Rithmm Player Prop Graphs
Transforming Raw Player Data to Increase In-App Engagement
Role:
Industry:
Tools:
Sole UX/UI designer
Sports betting tech / sports analytics
Figma, FigJam
TLDR;
The Product
Rithmm is a mobile app giving users the ability to personalize sports betting models.
Users
Dave — a beginner bettor who just wants quick, visual proof that a pick is backed by data.
Alex — advanced bettor who wants full transparency into past results and trends to validate (or challenge) Rithmm's predictions.
Scroll to Section
The Problem
Before this feature, Rithmm's pre-game view only displayed the pick itself with no supporting stats or trends. Users had no way to validate the recommendation, which led to lower engagement and trust. Many were leaving the app to research player performance elsewhere.
Goal
Create a visual in the pre-game view for player props that provides:
Clear, visual representation of past performance
Enough context for beginners to build confidence while still offering deep data for advanced users
A seamless, mobile-friendly experience that doesn't overwhelm users
Scroll to Section
Research
Analyzed competitors (DraftKings, Action, Outlier, Pikkit, Linemate) to study how they use graphs to present player data
Worked with CTO, backend, and frontend engineers to define available data and technical feasibility
Identified key data points: past performance, hit rate, average stats, timeframe toggles, and vertical list results
Outlined three main design challenges: tight mobile space, clarity in data presentation, fast design turnaround (1 week)
Scroll to Section
Proposed Solutions
Designed a bar graph with color-coded results (green = hit, red = miss) and a current line overlay to compare past performance
Added a timeframe toggle (last 5, 10, or 20 games) and a home/away filter for deeper analysis
Created a vertical list of past game results, including an up/down arrow indicator for quick scanning
Conducted graph library research and frontend testing to ensure smooth performance without impacting load times
Scroll to Section
Final Design & Outcome
The player props page became the most visited page in the app, surpassing the game predictions page
Received overwhelmingly positive user feedback — users felt more confident in their picks
Created a sticky experience, reducing the need for users to leave Rithmm for external research
Successfully balanced engagement, trust, and performance, making player props a core feature, not just an add-on
Scroll to Section
Introduction;
This project focuses on increasing user engagement and rebuilding trust in our player prop predictions by giving users quick, visual access to historical performance data — helping them validate picks at a glance. This fast-tracked feature was driven by myself, stakeholder feedback, and user frustration, created under a tight 2-week timeline with a strong focus on ensuring performance wasn't impacted after recent technical issues.
Rithmm is a mobile app giving users the ability to personalize sports betting models.
A sports betting model utilizes predictive analytics and historical data to predict the outcome of upcoming games and player props. You input what you think is important to the outcome of a game and from there, Rithmm's algorithm builds a unique model. This model is able to tell you what bet to take based on win probability and risk to reward ratio. All of this makes for a smarter, better bettor.
"The pick could be coming from a coin flip for all we know…"
Persona 1: Dow Jones Dave
Age: 26
Experience Level: Beginner to intermediate sports bettor
Sports Interest: Football and basketball
Tech Comfort: Moderate - loves DraftKings & FanDuel
Primary Goal: Get reliable picks that make money — fast
Dave wants to bet smarter but doesn't want to spend hours researching. He's interested in following picks from smart sources, not building his own models. He treats sports betting like an investment opportunity — looking for a sports stock tip.
Pain Points:
Lacks meaningful data leaving him feeling he has to blindly trust the pick
Doesn't want to spend time digging through other platforms
Seeing just the pick with no other context makes him wonder "Why this pick? Why should I trust it?", making him hesitant to fully trust it
Needs from Rithmm:
Trust at a glance — Dave wants quick, confidence-boosting data that makes him feel like the pick is credible without getting lost in the weeds
Simple summaries
Clear trends and records
Minimal jargon
From player prop graphs, Dave wants…
A simple visual story
Clear indicators so he can understand success/failure without overthinking
Embedded into the current flow
Ability to dig deeper
Persona 2: Analytical Alex
Age: 34
Experience Level: Advanced/Data-Driven Bettor
Sports Interest: All major sports, but especially ones with rich data
Tech Comfort: High - familiar with spreadsheets, Python, and advanced betting tools
Primary Goal: Use Rithmm as either a modeling playground or an extra data source
Alex loves digging into the numbers and building his own predictive models. He treats sports betting like a statistical puzzle. He wants Rithmm to be a transparent data partner, not a black box.
Pain Points:
Frustrated that the pre-game view only showed the pick with no supporting data
Lacking meaningful data leaving him having to blindly trust the pick or leave the app
The lack of granular data per game made it hard to spot patterns or trends, limiting his ability to fully trust or challenge Rithmm's pick and prediction
Needs from Rithmm:
Full transparency
Historical team and player performance data
Ability to spot trends and patterns on his own
Data should feel explorable, not static
From player prop graphs, Alex wants…
Full transparency — he wants to see 5-10 games' results
A way to spot trends quickly
He wants to understand if the player is trending up or down or if there's any clear pattern of variance
The Problem
Our player prop predictions lacked supporting data, leaving users feeling like they had to blindly trust the picks. This lack of transparency reduced confidence and trust, ultimately contributing to lower engagement with the player props feature, even though player props are a high-interest area for users. Without clear trends or past performance data, users had no way to validate the model's accuracy, making them less likely to engage and place bets.
Hypothesis
By adding player prop graphs showing historical performance, past lines, and visual trends directly into the pre-game view, we can:
Increase trust in the player prop picks by providing trend data
Drive higher engagement on the player props page by giving users a richer, more interactive experience
Create an entry point for future features (like player prop tracking) by making the prop data feel more reliable and useful
To kick of this project, I looked through some key players in the space including ESPN, DraftKings, Action Network, Outlier, Pikkit, and Linemate — focusing on how they present player performance data visually. I paid particular attention to graph usage, exploring how they balanced depth and clarity to avoid overwhelming users while still providing meaningful context. I also researched what graph libraries they were using, both for design inspiration and to understand what was technically achievable within our constraints.
Internally, I collaborated with the CTO, backend, and frontend engineers to outline what data was available for player props and how that data could flow from backend to frontend. I also worked to define the key data points that would meet the needs of both Dave (who wants quick, confidence-building visuals) and Alex (who wants full transparency and detailed context).
The result was clear of what the graphs and supporting data needed to display:
Timeframe toggles (last 5, 10, 20, season games) to give users control over the view
A bar graph showing past performance, with each bar color-coded (green = hit, red = miss)
A current line overlay to show how today's line compares to past performance
Averages for over %, under %, and average stat value
A vertical list view showing the line and result for each past game — giving users a fast way to scan detailed data
This helped me identify three core design challenges:
1 - Timeframe: with only 1 week allocated for design, speed was critical — requiring quick decision-making and close alignment with engineering
2 - Clarity: the graph and supporting data needed to be immediately understandable, especially for newer users who might not be familiar with interpreting sports data
3 - Limited mobile space: all of this content had to fit within the confines of a mobile action sheet, meaning clear hierarchy and scannability were essential
Proposed Solution;
From Sketch to Lo-Fi
With the data requirements and design challenges in mind, I started with quick sketches to explore layout ideas, testing ways to stack the graph, average stats, and list view into a small space without overwhelming the user. These sketches evolved into low-fidelity mockups, adjusted slightly based on subscriber tier (free vs core vs premium) to have upsell opportunities and keep in mind the personas for each tier (Dave's are usually core subscribers and Alex's are usually premium subscribers).
I also conducted graph library research to find a library that could handle our requirements (mobile-friendly, customizable, performant) and worked closely with the frontend developer to test integration into the codebase before moving into high-fidelity design.
Technical Collaboration
Once the initial lo-fi designs were ready, I met with the CTO, backend, and frontend engineers to review the design and confirm technical feasibility. We aligned on:
What data the backend needed to deliver
How that data would be structured
How the frontend would integrate the graph library I selected
Final Design;
A key revision made after testing was the introduction of a brief overview step immediately after collecting the user's email, password, and phone number. This was made to set clear expectations for what's coming next — addressing feedback from both users. The small adjustment helped make the process feel more transparent and less intimidating, especially for users who are less tech-savvy.
Key features in the final design:
Dynamic bar graph — displays player performance over the last 5, 10, or 20 games, with bars color-coded (green = hit, red = miss) to make success or failure instantly scannable
Current line overlay — visually connects past performance to today's line, helping users spot trends like streaks or consistent underperformance
Averages section — highlights over %, under %, and average stat totals, providing quick reference points for users who prefer summary stats over full history
Vertical game list — displays each individual game's line and result, giving users who prefer raw data a clear, fast way to review detailed numbers
Timeframe toggle — allows users to switch between last 5, 10, or 20 games, offering flexibility to zoom in/out based on their personal research style
Home/Away toggle — added to allow users to filter performance by location, providing extra context for props that may swing depending on home court advantage
Key revisions based on team feedback:
Remove the "season" timeframe option — it introduced too much data for the available space and created potential confusion when transitioning between seasons
Added a Home/Away toggle — based on user and internal feedback, this gave an extra layer of insight without overwhelming the primary graph view
Simplified hit rate and average displays — to reduce cognitive load, these were condensed into a more digestible visual summary
Added up/down arrow icons in the vertical list view — this small but powerful addition enabled users to quickly scan past results at a glance, improving usability for both quick-check and deep-dive users
Given Rithmm's recent performance challenges, the final design had to balance richness of data with technical difficulty — ensuring the new graphs loaded quickly and reliably. This required ongoing collaboration with the engineering team to optimize both the data payload and the way graphs were rendered in-app.
Outcome & Results
This feature had an immediate and measurable impact on engagement and user sentiment. After launch, the player props page surpassed the game predictions page as the most visited screen in the app — a clear sign that users were gravitating toward the richer data experience.
In addition to improved engagement, we received overwhelmingly positive feedback from users, who appreciated finally having the missing context they needed to trust and understand player prop picks. This layer of transparency helped shift the perception of Rithmm's player props from "just another pick" to a trusted, data-backed recommendation.
Critically, the addition of this feature also reduced the need for users to leave Rithmm to do their own research on external sites. By providing recent performance data, historical lines, and clear visual trends all in one place, we became a one-stop shop for player prop analysis — keeping users in the app longer and creating a more sticky experience.
This outcome directly supported the product's larger goals of driving:
Higher engagement on player props
Stronger trust in Rithmm's model
Increased retention by making us a destination for research, not just picks
Check Out ShiftForce's Scheduler Redesign