Poll Sports
Community-driven fantasy football polls with accuracy leaderboards, web + iOS

The Problem
Every week, millions of fantasy football players make lineup decisions based on unstructured advice scattered across Reddit threads, Discord servers, and group chats. There's no way to compare opinions, no accountability for bad advice, and no signal for whose take is actually worth listening to.
The Approach
I turned fantasy questions into lightweight polls. Instead of writing a long post and hoping for responses, a user asks a specific question and the community votes. I built an Angular web app with accuracy-based leaderboards, a native SwiftUI iOS app, and a Rails API that scrapes expert projections from fantasy sources using Selenium and processes them asynchronously with Sidekiq. The Angular app and iOS app both hit the same Firebase backend with Cloud Functions handling poll scoring, NFL data sync, and weekly accuracy percentiles.
Key Insight
“Trust is everything when the product touches advice. People aren't looking for more opinions. They're looking for confidence. That means the quality of the insights, the credibility behind each vote, and the logic behind any AI-generated guidance all have to feel reliable. If the product makes bad assumptions, users feel it immediately.”
3 Repositories
poll-sports-angular
Web app with accuracy leaderboards and NFL data sync
poll-sports-ios
Native iOS app with credibility badges and poll voting
poll_sports_rails_api
Projections scraper and analyst data pipeline