Basketball Scorekeeping App

User Research, Ideation, Prototypes, App Development

 

I joined a small sports tech start up at the end of 2012, a few months after the launch of their new basketball scorekeeping app. The company built its success on scorekeeping for baseball but determined multi-sport to be crucial to the business model. The launch of basketball did not gain traction in the way that was hope and the product did not reach expected goals. A few months later it was decided that we would try again. I lead the research portion of the project and the UX for the ultimate redesign. 

Note: All work is owned by GameChanger Media. Work shown was live to consumers at some point but is not currently still in use as is. You can check out their app at gc.io and see how it's changed. 

 

Why Basketball? 

Before diving into the research, I began by exploring the listed assumption that basketball should be our next sport. Based on current youth sport enrollment numbers, adoption rate of the sport, competition and its translation to other sports, I agreed with the decision to move forward with basketball. 

Graphic from ESPN’s “The Hidden Demographics of Youth Sports” based off of SFIA’s annual report. 

Graphic from ESPN’s “The Hidden Demographics of Youth Sports” based off of SFIA’s annual report. 

 

Original Solution

I had the benefit of working on a product where a first version had already been launched and been unsuccessful. I used this as a first version prototype and aimed to determine what worked well within the app and what failed. 

  A screenshot of the original solution launched by the company.

 

A screenshot of the original solution launched by the company.

1. Tap court location to score a play 

2. Select made or missed 

3. A play line will appear for the team that shot with players numbers showing 

 

4. Select a player to associate the shot with them 

5. Swipe to see last 6 plays by the team 

6. Tap the Foul button if a foul occurs 

7. Tap to sub players in and out 

To determine our successes and failures, I began testing the app with coaches at conventions, in our office with videos of games and on courts. My findings came down to this:

Successes

1. Tap the court for location and action is intuitive 

2. Court slide feels engaging. 

 

Failures

1. Lack of core features with no roadmap of how they will work. (Assists, Steals, Blocks, And 1s, Individual Rebound and Turnovers were all missing) 

2. Easy to get lost. 

3. Could barely keep up with the speed of the game. 

4. Fouls are confusing and a point where people fall behind 

5. Users are unable to keep up with substitutions 

6. Unable to track time. 

Coaches repeatedly suggested that they liked the start but the product wasn't done. At the time, we had no clear road to how new features would be added to the product. 

 

Research - Methods

After evaluating our current product, I decided to step back and look at basketball as a whole. This included: 

2 weekend long basketball clinics with coaches 

7 Phone Calls with coaches and league administrators 

4 in office user testing with potential or current users 

5 live game scoring experiences of various levels 

1 book on APBR Metrics 

Began reading Deadspin, ESPN, Grantland, SBNation and bleacher report daily to understand the culture of basketball 

Watched the full NBA Finals 

 

Research - Insights

The key finding for this project was understanding how possessions tell the time of the game. Like an "At Bat" in baseball, the possession is the smallest meaningful chunk of time in a basketball game and determines each game's rhythm. Most coaches inherently understand this. 

Dean Oliver's shorthand from "Scoring Basketball on Paper" provides a clear visual breakdown of what following possessions look like. Each line is a possession with a code for what happened in each. By following the possession, it makes it possible to scribe and watch a basketball game at the same time. 

The second major finding was that all stats can be described as either independent or dependent. Independent stats can happen at any times - a made or missed shot, a turnover or a foul. Dependent stats required an independent stat to occur first - assist, steals, blocks, rebounds and free throws. This dictated how these two types should be treated in our interface. 

 

Solution

Key Features

1. Vertical Orientation 

2. Single stream of game play, making dependent stats feasible 

3. Focus on current play 

4. All independent stats on court 

5. Solidified scoreboard 

Process

The vertical orientation was the breakthrough for our redesign. While seemingly obvious after the fact, there was a strong desire to give basketball a horizontal orientation to match the heuristics of the court and provided a larger court space. We found that the vertical orientation allowed for a court of the same size and the change in orientation let us to present the game as a stream, similar to play by play features standard on every fan based product. 

Our main goal with the vertical orientation was to associate the current play with the court so users wouldn’t get lost during the transition from watching the court to scoring on the device. 

Our process included

  • Initial Sketches
  • Wireframes
  • Wireframed Flows
  • Paper and code based prototypes

Due to the small screen size and of the intricacies of the interactions, our UX and UI were developed hand in hand. Often, visual solutions for simplifying the interface would lead to clearer interactions. For instance, using the stat labels themselves as buttons was initially done to make it easier to understand how various stats were being added. Once visualized, it became clear we could make editing plays after the fact much easier by letting users tap on old buttons, rather than using undo and redo. Our Undo/ Redo had cause tremendous problems for users, making this a huge win. 

 

Building it

After initial concept approval, we sat down with our mobile team and decided on a 3 sprint process

As the app was being built, the team released new alphas daily. We began testing the product 3-4 times a week. I sent our weekly emails with venues, directions, times of games as well as instructions on what to look for as you scored. I personally scored 3 games a week myself until we set up twice weekly tests on external users at live basketball games.

This massive testing effort let us catch issues early and have full team buy in to the process. It got the full team invested in the successes we had already made while showing us all where we still needed to improve.