
We conducted 2 Qualtrics surveys sent out to students across campus and through different group chats.
Fantasy Sports Survey
Goals
Understand interactions of fantasy sports users and identify valued features
Explore the potential of Generative AI for customizing the app and enhancing the drafting process
Structure - 3 Sections
General Information, Generative AI for Customization & Generative AI for Drafting
Survey Demographics
31 Responses
All male
Generative AI Survey
Goals
Examine user interactions with Generative AI and preferences
Assess reliance and trust in AI-generated information
Structure (2 separate paths)
Participants who have not used Generative AI
Participants who have used Generative AI
Survey Demographics
22 Responses
Roughly 50% females, 30% males, 2 non-binary, 1 prefer not to say, and 1 other
Surveys
We conducted 2 Qualtrics surveys sent out to students across campus and through different group chats.
Fantasy Sports Survey
Goals
Understand interactions of fantasy sports users and identify valued features
Explore the potential of Generative AI for customizing the app and enhancing the drafting process
Structure - 3 Sections
General Information, Generative AI for Customization & Generative AI for Drafting
Survey Demographics
31 Responses
All male
Generative AI Survey
Goals
Examine user interactions with Generative AI and preferences
Assess reliance and trust in AI-generated information
Structure (2 separate paths)
Participants who have not used Generative AI
Participants who have used Generative AI
Survey Demographics
22 Responses
Roughly 50% females, 30% males, 2 non-binary, 1 prefer not to say, and 1 other
Surveys
We conducted 2 Qualtrics surveys sent out to students across campus and through different group chats.
Fantasy Sports Survey
Goals
Understand interactions of fantasy sports users and identify valued features
Explore the potential of Generative AI for customizing the app and enhancing the drafting process
Structure - 3 Sections
General Information, Generative AI for Customization & Generative AI for Drafting
Survey Demographics
31 Responses
All male
Generative AI Survey
Goals
Examine user interactions with Generative AI and preferences
Assess reliance and trust in AI-generated information
Structure (2 separate paths)
Participants who have not used Generative AI
Participants who have used Generative AI
Survey Demographics
22 Responses
Roughly 50% females, 30% males, 2 non-binary, 1 prefer not to say, and 1 other
ANALYSIS
ANALYSIS
FINDINGS FROM FANTASY SPORTS SURVEY
FINDINGS FROM FANTASY SPORTS SURVEY
FINDINGS FROM GENERATIVE AI SURVEY
FINDINGS FROM GENERATIVE AI SURVEY
Website Breakdown
Goals
Systematically analyze the Fantasy Sports League app's UI and functionality
Identify areas of improvement to align the app with user expectations
Overview
3 main steps: creating a contest, drafting lineups, and analyzing results during game play
ANALYSIS
Conducted a game night contest among 3 members of our team to experience DFS gameplay
Documented screenshots of each step for evaluation purposes
Website Breakdown
Goals
Systematically analyze the Fantasy Sports League app's UI and functionality
Identify areas of improvement to align the app with user expectations
Overview
3 main steps: creating a contest, drafting lineups, and analyzing results during game time
ANALYSIS
Conducted a game night contest among 3 members of our team to experience DFS gameplay
Documented screenshots of each step for evaluation purposes
Website Breakdown
Goals
Systematically analyze the Fantasy Sports League app's UI and functionality
Identify areas of improvement to align the app with user expectations
Overview
3 main steps: creating a contest, drafting lineups, and analyzing results during game play
ANALYSIS
Conducted a game night contest among 3 members of our team to experience DFS gameplay
Documented screenshots of each step for evaluation purposes
POSSIBLE DESIGN IDEAS
Task Analysis
Goal
Understand where pain points are when using the FanDuel app
Overview
Followed each step of participating in a FanDuel app competition, documented steps, and visualized them with a flow chart diagram
ANALYSIS
The process involved 13 main steps
Selecting a player was a sub-process comprising 5 steps within the main steps
Task Analysis
Goal
Understand where pain points are when using the FanDuel app
Overview
Followed each step of participating in a FanDuel app competition, documented steps, and visualized them with a flow chart diagram
ANALYSIS
The process involved 13 main steps
Selecting a player was a sub-process comprising 5 steps within the main steps
Task Analysis
Goal
Understand where pain points are when using the FanDuel app
Overview
Followed each step of participating in a FanDuel app competition, documented steps, and visualized them with a flow chart diagram
ANALYSIS
The process involved 13 main steps
Selecting a player was a sub-process comprising 5 steps within the main steps
TASK ANALYSIS
Ideation & Design
Design Implications
Ideation & Design
Design Implications
Ideation & Design
Design Implications
Fantasy Sports User Persona
Ideation Process
Our design implications helped guide our ideation, concept sketches, and eventually our final design prototype.
Ideation Process
Our design implications helped guide our ideation, concept sketches, and eventually our final design prototype.
Ideation Process
Our design implications helped guide our ideation, concept sketches, and eventually our final design prototype.
Initial Design Concept Sketches
Sketches Feedback Session
We conducted 2 feedback sessions, each with 2 users who were fantasy sports players. We walked users through each of our concept sketches and asked specific questions:
How well do these concepts align with your user needs?
What modifications or improvements would you suggest?
Are there any accessibility issues you foresee?
Which features do you find most useful or relevant?
KEY FINDINGS
Users highly value customizable options and flexibility in a given system
Users prefer when information is shown in different ways
Users prefer features that do not interfere during game time
Users enjoy when a feature contains something fun and engaging, outside of just statistics
Users preferred when UI was kept simple and intuitive
Users expressed wanting accessible visual scaling to ensure an inclusive design
We extracted specific design requirements from their suggestions to identify what to prioritize in our next iteration of creating wireframes. From their feedback, we selected our final concept.
Sketches Feedback Session
We conducted 2 feedback sessions, each with 2 users who were fantasy sports players. We walked users through each of our concept sketches and asked specific questions:
How well do these concepts align with your user needs?
What modifications or improvements would you suggest?
Are there any accessibility issues you foresee?
Which features do you find most useful or relevant?
KEY FINDINGS
Users highly value customizable options and flexibility in a given system
Users prefer when information is shown in different ways
Users prefer features that do not interfere during game time
Users enjoy when a feature contains something fun and engaging, outside of just statistics
Users preferred when UI was kept simple and intuitive
Users expressed wanting accessible visual scaling to ensure an inclusive design
We extracted specific design requirements from their suggestions to identify what to prioritize in our next iteration of creating wireframes. From their feedback, we selected our final concept.
Sketches Feedback Session
We conducted 2 feedback sessions, each with 2 users who were fantasy sports players. We walked users through each of our concept sketches and asked specific questions:
How well do these concepts align with your user needs?
What modifications or improvements would you suggest?
Are there any accessibility issues you foresee?
Which features do you find most useful or relevant?
KEY FINDINGS
Users highly value customizable options and flexibility in a given system
Users prefer when information is shown in different ways
Users prefer features that do not interfere during game time
Users enjoy when a feature contains something fun and engaging, outside of just statistics
Users preferred when UI was kept simple and intuitive
Users expressed wanting accessible visual scaling to ensure an inclusive design
We extracted specific design requirements from their suggestions to identify what to prioritize in our next iteration of creating wireframes. From their feedback, we selected our final concept.
"Highlights Reel" Design
INFORMATION ARCHITECTURE
WIREFRAMES
Main Features
Pop-up after contest for user to share
Editing video reels
Creating a highlight reel from their competition history
WIREFRAMES
Main Features
Pop-up after contest for user to share
Editing video reels
Creating a highlight reel from their competition history
WIREFRAMES
Main Features
Pop-up after contest for user to share
Editing video reels
Creating a highlight reel from their competition history
Wireframes Feedback Session
We conducted a feedback session with nine FanDuel employees. We asked them questions framed to understand the overall usability and functions of specific elements of our design.
What feedback can you provide on the "Highlight Reel" design?
What type of content or AI observations would you prefer in the "Highlights Reel?”
Are there any accessibility issues within this design?
Based on the current design, how likely would you use this feature?
Prototype Design Requirements
Accessibility, customizable options, providing encouragement and feedback, and creating an infographic-like summary report and videos are the most important to users based on our findings.
Wireframes Feedback Session
We conducted a feedback session with nine FanDuel employees. We asked them questions framed to understand the overall usability and functions of specific elements of our design.
What feedback can you provide on the "Highlight Reel" design?
What type of content or AI observations would you prefer in the "Highlights Reel?”
Are there any accessibility issues within this design?
Based on the current design, how likely would you use this feature?
Prototype Design Requirements
Accessibility, customizable options, providing encouragement and feedback, and creating an infographic-like summary report and videos are the most important to users based on our findings.
Wireframes Feedback Session
We conducted a feedback session with nine FanDuel employees. We asked them questions framed to understand the overall usability and functions of specific elements of our design.
What feedback can you provide on the "Highlight Reel" design?
What type of content or AI observations would you prefer in the "Highlights Reel?”
Are there any accessibility issues within this design?
Based on the current design, how likely would you use this feature?
Prototype Design Requirements
Accessibility, customizable options, providing encouragement and feedback, and creating an infographic-like summary report and videos are the most important to users based on our findings.
Design Walkthrough
DESIGN SYSTEM
Our design system was based on FanDuel's existing UI components, typography, icons and colors.
DESIGN SYSTEM
Our design system was based on FanDuel's existing UI components, typography, icons and colors.
DESIGN SYSTEM
Our design system was based on FanDuel's existing UI components, typography, icons and colors.
"HIGHLIGHTS REEL" FEATURES
The Highlights page is a summary of all the highlights of the user’s most recent game in the form of cards with the appropriate graphics, videos, and stats. It also provides AI suggestions for improvement to help users strategize better for their next fantasy game.
The highlights can be shared either as a report or a video based on user preference. The user can share these on any social platform or choose to save them on their device.
"HIGHLIGHTS REEL" FEATURES
The Highlights page is a summary of all the highlights of the user’s most recent game in the form of cards with the appropriate graphics, videos, and stats. It also provides AI suggestions for improvement to help users strategize better for their next fantasy game.
The highlights can be shared either as a report or a video based on user preference. The user can share these on any social platform or choose to save them on their device.
"HIGHLIGHTS REEL" FEATURES
The Highlights page is a summary of all the highlights of the user’s most recent game in the form of cards with the appropriate graphics, videos, and stats. It also provides AI suggestions for improvement to help users strategize better for their next fantasy game.
The highlights can be shared either as a report or a video based on user preference. The user can share these on any social platform or choose to save them on their device.
Users can edit the visual style and the content of the highlight video. The user has the option to decide the style from the given options, which changes seasonally, and can also use an AI prompt to describe how they want the video to be generated.
This feature provides a way for the user to get a summary of highlights from their past games. The user can select games from the history tab and create their highlights compilation.
Users can edit the visual style and the content of the highlight video. The user has the option to decide the style from the given options, which changes seasonally, and can also use an AI prompt to describe how they want the video to be generated.
This feature provides a way for the user to get a summary of highlights from their past games. The user can select games from the history tab and create their highlights compilation.
Users can edit the visual style and the content of the highlight video. The user has the option to decide the style from the given options, which changes seasonally, and can also use an AI prompt to describe how they want the video to be generated.
This feature provides a way for the user to get a summary of highlights from their past games. The user can select games from the history tab and create their highlights compilation.
Evaluation Methods
Heuristic Evaluation
UX experts were asked to evaluate our prototype based on 8 metrics each on a 5-point scale:
Visibility of System Status
Match between the system & the real world
User control and reversibility
Consistency and navigational clarity
Error prevention
Memory Recognition Ease
Flexibility and Efficiency of Use
Simplicity and information architecture
Evaluation Methods
Heuristic Evaluation
UX experts were asked to evaluate our prototype based on 8 metrics each on a 5-point scale:
Visibility of System Status
Match between the system & the real world
User control and reversibility
Consistency and navigational clarity
Error prevention
Memory Recognition Ease
Flexibility and Efficiency of Use
Simplicity and information architecture
Evaluation Methods
Heuristic Evaluation
UX experts were asked to evaluate our prototype based on 8 metrics each on a 5-point scale:
Visibility of System Status
Match between the system & the real world
User control and reversibility
Consistency and navigational clarity
Error prevention
Memory Recognition Ease
Flexibility and Efficiency of Use
Simplicity and information architecture
SCORES
User Testing
Users were asked about immediate impressions of the prototype's functions, features, and design
Users were asked to complete 2 tasks
Create a highlight and share it with your group
Analyze your most recent game and tell us how you would change your lineup
Users were asked about features they liked and disliked
Users completed 2 questionnaires to gauge their experience
SUS Questionnaire
Word Choice Questionnaire
Findings
Provide stats and rationales that back up the insight (for more competitive users)
Make tiles clickable to view more details
Focus even more on competition and less on sharing
Provide practical suggestions for the upcoming game
Redesign information architecture and improve filtering features
Modify the report generation process to be more sequential
User Testing
Users were asked about immediate impressions of the prototype's functions, features, and design
Users were asked to complete 2 tasks
Create a highlight and share it with your group
Analyze your most recent game and explain how you would change your lineup
Users were asked about features they liked and disliked
Users completed 2 questionnaires to gauge their experience
SUS Questionnaire
Word Choice Questionnaire
Findings
Provide stats and rationales that back up the insight (for more competitive users)
Make tiles clickable to view more details
Focus even more on competition and less on sharing
Provide practical suggestions for the upcoming game
Redesign information architecture and improve filtering features
Modify the report generation process to be more sequential
User Testing
Users were asked about immediate impressions of the prototype's functions, features, and design
Users were asked to complete 2 tasks
Create a highlight and share it with your group
Analyze your most recent game and tell us how you would change your lineup
Users were asked about features they liked and disliked
Users completed 2 questionnaires to gauge their experience
SUS Questionnaire
Word Choice Questionnaire
Findings
Provide stats and rationales that back up the insight (for more competitive users)
Make tiles clickable to view more details
Focus even more on competition and less on sharing
Provide practical suggestions for the upcoming game
Redesign information architecture and improve filtering features
Modify the report generation process to be more sequential
Prototype
Prototype
Prototype
Highlights Summary Page
HI-FI PROTOTYPE
Highlights Summary Page
Highlights Summary Page
Share Report
Share Video Reel
6023 Research Methods for HCI
Georgia Institute of Technology | Fall 2023
Professor Carrie Bruce
6023 Research Methods for HCI
Georgia Institute of Technology | Fall 2023
Professor Carrie Bruce
6023 Research Methods for HCI
Georgia Institute of Technology | Fall 2023
Professor Carrie Bruce